THE CATHOLIC UNIVERSITY OF AMERICA

Reputations between Enemies: Examining Threat Credibility in the U.S.-North Korea Rivalry

A DISSERTATION

Submitted to the Faculty of the

Department of Politics

School of Arts & Sciences

Of The Catholic University of America

In Partial Fulfillment of the Requirements

For the Degree

Doctor of Philosophy

By

Van Allen Jackson

Washington, D.C.

2014

Reputations between Enemies: Examining Threat Credibility in the U.S.-North Korea Rivalry

Van Allen Jackson, PhD

Director: Wallace J. Thies, PhD

How do the past actions of rival states have consequences in present crises, if in fact they do? Central to this question is the concept of reputation; the idea that states take into account the past words and deeds of other states when assessing the credibility of those states’ future words and deeds. That scholars disagree about how to answer the above question, on theoretical and empirical grounds, is a problem because the answer may hold the key to explaining two puzzles in international rivalries: (1) Why do some rivalries manage not to go to war despite experiencing repeated crises, and (2) why would a weaker state in an asymmetric rivalry repeatedly challenge its stronger rival? This dissertation seeks to offer an explanation for these puzzles at the same time that it advances the academic debate about reputations in international politics.

Two hypotheses reflecting the logic of reputation are tested against four episodes in U.S.-

North Korea relations, an ideal-type asymmetric rivalry. The first hypothesis posits that backing down from a confrontation initiated by a rival challenger: (1) strengthens the future threat credibility of the state that backed down; (2) increases the likelihood of future rival challenges;

and (3) de-escalates the crisis in which backing down occurred. The second hypothesis posits that initiating challenges against a rival—as opposed to reacting to the challenges of a rival— does not increase the future threat credibility of the challenger.

In the U.S.-North Korea rivalry, I find strong but imperfect support for both hypotheses.

Despite a U.S. history of backing down from North Korean challenges, North Korea found U.S. threats credible in the rare instances when the United States made them. Each time the United

States backed down from North Korea, crisis conditions also abated, but additional North Korean challenges followed each instance of backing down. I also find that North Korea’s history of initiating challenges did not strengthen its reputation for resolve; U.S. officials instead attributed a reputation for bluffing to North Korea because the frequency and intensity of its threatening rhetoric surpassed its actual behavior.

This dissertation by Van Allen Jackson fulfills the dissertation requirement for the doctoral degree in world politics approved by Wallace J. Thies, PhD, as Director, and by Andrew Yeo, PhD and Maryann Cusimano Love, PhD as Readers.

______Wallace J. Thies, PhD, Director

______Andrew Yeo, PhD, Reader

______Maryann Cusimano Love, PhD, Reader

ii

Table of Contents

Signature Page ...... ii Preliminary Materials Table of Contents ...... iii List of Tables ...... vii Acknowledgements ...... viii Chapter I: Introduction ...... 1 The Problem ...... 1 The Argument ...... 2 The Way Ahead ...... 5 Reputation Optimism: The Classical View ...... 6 Reputation Pessimism: Contemporary Evidence ...... 9 Renewed Optimism on Reputations ...... 15 Resolve, Honesty, or Both? ...... 17 Limits of the Reputations Literature ...... 24 Bringing Rivalries Back In ...... 31 The Effects of Rivalry on Escalation ...... 32 Limits of the Rivalries Literature ...... 35 What is a Rivalry? ...... 36 Toward a Synthesized Research Agenda ...... 41 Chapter II: Research Design and Methodology ...... 43 Definitions and Assumptions ...... 43 Hypotheses ...... 48 Operationalizing Reputations ...... 51 Competing Hypotheses on Credibility ...... 53 Observing Ideal-Type Evidence ...... 55 Case Selection ...... 61 Why the U.S.-North Korea Rivalry? ...... 61 What Makes This Case a Rivalry? ...... 65 Criteria for Event Selection ...... 67 Analytical Method ...... 69 Sources of Evidence ...... 70 iii

A Methodological Clarification ...... 71 A Theoretical Clarification ...... 74 Chapter III: The USS Pueblo Crisis (1968) ...... 78 Historical Context ...... 78 North Korea v. United States ...... 79 Cheap Talk and Credibility, 1965-1968 ...... 84 Geopolitical Context ...... 88 Overview of the Incident ...... 92 Debating Retaliation...... 96 Orchestrating U.S. Actions ...... 100 Crisis Resolution through Conciliatory Bargaining ...... 110 The North Korea Perspective ...... 115 Analysis: Explaining Decisions and Outcomes ...... 124 Explanatory Power of Reputations for Honesty ...... 125 Explanatory Power of Reputations for Resolve ...... 133 Alternative Explanations about Credibility ...... 138 Assessing Hypothesis Performance ...... 143 Chapter IV: The EC-121 Shoot Down (1969)...... 147 Historical Context ...... 148 North Korea v. United States ...... 148 Cheap Talk and Credibility before the Shoot Down ...... 152 Geopolitical Context ...... 156 Overview of the Incident ...... 162 Debating Retaliation...... 163 Orchestrating U.S. Actions ...... 181 The North Korea Perspective ...... 188 Analysis: Explaining Decisions and Outcomes ...... 194 Explanatory Power of Reputations for Honesty ...... 194 Explanatory Power of Reputations for Resolve ...... 199 Alternative Explanations about Credibility ...... 205 Assessing Hypothesis Performance ...... 209

iv

Chapter V: The Panmunjom Crisis (1976) ...... 212 Historical Context ...... 213 North Korea v. United States ...... 213 Cheap Talk and Credibility before the Crisis ...... 217 Geopolitical Context ...... 223 Overview of the Incident ...... 231 Debating Retaliation...... 233 Orchestrating U.S. Actions ...... 240 The North Korea Perspective ...... 249 Analysis: Explaining Decisions and Outcomes ...... 263 Explanatory Power of Reputations for Honesty ...... 264 Explanatory Power of Reputations for Resolve ...... 268 Alternative Explanations about Credibility ...... 273 Assessing Hypothesis Performance ...... 278 Chapter VI: The North Korean Nuclear Crisis (1993-4) ...... 283 Historical Context ...... 284 North Korea v. United States ...... 284 Geopolitical Context ...... 288 Overview of the Incident ...... 295 Debating Coercion ...... 299 Orchestrating U.S. Actions ...... 310 The North Korea Perspective ...... 316 Analysis: Explaining Decisions and Outcomes ...... 324 Explanatory Power of Reputations for Honesty ...... 325 Explanatory Power of Reputations for Resolve ...... 329 Alternative Explanations about Credibility ...... 332 Assessing Hypothesis Performance ...... 336 Chapter VII: Conclusion: Findings and Implications ...... 343 Summary of Findings ...... 345 Pueblo Crisis ...... 345 EC-121 Shoot Down ...... 346 Panmunjom Crisis ...... 348 v

Nuclear Crisis ...... 349 Overall Summary ...... 352 Hypothesis Performance...... 353 Implications for Theory ...... 356 Optimism, Pessimism, and Coercion Theory...... 357 Resolve, Honesty, and Cheap Talk ...... 358 Rivalries ...... 360 Implications for Policy ...... 363 Backing Down is a Complex Decision ...... 363 Bullying May Not Pay...... 365 Bluffing Has Costs, But Not Always the Same Costs ...... 366 Areas for Future Research ...... 368 Bibliography ...... 374 Reputations ...... 374 Rivalries ...... 380 North Korea ...... 384

vi

List of Tables

1. Table 1: The Effects of a Reputation for Honesty versus Resolve …………..…...... 21

2. Table 2: Competing Hypotheses on Credibility: Interpreting Evidence …………….54

3. Table 3: Honesty, or Resolve? Hypothesized Effects of Reputation on Intra-Rivalry Crisis Behavior ………………………………………………………..56

4. Table 4: H1 Results: The Effects of Backing Down in a Rivalry Context ………...334

5. Table 5: H2 Results: The Temporal Effect of Challenge Initiation …..……………335

vii

Acknowledgements

If you see a turtle on a fencepost, you know it didn’t get there on its own. No man can accomplish a project this long without the help of others along the way. I’ve been lucky enough to owe a debt of gratitude to many people who have touched my life and made sure I saw this dissertation through to the end. I thank my parents for showing me that you don’t have to wait for your life to be perfectly situated before launching into a new venture. Their work ethic while raising three children was a constant reminder that no matter how hard I thought I worked, I was lucky to be doing so. I also owe a large debt to Alan Narz, Neil Devera, and the entire Big

League family. They not only taught me what it means to be entrepreneurial, but they kept me off the streets at a time in my formative years when gang life was pervasive. If not for them I could’ve gone down a very different path as a youth.

To the entire faculty in the Politics Department at Catholic University, thank you for supporting my intellectual development. I am particularly grateful for the support of Wallace

Thies, Andrew Yeo, and Maryann Cusimano Love for their support throughout the PhD program, but also during the dissertation process. Each of them brings a distinct perspective to the study of international relations and I have benefited from their intellectual eclecticism. Prof.

Thies has shown me what a truly great teacher and scholar looks like; the care he takes in mentoring students is matched only by the care he takes in producing rigorous scholarship. He

viii

hails from a generation of scholars that cares about the quality and historical contingency of research findings, regardless of what methods and topics may be de rigueur in political science.

Although any errors in this dissertation belong solely to me, my empirical attentiveness is due largely to his example. I must give thanks to Prof. Yeo for his friendship, mentorship, and for introducing me to constructivism and critical approaches to international relations. He’s enabled me to think differently about the world and international security. Most importantly, he’s been the greatest mentor on Asia and Korea issues a young scholar could hope for, patiently reviewing and critiquing everything I’ve thrown his way…much of which wasn’t very good. The significance of my research is due mostly to him.

While I couldn’t possibly mention all the people who have helped me along the way, there are a number of people who have made an impression on me for the better. I have greatly appreciated consulting with Patrick Bratton at Hawai’i Pacific University; his research has influenced mine, and our frank conversations about the “politics” of the Politics Department have been invaluable. I’m grateful to Terry Roehrig for our many discussions about Korea, and for twice giving me the opportunity to lecture at the U.S. Naval War College—the most enjoyable lectures of my life to date. I must extend my appreciation for the friendship of Mike

Mazarr at the U.S. National War College; he proves that it’s possible to make meaningful contributions to both the academic and foreign policy realms. A special thanks are due to Dan

Chiu and Mara Karlin, not only for giving me a great, stimulating place to work in the Office of the Secretary of Defense, but also for pushing me to finish my dissertation no matter how crazy the Pentagon got. Thanks are also due to Mike Horowitz, Chris Twomey, and Mike Glosny, all

ix

of whom are top notch scholars, and all of whom kicked me in the butt from time to time to make sure I finished my dissertation.

Finally, I am eternally indebted to my wife, Kristin Chambers, for making me the luckiest guy in the world. She believed in her core that I would be “Dr. Jackson” even when I had doubts about my ability to persevere. She’s my muse, my intellectual peer, my biggest supporter, and my best friend. I’d be nowhere without her. Many of “my” research ideas are actually due to our frequent, impromptu conversations about love, life, justice, and politics. She was an expert on the subject of this dissertation long before I was, and she’s been tireless in helping me get through every step of the PhD process. She has patiently critiqued my work, offered suggestions, and made me feel like anything is possible. This is only one of many adventures we’ll take together!

Van Allen Jackson, PhD Washington, D.C. February 19, 2014

x

Chapter I: Introduction The Problem

What, if any, are the consequences of backing down from a rival’s challenge? Does backing down in a rivalry context strengthen future threat credibility, encourage future challenges from rivals, or is it irrelevant? Does initiating challenges against rivals make future threats more credible? Is a firm response to a rival’s challenge the only way to make future threats more credible? Are there other ways to signal credibility? In short, how do the actions of rival states in a crisis have consequences beyond the crisis in which they occur, if in fact they do? Central to these questions is the concept of reputation; the idea that states take into account the past words and deeds of other states when assessing the credibility of their future words and deeds. That scholars disagree about how to answer the above questions, on theoretical and empirical grounds, is a problem because the answers have direct consequences for policymakers facing questions of war, peace, and crisis management.

The answers may also provide insights into two puzzles in a separate academic literature on rivalries: Why are some rivalries able to experience serial crises without escalation to war, and why do weaker states frequently challenge their stronger rivals? The generalized questions and puzzles described above can be particularized in the history of U.S.-North Korea relations.

Why did the U.S.-North Korea rivalry experience repeated crises without escalation to war?

Why did North Korea, a much weaker power, repeatedly unleash small-scale violence against the

United States? And on what basis did the United States and North Korea assess the threat credibility of one another?

1 2 The Argument

In this dissertation, I focus on four episodes in U.S.-North Korea relations in order to both understand the implications for competing theories about the causal effects of reputations in international rivalries and generate insights into credibility-related decisions and actions between the United States and North Korea. Based on the reputations literature, I hypothesize that backing down from a confrontation initiated by a rival state generates both reputational costs and benefits in the form of three effects: 1) it perpetuates rivalry by encouraging rivals to initiate future crisis-inducing challenges; 2) it makes rivals perceive future threats as more credible because it represents a costly form of honest communication; and 3) it de-escalates the crisis in which it occurs, explaining why hostile rivals who experience crisis recurrence nevertheless manage not to escalate to larger conflict. I also hypothesize that initiating challenges against rivals—as opposed to reacting against them—does not have a determinate effect on a state’s reputation for resolve, even though attempting to deter and defend against a rival’s challenge does contribute to a reputation for resolve. These are the hypotheses for which I expect to find evidentiary support in the U.S.-North Korea rivalry. Their conceptual foundation fits squarely with Reputation Optimists in the reputations debate described below.

I put these reputationally based hypotheses to the test against the claims of scholars who have rejected the idea that reputations matter in international politics to show that it is reputational theories that better account for a wider range of phenomena that we observe between the United States and North Korea. A narrative grounded in reputations for resolve, and to some extent reputations for honesty, performs superiorly compared to alternative explanations that

3 focus on present, circumstantial factors. I argue that introducing a reputational approach can help us improve our understanding of many aspects of U.S.-North Korea relations, including: repeated crisis occurrence and the frequency of North Korean challenges; why the United States did not believe North Korean threats before each crisis but did believe them following crisis onset; why North Korea was not deterred from conducting violent provocations by superior U.S. power alone; and why North Korea found U.S. threats credible in the rare occasions when it did make them. No alternative to reputational theories can account for this breadth of outcomes.

Given the surge of research arguing against the merits of reputational explanations described in the literature review below, the value of this research to the reputations literature is primarily its contribution as an important corrective to this wave of pessimism. It adds to a growing body of research helping to vindicate reputation, a causal mechanism crucial to deterrence theory, but it does so in a way that directly answers the challenges of Reputation Pessimists by focusing on threat credibility in a rivalry context—a major shortcoming of recent literature attempting to vindicate reputation, discussed in the next section.

Going beyond theory testing, these findings answer the puzzles raised at the beginning of this dissertation. The first puzzle that the U.S.-North Korea rivalry poses to the rivalries literature asks: if rivalries are analytically important because they are more prone to escalation and war than other types of relationships, then why do some rivalries—such as the U.S.-North

Korea rivalry—manage to control escalation despite experiencing repeated crises? Backing down de-escalates intra-rivalry crises at the same time that it encourages the prolongation of the hostile relationship. The second, related puzzle relates to asymmetric rivalries asks: why do

4 weaker states frequently challenge their stronger rivals? The hostility inherent to a rivalry provides motivation for the weaker state to challenge the stronger, but the weaker state is not deterred by the power asymmetry because the stronger state accrued a reputation for irresoluteness—the weaker state perceives it will not suffer consequences for its use of small- scale militarized violence.

Finally, gaining a better understanding of reputational dynamics between the United

States and North Korea also has important implications for an under-examined tension within the reputation optimism camp of tremendous importance to policymakers—the consequences of backing down. Based on my observations of the United States and North Korea, I show that states in a rivalry dyad can generate a reputation for honesty and for irresoluteness based on the same action. Backing down thus has temporally differentiated costs and benefits. Depending on the context, policymakers would be just as wrong to think they can back down without suffering future consequences as they would be to think they have to fight a war simply to preserve their image. Academic research dismissing the causal relevance of reputation has reached the simplistic conclusion that states do not suffer consequences for backing down; research focused on reputations for honesty have gone still further, arguing that backing down has both immediate and long-term benefits, not costs. I describe all these claims in the literature review below, but this research serves as a rebuttal against such a simplistic view of the consequences of policymaker actions.

5 The Way Ahead

Two distinct research programs in international politics share a common interest in the causal implications of past behavior: one is focused on reputation, and the other on rivalries. The first research program, on reputation, concerns itself with the extent to which a state’s future behavior will be based on its past behavior. This literature, divided between Reputation

Optimists and Reputation Pessimists, centers on a debate about whether and under what conditions states and statesmen develop reputations of various kinds. The former camp shares in common the belief that past words and deeds affect present calculations in some way; the latter, more recent camp has offered a variety of explanations that share a broad belief that past behavior is not relevant in present crisis situations.

The second research program, on rivalries, concerns itself with a subset of interstate relationships—rivalries—that share certain characteristics about which scholars lack consensus but that generally meet two criteria: (1) states in a rivalry harbor hostility toward one another; and (2) each state has high expectations of future conflict with respect to its rival. Despite a shared concern about the interdependence of past and present, the reputations and rivalries literatures have largely evolved independent of one another. This is not just an abstract epistemological problem: I argue in this chapter that each research program has limitations or deficiencies, some of which may benefit from developing a more integrated research agenda.

There are as yet untapped gains to be realized from the hybridization of these separate research

6 programs, consistent with Larry Laudan’s observation that “…there are a number of different research traditions which can, in principle, provide the suppositional basis in any given theory.”1

In this chapter, I review the insights and deficiencies of both the reputations and rivalries literatures. In so doing, I not only identify important avenues for future research, but develop a new theoretical proposition that commingles these heretofore separate bodies of research.

Specifically, I suggest the possibility that backing down from a confrontation generates a reputation for honest communication that increases the credibility of future threats, but it also generates a reputation for irresoluteness, which increases the likelihood of future challenges. In this way, backing down from confrontation might generate competing reputations that at once prevents escalation in a crisis from getting out of hand while in turn making rivalry perpetuation through future crises more likely. Crisis escalation control, rivalry perpetuation, and threat credibility may all be functions of backing down from a confrontation. Such a hypothesis represents only one testable example of the kind of insights that might be possible by drawing the reputations and rivalries research programs closer together.

Reputation Optimism: The Classical View

Early thinking on the requirements and strategy of deterrence during the Cold War recognized the importance of demonstrating resolve in past conflicts for making credible deterrent threats in the future. Following Thomas Schelling’s original articulation, this causal relevance of the past to the future was referred to as the “interdependence of commitments,” the idea being that a state’s resolve to prevent some proscribed action at one place and time will

1 Larry Laudan, Progress and Its Problems: Toward a Theory of Scientific Growth (Berkley: University of California Press, 1977), 85.

7 influence others’ perception of that state’s resolve later.2 In an anarchical international system, states lack information about other states’ intentions, which leads them to pay particular attention to states’ behavior to divine the intentions of others.3 Illustrating the point, Schelling writes,

“We lost thirty thousand dead in Korea to save face for the United States and the United

Nations…and it was undoubtedly worth it. Soviet expectations about the behavior of the United

States are one of the most valuable assets we possess in world affairs.”4 Similarly, Hermann

Kahn argued that a “willingness to incur casualties in limited wars” would sometimes be necessary to strengthen the U.S. position in future confrontations.5 A state’s reputation for resolve thus referred to the temporal and spatial interdependence of its commitments; its reputation was an image that other states held of it, based on its past conduct.6

The primary concern driving this theorizing was how to make credible threats to deter undesired behavior or compel desired behavior. For such Reputation Optimists, successful deterrence required that threats be credible, and credibility required not only the capability to carry out a threat, but also the believability that a state had the resolve—that is, the willingness— to incur the costs of blood and treasure necessary to implement a threat.7 Deterrence strategy, therefore, implicitly became a task of figuring out ways and means to manipulate one’s own reputation for the sake of making threats appear credible; standing firm in the face of aggression

2 Thomas Schelling, Arms and Influence (New Haven: Yale University Press, 1966), 124. 3 See, for example, Robert Jervis, Perception and Misperception in International Politics (Princeton, NJ: Princeton University Press, 1976), ch. 6. 4 Schelling, Arms and Influence, 124-5. 5 Hermann Kahn, On Thermonuclear War (Princeton: Princeton University Press, 1960), 566. 6 Robert Jervis, The Logic of Images in International Relations (Princeton: Princeton University Press, 1970). 7 William Kaufman, “The Requirements of Deterrence,” in Military Policy and National Security, eds. William Kaufman, (Princeton, NJ: Princeton University Press, 1956), 12-38; Schelling, Arms and Influence.

8 was considered essential to avoid developing a reputation for weakness in future conflicts.8

Taking this logic to its extreme, some believed that states could develop a reputation for resolve in high-stakes conflicts by demonstrating resolve during conflicts with lower stakes, essentially

“free riding” on a reputation for resolve earned previously and comparatively cheaply.9 But this claim was a rather unbounded extrapolation based on the logic of reputation.10

By contrast, although Schelling is often associated with this universal interdependence- of-commitments claim, he recognized limits to the temporal interdependence of events. For example, he argues that if the Soviets invaded California the United States would have no choice but to demonstrate resolve, for “there is no way to let California go to the Soviets and make them believe nevertheless that Oregon and Washington, Florida and Maine…cannot be had under the same principle.”11 This illustrates a bounded—as opposed to a universal—conception of the interdependence-of-commitments logic; an attack on California could not logically be considered as different in kind from an attack on any other part of the United States, and the issue at stake— like Korea—would be taking place in the specific context of the superpower rivalry. Schelling is sometimes caricatured as advocating that the U.S. confront challenges anywhere at any time for the sake of building a reputation for resolve, but this may be undeserved. It was Schelling, after

8 Jervis, The Logic of Images, 6. 9 Jervis, The Logic of Images, 46; Kahn, On Thermonuclear War; Glenn H. Snyder, Deterrence and Defense: Toward a Theory of National Security (Princeton: Princeton University Press, 1961), 36-7; Glenn H. Snyder, “The Security Dilemma in Alliance Politics,” World Politics 36 (1984): 469-70. 10 Although a logical extreme, a contemporary quantitative study has found strong support for this most liberal interpretation of the interdependence-of-commitments logic, arguing that State A’s past behavior with State B affects how States C, D, and E assess the credibility of State A’s threats against State B. A real-world example that illustrates this argument is that U.S. behavior in Libya is seen by North Korea, and affects North Korea’s assessment of the credibility of U.S. threats toward it. See Mark J.C. Crescenzi, “Reputation and Interstate Conflict,” American Journal of Political Science 51, no. 2 (2007): 382-96. 11 Schelling, Arms and Influence, 56.

9 all, who warned that preserving one’s reputation for resolve was “worth some cost and risk,” but this did “not mean that in every instance it is worth the cost or risk of that occasion.”12 Wallace

Thies similarly posited that “the U.S. would be better off not attempting coercion at all” in some instances; standing tough and demonstrating resolve could actually damage reputation if threats failed to achieve their objectives.13 Considerable empirical work on reputation and deterrence subsequently proved consistent with the claims of Schelling and Thies, finding that states are more likely to earn a reputation for resolve or weakness under certain scope conditions— specifically within rivalries with a history of prior conflicts—than in others, implying that it therefore did not make sense to stand firm and risk conflict in all instances.14

Reputation Pessimism: Contemporary Evidence

After the Cold War, a new wave of scholarship began challenging the notion that states assign reputations to other states. Such claims had important implications for scholarship and policymakers. If states could not earn a reputation for resolve, then the logic of rational deterrence theory had no empirical standing.15 Of perhaps greater consequence, the Reputation

Pessimists claimed that since states cannot develop a reputation for resolve, policymakers who entered or prolonged conflicts for the sake of making future threats more credible were wasting

12 Schelling, Arms and Influence, 125. 13 Wallace J. Thies, When Governments Collide: Coercion and Diplomacy in the Vietnam Conflict, 1964-1968 (Berkley: University of California Press, 1980), 416. 14 Paul K. Huth and Bruce R. Russett, “What Makes Deterrence Work? Cases from 1900-1980,” World Politics 36, no. 4 (July 1984): 496-526; Paul K. Huth, “Reputation and Deterrence: A Theoretical and Empirical Assessment,” Security Studies 7, no. 1 (1997): 72-99; John David Orme, Deterrence, Reputation, and Cold War Cycles (London: Macmillan, 1992); J. Shimshoni,, Israel and Conventional Deterrence (Ithaca, NY: Cornell University Press, 1988). 15 Although only partially related to the reputations debate on which I focus here, much of the third wave of deterrence theory challenged the deductive, largely untested logics of the prior wave of deterrence theory, which included the interdependence-of-commitments logic of reputations in deterrence. For a review of the various waves of deterrence theory and their consequences, see Jeffrey W. Knopf, “The Fourth Wave in Deterrence Research,” Contemporary Security Policy 31, no. 1 (April 2010): 1-33.

10 money and lives. Such conclusions directly and explicitly challenged the early waves of deterrence scholars like Schelling, and provided theoretical justification for a more pacifistic foreign policy. The claims of Reputation Pessimists therefore had a strong normative overtone, even if based on developed theoretical expectations and empirical observations.

As early as 1977 scholars found that states only rarely assigned a reputation for resolve to other states.16 Limiting his inquiry to U.S. crisis behavior in the Third World during the Cold

War, Ted Hopf found that U.S. decisions to back down from conflict in the Periphery did not alter Soviet perceptions of U.S. resolve, suggesting that state A is not likely to assign a reputation for resolve or weakness to state B based on state B’s behavior in crises with states C,

D, and E, as long as interests in the issues at stake are low.17 Following Hopf, Jonathan Mercer’s theoretically innovative research made an even stronger claim about the irrelevance of reputation, drawing on social psychology’s theory of in-group/out-group biases to explain why states can almost never accrue a reputation for weakness or lack of resolve.18 According to

Mercer, states explain other states’ behavior based on either dispositional or situational criteria; that is, either the circumstances of the situation are used to explain a state’s behavior, or the overall image of the state is used to explain its behavior.

The cognitive biases resulting from in-group/out-group distinctions lead states to perceive enemies as always having a reputation for resolve, from which he concludes that a state will not

16 Glenn Snyder and Paul Diesing, Conflict among Nations: Bargaining, Decision-Making, and System Structure in International Crises (Princeton: Princeton University Press, 1977), 188. 17 Ted Hopf, Peripheral Visions: Deterrence Theory and American Foreign Policy in the Third World, 1965-1990 (Anne Arbor, MI: University of Michigan Press, 1994). 18 Jonathan Mercer Reputation & International Politics (Ithaca, NY: Cornell University Press, 1996).

11 perceive its enemy as weak or lacking resolve. If an enemy demonstrates weakness or a lack of resolve, the behavior is rationalized as situation-specific, leading the state’s rival to believe that such weakness is not likely to recur; if an enemy demonstrates resolve by risking or engaging in conflict, out-group perceptual biases lead its rival to attribute this behavior to the dispositional character of the enemy.19 Although Mercer conducted a plausibility probe as an initial test of his theoretical explanation in three crises involving England, France, Russia, Germany, and Austria prior to World War I, from 1905-1911, his theory has not been tested against a broader universe of cases and has come under critical scrutiny for theoretical and empirical reasons.20

Mercer’s claim that states were unlikely to develop a reputation for weakness was supported by several scholars subsequently, though based on the rational, rather than the psychological, theoretical tradition. Advancing a structural realist explanation for why reputations do not matter, Shiping Tang argued that the uncertainty inherent to the anarchical structure of the international system meant that states could not trust other states, leaving states to take the conservative course of always assessing other states’ threats as credible.21 Like

Mercer, Tang explained that states nearly always assigned a reputation for resolve to other states.

Unlike Mercer, Tang’s explanatory mechanism was structural rather than psychological; states assumed other states made credible threats because to not judge threats credible would be to imprudently make one vulnerable. In a statistical study, Vesna Danilovic found that reputation

19 Mercer, Reputation & International Politics, 44-73. 20 Dale C. Copeland, “Do Reputations Matter?” Security Studies 7, no. 1 (1997): 33-71; Paul K. Huth, “Reputation and Deterrence: A Theoretical and Empirical Assessment,” Security Studies 7, no. 1 (1997): 72-99. 21 Shiping Tang, “Reputation, Cult of Reputation, and International Conflict,” Security Studies 14, no. 1 (2005): 34- 62.

12 played no causally significant role in states’ credibility assessments of other states, pointing instead to interests and material capabilities as the elements of credibility.22 Danilovic later modified this finding slightly, concluding that a reputation for resolve was not irrelevant but rather of secondary causal importance, identifying instead that the interests at stake were a necessary condition for a reputation for resolve, as well as the primary causal predictor of threat credibility.23 In a qualitative study of three major crises in the 20th century—British and German appeasement of Hitler in the interwar period, Kennedy’s Berlin crisis, and the Cuban missile crisis—Daryl Press found no evidence of reputation driving states’ credibility assessments in these crises, pointing instead to material capabilities and interests as the sources of threat credibility.24 The various arguments advanced by the Reputation Pessimists shared in common that they challenged the interdependence-of-commitments logic advanced by Reputation

Optimists like Schelling. Although a theoretical challenge, the policy implications are great: if each dispute is independent of all past disputes, then all causally relevant variables can be found within the confines of the geography, power distribution, issue stakes, or capabilities of the current dispute; states can back down from conflict without having to worry about developing a reputation for weakness, and by extension, inviting future aggression.

There are several deficiencies with the arguments of Reputation Pessimists, however.

When evidence of reputation as a causal factor was absent in the cases that Reputation Pessimists

22 Vesna Danilovic, “The Sources of Threat Credibility in Extended Deterrence,” Journal of Conflict Resolution 45, no. 3 (2001): 341-369. 23 Joe Clare and Vesna Danilovic, “Reputation for Resolve, Interest, and Conflict,” Conflict Management and Peace Science 29, no. 1 (2012): 3-27. 24 Daryl G. Press, Calculating Credibility: How Leaders Assess Military Threats (Ithaca, NY: Cornell University Press, 2005).

13 analyzed, all were quick to judge that reputations are simply not causally relevant, offering their own perspectives about what matters and why, instead of looking for factors that might have impeded reputations from accruing, or considering different approaches to observing reputations.

The absence of evidence, in other words, was taken to constitute evidence of absence. This may be inappropriate for multiple reasons. First, the vast literature on bureaucratic politics and organizational theory points to the complex and occasionally conflicting processes of government decision-making and action that can all too easily impede signaling and affect perceptual biases at the state and individual levels of analysis.25 It is more exception than rule, moreover, that all officials within a government share a consensus view about anything; who views what as credible matters because officials may disagree, and may have differing degrees of proximity to a government’s final decision. Walter and Tingley also found that under the conditions of controlled laboratory experiments, reputation often had strong effects, but that occasionally cognitive weaknesses and biases in test subjects impeded the effects of reputation.26

This suggests the possibility that when reputations of any type do not form in a specific case, it might be due to intervening variables in specific cases rather than the sweeping conclusion that reputations are not causally relevant. Notwithstanding Mercer and Press, however, most empirical studies by Reputation Pessimists have been statistical studies, which do not lend themselves to analyzing cognitive or bureaucratic intervening variables; in statistical studies,

25 Morton Halperin, “Why Bureaucrats Play Games,” Foreign Policy 2, no. 2 (Spring 1971):70-90; David A. Welch, “The Organizational Process and Bureaucratic Politics Paradigms: Retrospect and Prospect,” International Security 17, no. 2 (Autumn 1992): 112-46. For the application of the bureaucratic politics framework to explain foreign policy outcomes, see Thies’ explanation of the failure of the U.S. bombing campaign in Vietnam. Thies, When Governments Collide. 26 Dustin H. Tingley, and Barbara F. Walter, “The Effect of Repeated Play on Reputation Building: An Experimental Approach,” International Organization 65, no. 1 (Spring 2011): 343-65.

14 cases are reduced to data points, and complex intervening variables are difficult, if not impossible, to control for. And even in the qualitative case study approaches of Mercer and

Press, an unintentional case selection bias may be at play: the question of why reputations did not form in their respective case selections is not asked, presumably because both arguments require that a state’s past behavior does not affect the credibility of its threats in the current dispute.

Second, in the study of reputation in particular, it is extreme to reach the conclusion that reputations do not matter given that reputations are often an “unobservable”; statesmen can be influenced by the reputations of other states without verbalizing or documenting that influence.

Press’s analysis of the Cuban missile crisis is an example of such a shortcoming. Scholars have suggested that a motivation for Khrushchev to place offensive missiles in Cuba—initiating the

Cuban missile crisis—was his observation of U.S. concessions made in 1961 regarding the U.S.-

Soviet confrontation over Berlin.27 If Khrushchev’s decision to initiate what became the Cuban missile crisis was in some way caused by perceptions of American irresoluteness based on U.S. concessions on Berlin the year prior, then it makes little sense to conclude as Press does that reputation in this case did not matter; even if reputation did not play a causal role in how the crisis ended, the fact that there was a crisis at all may have been because of reputation—that is, beliefs that Kruschev formed about the United States based on Kennedy’s earlier behavior. Most troubling for Reputation Pessimists’ arguments is that it presents a puzzle for which no good

27 For a discussion of linkages between the Berlin and Cuban missile crises, see, for example, Marc Trachtenberg, Marc, A Constructed Peace: The Making of the European Settlement, 1945-1963 (Princeton: Princeton University Press, 1999), ch. 7-8.

15 explanation exists: why do statesmen invest so much in building reputation if the same statesmen do not assign reputations to others? As Press puts it, his “rationalist story seems to have a nonrationalist underpinning.”28

Renewed Optimism on Reputations

In response to this surge of pessimism about the relevance of reputations and the confirmation that reputations proved causally important in other fields such as cognitive psychology and economics, international relations scholars have begun something of a renaissance in the study of reputation in the political context, advancing the modern body of reputations research with refreshed theoretical grounding, typological diversity, and methodological plurality. The new wave of reputation optimism expanded beyond the traditional security focus of the classical deterrence thinkers, and has yielded innovative research showing that a variety of reputations have causal relevance in a variety of contexts.

Strong theoretical and empirical contributions come from Barbara Walter, who integrated reputations research in economics with the study of intra-state separatist conflicts. Using controlled laboratory experiments, Barbara Walter and Dustin Tingley were able to show that under conditions of incomplete information—an apt description of the condition facing states in international anarchy—individuals not only invested in reputation-building when they expect many repeated interactions, but they also found that building a reputation for toughness (as

28 Press, Calculating Credibility, 158.

16 opposed to a reputation for resolve)29 helped deter challenges.30 Walter established in an earlier work combining statistical analysis, case study analysis, and laboratory experiments that states found success deterring future violent self-determination movements when they fought or prolonged wars against separatists, which served as an effective costly signal to deter future challengers.31 Walter’s theoretical framework was built on an economics literature that explained how monopolies deter new firms from market entry, identifying the repeated play and incomplete information conditions that give reputation salience.32 She identified three defining criteria of repeated play: observable (as opposed to non-attributable clandestine) government action; repeated interaction; and issue specificity, which acknowledges that reputation building based on economic logic “should not have a consistent effect…across different issue areas, but it should have a strong effect when one looks at similar players, fighting over similar issues, against the same opponent over time.”33 The second condition necessary for reputation building, incomplete information, refers in particular to the uncertainty about the value that a government places on the stakes of a conflict.34

29 Reputations for toughness and resolve, while similar, are not the same, though the term is often used interchangeably. Toughness refers to a willingness to endure high costs to achieve a goal, while resolve refers to a willingness to risk conflict to achieve a goal. 30 Tingley, and Walter, “The Effect of Repeated Play,” 343-65. 31 Barbara F. Walter, Reputation and Civil War: Why Separatist Conflicts are So Violent (Cambridge: Cambridge University Press, 2009). 32 The foundations of reputation in the economics literature include David M. Kreps, and Robert Wilson, “Reputation and Imperfect Information,” Journal of Economic Theory 27 (1982): 245-79; Paul Milgrom and John Roberts, “Predation, Reputation, and Entry Deterrence,” Journal of Economic Theory 27 (1982): 280-312; Richard Selten, “The Chain-store Paradox,” Theory and Decision 9 (1978): 127-59; Richard Schmalensee, “Advertising and Entry Deterrence: An Exploratory Model,” Journal of Political Economy 91 (1983): 636-53. 33 Walter, Reputation and Civil War, 12-14. 34 Walter, Reputation and Civil War, 14-15.

17 Gregory Miller pointed out that the empirical focus of the arguments of Reputation

Pessimists was mostly limited to whether states form a reputation for resolve or credibility in a crisis context, arguing that resolve was only one type of reputation that states might form.

Drawing on pre-World War I alliance formations, Miller showed that not only do states form a reputation for reliability or unreliability based on their past behavior, but also that a reputation for reliability was causally relevant; states found it easier or more difficult to form military alliances depending on its past reputation for reliability.35 Michael Tomz reinforces Miller’s claims about a reputation for reliability using a mixed method approach of statistical analysis, case studies, and content analysis to show that states earn reputations for being reliable debtors, and that international creditors’ decisions to lend to sovereign governments, and at what rate of borrowing, is affected by a state’s history of repaying sovereign debt.36 Reputation Optimists have also combined the domestic audience costs literature37 with modeling and laboratory experiments to show that voter disapproval of statesmen who make empty threats occurs out of recognition that such a mismatch between words and deeds erodes the state’s future credibility.38

Resolve, Honesty, or Both?

35 Gregory D. Miller, The Shadow of the Past: Reputation and Military Alliances before the First World War (Ithaca, NY: Cornell University Press, 2012); Gregory D. Miller, “Hypotheses on Reputation: Alliance Choices and the Shadow of the Past,” Security Studies 12, no. 3 (Spring 2003): 40-78. 36 Michael Tomz, Reputation and International Cooperation: Sovereign Debt across Three Centuries (Princeton: Princeton University Press, 2007). 37 The seminal work on audience costs remains James D. Fearon, “Domestic Political Audiences and the Escalation of International Disputes,” American Political Science Review 88, no. 3 (1994): 577-92. 38 Alexandra Guisinger and Alastair Smith, “Honest Threats: The Interaction of Reputation and Political Institutions in International Crises,” Journal of Conflict Resolution 46, no. 2 (April 2002): 175-200; Michael Tomz,. “Domestic Audience Costs in International Relations: An Experimental Approach,” International Organization 61, no. 4 (October 2007): 821-40.

18 The new wave of reputation optimism has also included an attempt by some to shift focus from a reputation for resolve to a reputation for honesty, by which scholars usually mean consistency between the words and deeds of government. In her theoretical and empirical work to highlight how a reputation for honesty strengthen the credibility of threats, Anne Sartori positioned its logic as a competing alternative to that of the reputations for resolve work of earlier deterrence thinkers.39 The logic of reputations for honesty is important not simply because it has policy implications that partially contrast those of reputations for resolve, but also because it has been utilized to generate insights about the interaction effects of reputation and domestic institutions,40 as well as the conditions under which states can learn to cooperate with other states41 and credibly mediate disputes as third party honest brokers.42 For reasons discussed below, it is noteworthy that Sartori neither denies the existence of, nor tests for, reputations for resolve in her work; she only refutes its causal claims, to the extent they contradict her findings. The theoretical logic for why and how a reputation for honesty matters focuses on the communicative aspect of threats and threat credibility—whether the word and deed of a state match—rather than the content of communication. Whereas a reputation for resolve or irresoluteness develops as a function of whether a state responds firmly or acquiesces to a challenger in a dispute, a reputation for honesty or bluffing forms as a function of whether a

39 Anne E. Sartori, “The Might of the Pen: A Reputational Theory of Communication in International Disputes,” International Organization 56, no. 1 (2002): 121-149. 40 Guisinger and Smith, “Honest Threats.” 41 See, for example, Andrew H. Kydd, Trust and Mistrust in International Relations (Princeton, NJ: Princeton University Press, 2005). 42 Andrew H. Kydd, “When Can Mediators Build Trust?” American Political Science Review 100, no. 3 (2006): 449-462.

19 state communicates honestly or gets caught bluffing.43 Although Sartori only passingly acknowledges it in her work,44 the idea that consistency in a state’s words and deeds has causal implications for its future credibility is not new. Some among the previous waves of deterrence thinkers saw honesty as crucial to credible signaling and a mismatch between a state’s words and deeds as detrimental to future credibility.45 Moreover, despite Sartori’s attempt to distance a reputation for honesty from a reputation for resolve along theoretical and empirical lines, both types of reputation share an interdependence-of-commitments logic; both see past behavior as affecting credibility in the present. The difference between Sartori’s conception of the interdependence of commitments over space and time and that of earlier deterrence theorists is therefore more apparent than real, being largely one of degree. The real difference of consequence between a focus on honesty vice resolve is the former’s focus on only word-deed alignment while the latter focuses on both word-deed alignment and the importance of behaving in a way that demonstrates a willingness to fight when one’s interests are threatened. As Sartori acknowledges, there is considerable overlap in terms of the observable causal expectations of these alternative conceptions of reputation. Standing firm in response to a challenge is an honest communication of resolve, so both types of reputation would expect that future threat credibility increases as a result of standing firm.

43 Sartori defines “bluffing” as making a threat in response to a challenge that fails to deter the challenger, and that the defending state—despite making a threat—then acquiesces to the challenger. Anne E. Sartori, Deterrence by Diplomacy (Princeton, NJ: Princeton University Press, 2007), 85. 44 Sartori points to Jervis (1970) as the theoretical foundation on which she builds her theory, but does not acknowledge that the same work is also foundational to the reputations for resolve logic. Sartori, Deterrence by Diplomacy, 43. 45 In addition to Jervis, The Logic of Images, 80-81, see Thomas Schelling, The Strategy of Conflict (New York: Oxford University Press, 1963), 24-30, 36-38; Schelling, Arms and Influence, 35-91; Thies, When Governments Collide, 406-15.

20 There are, however, principally two observable causal differences between a reputation for resolve and one for honesty, both of which result from backing down or offering conciliation in response to the initiation of a challenge. First, for Sartori and likeminded others, a reputation for honesty is the causal mechanism that explains why backing down from a challenge is a form of honest communication about the importance (or rather unimportance) the conciliating state places on the issue at stake, strengthening the credibility of future threats, should the conciliating state make them. Deterrence theory, however, generally holds that backing down cultivates a reputation for weakness or lack of resolve, which means that in future challenges, a defending state’s threats are, ceteris paribus, less likely to be credible. The second observable difference when comparing the effect of a reputation for honesty with one for resolve is in the likelihood of subsequent challenges. Sartori finds that a challenger in one dispute is less likely to challenge the same defending state in subsequent disputes if the defending state develops a reputation for honesty.46 Sartori finds that the opposite also holds true; a challenger is much more likely to challenge a defending state in subsequent disputes if the defending state develops a reputation for dishonesty (that is, gets caught bluffing).47 By contrast, scholars focusing on a reputation for resolve expect that, based on the interdependence-of-commitments logic under conditions of repeated play and incomplete information, a state that backs down from a challenge is more

46 Sartori, Deterrence by Diplomacy, 111-19. 47 Ibid.

21 likely to experience future challenges.48 Table 1, below, illustrates the different causal expectations between a reputation for honesty and one for resolve.

Table 1 The Effects of a Reputation for Honesty versus Resolve Reputation for Honesty Reputation for Resolve

Threat Credibility If State A has a State B believes State A’s threats State B does not believe State A’s History of Backing threats Down from State B,

If State A has a State B does not believe State A’s State B does not believe State A’s History of Bluffing threats threats against State B,

If State A has a State B believes State A’s threats State B believes State A’s threats History of Standing Firm against State B, Likelihood of Future Challenges If State A has a State B is less likely to challenge State B is more likely to challenge History of Backing State A in the future State A in the future Down from State B,

If State A has a State B is more likely to challenge State B is more likely to challenge History of Bluffing State A in the future State A in the future against State B,

If State A has a State B is less likely to challenge State B is less likely to challenge History of Standing State A in the future State A in the future Firm against State B,

So why do the causal arrows for reputations for honesty and resolve point in opposite directions when it comes to backing down from a challenge? The answer is slightly different for explaining claims about threat credibility than for claims about conflict recurrence. From the

48 Although pervasive in the reputations for resolve literature, see especially Jervis, The Logic of Images, 6; Orme, Deterrence; Tingley and Walter, “The Effect of Repeat Play.”

22 honesty perspective, backing down from a challenge increases future threat credibility because the state that backed down during the last confrontation has proven that it is willing to back down when its interests in the issue at stake are not sufficiently engaged. If the state that previously backed down is now standing firm, the challenging state is more likely to conclude that its interests must now be sufficiently engaged, and its threats are therefore credible. From the resolve perspective, backing down from a challenge decreases future threat credibility because the challenger state has added incentive to engage in the same behavior repeatedly; if the challenger state mounted a successful challenge at no cost at one point in time, why would it not be able to do so again?49

Sartori’s explanation for her claim about the effect of honest communication is more circumspect when it comes to conflict recurrence. She makes the empirical observation that states that communicate honestly discourage future challenges, but she admits that her explanation for why this is so is only speculation: “potential challengers recognize that defenders with a reputation for bluffing are at a disadvantage, and challenge them more often because they are easy targets.”50 The somewhat tortured logic that Sartori displays in this speculative explanation represents her attempt to provide a logical underpinning for her finding, rather than

49 A large body of research that focuses on the difference between status quo and revisionist states, claiming that the causal implications of certain kinds of behavior—such as conciliation—depends on the character of the target state. Such distinctions, while important, are endogenous to the phenomena analyzed in the reputations and rivalries literatures. See, for example, Stephen R. Rock, Appeasement in International Politics (Lexington, KY: University of Kentucky Press, 2000); Randall L. Schweller, “Bandwagoning for Profit: Bringing the Revisionist State Back In,” International Security 19, no. 1 (1994): 72-107; Charles L. Glaser, Charles L., “Political Consequences of Military Strategy: Expanding and Refining the Spiral and Deterrence Models,” World Politics 44, no. 4 (July 1992): 497-538; Andrew H. Kydd, “Sheep in Sheep’s Clothing: Why Security Seekers Do Not Fight Each Other,” Security Studies 7, no. 1 (1997): 114-55; Evan Braden Montgomery, “Breaking Out of the Security Dilemma: Realism, Reassurance, and the Problem of Uncertainty,” International Security 31, no. 2 (2006): 151-85; James W. Davis, Threats and Promises: The Pursuit of International Influence (Baltimore, MD: Johns Hopkins University Press, 2000). 50 Sartori, Deterrence by Diplomacy, 118.

23 question it. Why would a state that gets caught bluffing be an easier target than a state that has proven it is willing to back down from a challenge? Even if backing down constitutes a form of honest communication, does it not also communicate a lack of resolve (that is, a lack of willingness to defend against challenges)? If so, would it not follow that, ceteris paribus, challengers are just as likely to initiate challenges in the future against bluffing states as they are against honest states if the latter’s honesty takes the form of acquiescence? It may be the case that Sartori’s logic is illogical because her empirical finding is spurious, struggling in search of an explanation other than that offered by reputations for resolve logic to support a finding whose difference from that expected by the reputations for resolve logic is perhaps merely illusory.

Because she does not control for reputations for resolve in her analysis, it is inappropriate for her to attribute a shift in the dependent variable (a would-be challenger becoming discouraged) to a reputation for honesty when the observable indicator for a reputation for honesty and a reputation for resolve are the same when the defending state stands firm. That is, if honesty takes at least two observable forms (standing firm and backing down), but one of these forms (standing firm) is also an observable form for resolve, then one cannot claim that honesty is the cause of the would-be challenger becoming discouraged. The reputations for resolve literature predicts that a would-be challenger becomes discouraged when the defending state stands firm. For Sartori to attribute this discouragement to her variable of choice (honest communication), she would need to control for reputations for resolve by reducing her sample to only those cases in which the causal arrows for resolve and honesty point in opposite directions (cases where the defending state backed down). Because the explanation for how a reputation for honesty affects conflict

24 recurrence differ from how such reputation affects threat credibility, it is possible that backing down increases the credibility of future threats at the same time that it encourages future challenges, just as the scholars emphasizing reputations for resolve would expect. Testing a reputation for resolve against reputation for honesty in additional cases would help clarify whether this is true.

The reputations literature demonstrates a progressive evolution: Reputation Pessimists’ challenge to early Reputation Optimists led to a new wave of reputation optimism with a stronger and clearer theoretical grounding, diversity in the types of reputations identified as causally relevant, a variety of methodological approaches to test for the effects of reputation, and clarified scope conditions for when reputations are most likely to matter.

Limits of the Reputations Literature

Despite progress, the reputations debate between Optimists and Pessimists remains far from conclusively resolved. As evidenced by the honesty vice resolve contest discussed above, even among Reputation Optimists there is disagreement about which reputations matter and why.

Overcoming several limitations of the extant reputations literature should bring further clarity and, hopefully, something eventually approaching a consensus regarding the appropriate intellectual guideposts for policymakers. There are at least four such shortcomings, which suggests where our analytical attention can add the greatest value to the reputations research program: (1) most scholars have failed to distinguish between reputation-building by initiating challenges from reputation-building by defending against challenges; (2) the key point of contention between reputations for honesty and reputations for resolve—the effects of states’

25 decisions to back down from confrontations—requires not simply tests against a larger number of cases, but a focus on longitudinal case studies in order to contextualize decisions to back down; (3) studies of a reputation for honesty have drawn linkages between honesty and the vast literature on audience costs, but there has been no attempt to separate the causal effects of audience costs from that of reputations for honesty, which is significant because both share the same causal expectations in certain circumstances; and (4) new-wave reputation optimism has tended to avoid analyzing interstate rivalries, the context that gave the reputations debate its consequentialness in the first place.

A largely overlooked issue within the research on reputation is the question of whether the same reputational effects can be achieved by initiating challenges as by defending against them. Deterrence and coercion theorists have argued that a reputation for resolve is earned by fighting, but fighting when confronted is different than initiating fights to demonstrate resolve.

Hermann Kahn and Bernard Brodie suggested that initiating challenges may be necessary to build a reputation for resolve,51 and Jervis’s earlier work might be construed to support such thinking because of how he suggests states can manipulate others’ image of itself;52 Schelling’s discussion of strategy that manipulates risk might be interpreted similarly.53 But William

Kaufman,54 Albert Wohlstetter,55 and much of the third-wave deterrence literature implicitly focused on gaining a reputation for resolve in the context of responding to—rather than

51 Kahn, On Thermonuclear War; Bernard Brodie, Escalation and the Nuclear Option (Princeton: Princeton University Press, 1966). 52 Jervis, The Logic of Images. 53 Schelling, Strategy of Conflict; Schelling, Arms and Influence. 54 Kaufman, “The Requirements of Deterrrence.” 55 Albert Wohlstetter, “The Delicate Balance of Terror,” Foreign Affairs 37, no. 2 (1959): 211-34.

26 initiating—a challenge. A central limitation of these early works is that they were deductive, and as such marshaled little empirical evidence in support of their claims. Later works by Jervis,

Thies, and George and Smoke, all of whom took a more inductive approach to the study of deterrence and reputation, pointed to cognitive bias,56 bureaucratic, 57 and contextual factors58 respectively that made coercion a troublingly difficult endeavor.

Deterrence literature more generally has also distinguished compellence from deterrence as distinct types of coercion on both theoretical and empirical grounds, the former being much more difficult than the latter.59 Since initiating a challenge (as opposed to responding to a challenge) often constitutes a form of compellence, it follows from the general wisdom of the deterrence literature that initiating a challenge poses considerable risks of failure, and therefore is something to be avoided in most circumstances; at a minimum one must be circumspect about decisions to threaten or use force because of the complexities inherent to government action and the social world more generally.60 But these conclusions were implications drawn from research that had a reputation for resolve as only part of its logic rather than an exclusionary research focus on reputation per se. Within the new wave of reputation optimism, Sartori’s work on

56 Jervis, Perception and Misperception, ch. 6. 57 Thies, When Governments Collide; Wallace J. Thies, “Compellence Failure or Coercive Success: The Case of NATO and Yugoslavia,” Comparative Strategy 22, no. 2 (203): 243-67. 58 Alexander L. George and Richard Smoke Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974). 59 For a discussion of the controversies and implications of distinguishing compellence from deterrence, see Patrick C. Bratton, “When is Coercion Successful? And Why Can’t We Agree on It?” Naval War College Review 58, no. 3 (Summer 2005): 99–120. For works arguing that compellence is more difficult than deterrence, see, for example, Schelling, Arms and Influence; Thies, When Governments Collide; Thies, “Compellence Failure”; and Daniel Byman and Matthew Waxman, The Dynamics of Coercion: American Foreign Policy and the Limits of Military Might (New York: Cambridge University Press, 2002). 60 This is a common theme throughout Thies’ explanation of the 1964-68 U.S. bombing campaign in Vietnam. Thies, When Governments Collide. See also, Robert Jervis, “Complexity and the Analysis of Political and Social Life,” Political Science Quarterly 112, no. 4 (1997/98): 569-93.

27 reputations for honesty stands out as the only contemporary attempt to distinguish reputational implications of being a challenger vice a defender in a confrontation. Although not marshaling case studies or developing a theoretical explanation distinct from her broader argument about the effects of honest communication on credibility, Sartori’s statistical analysis indicates that initiating challenges could achieve a similar effect as defending against challenges, but this finding only holds for a reputation for honesty; she did not seek to test the effects of challenge initiation on a reputation for resolve.

Also missing from the reputations literature are meaningful comparisons of a reputation for resolve with a reputations for honesty. We know that proponents of a reputation for honesty and resolve both expect that standing firm or making threats when challenged will generate the same causal effect—greater threat credibility and thus a higher likelihood of coercive success— but each type of reputation predicts consequences of backing down that point in opposite directions. What we need is to observe the competing expectations for honesty and resolve together, against new cases. Most effective in this regard would be a longitudinal case study approach that contextualizes and—to the extent possible—draws on process tracing evidence to explain decisions to stand firm against or back down from the challenge of another state, as opposed to statistical and case study tests that often treat crisis cases as discrete, temporally independent events. Of note, the enduring rivalries literature, discussed at greater length below, marshals strong empirical support for the causal relevance of the entire history of a rivalry relationship on present disputes, in contrast to studies asserting that events are either temporally

28 independent or that only the last event matters.61 As Sartori acknowledges, it is possible that a reputation for honesty and resolve both hold true. At first glance this would seem impossible: each predicts different effects from one another when a state backs down from confrontation.

However, if we see a defender who repeatedly backs down repeatedly challenged subsequently, it would be reasonable to conclude that the act of backing down is probably generating a reputation for irresoluteness; and if we see threats have a credible effect following past instances of backing down from a confrontation, the act of backing down may be generating a reputation for communicating credibly. So if both types of reputation are causally relevant, we should see backing down lead to subsequent challenges, but if the state that backed down makes future threats after having backed down, we should also see such threats judged as credible.

Studies arguing for the causal relevance of only a reputation for honesty face an additional challenge because of the need to separate the potential effects of audience costs in a specific event from the effect that honesty in one event has on threat credibility in subsequent events. Audience costs, by which scholars often mean domestic political consequences of observable foreign policy actions,62 are a causal mechanism that has proven important to debates within the democratic peace literature, but whose contributions most salient to this study consist of its use to explain why threatening signals are rare (audience costs create strong disincentives

61 See, for example, Gary Goertz, Bradford Jones, and Paul F. Diehl, “Maintenance Processes in International Rivalries,” The Journal of Conflict Resolution 49, no. 5 (October 2005): 742-69; Russell J. Leng, Bargaining and Learning in Recurring Crises: The Soviet-American, Egyptian-Israeli, and Indo-Pakistani Rivalries (Ann Arbor: University of Michigan Press, 2000); Paul Hensel, “An Evolutionary Approach to the Study of Rivalry,” Conflict Management and Peace Science 17 (1999): 175-206; Michael P. Colaresi and William R. Thompson, “Hot Spots or Hot Hands? Serial Crisis Behavior, Escalating Risks, and Rivalry,” The Journal of Politics 64, no. 4 (November 2002): 1175-98. 62 Fearon, “Domestic Political Audiences,” 581.

29 to bluff)63 and why clashing states would opt out of bargaining with a competitor in order to opt in to war (audience costs rigidify bargaining positions as crises evolve, making peaceful resolution more difficult).64 Reputation Optimists focused on honesty vice resolve have drawn on the logic of audience costs to reinforce claims both about why honesty is common and why it pays. Alexandra Guisinger and Alastair Smith posited that threats and promises are more likely to be credible when statesmen are accountable to domestic constituents; Sartori argued that domestic and international audience costs incentivize states to make threats and promises honestly; and Michael Tomz found that in controlled laboratory experiments domestic audiences punished statesmen who failed to execute threatened actions. Yet the concept of audience costs and its purported effects has recently come under great scrutiny. Although the existence of audience costs are not a necessary condition for a reputation for honesty to have causal effects, we must wonder whether statesmen are motivated to make threats and promises honestly in order to preserve reputation, or to avoid domestic audience costs. We must also question whether threat credibility in the reputation for honesty model derives from a track record for honest communication, or whether it derives from other states’ recognition that domestic audience costs are “tying the hands” of leaders.

But is it even possible to separate audience costs from a reputation for honesty given that the latter has incorporated the logic of the former? Like reputation, audience costs are a mechanism that is often not directly observable. Because the effects of such mechanisms are

63 Alexander B. Downes and Todd S. Sechser, “The Illusion of Democratic Credibility,” International Organization 66, no. 3 (2012): 457-89. 64 Fearon, “Domestic Political Audience,” 560; James D. Fearon, “Signaling Foreign Policy Interests: Tying Hands Versus Sinking Costs,” International Organization 41, no. 1 (1997).

30 observable, however, the existence of a reputation for honesty, as distinct from audience costs, would be most evident if a state chooses to communicate honestly (by either backing down or demonstrating resolve) even though domestic audiences express a strong preference for the opposite action. Rational models of audience costs have always suggested that domestic audiences would prefer honest communication (backing down or fulfilled threats) to bluffing, but this is a modeling assumption; it is quite conceivable that domestic audiences demand action or conciliation in a crisis but the state’s leaders decide to act contrary to the demands of domestic audiences. If statesmen choose to communicate honestly (by either backing down or fulfilling threats) in such instances even when domestic preferences expect the opposite communication, the importance of a reputation for honesty in international politics would be better insulated from the spate of recent research casting doubt on the implications of audience costs. Thus, research that finds support for a reputation for honesty (states maintaining consistency between word and deed) while finding a corresponding lack of support for the effects of audience costs (statesmen acting contrary to domestic preferences) would strengthen confidence in the former.

The final limitation of the reputations literature is a sin of omission among new-wave

Reputation Optimists: the absence of reputations research in a dyadic rivalry context. This limitation is especially important because it is within rivalries that we should be most likely to find a high degree of future conflict expectations and thus repeated interaction, which Reputation

Optimists have identified as an important scope condition for when and how reputation is most likely to matter. More innovative methods, such as controlled experiments, are insightful but face reasonable questions of external validity, given their attempts to extrapolate from the

31 behavior of college students in a laboratory setting to the behavior of states in the international system. Greater diversity in identifying types of reputations, such as a reputation for being tough against separatist movements, a reputation for being a reliable ally, and a reputation for being a reliable debtor, strengthens confidence that reputations are causally relevant, but sidesteps the relevance of a reputation for resolve in a rivalry context. Rivalries analysis would not only replicate the original context of the optimism-pessimism debate, but arguably focus on the most salient context for matters of war and peace. Moreover, one of the problems with reputations for honesty research generally has been its seeming theoretical and empirical aversion to examining the effects of honest behavior in a rivalry context, which, as discussed below, often strongly colors the intentions, behavior, and perceptions of each party to the rivalry.

Bringing Rivalries Back In

All dyadic relationships are not the same. Interstate rivalries account for a disproportionate share of crises and conflicts in international politics, and the entire research program on interstate rivalries has been premised on this empirical observation.65 It follows, at least as an assumption in the rivalries literature, that scholars investigating the causes of war would do well to look for answers in that relative handful of interstate relationships we call

“rivalries,” a category of relationship that has proven both crisis and conflict prone. The rivalries literature shares a similar ontological view of history and conflict with that of Reputation

Optimists. Rivalries scholars, for instance, have shown that multiple disputes between the same

65 Gary Goertz and Paul F. Diehl, “Enduring Rivalries: Theoretical Constructs and Empirical Patterns,” International Studies Quarterly 37 (1993): 147-71.

32 states must be interpreted in a temporal context, rather than as independent events,66 and that a history of conflict is a strong indicator of future conflict.67 The rivalries literature in general addresses three broad aspects of rivalries: initiation;68 maintenance and effects;69 and termination.70 Both puzzles that this dissertation addresses—why the U.S.-North Korea rivalry managed to avoid escalation to war despite theoretical expectations to the contrary and experiencing recurring crises, as well as why a weaker state so frequently challenges its stronger rival—contributes primarily to our understanding of the effects of rivalries and how they are maintained.

The Effects of Rivalry on Escalation

Within the sub-group of rivalries research that focuses on rivalry maintenance processes and the effects of rivalry, we find multiple reinforcing explanations for why rivalries are more prone to conflict escalation and crisis recurrence than other relationships. Russell Leng is one of many who have argued that dispute density—a given number of militarized disputes within a

66 For a fuller discussion of this argument in the rivalries literature generally, see Paul F. Diehl and Gary Goertz, War and Peace in International Rivalry (Ann Arbor, MI: University of Michigan Press, 2000): 17-28. 67 See, for example, Mark J.C. Crescenzi and Andrew J. Enterline, “Time Remembered: A Dynamic Model of Interstate Interaction,” International Studies Quarterly 45, no. 4 (2001): 409-31; Russell J. Leng, “When Will They Ever Learn? Coercive Bargaining in Recurrent Crises,” Journal of Conflict Resolution 27 (1983): 379-419; Colaresi and Thompson, “Hot Spots or Hot Hands?” 68 Gary Goertz and Paul F. Diehl, “The Initiation and Termination of Enduring Rivalries: The Impact of Political Shocks,” American Journal of Political Science 39, no. 1 (1995): 30-52; Diehl and Goertz, War and Peace; Zeev Maoz and Ben Mor, Bound by Struggle (Ann Arbor: University of Michigan Press, 2002); Douglas Stinnett and Paul F. Diehl, “The Path(s) to Rivalry: Behavioral and Structural Explanations of Rivalry Development,” Journal of Politics 63, no. 3 (2001): 717-40. 69 Leng, Bargaining and Learning in Recurring Crises; Goertz, Jones, and Diehl, “Maintenance Processes”; Mark J.C. Crescenzi, Jacob D. Kathman, and Stephen B. Long, “Reputation, History, and War,” Journal of Peace Research 44, no. 6 (2007): 651-67. 70 Scott D. Bennett, “Integrating and Testing Models of Rivalry Duration,” American Journal of Political Science 42 (1998): 1200-32; Michael Colaresi, “Shocks to the System: Great Power Rivalries and the Leadership Long Cycle,” Journal of Conflict Resolution 45 (2001): 569-93; Christopher Darnton, “Rivalry and Rapprochement in Early Cold War Central America,” Security Studies 20, no. 2 (2011): 198-237.

33 given period—is a reasonable predictor of future conflict; the more disputes recur, the higher the probability of war.71 The causal logic that translates rivalry into escalation is a rigidifying learning process that states and statesmen undergo as a result of experiencing dyadic disputes; each crisis within a dyad makes escalation to war more likely in each future crisis within the same dyad because each crisis experience increases pressures on statesmen to adopt realpolitik coercive bargaining strategies, and to doubt the efficacy of non-coercive measures.72 Thus the incentives for mistrust grow with each hostile encounter. For this reason, Leng and others find that while the analytical assumption that states operate as unitary rational actors can be a troubling oversimplification in many instances, it comes close to the truth when describing the behavior of rivals because statesmen in rivalries mistrust one another, impute malign intentions when interpreting the behavior of the other, and find themselves subject to the same pressures to adopt a realpolitik perceptual lens when deciding on the best strategy for coping with the rival.73

In such circumstances, a willingness to engage in high-risk strategies like brinkmanship becomes much greater; indeed, the primary purpose of the earlier coercive bargaining literature was to show the potential payoffs of a willingness to manipulate risk.74

71 John A. Vasquez, The War Puzzle (Cambridge: Cambridge University Press, 1993); Leng, “When Will They Ever Learn?”; Russell J. Leng, Interstate Crisis Behavior, 1816-1980: Realism vs. Reciprocity (Cambridge: Cambridge University Press, 1993); Leng, Bargaining and Learning in Recurring Crises; Crescenzi and Enterline, “Time Remembered.” 72 Russell J. Leng, “Escalation: Competing Perspectives and Empirical Evidence,” International Studies Review 6, no. 4 (2004): 51-64; Michael Colaresi, “When Doves Cry: International Rivalry, Unreciprocated Cooperation, and Leadership Turnover,” American Journal of Political Science 48, no. 3 (2004): 555-70. 73 Leng, “When Will They Ever Learn?”; Leng, Bargaining and Learning in Recurring Crises; Leng, “Escalation”; Colaresi, “When Doves Cry”; Crescenzi, Kathman, and Long, “Reputation, History, and War.” 74 Schelling, Strategy of Conflict; Schelling, Arms and Influence; Kahn, On Escalation.

34 Rivalries scholars have also hypothesized several additional paths to escalation. First, rivalries frequently give way to escalation because intra-rivalry mistrust and hostility can overwhelm factors thought to constrain conflict, such as democratic regime type or capability asymmetry. Bruce Russett provides an exemplary argument for why and how democracy constrains interstate violence, explaining that publicly transparent decision-making and institutionalized power sharing arrangements mediate the arbitrary exercise of power, which is often assumed to be a cause of conflict.75 Yet, as Karen Rasler and William Thompson have shown, the rivalry context erodes the pacifying effects of democracy, particularly in the context of a mixed regime dyad.76 While scholars disagree about the causal direction of capability asymmetry—whether it makes conflict more or less likely—most agree that capabilities are causally relevant. 77 However, to the extent that capability symmetry or asymmetry induces restraint, John Vasquez tells us that such cost-benefit calculations become warped in a rivalry context, which, he argues, leads states to prioritize hurting their rivals over benefiting themselves.78 Rivalries are also thought to lead to escalation because of how they lower the threshold for conflict. Outside of rivalries, only conflicts over high salience issues are sufficient triggers for escalation to violence; in a rivalry context, a rival’s words and deeds take on added meaning based on the extent to which they reinforce existing biases, which means low salience

75 Bruce Russett, Grasping the Democratic Peace: Principles for a Post-Cold War World (Princeton, NJ: Princeton University Press, 1993). 76 Karen Rasler and William R. Thompson, “Rivalries and the Democratic Peace in the Major Power Subsystem,” Journal of Peace Research 38, no. 6 (2001): 659-83. 77 For the classic argument that capability asymmetry eliminates one of the incentives for great power restraint, see Kenneth N. Waltz, Theory of International Politics (Reading, MA: McGraw Hill, 1979); Vasquez, The War Puzzle. For the argument that capability symmetry, not asymmetry, makes war more likely, see Daniel S. Geller, “Power Differentials and War in Rival Dyads,” International Studies Quarterly 37, no. 2(1993): 173-93. 78 Vasquez, The War Puzzle.

35 issues and otherwise ambiguous words and deeds can become triggers for violence when they might not in non-rivalry contexts.79 Finally, rivalries have been identified as a cause of escalation to war because rivals are more likely to take advantage of other rivals’ “windows of vulnerability.”80 Richard Ned Lebow has suggested that the absolute costs (in human and economic terms) of war and the desire of statesmen to avoid taking responsibility for such costs incentivizes states to not take advantage of windows of opportunity/vulnerability,81 but Randall

Schweller has argued that revisionist states are opportunistic, and will take advantage of favorable circumstances if they perceive that the costs of doing so are low.82 In a rivalry context, the combination of a state’s potential for prioritizing hurting its rival above benefiting itself with a willingness to exploit windows of opportunity/vulnerability is much more likely to lead to war than in a non-rivalry context.

Limits of the Rivalries Literature

The rivalries literature suffers from several shortcomings, the most serious of which is a continuing disagreement about what constitutes a rivalry, though this is more of a problem for large-N statistical studies than for longitudinal studies of specific rivalries. The rivalries

79 For scholars who have posited hostile and uncertain contexts as creating sufficient conditions for low salience issues to serve as triggers, see Leng, “When Will They Ever Learn?”; Colaresi and Thompson, “Hot Spots or Hot Hands?”; Paul R. Hensel, “One Thing Leads to Another: Recurrent Militarized Disputes in Latin America 1816- 1986,” Journal of Peace Research 31, no. 3 (1994): 281-98; Richard Ned Lebow, Between War and Peace: The Nature of International Crisis (Baltimore, MD: Johns Hopkins University Press, 1981). Robert Jervis’s discussion of conflict spirals relates closely to this phenomenon. See Jervis, Perception and Misperception, 58-84; Robert Jervis, “Perceiving and Coping with Threat,” in Psychology and Deterrence, eds. Robert Jervis, Richard Ned Lebow and Janice Gross Stein (Baltimore, MD: Johns Hopkins University Press, 1988), 1-31. 80 Colaresi and Thompson, “Hot Spots or Hot Hands?”; Geoffrey Blainey, Why Nations Go to War (New York: Free Press, 1973). 81 Richard Ned Lebow, “Windows of Opportunity: Do States Jump through Them?” International Security 9, no. 1 (1984): 147-86. 82 Schweller, “Bandwagoning for Profit.”

36 literature has also reached a point in its development where we should expect greater research emphasis on deviant or outlier cases that lead to refinements of theoretical claims.

What is a Rivalry?

If, as the Chinese proverb says, knowledge begins by calling things by their right names, then the rivalries literature has a major problem: a research program built on the causal effects of a phenomenon defined differently in different studies makes consolidating knowledge gains over time questionable at best. The two basic approaches to identifying rivalries, as William

Thompson outlines, are the interpretive approach relying on perceptual criteria and the empirical approach attempting to rely on objective criteria.83 The former approach, which risks sacrificing some reliability, requires that for a dyad to constitute a rivalry each side must identify the other as an explicit threat; the identification of a rivalry depends on the perception of the states comprising the dyad. The latter approach, better for purposes of scientific reliability but requiring questionable assumptions, identifies rivalries based on the presence of militarized interstate competition and the occurrence of a certain number of disputes over a given period of time.

The perceptual approach recognizes that the origin of rivalries are often found in disputes about ideas, influence, or territory, and advance three criteria as an attempt to systematize the perceptual identification of rivals: (1) states must see each other as competitors; (2) states must

83 William R. Thompson, “Identifying Rivals and Rivalries in World Politics,” International Studies Quarterly 45, no. 4 (2001): 557-587. Other studies that have used this approach for identifying rivalries include Karen A. Rasler and William R. Thompson, “Contested Territories, Strategic Rivalries, and Conflict Escalation,” International Studies Quarterly 50, no. 1 (2006): 145-68; Sumit Ganguly and William R. Thompson, Asian Rivalries: Domestic Politics, Conflict Escalation, and Limits on Two-level Games (Palo Alto: Stanford University Press, 2012).

37 see each other as enemies; and (3) states must see each other as the potential source of militarized threat.84 As Thompson himself admits, operationalizing these conceptual criteria requires “more than usual” subjective analyst interpretation, particularly since the “competitors” criterion does not require capability symmetry between rivals, and which at any rate lacks a clean conceptual distinction from the “enemy” criterion.85 The empirical approach shares common ground with the perceptual approach with respect to two necessary conditions for the concept of a rivalry: a shared hostility between dyads, and each state having high expectations of future conflict. The two approaches diverge on the issues of militarized competition, duration of hostility, and frequency of disputes. The perceptual approach does not require either the occurrence of a dispute or that interstate militarized competition exist for a rivalry to exist; but the perceptual approach does require that hostility exist following a militarized dispute, meaning that experiencing a military conflict is not in itself sufficient for the onset of rivalry.86 The perceptual approach tends to employ the term “strategic rivalry” to distinguish that such rivalries do not have to satisfy any specific time duration, nor do they need to have experienced an actual militarized dispute. By contrast, the empirical approach requires not only that militarized disputes occur for rivalries to exist, but that the relationship must realize a certain number of disputes within a certain time period. Paul Diehl and Gary Goertz define rivalries as hostile relationships with high expectations of future conflict in which six militarized disputes occur

84 Thompson, “Identifying Rivals.” 85 Thompson, “Identifying Rivals,” 567-8. 86 Thompson, “Identifying Rivals.”

38 within a twenty year period.87 Diehl and Goertz also distinguish between isolated (1-2 disputes), proto (3-5 disputes), and enduring (6 disputes) rivalries within twenty years, a gradation that other studies do not attempt.88 Also utilizing the empirical dispute-density approach but operationalizing it with different thresholds, Scott Bennett defines rivalries as those relationships experiencing five militarized disputes within a twenty-five year period.89 Bennett later used a different dispute-density threshold, requiring six militarized disputes within twenty years, as long as there is no more than a 15-year gap between disputes.90 The problem, as Thompson notes, is that it may be correct to consider all serial militarized disputes rivalries, but all rivalries are not serial militarized disputes, meaning the dispute-density approach yields an incomplete dataset.91

The dispute-density approach falls short on two counts: it has not yielded a common set of criteria for identifying rivalry, despite trying to be systematic in the study of rivalry, and it neglects analysis of those cases in which military conflict is latent.

As long as scholars lack consensus about the conceptual and operational criteria for rivalries, the challenges of rivalry identification will not be easily resolved for large-N studies.92

Thompson’s perceptual approach has more intuitive validity, but is difficult to replicate for large-

87 Diehl and Goertz, War and Peace in International Rivalry. 88 Ibid. 89 Scott D. Bennett, “Security, Bargaining, and the End of Interstate Rivalry,” International Studies Quarterly 40 (1996): 157-83; Scott D. Bennett, “Democracy, Regime Change, and Rivalry Termination,” International Interactions 22 (1997): 369-97. 90 Scott D. Bennett,. “Integrating and Testing Models of Rivalry,” American Journal of Political Science 42 (1998): 1200-1232. 91 Thompson, “Identifying Rivals,” 575. 92 Case study approaches to rivalries, while comparatively few in number, have started appearing in recent years, though few draw on insights from the rivalries literature. See Janice Stein, “Deterrence and Learning in an Enduring Rivalry: Egypt and Israel, 1948-73,” Security Studies 6 (1996): 104-52; Karen Rasler, “Political Shocks and the De- escalation of Protracted Conflicts: The Israeli-Palestine Case,” in Evolutionary Interpretations of World Politics, ed. William R. Thompson (New York: Routledge, 2001), 240-60; Leng, Bargaining and Learning in Recurring Crises; Ganguly and Thompson, Asian Rivalries.

39 N studies. The dispute-density approach is certainly more systematic, but as the different operationalizations above suggest, hardly objective, still requiring a degree of seemingly arbitrary interpretation. We should keep in mind that one “basis for judging any definition of enduring rivalries is that it match our intuition about what cases qualify as enduring rivalries and exclude those from historical knowledge that we think deserve to be excluded.”93 It would seem that if definitional consensus escapes us but that intuition should play an important role in identifying rivalries, an inductive, small-N approach to the study of rivalries may have a comparative advantage. To the extent that scholars have pursued case studies of specific rivalries, they have chosen rivalries that intuition tells us are ideal-type rivalries, such as between

Egypt and Israel, Pakistan and India, or China and Taiwan.94 Since the perceptual and empirical approaches share a belief that hostility and high expectations of future conflict are necessary conditions for identifying rivalry in concept, the rivalries research program would benefit from case analyses of those dyads that meet these conceptual criteria but that are only coded as rivalries in some datasets and not others. If we care about rivalries because a range of factors make conflict more likely if they are present in a rivalry context than not, then we should be finding and analyzing the causal mechanisms that make conflict more or less likely in the rivalry context; this is a task best done through case study analysis, and it need not wait for agreement about the conceptual or operational criteria used to define a relationship as a rivalry.

93 Goertz and Diehl, “Enduring Rivalries,” 164. 94 Leng, Bargaining and Learning in Recurring Crises; Stein, 1996; Andrew Scobell, “China and Taiwan: Balance of Rivalry with Weapons of Mass Democratization,” in Asian Rivalries: Conflict, Escalation, and Limitations on Two-level Games, eds. Sumit Ganguly and William R. Thompson (Palo Alto: Stanford University Press, 2012), 26- 43.

40 This discussion of a case study approach to addressing the problem of identifying rivalries brings us to a second limitation of the rivalries literature, which may also lend itself to being remedied by a case study approach: explaining deviant cases. As noted above, very few of the studies comprising the rivalries literature consist of case studies, and the few case studies that exist are vehicles for illustrations in support of claims generated from statistical studies, or to inductively generate new hypotheses.95 None of these studies seek to explain statistical anomalies in other analyses of rivalries, or to answer puzzles in which empirical observations run counter to theoretical expectations. Such studies represent the next logical step in advancing the rivalries research program. The evolution of theory development often reflects a dialectic process wherein findings from large-N statistical studies are refined and caveated by analyzing deviant cases, which then leads to adjustments in subsequent statistical and case studies.96 For instance, how are we to explain rivalry dyads whose interactions do not result in escalating militarized crises to war? As discussed, the rivalries literature is premised on the idea that rivalry dyads are highly prone to escalation for multiple reinforcing reasons—realpolitik-based coercive strategies become more common, hurting one’s rival can become more important than securing gains for oneself, the constraining effects of power distribution and regime type are eroded by mistrust and hostility, and otherwise low-salience issues can trigger broader conflict— so rivalries that deviate from this theoretical and empirical expectation merit explanation,

95 For example, Stein, “Deterrence and Learning in Enduring Rivalry”; Leng, Bargaining and Learning in Recurring Crises; Rasler, “Political Shocks”; Ganguly and Thompson, Asian Rivalries. 96 Alexander L. George and Andrew Bennett, Case Studies and Theory Development in the Social Sciences (Cambridge: MIT Press, 2005), 19-21.

41 particularly for generating potential insights about how rivalries can avoid escalating crises to war.

Toward a Synthesized Research Agenda

The reputations and rivalries research programs have similar views of conflict history.

Both are premised on the temporal interdependency of state behavior; the idea that words and deeds at one point in time can have causal implications for words and deeds at a later point in time. By focusing on the role of types of reputations in a rivalry context, we bring the study of reputations back to a context in which the stakes of research and policy prescriptions bear directly on matters of war and peace. Each body of literature may also provide greater insights into the other. Reputational mechanisms may be useful in explaining deviant cases in the rivalries literature, and the presence or absence of rivalries may affect the causal impact of reputational mechanisms.

As I articulate further in the following chapter, I generate two hypotheses by bringing together these distinct bodies of research.97 First, I expect that backing down from confrontation in a rivalry context yields three effects: 1) it perpetuates rivalry by encouraging rivals to initiate future crisis challenges; 2) it makes future threats more credible because it represents a costly form of honest communication; and 3) it de-escalates the crisis in which it occurs, explaining why hostile rivals who experience crisis recurrence nevertheless manage not to escalate to larger conflict. Second, I hypothesize that initiating challenges against rivals—as opposed to defend

97 An early attempt to integrate the reputations and rivalries literature includes Crescenzi, “Reputation and Interstate Conflict.” Krista Wiegand also makes such an attempt, but does so by considering reputations for resolve in the context of disputes between general conflict dyads, as opposed to explicit sensitivity to the effects of rivalries per se. Krista E. Wiegand, “Militarized Territorial Disputes: States’ Attempts to Transfer Reputations for Resolve,” Journal of Peace Research 48, no. 1 (2011): 101-13.

42 against them—does little to contribute to a reputation for resolve. If both of these hypotheses are correct, then reputations provide a valuable explanation for why some rivalry dyads can experience recurring crises without escalation to war, and why a weaker power would be willing to repeatedly challenge a stronger rival. Additional possible avenues for future research based on integrated reputations and rivalries literatures includes the question of whether a reputation for honesty is more difficult to cultivate in a rivalry context, and to what extent and how other types of reputation—such as a reputation for reliability—might result from within-rivalry behavior.

Chapter II: Research Design and Methodology

This chapter explains the “what,” “why,” and “how” of this dissertation. I define the key terms used throughout and specify the hypotheses for which I expect to find support, I consider alternative explanations, I describe the criteria for case selection, how causal terms will be operationalized, and the analytical method I employ to produce my findings.

Definitions and Assumptions

Although touched upon in the previous chapter, it is necessary to explicitly define the key conceptual and operational terms used throughout this dissertation. The concepts of reputation and credibility need to be unpacked in order to understand their distinctions and relations to one another. As discussed above, scholars posit many types of reputation, though I focus here on the subset that has been suggested to bear the most relevance in a dyadic rivalry context, which is the focus of this inquiry. Distinguishing these concepts is important because Reputation Optimists tell us that reputations can have causal impact on credibility.

Reputation and credibility are relational concepts, meaning that each “…depends entirely on the perception of others.”1 Offering a definition of reputation for resolve common to both

Reputation Optimists and Pessimists, Shiping Tang describes reputation for resolve as “others’ perception of that state’s resolve to risk war in a given situation.”2 The limitation of this definition is its applicability only to reputation for resolve; for purposes here we need a definition that applies to reputations for resolve and reputations for honesty. I therefore use the term “reputation” throughout as referring to the perceived likelihood of a state’s future behavior

1 Mercer, Reputation & International Politics, 27. 2 Tang, “Reputation, Cult of Reputation, and International Conflict,” 38. 43 44 based on its past behavior.3 Although authors assert various strengths and types of reputations,4 the general logic of reputation that this definition captures is the temporal interdependence of events; threats, signals, and actions in one crisis can affect how an opposing state interprets the credibility of future threats, both within the same crisis and in future crises. Credibility is the extent to which others believe that a state will fulfill the threats and promises it makes if identified triggering conditions are met.5 Reputation Pessimists such as Shiping Tang, Daryl

Press, and Vesna Danilovic argue that credibility is determined by current capabilities and current interests, not past behavior.6 Reputation Optimists do not deny that current capabilities or interests matter, but rather argue that they are insufficient determinants of credibility, pointing primarily to observations of past words and deeds. For Reputation Optimists, uncertainty about the intentions of other states in an anarchical international system incentivizes states to infer meaning from behavior that it has observed.7

3 This definition of reputation hews closely to Schelling’s definition. Schelling, Arms and Influence, 124. 4 Reputational types that have been explored include resolve and toughness, cooperation, honesty, and reliability. Jervis, The Logic of Images; Walter, Reputation and Civil War; Tomz, Reputation and International Cooperation; Sartori, Deterrence by Diplomacy; Miller, The Shadow of the Past. The narrowest claim about the strength of the interdependence of commitments holds that only behavior in the previous event matters for the next event with the same opponent (Sartori, Deterrence by Diplomacy). The most sweeping claim about the strength of the interdependence of commitments holds that states are able to observe the behavior of others and that therefore past behavior with one opponent can impact future events with all other observers (Kahn, On Thermonuclear War; Snyder, Deterrence and Defense; Crescenzi, “Reputation and Interstate Conflict”). 5 Press, Calculating Credibility, 10. 6 Tang, “Reputation, Cult of Reputation, and International Conflict,”; Press, Calculating Credibility; Danilovic, “The Sources of Threat Credibility in Extended Deterrence.” 7 Jervis, The Logic of Images; Jervis, Perception and Misperception; Walter, Reputation and Civil War, 12-15; Morgan, “Saving Face for the Sake of Deterrence,” 134-6. For a review of how different approaches to the study of international relations understand the implications of uncertainty, see Brian C. Rathbun, “Uncertain about Uncertainty: Understanding the Multiple Meanings of a Crucial Concept in International Relations Theory,” International Studies Quarterly 51 (2007): 533-57.

45 The two types of reputation I am concerned with here for their effects on credibility and the likelihood of future challenges are reputations for resolve/irresoluteness and reputations for honesty/bluffing. As explained in the previous chapter, the operational distinction between these two concepts lies with their differing predictions about the consequences of backing down from confrontation: the reputation for honesty literature expects that (1) any threats made in the next crisis will be more credible and (2) future challenges will be less likely,8 while the reputation for resolve literature expects that backing down invites future aggression.9 I argue that while these projections are different, they are not necessarily completely incompatible, and to the extent that they conflict, I expect to find that the reputations for resolve hypothesis has explanatory superiority.

In order to observe whether either or both of these sets of predictions hold in specific cases, I introduce terms generally consistent with their usage in the literature on reputations and rivalries. In each crisis, the “challenger” is the state that initiates a confrontation with a rival, either through verbal threat, a signal such as military mobilization or a show of force, or a physical attack. The “defender” is the target of the initiated threat, signal, or attack. 10 Coercive contests are fluid in the sense that both sides make decisions and take actions often simultaneously and in environments of incomplete information, reacting not only to the

8 Sartori, Deterrence by Diplomacy; Guisinger and Smith, “Honest Threats.” 9 Kahn, On Thermonuclear War; Brodie, Escalation and the Nuclear Option; Schelling, The Strategy of Conflict; Schelling, Arms and Influence; Jervis, The Logic of Images; Orme, Deterrence, Reputation, and Cold War Cycles. 10 See, for example, Schelling, Arms and Influence; Huth, Extended Deterrence and the Prevention of War; Sartori, Deterrence by Diplomacy.

46 perceived words and deeds of the rival but also to the anticipated words and deeds of the rival.11

Given the dynamism inherent in rival crisis behavior, it is plausible—even likely—that because each side in a coercive contest makes many moves, the challenger winds up becoming the defender as the crisis unfolds, or vice versa. Because the case study approach I employ allows me to track in fine-grain detail the moves of each side throughout each crisis episode examined here, I assign the challenger and defender labels based on the initiation of the crisis, and reassign them subsequently based on the context of subsequent actions within the crisis.

The behavior of challenger and defender also requires explication. For the sake of hypothesis testing, I need to be able to classify behavior according to a limited number of categories. In keeping with common usage, I define behavioral categories following the initial challenge as: (1) “stand firm” or “defend,” which I use interchangeably, and (2) “back down,”

“conciliate,” or “acquiesce,” which I also use interchangeably. Although “fight” or other synonyms might apply to the former category, a defender need not actually engage in conflict in order to stand up to a challenger. Similarly, I avoid the term “appeasement” to describe the latter category, not only because it tends to be a politically loaded term, but also because it often implies the transfer of something tangible to the adversary.12 The former category is intended to convey that the defender chooses to demonstrate a willingness to risk conflict, either by escalating the confrontation, refusing to comply with challenger demands, or making counter threats in response to challenger demands. The latter category is intended to convey that the

11 For arguments that coercion is not linear but dynamic, see Schelling, Arms and Influence, 98-9; Thies, When Governments Collide, 397-99; Byman and Waxman, The Dynamics of Coercion. 12 For a comprehensive definition of appeasement, see Rock, Appeasement in International Politics, 10-12.

47 defender responds to a challenger’s initiation of a confrontation by choosing not to retaliate when threatened or attacked, which includes conceding to deterrent threats or complying with compellent demands.

The final term requiring definition is “escalation,” defined here as “an increase in the intensity or scope of conflict.”13 Herman Kahn popularized conceptions of escalation as existing as a “ladder” that culminated with a nuclear exchange; it was an illustrative depiction of a basic premise of the earlier coercive bargaining and limited war literature. 14 The metaphor fell out of favor not only because researchers began finding that it conveyed an inaccurate image of a state that could move up and down an “escalation ladder” at will, but it also made the unrealistic assumption that coercive actions were not necessarily context dependent; the very idea of a ladder—that is, a behavioral hierarchy based on the risk of escalation—wrongly implied that the defender and challenger would view the significance of a given coercive action the same way.15

To seek to “control” escalation is a task fraught with tremendous risk because escalation is itself often the unintentional product of prior coercive failures.16 When I refer to “escalation control,”

I thus do not imply anything about the utility of coercive bargaining techniques or manipulating risk in any deliberate sense; rather, I use the term “escalation control” synonymously with “de-

13 Forrest E. Morgan, Karl P. Mueller, Evan S. Medeiros, Kevin L. Pollpeter, and Roger Cliff, Dangerous Thresholds: Managing Escalation in the 21st Century (Santa Monica, CA: Rand Corporation, 2008), 1. 14 Herman Kahn, On Escalation: Metaphors and Scenarios (Baltimore, MD: Penguin, 1965). 15 That coercion is context dependent and that ascending an escalation ladder is much easier than climbing down one is vividly illustrated in Thies’ analysis of the U.S. bombing campaign in Vietnam. Thies, When Governments Collide. See also Alexander L. George and William E. Simons, The Limits of Coercive Diplomacy (Boulder, CO: Westview Press, 1971). 16 Thies, When Governments Collide, 71-142; Thies, “Deterrence Failure or Coercive Success?”; Walter J. Petersen, “Deterrence and Compellence: A Critical Assessment of Conventional Wisdom,” International Studies Quarterly 30, no. 3 (1986): 269-94; Barry R. Posen, Inadvertent Escalation: Conventional War and Nuclear Risks (Ithaca: Cornell University Press, 1991).

48 escalation,” “crisis termination,” and “escalation prevention” to convey that specific actions did or did not lead to an increase in the scope or intensity of conflict in specific instances.

Hypotheses

In this dissertation I generate two hypotheses based on the current state of the literature on reputations, which I test against multiple crises within the U.S.-North Korea rivalry.

Hypothesis 1: Backing down from a confrontation initiated by a rival challenger strengthens the future threat credibility of the state that backs down, increases the likelihood of facing future rival challenges, and de-escalates the current crisis.

Hypothesis 2: Initiating coercive challenges against rivals has an indeterminate effect on a reputation for resolve.

The first hypothesis consists of a combination of three hypotheses, two of which come directly from the literature on reputations, though in previous literature they have been contrasted as alternatives to one another. First, the literature on reputations for honesty have shown that states can make their future threats more credible not only by standing firm when confronted by a challenger, but also by backing down when confronted by a challenger; the only action that would hurt the credibility of future threats would be to make a threat that the challenger tests but that then goes unfulfilled.17 The other aspect of the reputations for honesty claim—contra

Schelling and others—is that honest signals make future challenges less likely, even though one way of accruing a reputation for honesty is by backing down from a challenger.18 The reputations for honesty literature thus tells us that backing down should generate a reputation for honesty, but it also tells us that reputations for honesty make future challenges less likely.

17 Sartori, “Mightier than the Sword”; Sartori, Deterrence by Diplomacy; Guisinger and Smith, “Honest Threats.” 18 Sartori, Deterrence by Diplomacy.

49 Although I expect to find support for the former claim, the latter claim contradicts a long history of findings in the deterrence literature, leading me to expect that earning a reputation for honesty by backing down makes future challenges more likely, not less. Second, the early work on reputations for resolve by deterrence theorists posits that backing down from a rival challenger generates a reputation for irresoluteness.19 Third, backing down is hypothesized to de-escalate crises. Stephen Rock, using the term “appeasement” to describe acquiescing to a challenger’s demands, argued that one of the uses of appeasement is to return crises to the status quo.20 That acquiescence or “appeasement” seems to be a rarer phenomenon than coercion in world politics is, according to David Baldwin, because the use of such positive incentives are always costly while coercion is only costly when the other party decides to challenge the coercer.21 Similar to

Rock, Snyder and Diesing found that in an effort to avoid signaling irresoluteness, states that back down in a crisis will often communicate conditional threats in tandem with acquiescence.22

Patrick Morgan reinforced this point, arguing that states that back down have strong incentives to do so in a way that seeks to avoid cultivating a reputation for irresoluteness.23 Using a game- theoretic approach, Martin Patchen also showed that conciliation defuses crises when the content of the accommodating act is perceived as addressing the grievance of the challenger.24 Even

19 This claim is pervasive in early work on coercion. See, for example, Schelling, Arms and Influence; Kaufman, “The Requirements of Deterrence”; Jervis, The Logic of Images; Huth, Extended Deterrence and the Prevention of War; Robert Jervis, “Deterrence Theory Revisited,” World Politics 31 (1979): 289-324. 20 Rock, Appeasement in International Politics, 12-15. See also Davis, Threats and Promises. 21 David A. Baldwin, “The Power of Positive Sanctions,” World Politics 24 (1971): 19-38. 22 Snyder and Diesing, Conflict Among Nations. 23 Patrick M. Morgan, “Saving Face for the Sake of Deterrence,” in Psychology and Deterrence, eds. Robert Jervis, Richard Ned Lebow, and Janice Gross Stein (Baltimore: Johns Hopkins University Press, 1986) 125-52. 24 Martin Patchen, Martin, Resolving Disputes between Nations: Coercion or Conciliation? (Durham, NC: Duke University Press, 1988), 123-68, 261-75.

50 early deterrence theorists like William Kaufmann argued in favor of occasional accommodation of enemy objectives, as one of many potential strategies for preserving deterrence against an adversary.25 The first hypothesis I pose is thus a synthesis of these disparate threads of knowledge: backing down from confrontation with a rival hurts reputation for resolve, helps reputation for honesty in a caveated form (that is, future threats will be more credible), and helps de-escalate the crisis in which it occurs.

The second hypothesis distinguishes between reputational effects that might accrue to challenger behavior from those that accrue to defender behavior. In the first chapter I described theoretical reasons to expect that a challenger initiating a threat or the demonstrative use of force might have different consequences than a defender reacting to a threat or demonstrative use of force with coercion in kind. In addition to theoretical claims that initiating a coercive contest can enhance future threat credibility,26 there are also empirical findings that suggest it is possible.

Sartori’s statistical analysis on reputations for honesty finds that initiating coercive challenges does contribute to a reputation for honesty.27 Paul Huth has similarly found that “bullying”—his term for describing the act of initiating coercive challenges—helps build a reputation for resolve, stating “…a record of intransigence and bullying will have already imposed domestic and international costs on the potential attacker and left it determined to protect its bargaining reputation and avoid another retreat before the defender.”28 In other words, bullying deters

25 Kaufmann, “The Requirements of Deterrence,” 31-2. 26 Kahn, On Thermonuclear War; Kahn, On Escalation; Brodie, Escalation and the Nuclear Option; Jervis, The Logic of Images. 27 Sartori, Deterrence by Diplomacy. 28 Huth, Extended Deterrence and the Prevention of War, 55.

51 states from initiating future challenges. But should we believe this? Given the well documented challenges in achieving coercive successes in international politics due to issues of misperception, cognitive biases, and complexities of government decision-making, it is hardly surprising that coercion tends to fail far more than it succeeds.29 A failure to align word and deed has been shown to have a negative impact on future credibility,30 and coercive failure is by its nature a failure to match word and deed. One scholar has even gone as far as saying that initiating compellence challenges is “reckless” because of the likelihood that it will lead to escalation.31 And, as Robert Jervis has argued, challengers initiating coercive confrontations are likely to back down when defenders choose to stand firm because they have difficulty convincing others that living with the status quo is no longer acceptable.32 Despite some evidence and arguments to the contrary, then, we find that when it comes to casting reputations for resolve as a dependent variable, there are theoretically and empirically founded reasons to expect that the behavior of a challenger does not have a clear, determinate effect on how others view its reputation, and initiating a coercive challenge may in some instances actually harm a state’s reputation for resolve.

Operationalizing Reputations

Thus far I have explained the causal arguments for which I expect to find support by examining episodes from the U.S.-North Korea rivalry. I have also situated my hypotheses

29 Several statistical studies have shown that coercion often fails. For a summary discussion, see Todd S. Sescher, “Goliath’s Curse: Coercive Threats and Asymmetric Power,” International Organization 64, no. 3 (2010): 627-60. 30 Sartori, Deterrence by Diplomacy; Thies, When Governments Collide; Jervis, The Logic of Images. 31 Jack Snyder, “Rationality at the Brink,” World Politics 30 (1978): 345-65. 32 Jervis, “Deterrence Theory Revisited,” 298.

52 within the existing literature by breaking down my causal arguments to show from where in the literature they derive. But what does it mean to say that the behavior of defender and challenger contribute to the development of reputational characteristics? What does that look like? To make testable sense of causal arguments about reputations, I must operationalize the concepts of reputations for honesty and resolve, in essence translating the hypotheses into observable arguments.

Recalling the discussion from the previous chapter in which I described the causal similarities and differences between the logics of reputations for honesty and resolve, there are two relevant causal claims associated with states having a reputation for honesty: (1) they make threats that others deem credible, and (2) they are less likely to experience future challenges.

The reputations for honesty hypothesis claims that both effects are achieved in one of two ways:

(1) backing down from a rival in a confrontation, or (2) standing firm against a rival’s challenge.

By contrast, states with a reputation for resolve: (1) make threats that others deem credible, and

(2) are less likely to experience future challenges. The reputations for resolve hypothesis claims that both effects are achieved by standing firm against a rival’s challenge; according to this hypothesis, backing down from a rival in a confrontation makes future challenges more likely.

Thus, when operationalized, the hypotheses on reputations for honesty and resolve agree on the effects of standing firm (strengthens threat credibility, deters future challenges), but disagree on the effects of backing down. Theoretically, confirming Hypothesis 1 strongly supports the reputations for resolve hypothesis, and partially supports the reputations for honesty hypothesis.

Empirically, confirming Hypothesis 1 reveals the decision to back down has costs and benefits: it

53 strengthens future threat credibility, makes future rival challenges more likely, and prevents crises from escalating.

Hypothesis 2 challenges the limited literature suggesting that states can positively manipulate their reputations by initiating coercive contests. To be fair, I know of no scholars who argue that initiating coercive contests is the only way to build a reputation for resolve, though as I have noted, some clearly believe it is one of multiple ways for doing so. If initiating coercive contests strengthens a reputation for resolve, then states that initiate challenges will be more likely to face defenders that believe the challenger’s subsequent threats and back down as a result. By contrast, if Hypothesis 2 is correct, then states initiating challenges will be more likely to face defenders that stand firm, responding to the challenger with counter-threats, retaliatory responses, or escalation. Theoretically, confirming Hypothesis 2 would suggest a scope condition for behavior that contributes to reputation-building: demonstrating a willingness to risk conflict as a challenger has different implications than demonstrating a willingness to risk conflict as a defender. Empirically, confirming Hypothesis 2 reveals that even if a challenger achieves proximate concessions by initiating coercion against a rival, such behavior does not reliably affect the decision-making calculus of defenders in a favorable way for a challenger.

Competing Hypotheses on Credibility

Although I have articulated the arguments for which I expect to find support and the theoretical and empirical reasons for harboring them, such expectations do not exist in a vacuum.

Competing alternative expectations exist, as discussed in the previous chapter, and for my arguments to be convincing, I not only need to evaluate evidence for the extent to which it

54 supports my arguments, but I should also evaluate how my arguments “perform” against competing alternative explanations. There are two major variants of the reputation pessimism arguments from which I construct alternative hypotheses about threat credibility.

The first variant of reputation pessimism argues that threat credibility is a function of material capabilities to implement a threat and the relative balance of national interests at stake;

Daryl Press called this hypothesis “current calculus theory” because he believes that states’ decisions to back down or stand firm in previous crises has no bearing on threat credibility in the current crisis.33 Current calculus theory addresses only one portion of the causal claims by

Reputation Optimists; that is, the scope of current calculus theory’s causal claim, despite a total rejection of the validity of the concept of reputations, is limited to threat credibility in future crises. Thus, even if empirical analysis fully supports current calculus theory, it would not necessarily serve as a total repudiation of the causal claims related to reputations for honesty and resolve. Finding support for current calculus theory leaves unaddressed the causal claim that states with a reputation for honesty and resolve are less likely to face future challenges, while states with a reputation for bluffing or irresoluteness are more likely to face future challenges from rivals. So by operationalizing current calculus theory, we find clear contrasts in expectations regarding the threat credibility dimension among reputations for honesty, reputations for resolve, and current calculus theory. Whereas the reputations for honesty hypothesis states that threats are credible if the state making the threat either backed down or stood firm in the previous crisis, and the reputations for resolve hypothesis states that threats are

33 Press, Calculating Credibility. See also Danilovic, “The Sources of Credibility in Extended Deterrence.”

55 credible if the state making the threat stood firm in the previous crisis, current calculus theory states that threats will be credible only when the balance of power favors the threatening state and vital security interests are at stake.

The second credibility-relevant variant of reputation pessimism, to which I refer as

“constant calculus theory,” can be traced to social psychology and structural realism. Jonathan

Mercer’s theory of in-group/out-group bias posits that rival states’ threats are always credible because reputations do not exist; for Mercer, the out-group lens through which states view rivals attributes malign intentions to the behavior of rivals, meaning states will not assign reputations for irresoluteness to its rivals.34 Taking a structural realist approach, Shiping Tang posits that because uncertainty about the intentions of other states is pervasive in an anarchical international system, states cannot afford to let their guards down. Tang argues that this mistrust of other states’ intentions means that states will always judge other states’ threats as credible because to do otherwise risks incurring the costs of security vulnerability.35 Although Mercer and Tang operate in different paradigms and levels of analysis, both expect to see the same outcome with regard to threat credibility: no variance. So whereas the other three perspectives on threat credibility—reputations for honesty, reputations for resolve, and current calculus theory—all anticipate variation in threat credibility based on different criteria, the constant mistrust hypothesis predicts that threats will always be credible.

Observing Ideal-Type Evidence

34 Mercer, Reputation & International Politics. 35 Tang, “Reputation, the Cult of Reputation, and International Conflict.”

56 Since the dispute among the four competing perspectives—reputations for honesty, reputations for resolve, current calculus theory, and constant calculus theory—centers on threat credibility, we need to clarify what constitutes ideal-type evidence in support of each perspective, which can be satisfied in either of two ways, and both of which derive from the criteria already used as thresholds for evidence in the extant literature representative of both reputation optimism and pessimism. The first way is statements by decision-makers in the crisis who express belief that the rival’s threats—explicit or implied—will be executed; such statements of belief would be attributable to something, and that something should help us determine whether and how threat credibility is a function of past words and deeds or strictly on current factors.

The second way to assess threat credibility is based on the degree to which government actions in response to threatening signals match what we would expect to see when a state’s threats are believed to be credible. As Sartori and others note, states rarely bluff.36 We should therefore expect that in the midst of a crisis, a threatened state will back down (that is, not counter-threaten or escalate) only when it deems the other state’s threat is credible; it would make little sense to back down from a confrontation if the defender thought there would be no negative consequence for standing firm by making a counter threat. In this way, observing that a defender backed down from a coercive challenge indicates that it found the challenger’s threat to be credible, even if there are no recorded verbal or written statements to indicate as much.

36 Sartori, Deterrence by Diplomacy. John Mearsheimer reached the same conclusion, though he notes that while states rarely bluff to other states, they frequently lie to their domestic constituencies. See John J. Mearsheimer, Why Leaders Lie: The Truth About Lying in International Politics (New York: Oxford University Press, 2011).

57 Observing that State A backed down from State B: supports the reputations for honesty hypothesis on threat credibility if in the confrontation prior to that State B backed down from

State A; supports the reputations for resolve and reputations for honesty hypothesis on threat credibility if in the confrontation prior to that State B stood firm against a threat from State A; and supports current calculus theory if the balance of power at the time of the confrontation favors State B or the issue at stake is not a sufficiently vital national interest to State A. Because constant calculus theory says that states will always judge the threats of their rivals as credible, backing down always supports constant calculus theory.

Alternatively, if a defender stands firm in response to a challenger’s threat, though far from a conclusive indicator on its own, then we might have reason to suspect that the defender did not find the challenger’s threat sufficiently credible.37 Observing State A stand firm against a challenge from State B indicates: support for the reputations for honesty hypothesis on threat credibility when State B failed to fulfill a threat it made in the previous confrontation, or when

State B has vital interests at stake in the current crisis; support for the reputations for resolve hypothesis on threat credibility when State B backed down in the previous confrontation; and support for current calculus theory when the balance of power and balance of interests favor

State A. Because constant calculus theory says that states will always judge the threats of their rivals as credible, decisions to stand firm always reflects the judgment that their rivals’ threats

37 One of the virtues of a case study approach is that it allows us to seek an answer to why a state chooses to stand firm in a specific case rather than simply assuming why. It should be noted, however, that in most large-N studies on reputations, standing firm is sufficient evidence that the defender viewed the challenger’s threat as lacking sufficient credibility. See, for example, Sartori, Deterrence by Diplomacy.

58 are credible. The table below summarizes the above discussion addressing the ideal predictions for each of the competing hypotheses on credibility.

Table 2 Competing Hypotheses on Threat Credibility: Interpreting Evidence Evidence Supports Evidence Supports Reputational Hypotheses Alternative Hypotheses

Observation Reputations for Reputations for Current Constant Honesty Resolve Calculus Theory Calculus Theory

State B has State A State B has State B has favorable balance A rivalry exists; Believes history of history of of power and/or State A cannot State B’s backing down or standing firm vital interests at risk incredulity Threats standing firm stake

Balance of power State A State B backed State B stood favors State B A rivalry exists; Backs Down down or stood firm in prior and/or balance of State A cannot

from State B firm in prior confrontations interests favors risk incredulity

confrontations State B

State B has a Balance of power State A Does State B has a history of favors State A Cannot explain Not Believe history of getting backing down or and/or balance of without State B’s caught bluffing a history of interests favors referencing other Threats getting caught State A theories bluffing

State A State B has a State B has a Balance of power Cannot explain Stands Firm history of getting history of favors State A without against State caught bluffing backing down or and/or balance of referencing other B a history of interests strongly theories getting caught favors state A bluffing

States look to States look to The balance of past behavior to past behavior to power and determine determine interests within a If a state is a rival, current current dyad are the only its threats are Causal Logic credibility; a credibility; a determinants of believed because track record of track record of threat credibility; incredulity risks matching word standing firm states in a crisis being too costly and deed makes makes current do not care about current word and word and deed the past deed believable believable

59 But threat credibility is only one of the causal consequences expected by the reputations literature. The other causal claim by hypotheses on reputations for honesty and resolve respectively is, as discussed at length above, that states with either reputation are less likely to be challenged by other states; conversely, states with a reputation for bluffing or irresoluteness are more likely to be challenged by other states. So putting aside the causal claim regarding threat credibility in the midst of a crisis, we must also scrutinize the rivalry dyad longitudinally, both within crises (multiple interaction observations within a single crisis) and between crises. As mentioned in the first chapter, reputation is a phenomenon that is often an “unobservable”; finding direct evidence in the form of private statements in which decision-makers acknowledge that they are assessing a threat in terms of past words and deeds is sufficient but not necessary to establish that reputations are having a causal effect. Indeed, one of the limitations of the qualitative studies with outcomes that are pessimistic about reputations is that they set the bar for minimally necessary evidence at primary sources in which a statesman explicitly says something to the effect of “I believe the threat my rival just made because he has fulfilled his threats in the past”;38 this is an unreasonably rigid formulation and as a consequence an unreasonable evidentiary threshold. While I scrutinize decision-makers’ statements to the extent possible, another complementary approach is to observe whether states are more or less likely to be challenged (dependent variable) following certain behaviors (independent variable) is to do just that. If after a state backed down in a previous crisis we find a reduced frequency in challenges to it by the same challenger from the previous crisis, then we have found support for the

38 Mercer, Reputation & International Politics; Press, Calculating Credibility.

60 reputations for honesty hypothesis and against the reputations for resolve hypothesis. If after a state backed down in a previous crisis we find an increased frequency in challenges to it by the same challenger as previous, then we have found support for the reputations for resolve hypothesis and against the reputations for honesty hypothesis. The table below summarizes how backing down, standing firm, or initiating challenges in a rivalry (the independent variables) affects threat credibility and the likelihood of future challenges (the dependent variables), as well as how these empirical observations would support the claims of reputations for honesty and reputations for resolve respectively.

Table 3: Honesty, or Resolve? Hypothesized Effects of Reputation on Intra-Rivalry Crisis Behavior Note: Effects expected in Hypothesis 1 and 2 highlighted in grey

Reputations for Reputations for Reputations Reputations Honesty (Threat Honesty for Resolve for Resolve Credibility) (Probability of (Threat (Probability of Future Credibility) Future Challenges) Challenges)

State B backs State B does not State B State B If State A down at time 2 if challenge State challenges or challenges backs down State A makes A at time 2 stands firm State A at time threat against State A 2 at time 2

State B backs State B does not State B backs State B does If State A down at time 2 if challenge State down at time 2 not challenge stands firm State A makes A at time 2 if State A State A at time threat makes threat 2

State B backs State B does not Indeterminate Indeterminate If State A down at time 2 if challenge State initiates State A makes A at time 2 challenge threat

61 Case Selection

I have up to this point described my hypotheses, how I intend to operationalize the variables I will observe, competing hypotheses and their operationalization, and the thresholds for ideal-type evidence in support of my anticipated observations. There are several issues of consequence relating to case selection that I will address here before describing the analytical method of my research design in the section below.

Why the U.S.-North Korea Rivalry?

The goals of this research are idiographic and nomothetic. Theoretically, a focus on the

U.S.-North Korea rivalry is ideally suited to begin remedying shortcomings of the reputations and rivalries literatures identified in the previous chapter. The two outstanding puzzles in this dissertation are an ideal fit for the U.S.-North Korea context because it offers a set of cases in which 1) the weaker power frequently challenged the stronger rival and 2) the rivalry dyad experienced repeated crises without escalation to war. And as discussed previously, this study is motivated in part by the need to advance the debate about the relevance of reputations in international politics. The concept of reputations for resolve—a crucial element in deterrence and coercion theory—has faced challenges from Reputation Optimists and Pessimists alike, yet within the reputations literature, no scholar since the wave of reputation pessimism arose has evaluated claims relating to reputations for resolve against new cases, neither within an interstate rivalry context, nor against competing theoretical claimants (that is, the reputations for honesty hypothesis). Within the context of the reputations literature, therefore, the U.S.-North Korea rivalry gives us multiple opportunities to, for instance, observe challenger-defender dynamics

62 over time, which may give us an indication if there is merit in distinguishing between the behavior of challenger and defender for the purposes of reputation building.

A focus on the U.S.-North Korea dyad also provides an opportunity to weigh in with new empirical evidence in the debate between advocates of reputations for honesty and advocates of reputations for resolve; Reputation Optimists of any type have yet to examine the crises in this rivalry. Another potential payoff that studying this rivalry has for the reputations literature is that it brings the study of reputation back to the context that first gave the issue salience as a research topic in international relations—the rivalry context. Reputation Pessimists have principally challenged the relevance not of every type of possible reputation but of reputations for resolve, yet new-wave reputation optimism has consisted almost entirely of showing the importance of various types of reputation other than resolve.39 As Paul Huth observed, rivalries are “ideally suited for testing propositions about reputations for resolve,” which is one of the aims of this study.40 But observation of the U.S.-North Korea rivalry not only helps rectify some of the shortcomings of the reputations literature, but also advances the rivalries literature by examining a real-world case that deviates from the theoretical expectations of the research program, which holds that rivalries are highly prone to escalation and war, particularly when they experience recurring crises.41 Importantly, analysis of this rivalry also addresses the

39 Perhaps the lone exception to this is Barbara Walter, but her focus on reputations for resolve was not between rival states but between states and separatists. Walter, Reputation and Civil War. 40 Huth, “Reputation and Deterrence,” 98. 41 This claim is a defining characteristic of the rivalries literature. See, for example, Vasquez, The War Puzzle; Leng, “When Will They Ever Learn?”; Leng, “Escalation”; Leng, Bargaining and Learning in Recurring Crises; Crescenzi and Enterline, “Time Remembered.”

63 literature’s need to examine cases in which some rivalries scholars code it as a rivalry and some do not, which I address in further detail in the sub-section below.

Selection of the U.S.-North Korea dyad is driven by another set of purposes that are specific to this rivalry. The decision-making and intergovernmental dynamics between the

United States and North Korea, particularly during crises, is intrinsically interesting because prolonged periods of hostility punctuated by occasional crises continues to characterize this rivalry, and few studies exist of U.S. decision-making on North Korea policy, particularly during the Cold War; to the extent that any studies exist on U.S.-North Korea relations, they are historical accounts, not theoretically informed in any explicit sense.42 It is curious that the series of crises in U.S. foreign policy examined here are generally under studied, given that it is one of the few security challenges from the Cold War that endures today. As a result, there are many questions in the U.S.-North Korea rivalry still unanswered. Why did North Korea, the weaker power, repeatedly challenge the United States in the 1960s and 1970s? Why, after 1976, did

North Korea dramatically curb the frequency and intensity of its challenges against the United

42 Atheoretical treatments of specific episodes in U.S.-North Korea relations include: Narushige Michishita, Narushige, North Korea’s Military-Diplomatic Campaigns, 1966-2008 (New York: Routledge, 2011); Richard A. Mobley, Flash Point North Korea: The Pueblo and EC-121 Crises (Annapolis, MD: Naval Institute Press, 2003); Don Oberdorfer, The Two Koreas: A Contemporary History (Reading, MA: Addison-Wesley, 1997); Leon V. Sigal, Disarming Strangers: Nuclear Diplomacy with North Korea (Princeton, NJ: Princeton University Press, 1998); Joel S. Wit, Daniel B. Poneman, and Robert L. Gallucci, Going Critical: The First North Korean Nuclear Crisis (Washington, DC: Brookings Institution, 2004). One of the few studies taking a theoretically guided approach to U.S. decision-making toward North Korea is William Drennan, “Nuclear Weapons and North Korea: Who’s Coercing Whom?” in The United States and Coercive Diplomacy, eds. Robert J. Art and Patrick M. Cronin (Washington, DC: U.S. Institute of Peace Press, 2003), 157-223.

64 States but continued and even increased the intensity of its challenges against South Korea after

1976? How did the United States and North Korea experience multiple crises with one another—crises that harbored the risk of war—without escalation to major conflict? In repeated crises, and within crises with repeated interaction, how did the United States interpret the credibility of North Korea’s threats, how did North Korea respond to U.S. threats, and what determined either side’s decisions to back down or escalate? Anyone interested in U.S. foreign policy decision-making, escalation control, the study of rivalries, North Korea or threat credibility should be interested in the answers to these questions, some of which may have generalizable implications, but all of which are specific to the U.S.-North Korea rivalry.

In terms of research design, the recurring crises within the rivalry allow us to compare each crisis to the next in a way that controls for culture, regime type and in this case relative power distribution, all of which come closer to approximating experimental conditions than social science normally allows.43 The interests at stake also remain fairly consistent. The first three episodes examined all include issues in dispute that take on the appearance of low salience but that are imputed with high security significance because of the rivalry context in which they occur. The final episode examined—the North Korean nuclear crisis in 1993-4—has a very different issue at stake driving the crisis as compared to the other three episodes, but it also contains multiple observations because of the large number of interactions that took place; within that single episode, the issue at stake remained constant, across what amounts to multiple

43 Gary King, Robert O. Keohane, and Sidney Verba, Designing Social Inquiry: Scientific Inference in Qualitative Research (Princeton: Princeton University Press, 1994), 37-41; Arend Lijphart, “Comparative Politics and the Comparative Method,” American Political Science Review 65, no. 3 (1971), 687.

65 observations. Moreover, scrutiny of an asymmetric rivalry like the U.S.-North Korea dyad structures Reputation Optimist and Pessimist hypothesis comparisons in a way that favors

Reputation Pessimist predictions even though the findings here mostly support Reputation

Optimist claims. Current calculus theory predicts that the balance of capabilities and interests determines the credibility of a threat, so the extreme power disparity of an asymmetric rivalry provides a relatively easy case for current calculus theory to explain as a result. It is also within crises, as opposed to under normal or routine conditions, that decision-makers should be least likely to consider the adversary’s behavioral history when making credibility assessments and most likely to simply focus on the present.

What Makes This Case a Rivalry?

Some might question my characterization of the relationship between the United States and North Korea as a rivalry. After all, the vast power asymmetry and geographical space between the two states would seem to eliminate the “need” to engage in conflict. What can a small power like North Korea really hope to do to the United States, and why should the United

States feel anything approaching an existential threat from a small power like North Korea?

Recalling my discussion of rivalry definitions in the prior chapter, there are at least four reasons for coding this relationship as a rivalry. First, power parity in a dyad is neither necessary nor sufficient to determine a rivalry, which means that power asymmetry is no obstacle to defining a rivalry. As of 2001, there were around 20 major-minor power rivalry dyads, what Brantley

66 Womack and others refer to as “asymmetric rivalries.”44 Second, other rivalries research, using the dispute-density approach, already classifies the United States and North Korea as a rivalry.45

Following the 1953 Armistice Agreement that effectively ended the Korean War, intermittent crises in which either or both sides experienced casualties and which contained the real prospect of war continued to erupt throughout the Cold War; the period from 1966 through 1976 was particularly “dispute-dense,” to use the term of rivalries scholars, which I will discuss more below. Third, the U.S. relationship with North Korea was entirely predicated on competing claims regarding the legitimate existence of South Korea (which North Korea views as a

“Korean” matter) and competing claims regarding acceptable U.S. and North Korean behavior as international actors. The perpetual threat of conflict constituted the only basis for their interaction throughout the Cold War; they did not cooperate with each other within multilateral institutions, they had no diplomatic relations, and they had no economic or cultural relations.

What is more, while North Korea’s hostile behavior continuously posed a threat to certain U.S. interests, such as the protection of its citizens, its allies, and its preferred rules and norms, the antipathy that North Korea held for the United States ran much deeper; fear and hatred of the

United States was a part of North Korean identity starting from the formidable experience of the

Korean War, but fitting a strong historical xenophobic element in Korean nationalism that came

44 The precise number of asymmetric rivalries depends, again, on how one defines a rivalry. William Thompson puts the range of asymmetric rivalries from seven to 22 depending on whether you use the dispute density or perceptual approach to coding rivalries. Thompson, “Identifying Rivals,” 578-80. See also Brantley Womack, “Asymmetric Rivals: China and Vietnam,” in Asian Rivalries: Conflict, Escalation, and Limitations on Two-Level Games, eds. Sumit Ganguly and William R. Thompson (Palo Alto: Stanford University Press, 2012), 176-94. 45 See, for example, Diehl and Goertz, War and Peace in International Rivalry; Bennett, “Integrating and Testing Models of Rivalry Duration.”

67 from centuries of subordination to outside great powers.46 The fourth, most important reason for coding the U.S.-North Korea rivalry as a rivalry is because it meets the conceptual criteria I advanced for rivalries in the first chapter: a shared hostility, and a high expectation of future conflict. If either of these criteria were not true, there would be little basis for the long-term, sustained U.S. military presence on the Korean Peninsula. For all these reasons, it is apt to describe the U.S.-North Korea rivalry as such, both for the purposes of this dissertation as well as the coding of this dyad in the study of rivalries generally.

Criteria for Event Selection

The final issue to address regarding case selection is the criteria used to determine which aspects of the U.S.-North Korea to examine. This rivalry has endured for more than 60 years, providing far too many interactions to examine comprehensively. Parsing through the history, I narrowed the range of interactions to consider, including only those in which the United States and North Korea were involved in a direct confrontation with one another, as opposed to a confrontation between North and South Korea or other proxy. Because numerous small skirmishes occurred along the Demilitarized Zone (DMZ) separating North and South Korea and involved U.S. soldiers, I further refined the focus of my analysis to include those episodes that fit the conditions for crisis as defined by Charles Hermann in one of the seminal works on crisis and foreign policy: surprise; abnormal or compressed decision timelines; and a high perceived threat

46 On North Korean identity and the role foreign powers play in constituting it, see Brian R. Myers, The Cleanest Race: How North Koreans See Themselves and Why It Matters (Brooklyn, N.Y.: Melville House Publishing, 2011). On the historical anti-foreigner strain that partially constitutes Korean nationalism, see Gi-wook Shin, Ethnic Nationalism in Korea: Geneology, Politics, and Legacy (Palo Alto: Stanford University Press, 2006); Mark Caprio, Japanese Assimilation Policies in Colonial Korea, 1910-1945 (WA: University of Washington Press, 2009).

68 to values.47 Translating these attributes of a crisis contextually in the U.S.-North Korea case, I selected events that met at least three of the four following criteria: (1) one or both sides suffered casualties as the result of an action of the other side; (2) the crisis was of a sufficiently serious nature that it involved the attention, deliberation, and at least one decision by either the U.S. president or Kim Il Sung; (3) the alert levels in either North Korea or for the U.S.-Republic of

Korea (ROK) alliance increased abruptly; and (4) at least one side threatened war or retaliation.

Applying these criteria to the U.S.-North Korea rivalry yields four major episodes: the USS

Pueblo incident of 1968; the EC-121 shoot down incident of 1969; the Panmunjom axe murder crisis of 1976; and the nuclear crisis of 1993-4. Incidentally, these are also the four periods that

North Korea’s most high-ranking defector to South Korea claims were viewed by North Korean leadership at each time as crises most likely to result in war.48 To the extent that we can tell from the available evidence, which is limited given its recentness, U.S.-North Korea relations during the George W. Bush administration and since are excluded from inquiry here not only because of challenges with collecting evidence on such a recent period, but also because at no point did alert levels, casualties, war threats, and presidential decision-making converge in a way that war seemed likely, particularly given that the United States was in the middle of conducting two wars during most of the first decade of the 21st century.

47 Charles F. Hermann, Crises in Foreign Policy (Indianapolis: Bobbs-Merrill, 1969). 48 “North Korean Defectors 27 July News Conference,” Choson Ilbo (July 28, 1994): 3-4; “Defector Hwang Chang- yop Interviewed,” Sindong-a (July 1998): 328-345.

69 Analytical Method

This dissertation serves the positivist purposes that Arend Lijphart describes as “theory- confirming” and “theory-informing,” testing competing hypotheses against unexploited cases as a means of both assessing how well extant theories explain specific historical episodes, and in turn using specific historical episodes to affirm, revise, or circumscribe extant theories.49

Individual case studies have implications for claims made by extant theories. If a theory cannot explain a given case well or at all, such an anomaly may not be sufficient to falsify the theory, but it probably is sufficient to generate a new scope condition for the theory, to falsify some part of the theory, or to generate new hypotheses that complement or contradict the theory.

By setting up a Lakatosian “three-cornered fight” in which rival theories are pitted against one another in observing the same data, we are not only able to help rank the relative performance of competing theories under specific scope conditions, but also to suggest refinements to existing theories based on the observations made in specific cases.50 Within this setup, I employ the congruence and process tracing methods to a structured, focused comparison of cases. The congruence method is useful because it facilitates a direct comparison of the relative “performance” of competing theories’ expectations against empirically observable outcomes; it allows for logical and evidentiary inference even when direct evidence of causal mechanisms is lacking. Process tracing is a useful supplement to the congruence method because it helps orient causal inferences and helps reject alternative explanations, which is

49 Lijphart, “Comparative Politics and the Comparative Method,” 691. 50 Imre Lakatos, “Falsification and the Methodology of Scientific Research Programmes,” in Criticism and the Growth of Knowledge, eds. Imre Lakatos and Alan Musgrave (Cambridge: Cambridge University Press, 1970), 91- 196.

70 particularly important when multiple logics anticipate the same outcome.51 In this way, process tracing is the act of bringing direct, within-case evidence to bear on the explanation for why a given correlation is observed. If, for example, a threat is found credible, the reason is unclear without within-case evidence that tells us why it was found credible. It is this process tracing evidence that in turn tells us which theory best explains the observed outcome.

Sources of Evidence

The evidence I marshal in this dissertation comes from a variety of primary and secondary source materials. Evidence of U.S. deliberations and actions comes from declassified meeting transcripts and summaries as part of the U.S. Department of State Foreign Relations of the United States series, U.S. presidential library archives, congressional testimonies, and memoirs from individuals directly involved in the crises under analysis, supplemented by a variety of secondary sources where appropriate. As one might expect, the still-closed nature of

North Korean deliberations and internal decision-making poses a challenge to even the most enterprising researcher. The problem of evidence on North Korea is not intractable, however.

First, U.S. archival documents and memoirs contain documentation of U.S. interactions with

North Korea in instances where direct negotiations took place, giving us one angle from which to gain insights. Second, I draw extensively from the recently opened archives from Russia and states of the Former Soviet Republics maintained by the Woodrow Wilson International Center as part of its Cold War International History Project. Many of these documents capture meeting transcripts from Soviet interactions with North Korea during and after crises between the United

51 George and Bennett, Case Studies and Theory Development in the Social Sciences, 181-232.

71 States and North Korea, as well as third-party reflections on North Korean intentions during the crises. Finally, we know North Korea’s externally visible actions as part of each crisis because they are part of the historical record; we know what government actions North Korea took in each crisis, lacking only an understanding of the reasons for their actions. Although each type of evidence is perhaps suboptimal for documenting evidence of North Korean intentions on its own, taken together these sources should give us a sufficiently clear picture to satisfy the objectives of this study—that is, to determine the role of reputations for honesty and resolve in the U.S.-North

Korea rivalry, and to offer an explanation of the puzzles of why repeated crises managed not to escalate to broader conflict despite theoretical expectations to the contrary, and why North Korea would repeatedly challenge the United States despite being the weaker power. The availability of evidence on U.S. decision-making will support the congruence method, as well as facilitate process tracing to a large extent. The dearth of direct evidence on North Korean decision- making, while somewhat problematic for purposes of process tracing, is sufficient to facilitate the congruence method; I will compare the theoretical expectations for North Korea’s behavior with its actual words and deeds within and across crises.

A Methodological Clarification

I anticipate the possibility of two general types of angst from readers that I seek to address before proceeding to the empirical section of the dissertation. The first is methodological, and the second is theoretical. Regarding the first, some may question attempting a structured, focused comparison of cases because part of my argument rests on the idea that events over time are interdependent, not independent; therefore, the questioning goes,

72 my individual cases cannot logically be considered independent observations. Such a concern reflects the logic of statistics in which a case study represents a single observation, which is to say n=1. Viewing this study through such a lens, however, would miss both the point of this study and a fundamental truism of the rivalries literature: dyadic rivalries (such as the United

States and North Korea) should be viewed longitudinally; rivalry conflicts do not take place in a vacuum, but rather with the weight of historical baggage that pre-frames actions with meaning and a certain constraining logic.52 A comparison of one crisis to the next is useful for observing similarities and differences, and should be pursued in a consistently structured way, but such is not intended to suggest that each conflict consists of an independent observation in a methodological sense. The reputations (and rivalries) literature implicitly situates itself ontologically within the historical institutionalist tradition of interpreting political life. Charles

Tilly best summarizes the most relevant defining characteristic of the historical institutional perspective for the purposes of this dissertation as the understanding that, “…when things happen within a sequence affects how they happen.”53 Although historical institutionalism is compatible with rationalist modes of analysis,54 a major difference is that the former asserts that variables must be analyzed in their historical context to be properly understood.55

52 Diehl and Goertz, War and Peace in International Rivalry. 53 Charles Tilly, Big Structures, Large Processes, Huge Comparisons (New York: Russell Sage Foundation, 1984), 14. 54 Since historical institutionalism is not a paradigm in the Kuhnian sense, it need not be (though sometimes is) a competing alternative to rational or sociological paradigms. For a full discussion, see Orfeo Fioretos, “Historical Institutionalism in International Relations,” International Organization 65, no. 1 (Spring 2011): 367-99. 55 Paul Pierson, “Not Just What But When: Timing and Sequence in Political Processes,” Studies in American Political Development 14 (Spring 2000): 72-92.

73 A structured, focused comparison in this dissertation is a means of systematically structuring a narrative for each event in a logical way; it does not mean I seek to reduce each case to taking on the value of a single data point/observation of a single variable. Scholars have long since established that qualitative case studies allow for many observations within a single case, and that the reductionist logic of large-N statistical studies does not necessarily apply to case study analysis; the “degrees of freedom” problem, in other words, is not necessarily an issue in qualitative methodology.56 As I specified in Tables 1 and 2 earlier in this chapter, the temporal interdependence of events in a rivalry (that is, the causal effects of reputations) can be observed within a crisis (in the case of a crisis with multiple interactions) or from one crisis to the next; the narrative structure of each case is therefore not limited to the temporal boundaries of a specific crisis, but rather includes an interpretation of the historical context of a crisis as part of each “case.” Methodologically, it would be most appropriate to view my analysis of the U.S.-

North Korea rivalry as a single longitudinal “case” defined by four particular historical episodes selected based on criteria discussed above; from within this “case” I derive as many useful observations as possible to inform, affirm, or repudiate the aforementioned hypotheses guiding my inquiry.

Some might also be concerned with multiple collinearity because I explain three dependent variable outcomes using one independent variable. In large-N statistical studies, multiple collinearity is indeed a major concern because each case has been reduced to a single

56 Jack S. Levy, “Case Studies: Types, Designs, and Logics of Inference,” Conflict Management and Peace Science 25, no. 1(2008): 1-18.

74 data point stripped of any contextual meaning not explicitly controlled for.57 Using such a method of inquiry, it is difficult to know whether one of the claimed dependent variables is actually the cause of one or more of the other independent variables. The problem of collinearity does potentially exist in case study research as well, but the provision of detailed context and the marshaling of direct process tracing evidence potentially offers much greater confidence in the causal observations made than the mere correlation between two data points observed in statistical studies. In the chapters that follow I will make clear that my dependent variables do not cause one another, but are caused by my identified independent variable.

A Theoretical Clarification

Another type of clarification I seek to offer is of a theoretical nature, involving two different but overlapping issues: distinguishing the level of my analysis (whether reputations accrue to people or states) and the scope of my theoretical claims. Regarding the first issue, an unfinished task in the literature on reputations remains determining whether and when reputations accrue to individuals, states, or both.58 This is an important theoretical question, but not one to be resolved here. The possible distinction is relevant for the purposes of research design, however; specifying whether you are looking for reputational effects at the individual or state level can affect where you look for evidence and how you interpret it. For the purposes of this dissertation I assume that reputations accrue at the state level for at least four reasons: because the actions of statesmen are part of what constitutes the actions of states; because there

57 King, Keohane, and Verba, Designing Social Inquiry. 58 Paul K. Huth, Paul K., “Reputation and Deterrence: A Theoretical and Empirical Assessment,” Security Studies 7, no. 1 (1997): 75-8.

75 is not enough leadership turnover within the U.S.-North Korea rivalry for the time periods observed to resolve the issue; because, as discussed in the previous chapter, the logic of rivalries tends to encourage the statesmen within them to view their rival through a realpolitik lens;59 and because the same assumption is consistently used in the research of Reputation Optimists and

Pessimists alike.60 Such an assumption does not mean, however, that I treat the state as a “black box.” As should be evident from my discussion in the previous section of how I operationalize reputations and marshal evidence to evaluate my hypotheses, decision-maker statements and decisions about the credibility of the rival state’s threats is the principal way to determine which hypotheses best explain crisis behavior in this rivalry. Decision-makers and decision-making processes are crucial to reliable process tracing.

The second theoretically relevant issue worth clarifying is the scope of my theoretical claims and generalizability. In this dissertation I make two arguments: 1) that backing down from a confrontation with a rival has three effects (increased threat credibility, increased likelihood of future challenges, and escalation control), and 2) initiating a challenge against a rival does not have a clear effect on building a reputation for resolve. The causal logic for these arguments comes from the reputations literature, and testing these arguments requires testing their theoretical underpinnings against plausible alternatives. Although I expect to find that the reputations for resolve hypothesis and part of the reputations for honesty hypothesis will best

59 Leng, Bargaining and Learning in Recurring Crises. 60 Reputation Optimists whose research assumes that reputations accrue to the state are Huth, Extended Deterrence and the Prevention of War; Sartori, “Mightier than the Sword”; Sartori, “Deterrence by Diplomacy”; Walter, Reputation and Civil War; Miller, The Shadow of the Past. Reputation Pessimists who assume the same in their research designs are Mercer, Reputation & International Politics; Danilovic, “The Sources of Credibility in Extended Deterrence”; Press, Calculating Credibility.

76 explain how threat credibility was determined and escalation controlled within this rivalry, such does not in itself validate the role of reputations in international politics generally and the corresponding interdependence-of-commitments logic underpinning it, nor does it falsify the claims of Reputation Pessimists. However, such findings do suggest that under certain scope conditions Reputation Optimists have a very strong case for their claims and Reputation

Pessimists have a very weak case for theirs. Confirming my theoretical expectations regarding reputations for honesty and resolve respectively helps validate the “most likely” scope conditions for when these specific types of reputations are likely to be causally relevant, as well as the extent of the validity of their theoretical claims. What this dissertation does not do is to test claims about broader interpretations of the interdependence-of-commitments logic; that is, it makes no claims about whether reputations that developed through interactions with one state, like North Korea, carry over and have causal effect in interactions with another state, like Libya or Iran. So from this analysis I anticipate that the U.S.-North Korea rivalry will suggest that backing down has at least three consequences when occurring in the context of dyadic rivalries in which there are: high expectations of future conflict; uncertainty about the other side’s intentions; and similar issues in dispute. Under these conditions, I also expect to find that initiating challenges against a rival—whatever the tactical benefit of a first mover advantage— does not have a determinate impact on building a reputation for resolve. These consequences (of backing down and initiating challenges respectively) are theoretically grounded in a hybrid framework that incorporates part of the reputations for honesty logic with that of reputations for resolve. To the extent that these findings hold in my examination of the U.S.-North Korea

77 rivalry, generalizability to a broader universe of cases will depend on testing the hypotheses against other cases.

Chapter III: The USS Pueblo Crisis (1968)

This chapter conducts a temporally contextualized analysis of an incident that took place toward the end of the presidency of Lyndon Johnson in which North Korea captured the USS

Pueblo, a U.S. navy intelligence collection vessel, and held its 82-person crew—one crewmember died during the capture of the ship—hostage for close to a year. Particularly during the early moments, the crisis drew significant amounts of attention from President Johnson and his administration, even as the conflict in Vietnam continued as the primary line of foreign policy focus for the White House. The narrative for this episode is built on the answers to several questions that help us identify the effects of backing down, the sources of threat credibility, and the reasons the incident did not escalate to larger conflict. What was the recent conflict history of the dyad prior to this incident? What was the geopolitical context of the crisis; that is, what role, if any, did China, the USSR, ROK, and Vietnam play in defining the circumstances surrounding the Pueblo crisis? Who initiated coercion and why? How did each view the credibility of the other’s threats during the crisis and why? What actions did each side actually take during the crisis and what can be inferred about the motivations of each side from these actions, and to what extent does process tracing evidence support these inferences?

Finally, how did the crisis end?

Historical Context

U.S.-North Korea relations were increasingly hostile and violent throughout most of the

1960s. North Korea’s capture of the Pueblo stood out amongst the hundreds of violent incidents that preceded it because of the unprecedented boldness of the act, the fact that dozens of U.S. soldiers were at risk, and the total surprise by which it took U.S. officials. North Korean

78 79 harassment and capture of South Korean fishing vessels had, by 1968, become routine.1 More troubling was that violent incidents across the demilitarized zone (DMZ)—the ironically named, heavily armed area separating North and South Korea—had increased markedly since 1964.

Because U.S. and South Korean soldiers manned the southern side of the DMZ, violent incidents often involved the United States, sometimes exclusively so. One estimate of DMZ violence showed a rise from 32 incidents in 1964 to 435 incidents by1967; an annual increase of more than 10 fold.2 Another estimate places the number of incidents in 1967 at around 500;3 a third estimate identifies 348 incidents in 1967;4 yet another source claims 445 violent incidents occurred along the DMZ in 1967.5 It remains unclear why the estimates vary, though part of the distinction is likely that a tabulation of known events does not mean that a given event tabulated actually constituted a violation of the 1953 Armistice Agreement as investigated and documented by the United Nations Command Military Armistice Commission (UNCMAC).6 The numerical discrepancy among sources is not problematic, however, because all sources tell the same story: violence in and around the DMZ, often but not always instigated by North Korea, escalated by many orders of magnitude starting in late 1966, nearly two years before the Pueblo was captured.

North Korea v. United States

1 Trevor Armbrister, the first documentarian of the Pueblo incident, including from hundreds of first-person interviews, documents that in 1967 alone, North Korea captured 50 South Korean fishing boats and detained 353 South Korean crewmen. Trevor Armbrister, A Matter of Accountability: The True Story of The Pueblo Affair (Guilford, CT: The Lyons Press, 1970), 187. Since the publication of his account in 1970, hundreds of previously classified sources, including archival documents and oral histories, have been made public, many of which were released only in the last decade. 2 Mobley, Flashpoint North Korea, 14. 3 Christian F. Ostermann and James Person, Crisis and Confrontation on the Korean Peninsula, 1968-69: A Critical Oral History (Washington, DC: Woodrow Wilson International Center, 2011), 84. 4 Ostermann and Person, Crisis and Confrontation on the Korean Peninsula, 85. 5 Armbrister, A Matter of Accountability, 187. 6 UNCMAC investigates all alleged violations of the Armistice Agreement, but investigations do not always conclude that a violation of the Armistice Agreement occurred.

80 There are several incidents of note that took place between the United States and North

Korea as part of this landscape of escalating violence. Hundreds of small-scale, localized conflagrations occurred between the United States and North Korea in the 1960s, but a few of these became historical reference points for the North Korean government and U.S. decision- makers as the Pueblo crisis evolved. The first is the May 17, 1963 North Korean capture of an

OH-23 Raven, a U.S. Army helicopter. The helicopter’s two-man crew was on a routine boundary marker observation mission along the South Korean side of the DMZ when it accidentally strayed into North Korean airspace, where it conducted a forced landing. The North

Koreans held the two crewmembers and demanded an apology from the U.S. Government and an admission of espionage. After several rounds of direct, quiet negotiations between the United

States and North Korea, the United States capitulated to North Korea’s demands, signing something called an “overwrite receipt.”7 The “overwrite receipt” method of apology constituted

North Korea drafting an embarrassing apology and admission of guilt for the United States to sign; the commander of U.S. Forces Korea (USFK)—who is “dual-hatted” as the commander of

United Nations Command in Korea—then signed the North Korea-drafted apology, not admitting to the contents of the apology, but signing acknowledging the receipt of the apology as reflecting North Korea’s view; he physically wrote on the North Korea-drafted document stating that he acknowledges receipt of the document, and then signed his name under this hand-written acknowledgement. One year after the incident, on May 17, 1964, North Korea finally released the U.S. helicopter crew, two days after both sides agreed to settle the helicopter incident through

7 Armbrister, A Matter of Accountability, 274, 299.

81 an overwrite receipt of an admission of guilt.8 Later, during the USS Pueblo crisis, North Korea and the United States would both refer to the incident; as I discuss below, this event played a role in resolving that crisis, as well as a similar incident during the early part of the Nixon administration.

The helicopter incident was followed by another, in April 28, 1965, when North Korea made an unsuccessful attempt to shoot down an RB-47, a U.S. Air Force reconnaissance aircraft, that was conducting surveillance off of North Korea’s east coast, near the port of Wonsan—the same vicinity within which the Pueblo was conducting maritime intelligence collection when it was captured.9 North Korea’s attack on the RB-47 took place 35-40 nautical miles from North

Korea’s eastern coast, and “severely damaged” the aircraft.10 Although many in the U.S. government would subsequently remember and reference the RB-47 attempted shoot down

(which I discuss below), the incident did not lead to U.S. counter-coercion or retaliation of any kind.

Another incident representative of the uptick in North Korean aggression that took place in the 1960s was a North Korean infiltration attack that coincided with a visit by President

Johnson to Seoul, South Korea in November 1966. On November 2, while President Johnson was meeting with South Korea’s President Park Chung-hee, North Korean soldiers infiltrated a

8 U.S. Department of State, September 4, 1968. Memorandum from Alfred Jenkins of the National Security Council Staff to the President’s Special Assistant (Rostow), “Status of Pueblo Talks at Panmunjom,” Foreign Relations of the United States (hereinafter “FRUS”), 1964-68, Vol. XXIX, Part 1, Korea, document 312: 697-700. 9 Memorandum, Korea Task Force, “Fulbright Letter on USS Pueblo,” March 20, 1968. Folder POL 33-6 KOR N- US 3/15/68, Central Foreign Policy files, box 2269, RG 59 (College Park, MD: National Archives and Records Administration “NARA”). 10 U.S. House Armed Services Committee, Special Subcommittee on the USS Pueblo, Inquiry in to the USS Pueblo and EC-121 Plan Incidents. 91-12 (Washington, DC: U.S. Government Printing Office, 1969), 1654; Armbrister, A Matter of Accountability, 197.

82 U.S. patrol unit along the DMZ, killing one South Korean, six U.S. soldiers, and injuring others, constituting the twenty-ninth violent incident along the DMZ since October 15 that year alone.11

Prior to the incident, the CIA “dismissed” the prospect of North Korean attacks during the

President’s visit because it assessed that until that point, North Korean missions appeared not to aim at “deliberate violence but rather intelligence collection and recruitment.”12 Following the incident, an intelligence estimate subsequently found that the timing of the North Korean raid did not intentionally coincide with President Johnson’s visit to Seoul, and that it deliberately targeted what it knew to be a U.S. patrol unit.13 The only U.S. response to this specific incident was the initiation of an UNCMAC investigation, which concluded that the North Korean raid was likely instigated as a retaliatory response to a South Korean raid into North Korean territory that had taken place only days early, on October 26, though it offered no insight regarding why North

Korea targeted the United States rather than the ROK.14 Although U.S. servicemen stationed in

Korea had died at the hands of the North Korean People’s Army (KPA) prior to 1966, this attack was significant as it was the first premeditated attack in which the United States was the specific target since 1953; it erased any doubts that North Korea was willing to use coercive violence against the United States as a target distinct from South Korea.

On May 22, 1967, North Korea initiated yet another attack in which it explicitly targeted the United States, by crossing the DMZ and destroying two U.S. infantry barracks. The incident,

11 U.S. Department of State, November 1, 1966. Memorandum of Conversation between President Johnson and President Pak, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 96: 205-7. 12 U.S. Department of State, June 23, 1967. Report Prepared by the Office of National Estimates of the Central Intelligence Agency, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 120: 257-9. 13 U.S. Department of State, November 8, 1966. Intelligence Memorandum, “Armed Incidents along the Korean DMZ,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 98: 209-10. 14 Ibid.

83 which killed two U.S. servicemen and wounded 19 others, once again reflected the deliberate targeting of the United States, illustrating an increased willingness to initiate violence against the

United States.15 By the time of Vice President Humphrey’s July 1967 visit to South Korea, the

CIA had adjusted its assessment of North Korean intentions, noting North Korea’s newfound risk acceptance with regard to initiating violence against U.S. and ROK forces, together or apart.16

The United States took no actions against North Korea in response to the incident.

The overall trend of growing hostility and danger that these incidents represent, particularly having taken place in a context of hundreds of incidents over the period of only a few years, resulted in the “collections paradox”: the more violent North Korea became toward the United States, the greater the demand for intelligence, surveillance, and reconnaissance missions to anticipate North Korean behavior and divine its intentions.17 Much like today, the

United States in the 1960s had trouble determining North Korean motivations. As the State

Department’s Bureau of Intelligence and Research stated in a September 1967 Special National

Intelligence Estimate, “While we know of North Korea’s political objectives, we do not know its immediate intentions and plans.”18 The aggregate U.S. response to North Korea’s greater lethality compared to previous years amounted to little more than searching for more intelligence, even though increases in intelligence missions to meet policymaker demands put more intelligence collection assets in harm’s way. The United States also increased its

15 U.S. Department of State, June 23, 1967. Report Prepared by the Office of national Estimates of the Central Intelligence Agency, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 120: 259. 16 Ibid. 17 Mobley, Flashpoint North Korea, 22. 18 Memorandum from INR/REA Fred Greene to Claus Ruser, August 1, 1968, “State of Intelligence on North Korea,” Folder Korea 2 of 2, Senior Interdepartmental Group Files 1966-72, box 8, RG 59, NARA.

84 commitment to develop U.S.-ROK alliance counter-infiltration capabilities, but much of this investment was offered to help offset the large ROK contributions to the U.S.-led conflict in

Vietnam.19 It remains unclear the rate at which annual increases in U.S. intelligence collection missions against North Korea occurred leading up to the Pueblo crisis because the numbers are not publicly available, but the total number of missions in 1967 had increased from prior years to around 1500 total.20

Cheap Talk and Credibility, 1965-1968

The previous section makes clear that relations on the Korean Peninsula were hostile throughout the 1960s. The intra-rivalry behavioral pattern during this period was also quite clear: North Korea initiated challenges and the United States only responded through alliance military investments to defend against such challenges, mostly investments in counter-infiltration capabilities along the DMZ, and by placing greater emphasis on intelligence collection against

North Korean military and leadership targets. The United States initiated no discrete challenges of its own, and never responded to any North Korean provocations with counter-coercive or retaliatory measures.

19 Between 1964 and 1973, the ROK sent more than 300,000 troops to Vietnam at U.S. request. For a history of ROK contributions to Vietnam, see Stanley Robert Larsen and James Lawton Collins Jr., Vietnam Studies: Allied Participation in Vietnam (Washington, DC: U.S. Army Center of Military History, CMH Pub 90-5, 1975), 120-59. On several occasions, including between President Johnson and President Park, the United States and ROK discussed the continued and rising demand for ROK contributions to the U.S. effort in Vietnam, while in the same conversation discussing the need for greater U.S. military assistance and counter-infiltration investments to secure the ROK against ever-increasing North Korean belligerence. See Memorandum, “Notes on Conversation between President Johnson and President Park,” December 21, 1967, Folder POL KOR S-US, Central Files 1967-69, RG 59, NARA. 20 Armbrister, A Matter of Accountability, 183. For an investigation of the intelligence program of which the USS Pueblo mission was a part and sporadically declassified data on U.S. intelligence missions in and around the Sea of Japan, see U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121.

85 North Korea’s threat credibility problem during the mid-1960s stemmed from the incessant use of general threats that muted the ability to discern proximate threats. Throughout this period, North Korea’s threatening rhetoric increasingly outpaced its violent activities. Prior to the Pueblo crisis, however, at no point did ideal-type game-theoretic circumstances arise in which Pyongyang drew a “red line” that the United States or ROK dare not cross at the cost of conflict escalation; reality was more complex. North Korea did send more subtle signals, such as a public statement (discussed below) shifting to a more aggressive footing against the United

States. But much of North Korea’s provocative behavior during this period consisted of guerilla activities like infiltration and sabotage missions, which are the kinds of activities whose success requires secrecy, not advertisements to the enemy in the form of a big red line. U.S. analysts, at any rate, had difficulty distinguishing Pyongyang’s threats from its rhetorical flourishes, typically discounting statements as nothing more than “cheap talk.” North Korean threats became white noise that even made tracing actions to specific prior threats through post hoc analysis difficult for the United States; their frequency and consistently hostile content, particularly with regard to DMZ activity, made North Korean statements a poor indicator of

North Korean intent with regard to specific actions.21 A State Department Bureau of Intelligence and Research assessment of North Korean statements summarized the cheap talk problem in this way: “Pyongyang has been making intensely belligerent statements for years and has left little room for measurable increases in that direction.”22

21 Memorandum from Henry Kissinger to Richard Nixon, April 21, 1969, “Pueblo,” National Security Council Files, Pueblo Folder, box 379, Richard M. Nixon Collection, NARA. 22 Memorandum from INR/REA Fred Greene to Claus Ruser, August 1, 1968, “State of Intelligence on North Korea,” Folder Korea 2 of 2, Senior Interdepartmental Group Files 1966-72, box 8, RG 59, NARA.

86 The example of greatest consequence in this regard was a pivotal speech on October 5,

1966 by North Korean leader Kim Il-sung to the Korean Workers’ Party conference. The speech was significant not only because it publicly signaled growing distance between Pyongyang and

Beijing, but primarily because it was a public signal of a sea change in North Korea’s willingness to directly challenge the United States, articulating that it would “wage a positive struggle” against “U.S. imperialism” at every front, beginning a policy of some kind of limited warfare against the United States.23 It was a rare instance during this period in which the United

States was able to eventually look back and realize that later actions could be traced back to a specific threatening signal. The speech, which preceded President Johnson’s visit to Seoul by less than a month, was immediately followed with a surge in DMZ-related provocations, including the aforementioned attack on a U.S. patrol during President Johnson’s visit.24

Although months into the Pueblo crisis in 1968 the CIA would realize the significance of Kim’s

1966 declaration as presaging North Korea’s increasingly aggressive provocations,25 at the time of President Johnson’s visit to Seoul the United States had not yet recognized that the speech signaled a shift in North Korea’s willingness to use force, particularly against the United States.26

North Korea’s incessant use of cheap talk may have muted the U.S. Government’s ability to discern North Korea’s threatening signals in a timely manner, but by 1967 the U.S.

Government had come to recognize that North Korea’s general threats reflected a greater

23 Kim, Il Sung, “Report to the Conference of the Workers’ Party of Korea, October 5, 1966,” in Kim Il Sung Works Vol. 20. (Pyongyang: Foreign Language Publishing House, 1984). 24 U.S. Department of State, November 1, 1966. Memorandum of Conversation between President Johnson and President Pak, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 96: 205-7. 25 CIA Directorate of Intelligence, November 26, 1968, Kim Il-Sung’s New Military Adventurism, Freedom of Information Act request, released May 2007. 26 U.S. Department of State, June 23, 1967. Report Prepared by the Office of national Estimates of the Central Intelligence Agency, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 120: 257-9.

87 willingness to incite conflict and risk war through more aggressive action. USFK Commander,

General Charles Bonesteel, wrote in a July 1967 memo that two new themes arising in North

Korea’s threatening rhetoric suggested danger on the horizon: (1) the “Yankees and their ROK puppets” were plotting to start a second Korean war and that North Korea was prepared to destroy them in response, and (2) North Korea’s disavowal of its aggression along the DMZ and

ROK interior, blaming such acts on an indigenous South Korean revolution against “U.S. imperialism.”27 General Bonesteel argued that a new Special National Intelligence Estimate and a review of U.S. and ROK posture and policy toward North Korea were necessary given that this thematic shift in North Korea’s threatening rhetoric was taking place against a backdrop of increased North Korean guerilla activity; within a week of Bonesteel’s memo to the commander of U.S. forces in the Pacific, the State Department, Department of Defense (DoD), and the CIA had agreed to General Bonesteel’s recommendation.28 According to General Bonesteel, what made these general threats credible was their marriage with increasingly violent actions, even if one clearly outpaced the other; Kim Il-sung’s October 5 clear foreshadowing of events to come was not identified as a credible threat until many violent incidents had already occurred (that is,

North Korean aggression surged to match the policy decree), even though violence had already been on the ascent for the past year. As aggressive as North Korean discourse was generally,

Bonesteel was a minority in being able to distinguish an aggressive shift in North Korean discourse that to him seemed to be commensurate with observed behavior, but the connection

27 U.S. Department of State, “Telegram from the Commander in Chief, United Nations command, Korea and the Commander of United States Forces, Korea (Bonesteel) to the Commander in Chief, Pacific (Sharp),” July 21, 1967, FRUS, Vol. XXIX, Part 1, Korea, document 123: 262-6. The Pyongyang Times, written largely for foreign audiences, regularly carried these themes in their news coverage from November 1966 through at least late 1967. 28 Ibid.

88 took several months to make and was not, according to publicly available documents, immediately apparent to anyone in the U.S. Government.

Geopolitical Context

The geopolitical context of North Korea’s military activism against the United States and

ROK merits brief discussion here, though fuller histories of geopolitical factors can be found elsewhere. The major factors of note beyond the U.S.-North Korea rivalry itself were China, the

Soviet Union, the ROK, and most importantly, the U.S. war in Vietnam. By the start of 1966, at least 10 months before Kim Il-sung’s October 5 speech signaling a new approach to the United

States and deriding the Chinese (albeit without specifically naming them) as “nihilistic” and encouraging “violent extremes,”29 North Korea increasingly embodied the juche (self-reliance) philosophy that it had proclaimed for more than a decade.30 It began intensive military procurement and training for guerilla actions against the United States and ROK, spending the year marshaling the military capacity it would need to undertake the October 5 pivot toward violent militarism against the United States and the ROK.31

As China underwent the early stages of its bloody Cultural Revolution, it parted ways ideologically, diplomatically, and militarily with North Korea, openly calling Kim Il-sung a

“bourgeois revisionist” in its public propaganda. Sino-North Korean relations were so poor during the latter half of the 1960s that each side massed forces along the Yalu River separating

29 Kim Il-sung, “Report to the Conference of the Workers’ Party of Korea.” See also CIA Directorate of Intelligence, Kim Il Sung’s New Military Adventurism. 30 Charles K. Armstrong, “Juche and North Korea’s Global Aspirations,” NKIDP Working Paper #1 (Washington, DC: Woodrow Wilson International Center, 2009). 31 CIA Directorate of Intelligence, Kim Il Sung’s New Military Adventurism, 28-40.

89 China from North Korea, eventually engaging in armed clashes with one another.32 Taking advantage of what had by then become an evident split between China and the Soviet Union,

North Korea sought arms and military aid from the latter, achieving limited success in acquiring new Soviet MiG-21 aircraft and air defense technology, though relations with the Soviet Union too were somewhat distant and cool.33 By 1968 the U.S. intelligence community accurately assessed that North Korea maintained a foreign policy independent of Chinese and Soviet influence, and that the increase in North Korean aggression starting in 1966 was entirely a North

Korean initiative.34 Throughout the 1960s, as with the Pueblo crisis, China and the Soviet Union played a marginal and aloof role in U.S.-North Korea relations because each was to a great extent marginal and aloof when it came to influencing or even being aware of North Korean behavior. As a Soviet diplomat assigned to Pyongyang lamented just days after the Pueblo was captured, “When we try to moderate this warmongering state of mind…our position is not taken into account.”35

The other, much more significant, factors informing the context of the U.S.-North Korea rivalry were the U.S. war in Vietnam and the ROK’s anti-communist policy under its

32 Mitchell B. Lerner, “A Failure of Perception: Lyndon Johnson, North Korean Ideology, and the Pueblo Incident,” Diplomatic History 25, no 4 (2001): 663-5. 33 CIA Directorate of Intelligence, Kim Il Sung’s New Military Adventurism, 19. At least one scholar has argued that Kim Il-sung was motivated to militarization by a feeling of abandonment by its allies during the 1960s. See Dae Sook Suh, Kim Il Sung: The North Korean Leader (New York: Columbia University Press, 1988), 247-8. For a discussion of the implications of the Sino-Soviet split for North Korea and communist alliances generally, see Zhihua Shen and Danhui Li, After Leaning to One Side: China and Its Allies in the Cold War (Palo Alto: Stanford University Press, 2011). 34 Director of Central Intelligence, February 29, 1968, Update to “North Korean Intentions and Capabilities with Respect to South Korea,” folder NIE 14.2 North Korea,” National Security Files Country Files, NIEs, box 5, (Austin, TX: Lyndon Baines Johnson Presidential Library “LBJ”). 35 “Telegram from Pyongyang to Bucharest,” January 25, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5, 14-15.

90 authoritarian president, Park Chung-hee. In 1964 Lyndon Johnson initiated and gradually escalated a bombing campaign against the communist nationalists in North Vietnam. Between

1964 and 1968, the United States continued to spend lives and resources to pursue what was increasingly obviously a failed policy against the Viet Cong based on theories of limited war that had questionable mooring to the reality of modern warfare and the structures of modern governments.36

As the hottest front in the Cold War at the time, the Vietnam conflict tended to command not only the lion’s share of U.S. policymaker attention but global attention as well. Popular opinion in the United States and globally rose against the war as the death toll increased and the fighting dragged on.37 Although the United States struggled to gain troop contributions to the unpopular Vietnam war effort from its allies and partners, the ROK ultimately became one of the largest troop contributing governments.38 As indicated above, the ROK had instrumental reasons for dispatching so many of its own forces to fight what appeared to be the spread of communism to Vietnam, such as an implicit quid pro quo that would keep U.S. troops on the Korean

Peninsula, but the South Korean narrative of the ROK Government’s contribution was that it was a means of repaying the 16 nations that contributed to South Korean independence under the

36 The story of U.S. coercion against Vietnam, its causes and consequences, and its failure is best told in Thies, When Governments Collide. 37 William L. Lunch and Peter W. Sperlich, “American Public Opinion and the War in Vietnam,” The Western Political Quarterly 32, no. 1 (1979): 21-44; Andrew W. Daum, Lloyd C. Gardner, and Wilfried Mausbach, America, the Vietnam War, and the World: Comparative and International Perspectives (Cambridge: Cambridge University Press, 2003). 38 Charles K. Armstrong, “America’s Korea, Korea’s Vietnam,” Critical Asian Studies 33, no. 4 (2001): 527-39; Robert M. Blackburn, Mercenaries and Lyndon Johnson’s “More Flags”: The Hiring of Korean, Filipino and Thai Soldiers in the Vietnam War (Jefferson, NC: McFarland & Company, 1994).

91 auspices of the United Nations during the Korean War.39 For its part, North Korea contributed to

Ho Chi Minh’s struggle in Vietnam against the United States, though nothing on the scale of

South Korea’s contributions in support of the United States.40 North Korea perceived strong parallels between its own situation and that of North Vietnam, struggling as a small power against the United States, publicly condemning U.S. and ROK efforts accordingly.41 U.S. and

ROK “aggression” against the Viet Cong played a central role in Kim Il-sung’s October 5, 1966 speech, serving as the proximate justification for North Korea’s more risk acceptant behavior in the years to follow.42 In what came to be known as the “Tet offensive,” North Vietnam launched a large series of attacks throughout South Vietnam during what was an agreed upon cease fire period. The Tet offensive proved a critical juncture in the Vietnam war; for purposes here and discussed below, the timing was important as it occurred on January 30, 1968, one week after the

Pueblo crew was captured (on January 23, 1968).43

On January 21, 1968, two days before the seizure of the Pueblo, North Korea launched its boldest attack against the ROK since the Korean War, sending in a 31-man team of commandos to assassinate ROK President Park Chung-hee in Seoul.44 Posing as South Korean military

39 Jinim Pak, Narratives of the Vietnam War by Korean and American Writers (New York: Peter Lang, 2007), 78; Chung-hee Park, “To Help a Neighbor is to Defend Oneself,” in Major Speeches by Korea’s Park Chung-hee, compiled by Shin Bum Shik (Seoul: Hollym Corporation, 1970), 37-40. 40 Barry K. Gills, Korea versus Korea: A Case of Contested Legitimacy (London: Routledge, 1996), 110. 41 Armstrong, “Juche and North Korea’s Global Aspirations,” 8. 42 Kim, “Report to the Conference of the Workers’ Party of Korea.” For an analysis of the speech, see CIA Directorate of Intelligence, Kim Il Sung’s New Military Adventurism. 43 For an overview, see James J. Wirtz, The Tet Offensive (Ithaca NY: Cornell University Press, 1991). 44 For a fuller account of the Blue House raid, see Ostermann and Person, Crisis and Confrontation on the Korean Peninsula; Mitchell Lerner and Jong-Dae Shin, “New Romanian Evidence on the Blue House Raid and the USS Pueblo Incident,” NKIDP E-Dossier #5 (Washington, DC: Woodrow Wilson International Center, March 2012). Available online: http://www.wilsoncenter.org/sites/default/files/NKIDP_eDossier_5_The_Blue_House_Raid_and_the_Pueblo_Incide nt.pdf.

92 officers, the team infiltrated the DMZ on January 17, coming within 800 meters of the Blue

House (the official residence of the South Korean president) before finally being challenged by

South Korean police. Upon confrontation, a firefight ensued, the group scattered, and all but one of the commandos was killed in action, having come close but ultimately failing in their assigned mission. Under interrogation, the sole surviving commando from the assassination attempt, a

North Korean Peoples’ Army officer named Kim Shin Jo, said that the primary purpose of the assassination attempt was to stimulate revolution in South Korea: “…would agitate the South

Korean people to fight with arms against their government and the American imperialists.”45

The United States and ROK would almost immediately connect the Blue House raid with the capture of the USS Pueblo, identifying the timing as deliberately coordinated, with complementary purposes, discussed below.46

Overview of the Incident

On January 23, 1968, two days after North Korea launched the Blue House raid to assassinate ROK President Park, North Korea dispatched naval and air forces to capture the

Pueblo, a poorly armed U.S. signals intelligence collection vessel. Assigned to the Commander of U.S. Naval Forces Japan (NAVFORJAPAN), the Pueblo departed its base in Sasebo on

January 10, 1968 for its mission in the Sea of Japan to collect information about the order of battle, capabilities, and intentions of North Korean military forces along its eastern coastline.47

45 John J. Stefans, “Confessions of a Red Agent,” Army Digest 23 (May 1968): 19. 46 “Telegram from the Embassy in Korea to the Department of State,” January 24, 1968, Folder POL 33-6 KOR N- US, Central Files 1967-69, RG 59, NARA. 47 The minute by minute account of the capture portion of this incident can be found in U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121. The fully chronology of the incident can be found in the Lyndon Baines Johnson Presidential Library. National Security File, National Security Council History, Pueblo Crisis 1968, Vol. 1, Basic Study and Presidential Decisions.

93 Its planned path was to go to the top of North Korea, and then to meander down its coast and past the North Korean port of Wonsan as it loitered in several predefined patrol boxes. The crew and its commander were under strict orders not to come closer than 13 nautical miles to the

North Korean coast.48 The USFK commander, General Bonesteel, would not find out about the existence of the Pueblo mission until it was being captured because the vessel was assigned to

NAVFORJAPAN and performing missions on behalf of the U.S. intelligence community, both of which put the Pueblo outside General Bonesteel’s formal jurisdiction.49

Prior to the seizure, the collection analysts onboard the Pueblo had intercepted mostly static; North Korean forces along the eastern coast were quiet and inactive in the days prior to the ship’s January 23 capture.50 On January 21, two days before the Pueblo’s capture, the North

Korean navy identified it loitering in international waters off its coast, taking no immediate actions.51 When North Korea did react to the Pueblo on the afternoon of January 23, it dispatched MiG fighter aircraft and naval patrol boats from Wonsan to challenge the ship, board it, and drag it into Wonsan harbor with its 82-man crew. During the ship boarding process, the

North Koreans fired on the ship, killing one of the U.S. crew and wounding others; under orders

48 Armbrister, A Matter of Accountability, 6. 49 Interview with General Bonesteel, quoted in Daniel P. Bolger, “Scenes from an Unfinished War: Low-Intensity Conflict in Korea, 1966-69,” Leavenworth Papers, No. 19. (Ft. Leavenworth, KS: Combat Studies Institute, 1991), 66. 50 Armbrister, A Matter of Accountbility, 21. This characterization is somewhat in tension with Richard Mobley’s analysis claiming that the commander of NAVFORJAPAN received a January 19 report from a Japanese merchant vessel captain who regularly did business in North Korea that the port of Wonsan in the weeks prior to the Pueblo’s seizure had “extraordinary naval activity.” Mobley, Flashpoint North Korea, 41. Whichever characterization is more accurate, it is clear that decision-makers in Washington were unaware of increased activity at Wonsan. 51 Armbrister, A Matter of Accountability, 30; Mobley, Flashpoint North Korea, 41-2.

94 of their commander, and contrary to U.S. Navy norms of behavior, the U.S. crew did not resist

North Korean orders and did not fight back.52

By dusk on January 23, North Korea had moored the Pueblo at Wonsan and moved its crew to a prison where they would be held for nearly a year, frequently receiving beatings, undergoing various forms of torture, and being interrogated and coerced into giving false confessions. Through brutal coercion, the North Koreans were able to compel Commander

Bucher to sign a document apologizing for and admitting to unlawfully spying and doing so 7.6 nautical miles from the North Korean coast.53 Though U.S. decision-makers in Washington would be unsure during the initial moments of the crisis, U.S. intercepts of North Korean communications, coupled with detailed navigation logs from the Pueblo, prove that not only did the ship never come closer to North Korean territory than 13 nautical miles, but also that North

Korean military units tracking the Pueblo never identified it as being any closer than 13 nautical miles from North Korean territory; at the moment of capture, North Korean radio operators called out the position of the Pueblo as being 15.3 nautical miles from North Korea’s easternmost island (Ung-do) and 25 miles from Wonsan harbor.54

Although the U.S. Government was heavily invested in Vietnam at the time, the State

Department, DoD, the intelligence community, and the White House mobilized quickly to determine what to do about the Pueblo’s capture. The first several meetings, including of the

National Security Council at the White House, were mostly spent trying to ascertain the facts

52 The commander of the USS Pueblo, Lloyd Bucher, detailed his thinking during the capture process in his memoir. Lloyd M. Bucher with Mark Rascovich, Bucher: My Story (New York: Doubleday, 1970). 53 Bucher, Bucher; Armbrister, A Matter of Accountability, 254-6. 54 U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121, 1661-2; Armbrister, A Matter of Accountability, 263-4.

95 about what happened and the current state-of-play, though response options were raised and eventually rejected, discussed below. When the Pueblo, which dominated conversation, was raised at President Johnson’s Tuesday National Security lunch with key advisors and cabinet secretaries on the day of the Pueblo’s seizure, it was clear that of CIA Director Richard Helms,

Chairman of the Joint Chiefs of Staff General Earle Wheeler, and Secretary of Defense

McNamara, nobody could confirm why the incident happened, what North Korea wanted, or even whether the Pueblo had violated North Korea’s claimed maritime boundary, though DCI

Helms characterized the incident as part of a longer term pattern of North Korean provocation against the United States.55 Secretary of State Dean Rusk briefed the President that U.S.

Ambassador to the Soviet Union, Llewelyn Thompson, approached the Soviets on his instructions in order to secure Soviet cooperation in gaining the release of the U.S. crew, to which the Soviets responded, “this was not their problem…They said they would not take action on the matter.”56 President Johnson was “anxious” to learn more about the situation, immediately noting that at least three alternative courses of action were open to the United States and he wanted others as well: attacking North Korean forces; “getting a thorough explanation”; and capturing a North Korean ship.57 No decisions were made at this initial meeting, but the use of force remained on the table, pending the facts of the incident. As one participant observed,

55 In a memo from DCI Helms to Secretary McNamara later that day, Helms expounded on this point in greater detail, though he initially argued that the Blue House raid and Pueblo incidents were not coordinated with one another; within 24 hours he and the rest of the U.S. Government would realize the incidents were linked. U.S. Department of State, January 23, 1968. Memorandum from Director of Central Intelligence Helms to Secretary Defense McNamara, “North Korean Intentions,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 215: 464-5. 56 U.S. Department of State, January 23, 1968. “Notes of the President’s Tuesday National Security Lunch,” drafted by Tom Johnson, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 213: 260-63. 57 Ibid.

96 “No one ruled out a military response at that first meeting…[but] we had to have more information before we committed ourselves to a course that would start another.”58

Debating Retaliation

The following day, January 24, U.S. decision-makers gathered for four separate principals-level meetings,59 two of which included President Johnson. An early consensus emerged in these meetings that diplomatic pressure would be insufficient to secure the return of the Pueblo and its crew, though none could agree on which of several retaliatory options bandied about should be taken.60 Among the riskier retaliatory options considered were:61 seizing a

Soviet or North Korean ship and holding it hostage; laying mines in the area surrounding

Wonsan harbor; limited air strikes against North Korean naval and air bases along its eastern coast; a naval blockade of Wonsan; unleashing ROK forces for a “punishment” raid across the

DMZ; “spooking” North Korean radars by flying aircraft close enough to activate North Korean radars and raising their alert level; and explicit deterrence actions, such as overtly placing another intelligence collection ship in the location of the Pueblo’s capture as a show of resolve or signaling indirectly through the Soviets that North Korean conciliatory actions were necessary to

58 As quoted in Armbrister, A Matter of Accountability, 239. 59 Attendees, all of whom would become regular faces in managing the nearly year-long crisis, included McNamara, Helms, future Secretary of Defense and Johnson confidant Clark Clifford, General Wheeler, National Security Adviser Walter Rostow, Deputy Secretary of Defense Paul Nitze, Undersecretary of State Nicholas Katzenbach, Press Secretary George Christian, and several Assistant and Deputy Assistant Secretaries (Richard Steadman, Paul Warnke, and Samuel Berger). 60 U.S. Department of State, January 24, 1968. “Summary Minutes of Pueblo Group,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 217: 468-75. 61 These options and other, less risky but also less effective, options are ably summarized in Mobley, Flashpoint North Korea, 59-62. For primary sources discussing and rejecting these options, see Clark Clifford Papers, folder “North Korea-USS Pueblo Incident,” box 17, LBJ; U.S. Department of State, January 24, 1968. “Summary Minutes of Pueblo Group,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 217: 468-75; Memorandum from T.H. Moorer to Secretary of the Navy, January 25, 1968. “Draft Message OPLAN,” Wheeler records, file “Korea (Pueblo) February 21, 1968 Vol. 3,” box 29, RG 218, NARA.

97 avoid a “deterioration” of the situation. General Bonesteel, in a fit of frustration, recommended that the United States issue a nuclear ultimatum to North Korea in exchange for the crew; his suggestion was ignored.62 By the end of the four meetings on January 24, the President and his advisors had ruled out the blockade, capture of a Soviet ship (supported by Rostow), and

“spooking” North Korean radars (supported by Katzenbach). Rusk, McNamara, and Clark

Clifford were quick to kill each of these ideas before even putting them in front of the President because of the likelihood that North Korea would retaliate in kind and that none would likely get the Pueblo crew back.63 Other options, including airstrikes, at Walt Rostow’s insistence, initially remained on the table.64 The other options remained a part of subsequent discussions over the next several meetings, but as Undersecretary Katzenbach explained in a later interview, “We kept them alive, I suppose, because we felt so frustrated…We all thought something had to be done.”65 The NSC met with President Johnson two more times on January 25th to discuss the

Pueblo.66 The first meeting reached no decisions; the second would be the last time retaliation options would be discussed in-depth with the President. In this second meeting, President

Johnson was keen on finding some way to retaliate if possible, though he feared that any retaliation would lead to North Korean escalation to a full-blown conflict that would involve the

62 Interview with General Bonesteel in Bolger, An Unfinished War, 68. 63 Armbrister, A Matter of Accountability, 260-1. 64 U.S. Department of State, January 24, 1968. “Meeting on Korean Crisis Without The President,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 220: 483-91; U.S. Department of State, January 24, 1968. “Notes of the President’s Meeting,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 221: 492-5; Armbrister, A Matter of Accountability, 258-60. 65Quoted in Armbrister, A Matter of Accountability, 261. 66 U.S. Department of State, January 25, 1968. “Notes on the President’s Thursday Night Meeting on the Pueblo Incident,” drafted by Tom Johnson, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 225: 505-13.

98 Soviets or the Chinese.67 Secretary McNamara agreed with the need for reprisal actions, claiming that “The great danger that we must avoid is that the Soviets and the North Vietnamese will interpret something that we do as a sign of weakness.”68 Katzenbach also urged for standing firm against North Korea, asserting that “We should not let the North Koreans think that the

Security Council is the way we planned to deal with this. They must know we are prepared to take further actions, military ones if necessary.”69 Clark Clifford disagreed, arguing in front of the President that a second Korean War was not worth the lives of the crew, or U.S. reputation.70

The following day, January 26, saw a shift in the President’s emphasis from retaliatory options to diplomatic tactics that would help get the Pueblo back; he did not explicitly reject the future use of military reprisals of some kind, but instead focused the discussion on getting back the crew through diplomatic means. Referencing the system used during the Cuban Missile

Crisis, President Johnson directed McNamara and Katzenbach to put together working groups to sift through the information necessary to piece together recommendations.71 That same day, an interagency Korea working group and a State Department-led Korea task force were formed, each with an aim to produce information and advice relating to the Pueblo, focusing more on diplomatic, intelligence, and forensic information rather than recommendations for retaliatory measures.72 By January 29, all retaliatory options had been eliminated and subsequent months-

67 U.S. Department of State, January 25, 1968. “Notes on the President’s Thursday Night Meeting on the Pueblo Incident,” drafted by Tom Johnson, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 226: 514-19. 68 Ibid. 69 Ibid. 70 Ibid. 71 U.S. Department of State, January 26, 1968. “Notes of President’s Friday Morning Meeting on the Pueblo Incident,” drafted by Tom Johnson, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 228: 521-9. 72 Memorandum for the Korean Working Group, January 26, 1968, National Security File, Country file “Korea- Pueblo Incident,” Miscellaneous, Vol. 1, LBJ.

99 long discussions centered instead on the tactics of the U.S. negotiating position as it sought to seek a face-saving compromise that would achieve the goal of bringing the crew and vessel home. That day, a senior level advisory group73 reviewed the products of the interagency working group and explicitly ruled out most of the retaliatory actions for fear that North Korea would escalate: the cabinet secretaries, undersecretaries, and deputy assistant secretaries agreed that North Korea may risk war if the United States pursued stern responses such as mining

Wonsan harbor, air strikes, or deterrent threats in the form of replacement intelligence collection vessels in the same spot where the Pueblo was captured.74 The group rejected the blockade option because its results with regard to the crew would be unclear yet would still risk escalation; it supported seizing a North Korean ship for reasons of proportionality, but Korea’s disconnection from the global economy and corresponding lack of maritime commerce made the option difficult to execute, in addition to being of questionable efficacy.75 On March 15, the

State Department’s Korea task force led by Deputy Assistant Secretary of State Winthrop Brown made a final recommendation to commit to diplomacy and de-escalatory pressures rather than any forms of coercion, even if North Korea escalated tensions further by putting the Pueblo crew on trial. He argued that military actions of any kind, including the “threatening deployment of

73 The group was led by Walter Rostow and Dean Rusk, and was attended by Richard Helms and several State Department officials. Other than General Maxwell Taylor (at the time a consultant to President Johnson), it is unclear if any DoD officials were in attendance. 74 “Report on Meeting of Advisory Group, January 29, 1968,” February 1, 1968. Folder “Material regarding Vietnam and Pueblo, January-February 1968,” National Security File, Walter Rostow files, box 10, LBJ. 75 Ibid.

100 additional U.S. forces to the Korean area” would not produce a desirable effect on North Korean behavior.76

Orchestrating U.S. Actions

Amidst these deliberations about potential reprisal options, U.S. officials were also directing and attempting to oversee the various actions that taken together would represent the external appearance of the U.S. government response. The actions actually taken, as opposed to those from the previous section that were considered and rejected, can be grouped into four types: (1) military and intelligence preparations, including multiple steps to avoid confrontation with North Korea; (2) seeking support at the United Nations; (3) failed entreaties to the Soviets to influence North Korea; (4) and direct diplomacy with North Korea through the United Nations

Command Military Armistice Commission (UNCMAC). The first of these, military preparations, would require marshaling massive resources on a short timeline, would be only partially visible to other governments, and would eventually be largely curtailed by decision- makers in Washington seeking to prevent any U.S. actions that might incite a North Korean response. The second and third types of U.S. action, approaching the United Nations and the

Soviet Union, yielded little more than frustration. After the first month of the crisis had passed, it had become clear to all by the momentum of events that diplomacy would be the preferred

U.S. path to securing the Pueblo and its crew; virtually all subsequent senior level deliberations in the U.S. government centered on the tactics of U.S. diplomacy through UNCMAC. As the section addressing North Korea’s perspective makes clear below, there is little evidence to

76 Memorandum from the Director of the Korean Task Force (Brown) to the Under Secretary of State (Katzenbach), March 15, 1968,“Possible Actions in the Event the North Koreans Announce that They Intend to Try the Pueblo Crew,” Central Files 1967-69, RG 59, folder POL 33-6 KOR N-US, NARA.

101 suggest that most of these U.S. actions were noticed by North Korea. The U.S. movement of naval forces was noticed, but North Korea cited it as evidence for why Soviet Bloc countries should support the North Korean view.

The United States took three basic types of military action in response to the Pueblo: (1) force posture and readiness adjustments to the immediate situation in the first few days of the crisis, most of which were on existing authorities as opposed to being centrally directed from the

White House; (2) actions in direct response to White House instructions; and (3) ramped up intelligence operations, most of which had White House visibility and approval. The U.S. military buildup in the early weeks of the crisis would be noticed and viewed negatively by the

Soviets and North Korea. Combined with the intense resource pressures that Vietnam was imposing on DoD, the negative responses from the Soviet Union and North Korea were sufficient to lead the President to call off or unwind many of the military measures directed in the immediate aftermath of the Pueblo’s capture.

Because of a “low” risk assessment issued for the Pueblo’s mission,

COMNAVFORJAPAN did not seek to have any forces on-call and alert to support the Pueblo should it run into trouble.77 From a “cold start,” therefore, General McKee, the commander of

Fifth Air Force, immediately ordered available F-105 Thunderchiefs to fly from Japan to Korea; under a longstanding U.S. agreement with Japan, the United States cannot send forces into combat directly from Japanese territory, necessitating in this case first landing at air bases in

Korea for refueling. This decision started what would become more than three hundred aircraft flowing into the Pacific theater and the Korea operating area over the course of several weeks.

77 The surprisingly low risk assessment is a larger issue discussed in the analysis of credibility, below.

102 Upon news that the Pueblo had been captured, the commander of Seventh Fleet dispatched the

USS Enterprise (CVN-65), a U.S. aircraft carrier, and three destroyer vessels to the Tsushima

Strait. Several hours later, well after the Pueblo was docked at Wonsan, the commander of the

Pacific Fleet ordered the commander of the Seventh fleet to position a destroyer and any necessary accompanying support “off Wonsan immediately outside 12-mile limit. Be prepared to engage in operations that may include towing Pueblo and or retrieval of Pueblo crew/provide air cover as appropriate.” The Seventh Fleet in turn directed the USS Enterprise aircraft carrier to the Sea of Japan, and ordered USS Highbee (DD-806), a destroyer, to move in the direction of

Wonsan. On orders from Pacific Command, the Pacific Fleet readied photo surveillance missions over Wonsan, and had started the process of diverting the USS Kittyhawk, another aircraft carrier, away from Vietnam to support Korea. In Seoul, General Bonesteel ordered all of his forces to a heightened alert posture, ordered his I Corps near the DMZ to operating strength, increased air defense readiness, briefed the ROK Minister of National Defense about U.S. measures to enhance alert levels, and considered raising the Defense Condition (DEFCON) level of readiness for war from DEFCON 4 to DEFCON 3. War and intelligence planners assigned to the Joint Staff and U.S. Pacific Command elements surged efforts to draw up the logistical requirements and timelines in order to flow forces into the region to conduct intelligence and operational missions relating to the various reprisal options discussed in the previous section.

Although President Johnson had effectively ruled out any military actions against North Korea by January 29, the Pentagon and U.S. Pacific Command continued planning throughout February on details for military reprisal campaigns, including limited nuclear attacks.78

78 See National Security Council History, Pueblo Crisis 1968, Vol. VII, Day-by-Day Documents, Part 13, LBJ;

103 Washington’s involvement relative to these actions involved restraining the military to avoid conflict with North Korea and generally approving several specific measures to generate a more robust military presence. When the aforementioned activities and planning proposals were briefed to President Johnson and the NSC, Clark Clifford was the only voice openly opposed to a large buildup; by the end of the evening in Washington (early the next morning in Korea)

President Johnson had authorized 28 fighters and 14 reconnaissance aircraft for immediate deployment, and in the following days would approve McNamara and General Wheeler’s proposals for the military buildup, as long as deployments were “spaced out” to make it appear gradual.79 The prevailing logic was that having the supplemental military capabilities in place would be necessary to defend against another attack or initiate reprisals as a means of standing firm and signaling resolve should the President wish to move in that direction. One of the most publicly noticed decisions that President Johnson made, at McNamara and General Wheeler’s suggestion, was to mobilize selected reserve components, something President Johnson managed to avoid throughout the Vietnam conflict up to that point. By the end of the week, on January

25, the President had authorized calling up 14,787 individuals to active reserve, including 14 Air

National Guard units, eight air reserve units, and six navy reserve units.80 Secretary McNamara informed the full NSC that DoD could send 361 aircraft to the Korean theater without “pulling

Memorandum from C. Berger to N. Katzenbach, February 2, 1968, Central Files 1967-69, folder POL 33-6 KOR N- US, RG 59, NARA. 79 The military buildup “package” included deployment of 26 B-52 bombers (which are nuclear capable) and a mix of several hundred carrier- and land-based aircraft to the Pacific theater. McNamara also secured approval for increasing ROK military assistance by $100 million. 80 Armbrister, A Matter of Accountability, 263. Although some might be inclined to attribute the President’s decision to mobilize the reserves to Vietnam, particularly in light of the January 30 Tet Offensive, both the President and Robert McNamara stated in a private meeting on January 29 in unambiguous terms that the reserves were not needed for Vietnam. See U.S. Department of State, January 29, 1968. “Notes of the President’s Luncheon Meeting with Senior American Advisers,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 244: 562-8.

104 any away from Vietnam or Southeast Asia or Western Europe.”81 As might be expected, the

United States enhanced overhead satellite intelligence coverage of North Korea immediately following the capture of the Pueblo, focusing most intensely on the Wonsan area.82 At the direction of the Joint Chiefs of Staff, the Seventh Fleet deployed 17 submarines off the coast of

North Korea to conduct surveillance patrols; like the enhanced satellite imagery coverage, these actions were briefed to President Johnson, giving him an opportunity to veto the actions, but they were conducted without his specific guidance.83 The White House did explicitly approve intelligence sorties flown by the A-12 Oxcart—a high-altitude, high-speed aircraft that preceded the SR-71—over North Korea, assured by the CIA that even experienced North Korean surface- to-air missile units would have a low likelihood of being able to reach the A-12.84

But simply approving broad military readiness measures and intelligence collection, and mobilizing resources accordingly, was only one aspect of the U.S. military response to North

Korea’s actions. Indicative of U.S. intentions, several steps were also taken in the initial moments after the Pueblo’s capture that illustrated the desire of U.S. leadership to avoid confronting North Korea head on. From the outset of the crisis, DoD sought to position a

SIGINT collection aircraft along the DMZ for the purpose of providing indicators and warning.

81 Ibid. 82 U.S. Department of State, January 25, 1968. “Notes on the President’s Thursday Night Meeting on the Pueblo Incident,” drafted by Tom Johnson. FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 225: 505-13. 83 Memorandum from Russell Murray to Secretary of Defense, “Record of Meeting on January 25, 1968,” folder “Pueblo Incident 1968,” McNamara Papers, RG 200, NARA. 84 U.S. Department of State, January 24, 1968. “Summary Minutes of Pueblo Group,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 217: 468-75; U.S. Department of State, January 24, 1968. “Meeting on Korean Crisis Without The President,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 220: 483-91; U.S. Department of State, January 24, 1968. “Notes of the President’s Meeting,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 221: 492-5. For details of North Korea A-12 Oxcart missions and their approval, see Memorandum for the Chairman, Joint Chiefs of Staff, January 29, 1968, “Requirement for a Second Black Shield Mission over North Korea,” National Security File, Intelligence File “NRO,” box 9, NARA.

105 Concerned that the aircraft, if detected, would prove provocative, the State Department managed to prevent DoD from executing this operation through at least January 26. 85 Although extant evidence does not reveal definitively whether the State Department continued to successfully block this operation, in a lunch meeting with President Johnson on January 29, Walt Rostow, who was often quite hawkish in discussing Pueblo options, gained the President’s agreement not to pursue airborne reconnaissance missions against North Korea for the time being, arguing that the Oxcart’s high-altitude intelligence collection missions had proven effective.86 In that same meeting, when the President asked about the use of unmanned aerial drones, Rostow replied:

“For the moment we do not need it. The group is against the use of drones. To have one shot down would be very bad.”87 Fear of North Korea attacking another intelligence asset stymied the use of aerial reconnaissance drones, which were otherwise a regular feature in the U.S. intelligence collection arsenal against North Korea.88

At 1:38 am on January 24, only hours after Pacific Fleet and Fifth Air Force began mobilizing intelligence and surveillance operations and scrambling the deployment of aircraft and naval assets to the Korean theater of operations, the White House and the Joint Chiefs of

Staff reined in U.S. military actions in the Pacific, which was reflected in Pacific Fleet direction to all naval forces in its control to: remain south of a latitude of 36 degrees (two degrees latitude below the separation of North and South Korea); keep U.S. destroyers away from Wonsan; and

85 Memorandum from Russell Murray to the Secretary of Defense, “Record of Meeting on January 25, 1968,” folder Pueblo Incident 1968, RG 200, McNamara Papers, NARA. 86 U.S. Department of State, January 29, 1968. “Notes of the President’s Luncheon Meeting with Senior American Advisers,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 244: 562-8. 87 Ibid. 88 Memorandum from Commander, U.S. Forces Korea, January 28, 1968, Folder “091 Korea-Pueblo 23 January 1968-7 February 1968, February 1968, Vol. 1,” box 29, RG 218, NARA.

106 avoid all shows of force.89 Over the course of the day on January 24, several other pullback directives were issued from Washington. At Washington’s direction, the Pacific Fleet commander halted SIGINT flights off the east and west coasts of the Korean Peninsula.90 This was followed by the commander of USPACOM directing the across-service action that the

Pacific Fleet had directed for its forces, to explicitly prohibit shows of force, and for all naval and air force units to come no closer than 80 nautical miles from the North Korean coast.91 The commander of USPACOM also instructed all Pacific Fleet naval assets that moved into the

Korean theater of operations to halt and move no closer to North Korea than their positions at the time of instruction, at 7:30 am on January 24.92 Of note, the commander of USPACOM ordered the USS Enterprise aircraft carrier, which Pacific Fleet had initially ordered to proceed in the direction of Wonsan, to withdraw in a southwesterly direction, to the East China Sea.93 The many conservative, risk-averse actions directed by the commander of USPACOM occurred in spite of the fact that he advocated a more hawkish approach to North Korea. In the early moments of the crisis he expressed a preference for sailing the destroyer USS Highbee into

89 CINCPACFLT January 24, 1968, “USS Pueblo Incident,” Folder “Korea-Pueblo Incident-Military Cables, Vol. 1,” National Security Files Country Files, Korea, boxes 263, 264, LBJ. 90 Ibid. 91 ADMIN CINCPAC, January 23, 1968, “Pueblo incident,” folder “Korea –Pueblo Incident, Vol. 1, Pt. A,” NSF Country Files, Asia and Pacific, box 257, LBJ; ADMIN CINCPAC, January 24, 1968, “Operating Instructions for Pueblo Incident,” folder “Korea-Pueblo Incident, Vol. 1, Pt. A,” NSF Country Files, Asia and Pacific, box 257, LBJ. 92 ADMIN CINCPAC, January 23, 1968, “Pueblo incident,” folder “Korea –Pueblo Incident, Vol. 1, Pt. A,” NSF Country Files, Asia and Pacific, box 257, LBJ. 93 National Military Command Center, January 23, 1968, “Current Information Concerning USS Pueblo Incident,” folder “Korea-Pueblo Incident, Vol. 1, Pt. A,” NSF Country Files, Asia and Pacific, box 257, LBJ.

107 Wonsan, with rear shadowing by the USS Enterprise aircraft carrier, to pose an ultimatum to the

North Koreans to release the Pueblo or face retaliatory strikes.94

As the United States pursued military and intelligence responses, and imposed many constraints on those responses to prevent escalation, it pursued in parallel largely conciliatory diplomatic efforts along three tracks in hopes of resolving the crisis without resort to force: at the

United Nations; with the Soviet Union; and via UNCMAC directly with North Korea. In the initial meetings the day after the Pueblo was captured, President Johnson decided that the United

States needed to make its case to the UN Security Council. On Friday January 26, President

Johnson addressed the nation to inform them of the situation with the Pueblo, characterizing it as

“…yet another wanton and aggressive act.” While he claimed that “Clearly, this cannot be accepted,” he also stated that the United States was committed to conciliatory resolution of the situation: “We shall continue to use every means available to find a proper and peaceful solution.”95 Concurrent with the President’s national address, U.S. Ambassador to the United

Nations, Arthur Goldberg, delivered the President’s message to the UN Security Council. His plea for UN action was clear, calling the Pueblo seizure “no mere incident, no case of mistaken identity, no case of mistaken location…nothing less than a deliberate, premeditated armed attack on a U.S. vessel on the high seas.”96 Although Goldberg’s speech hinted at the right of unilateral self-defense under Article 51 of the UN Charter, the goal of taking the Pueblo case to the United

Nations was to secure a Security Council resolution demanding that North Korea return the

94 James P. Finley, The U.S. Military Experience in Korea, 1971-1982: In the Vanguard of ROK-U.S. Relations (San Francisco, CA: Command Historian’s Office, Secretary Joint Staff HQ, USFK/EUSA, 1983), 127. 95 Quoted in Armbrister, A Matter of Accountability, 265. 96 Quoted in Ibid.

108 vessel and its hostage crew. A phone conversation between President Johnson and Amb.

Goldberg two days after Goldberg’s presentation to the UN Security Council captured the frank determination of President Johnson to not respond to North Korea with force for fear of sparking a larger conflict and opening a second front in Asia.97 Even if the United States failed at the

United Nations, President Johnson agreed, the United States might take the case to an international court or tribunal; third party mediation was preferable to both demonstrating resolve on the one hand and to total capitulation to North Korean demands on the other.98

In no small part due to Soviet influence and the communist bloc’s presence in the United

Nations, the United States failed to gain UN support. From the outset of the crisis, the United

States sought Soviet support to resolve the crisis to no avail. Walt Rostow was perhaps a minority but not alone in thinking that North Korea was acting as an agent of the Soviet Union, or at least with tacit Soviet approval; this misperception reinforced extant arguments discussed among the President and his advisers that the Soviets could and should help the United States restore stability on the Korean Peninsula.99 Had the Soviet Union been willing to play a constructive role in the resolution of the crisis, U.S. entreaties to the United Nations may have been more successful.

As with the U.S. failure to persuade the UN Security Council, however, the Soviet Union was unmoved by the U.S. predicament and conveyed to the United States that it was content to

97 Telephone Conversation between President Johnson and the Representative to the United Nations (Goldberg), January 28, 1968, Recordings and Transcripts, Recording of Telephone Conversation between President Johnson and Goldberg, Tape F68.01, PNO 2, LBJ. 98 Ibid; U.S. Department of State, January 29, 1968. “Notes of the President’s Luncheon Meeting with Senior American Advisers,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 244: 562-8. 99 Ostermann and Person, Crisis and Confrontation on the Korean Peninsula.

109 remain aloof of the situation. Secretary Rusk, Amb. Thompson, and President Johnson all made requests to their Soviet counterparts for Soviet mediation of the crisis during the first three days following the Pueblo’s seizure, but were in each instance rebuffed. The Soviet Union claimed that North Korea presented it convincing evidence that the United States violated North Korean territorial waters, and that the United States needed to swiftly enter a political settlement.

Discouraging coercion, the Soviet Union suggested that “Any attempt to apply methods of pressure [to North Korea]…can only complicate the possibilities for a settlement.”100 On

January 28, Dean Rusk again conveyed a desire for Soviet mediation of the Pueblo capture to

Soviet Foreign Minister Gromyko and assured him that the United States intended to resolve the crisis “peacefully and quietly.” Gromyko responded by stating that “The USSR could not be an intermediary in this case,” and also that the United States should not even pay attention to Soviet opinions on this matter.101 In the Sea of Japan off of North Korea’s eastern coast and the far eastern region of Soviet Russia, the Soviet Union mobilized five surface warfare ships to the general area as a precaution against a U.S.-North Korea conflict spilling into Soviet territory.102

In a February 3 letter to President Johnson, Soviet Premier Kosygin warned that the

“unprecedented” U.S. military buildup in Northeast Asia was concerning, justified Soviet mobilization of military forces to the region, and was the opposite of what was needed to resolve the situation. President Johnson replied on February 5 that although the U.S. military buildup

100 Telegram from the Department of State to the Embassy in the Soviet Union, January 25, 1968, Central Files 1967-69, folder POL 33-6 KOR N-US, RG 59, NARA; Telegram from the Embassy in the Soviet Union to the Department of State, January 27, 1968, Central Files 1967-69, POL 33-6 KOR N-US, RG 59, NARA. 101 Telegram from the Embassy in the Soviet Union to the Department of State, January 29, 1968, Central Files 1967-69, POL 33-6 KOR N-US, RG 59, NARA. 102 Speech by Leonid Brezhnev at the CC CPSU Plenum, April 9, 1968, Russian State Archive of Recent History, obtained and translated for the Cold War International History Project by Sergey Radchenko (Washington, DC: Woodrow Wilson International Center for Scholars).

110 was a direct response to North Korean provocations, he was prepared to ease tensions in the region by limiting naval and air force units to current levels and by withdrawing the USS

Enterprise aircraft carrier from the area.103 These U.S. actions were not immediately met with any reciprocal tension-reducing actions by North Korea or the Soviet Union. The withdrawal of the USS Enterprise from the area, which was ordered to the Philippines on February 16, enraged

South Korean President Park, who thought more, not less, aggressive action was needed against

North Korea.104

Crisis Resolution through Conciliatory Bargaining

Ultimately, the only crisis management approach to yield anything meaningful for the

United States was direct communication with North Korea through the UNCMAC channel at

Panmunjom, though the channel was used to seek an accommodative solution to the crisis that

North Korea initiated, not for the purpose of conveying threats. The morning after North Korea seized the Pueblo, on January 24, the U.S. representative to UNCMAC, Rear Admiral John

Victor Smith, sat across from his North Korean counterpart, Major General Park Chung-kuk, for the first of what would be 29 meetings over the course of the Pueblo crisis. Accordant with specific instructions from the State Department that came in overnight, Rear Admiral Smith protested the Blue House raid, and conveyed the following relating to the Pueblo: “One, return the vessel and crew immediately; two, apologize to the Government of the United States for this illegal action. You are also advised that the United States reserves the right to ask for

103 Telegram from the Department of State to the Embassy in the Soviet Union, February 6, 1968, Central Files 1967-69, POL 33-6 KOR N-US, RG 59, NARA. 104 “Aircraft Carriers in the Sea of Japan,” March 9, 1968, File “Pueblo Crisis 1968, Vol. 15,” NSF, NSC Historical Archives, Pueblo 1968, box 34, LBJ.

111 compensation under international law.”105 Rear Admiral Smith made no threat of violence or reprisal. Major General Park responded by laughing, calling President Johnson a “war maniac,” and noting that North Koreans have a saying: “A mad dog barks at the moon.”106 Park then claimed that the Pueblo illegally infiltrated North Korean territorial waters, and that therefore the

North Koreans demanded what came to be called the “Three As”: admit that the Pueblo violated

North Korean territorial space; apologize for the transgression; and assure North Korea that this would not happen again.107 North Korea would not move on these demands throughout the course of the crisis.

When the two sides met again on February 4, Major General Park demanded that, in addition to the “Three As,” the United States would be required to take U.S. and ROK forces in

South Korea off of alert, remove “U.S. airpower” in the vicinity of North Korea, and withdraw the Seventh fleet ships, including the USS Enterprise, that were dispatched to the Sea of

Japan.108 The Korea Task Force set up within the State Department recommended to Dean Rusk that the USS Enterprise be moved out of the area as a conciliatory gesture to help along negotiations.109 While President Johnson would inform the Soviets of his decision to withdraw the USS Enterprise the following day, it is unclear whether that decision was a point that was deliberated upon by the NSC.

105 Armbrister, A Matter of Accountability, 246; U.S. Department of State, January 23, 1968. “Telegram from the Department of State to the Embassy in Korea,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 216: 466-7. 106 Armbrister, A Matter of Accountability, 246. 107 Armbrister, A Matter of Accountability, 247. 108 U.S. Department of State, February 4, 1968. “Memorandum from the Director of the Korea Task Force (Berger) to the Secretary of State (Rusk),” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 264: 600-03. 109 Ibid.

112 For the next several months meetings would yield little more than frustration. On

February 16, Major General Park presented what he claimed was evidence of American guilt, including what would later prove to be coerced confessions and doctored navigational log books.

Occasionally North Korea floated the possibility of staging a trial of the Pueblo crew, which the

United States viewed as escalatory, though the United States viewed responding in kind as undesirable.110 By March 4, the North Koreans had provided the names of injured and wounded among the Pueblo crew, but little else. In a meeting toward the end of February, the United

States proposed resolving the matter by submitting it to a neutral investigation body, something that President Johnson had discussed in a conversation with UN Ambassador Goldberg at the outset of the crisis. If an impartial body found that the United States was guilty of violating

North Korean territory, then the United States would be willing to apologize.111 The North

Koreans rejected this proposal outright.112 During the tenth meeting, on March 4, Major General

Park alluded to the 1964 helicopter incident as a means of resolving the incident through U.S. contrition in the form of an “overwrite receipt.” In a cable from William Porter, U.S.

Ambassador to the ROK, to Washington, he acknowledged Park’s suggestion that there were

“precedents of having settled similar cases in the past under the terms mutually agreeable to both

110 U.S. Department of State, March 4, 1968. “Memorandum from Alfred Jenkins of the National Security Council Staff to the President’s Special Assistant (Rostow),” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 289: 649-51. 111 Armbrister, 1970: 284-5; U.S. Department of State, February 23, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 283: 637-40. 112 Armbrister, A Matter of Accountability, 285.

113 sides,” which Porter specified was a clear reference to the overwrite receipt method of apology from the helicopter incident.113

By the 16th meeting, Rear Admiral Smith had moved on to his next assignment, replaced by Major General Gilbert Woodward. Major General Park used this occasion to present a document for Woodward to sign that would resolve the crisis, but that contained the “Three As,” couched in the embarrassing and inflammatory language for which North Korea has become famous, including explicitly “acknowledging the validity of the confessions of the crew.”114

Assistant Secretary Warnke and Undersecretary Katzenbach from the State Department thought the United States should simply sign the document and be done with it while Rostow and Rusk were opposed; ultimately President Johnson agreed with Rostow and Rusk.115 While the United

States decided that the North Korean document was unacceptable, the State Department did instruct Woodward to propose in the following meeting the same overwrite receipt method of apology used in the 1964 helicopter incident, which Park rejected.116 On October 31, in a meeting that lasted only 20 minutes, Park announced that if the United States would not apologize, then there was nothing more to discuss; to the United States talks had not stalled but possibly appeared to have stopped, leaving them in a position of total surrender to North Korean demands, abandoning the Pueblo crew and writing them off as a foreign policy loss, or

113 U.S. Department of State, March 4, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 288: 647-9. 114 U.S. Department of State, May 8, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 306: 683-6. 115 Armbrister, A Matter of Accountability, 299. 116 U.S. Department of State, May 28, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 309: 690-1; U.S. Department of State, June 27, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 310: 691-3.

114 reconsidering military action.117 Amb. Porter assessed in a cable to Washington that North

Korea might be seeking to wait out the Johnson administration, hoping that a new administration would bring a new negotiating partner who might concede to North Korean demands.118 The solution to the crisis came from James Leonard, then the Country Director for Korea in the

Office of the Secretary of Defense.119 From the time of the initial UNCMAC meetings in

January, the United States had two basic problems with apologizing. First, aside from looking weak for being coerced into granting an apology, the United States could not bring itself to confess to a “crime” it did not commit; doing so would undermine its credibility.120 Second, even if the United States offered a transactional apology purely for instrumental purposes and then renounced the apology, the United States would face the same credibility problem.121 How could U.S. threats and promises amount to anything more than cheap talk if North Korea and the world see that the United States will simply say anything to achieve the influence it desires?

Leonard’s solution was thus reconciling the need to apologize with the need to be honest.

Woodward would sign the North Korean apology statement, but would gain North Korean agreement prior to signature that it would be acceptable for him to publicly renounce the apology after giving his signature. In a December 3 memo from Katzenbach to President Johnson, a time when the world already knew Richard Nixon would be taking over the presidency the following month, Katzenbach noted the obvious time pressures to have the Pueblo crew returned before the

117 Armbrister, A Matter of Accountability, 328. 118 U.S. Department of State, October 31, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 323: 725-6. 119 Leonard would later acknowledge that the idea actually came from his wife. 120 Armbrister, A Matter of Accountability, 334. 121 Ibid.

115 end of his administration, raising the Leonard proposal.122 At the 26th meeting, on December 17, and with President Johnson’s approval, Woodward met with Park and made the Leonard proposal; the United States would sign the North Korea-drafted apology, but only if North Korea agreed to allow the United States to immediately and publicly withdraw the apology.123

Although Park did not initially provide an answer, both sides reconvened three more meetings in less than a week, ending the crisis according to the terms of the Leonard proposal on December

23, 1968.

The North Korea Perspective

In this and subsequent cases, insights into North Korean deliberations and actions pose a greater challenge to discovery and documentation than do those of the United States.

Nevertheless, we are able to glean a good deal of insight from the evidence that exists of North

Korean words and deeds, much of which has only been made available in recent years through the opening of classified archives in former Soviet Bloc countries. In addition to the actions it took against the USS Pueblo, its crew, and in negotiations with the United States through

UNCMAC discussed above, we can further detail North Korean word and deed in three ways: coercive threats it issued immediately prior to its seizure of the Pueblo; military actions it took in the context of the crisis; and the perceptions, threats, and requests it shared with communist governments. We now know that North Korea acted alone in the Blue House raid and the

Pueblo seizure, both of which were conducted without China’s or the Soviet Union’s

122 U.S. Department of State, December 3, 1968. “Action Memorandum from the Undersecretary of State (Katzenbach) to President Johnson,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 324: 726-31. 123 U.S. Department of State, December 17, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 326: 724-5.

116 cognizance. And although some in the United States, like Walt Rostow, saw the possibility of

Soviet involvement where there was none,124 U.S. intelligence analysis also reached the conclusion that these provocations were precipitated as part of an independent North Korean foreign policy.125 We now know that the Blue House raid was intended to have the appearance of an indigenous South Korean insurrection that would either catalyze a communist revolution in the South or lead to sufficient civil unrest and confusion that the South might ask North Korea to intervene on behalf of the people.126 We also now know that North Korea likely had multiple and complex motivations for capturing the Pueblo, including halting U.S. intelligence collection missions,127 deterring U.S. intervention on behalf of the ROK in the event of a renewed war outbreak on the Peninsula,128 and contributing to a geostrategic battle against American imperialism in solidarity with North Vietnam.129 A telegram from Amb. Porter back to

Washington the day after the Pueblo was seized pointed to all three of these as forming a

“complex of reasons” for the act. Porter also judged North Korean resolve relative to its past behavior, pointing specifically to the 1964 helicopter incident and its provocative behavior toward ROK fishing vessels in the Sea of Japan, assessing that North Korea “will not permit any

124 Mitchell Lerner in particular places great emphasis on the proclivity of some U.S. decision-makers to see Soviet involvement where it did not exist. See Lerner, “A Failure of Perception.” 125 Director of Central Intelligence, February 29, 1968. Update to “North Korean Intentions and Capabilities with Respect to South Korea,” folder NIE 14.2 North Korea,” National security Files Country Files, NIEs, box 5, LBJ. 126 Stefans, “Confessions of a Red Agent,” 19; Ostermann and Person, Crisis and Confrontation on the Korean Peninsula. This is also discussed in great detail in a declassified CIA report written while North Korea still held the Pueblo crew. CIA Directorate of Intelligence, Kim Il-Sung’s New Military Adventurism. 127 Michishita, North Korea’s Military-Diplomatic Campaigns. 128 Director of Central Intelligence, February 29, 1968, Update to “North Korean Intentions and Capabilities with Respect to South Korea,” folder NIE 14.2 North Korea,” National security Files Country Files, NIEs, box 5, LBJ. 129 Armstrong, “Juche and North Korea’s Global Aspirations.” In his memoir, Lyndon Johnson expressed the incorrect belief that North Korea seized the Pueblo to put pressure on the ROK to withdraw its forces from the Vietnam conflict. Lyndon B. Johnson, The Vantage Point: Perspectives of the Presidency, 1963-1969 (New York: Holt, Rinehart and Winston, 1971), 333-5.

117 action by us to go unchallenged,” and that North Korea believes that the United States had

“…neither capability nor determination to deal with them while so heavily engaged in

Vietnam.”130 Once North Korea captured the Pueblo, it continued to hold it and its crew in order to exploit the incident for propaganda purposes.131

Prior to the Pueblo’s seizure, North Korea issued multiple public and private warnings that the United States must cease the intelligence collection missions it was operating in the Sea of Japan against North Korea. In the Fall of 1967, North Korea’s UNCMAC interlocutor, Major

General Park, leveled persistent accusations against his American counterpart for U.S. violations of North Korean territorial waters.132 On December 22, 1967, Radio Pyongyang announced that it seized what it called “armed espionage boats disguised as fishing vessels.”133 Radio

Pyongyang provided additional forewarning on January 6, 1968 when it claimed that the U.S. army “has been incessantly committing provocative acts lately on the sea off the eastern coast,” and that “this morning it again dispatched many armed boats, mingled with fishing boats, under the escort of armed warships into the coastal waters of our side.”134 A January 8, 1968 Radio

Pyongyang statement declared that “U.S. imperialist aggressors” were “causing a grave situation

130 “Telegram from the Embassy in Korea to the Department of State,” January 24, 1968, Central Files 1967-69, folder POL 33-6 KOR N-US, RG 59, NARA; U.S. Department of State, January 24, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 219: 481-3. 131 Frederick Schumacher, one of the captured crew, said North Korea was uninterested in exploiting the military and intelligence value of the spy ship and its crew, focusing instead on interrogating them in order to elicit confessions of wrongdoing: “Our value to them was apparently as propaganda pawns only.” See Ostermann and Person, Crisis and Confrontation on the Korean Peninsula, 18. Throughout the crisis, North Korea broadcasted daily to the South through loudspeakers the guilt of the Pueblo crew and the resolve of North Korea in opposing U.S. imperialism. See Bolger, An Unfinished War, 89. 132 Armbrister, A Matter of Accountability, 27. 133 Robert R. Simmons, The Pueblo, EC-121, and Mayaguez Incidents: Some Continuities and Changes. Occasional Papers/Reprints Series in Contemporary Asian Studies, 8, no. 20 (1978): 3. 134 Armbrister, A Matter of Accountability, 27; British Broadcasting Corporation, Summary of World Broadcasts, Far East/Issue 2663 (January 8, 1968): A3/6.

118 in which a war may break out at any moment.”135 A New York Times article on January 27, 1968 cited reporting from Japan’s Sankei Shimbun newspaper that the North Korean Labor Party publicly warned on January 9, 1968 that it might take action against the Pueblo if “the Pueblo continued its intelligence activities for more than two weeks.”136 Radio Pyongyang followed this charge on January 11 by repeating its January 6 accusation and adding that “As long as the U.S. imperialist aggressor troops conduct reconnaissance by sending spy boats, our naval ships will continue to take determined countermeasures.”137 On January 19, during the 260th meeting between North Korea and the UNCMAC representative at Panmunjom, Major General Park issued what may have been one last warning, telling his American counterpart, Rear Admiral

Smith, that U.S. “spy boats” were repeatedly “infiltrating” North Korean waters, and demanding that they “immediately stop such criminal acts.”138 Illustrating that North Korean talk had indeed become cheap to the United States, when Rear Admiral Smith was asked about these warnings, he called them “The usual Communist garbage.”139 The U.S. embassy in Seoul, the

State Department country desk officer for Korea, and the intelligence analysts at

COMNAVFORJAPAN had all dismissed the January warnings on similar grounds; to them

Pyongyang’s specific threats were indiscernible from the general background noise of threats heard from North Korea.140

135 British Broadcasting Corporation, Summary of World Broadcasts, Far East/Issue 2664 (January 9, 1968): A3/16. 136 “Warning by North Reported,” New York Times (January 27, 1968): 7. 137 As quoted in Armbrister, A Matter of Accountability, 27. 138 As quoted in Armbrister, A Matter of Accountability, 246-7. 139 As quoted in Armbrister, A Matter of Accountability, 28. 140 Ibid.

119 Less proximate warnings of another kind, based on North Korea’s past behavior, existed as well. As early as 1964, North Korea issued public threats for U.S. and ROK “spy boats” to back down, though U.S. vessels rarely operated against North Korean coastlines. More common were ROK naval and fishing vessels in the vicinity of North Korean territorial waters, which were not only routinely harassed, but occasionally captured and shot at; North Korea even sank a

ROK naval vessel in the Sea of Japan in 1967.141 The East German Ambassador to North Korea observed in December 1967 that “incidents at sea are occurring more and more, in the course of which fishing boats are being seized,” adding that “the incidents will occur increasingly in the coastal areas and at sea.”142 As further proof of sensitivity to reconnaissance off of its eastern coast, North Korea attempted to shoot down an RB-47 U.S. reconnaissance aircraft in 1965, something that U.S. intelligence analysts and General Bonesteel warned demonstrated North

Korean resolve, particularly against “interference” in or around Wonsan, where the Pueblo was also seized.143 These types of warning statements became more common in November 1967, just two months before the Pueblo crisis.144 North Korea had thus demonstrated over the years in a general sense that it was sensitive to activity on its eastern coast, in addition to proving its willingness to directly target Americans with provocative attacks, not just the ROK. Although the rise of North Korea-instigated DMZ violence and the Blue House raid signaled an

141 Armbrister, A Matter of Accountability, 187. 142 Declassified Memorandum from Ambassador of the GDR in the DPRK to State Secretary and First Deputy Minister of Foreign Affairs, December 8, 1967, Obtained and translated for the North Korea International Documentation Project (NKIDP) by Karen Richert (Washington, DC: Woodrow Wilson International Center for Scholars). 143 Armbrister, A Matter of Accountability, 120-1; U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121, 1654-5. 144 Responses to Senator William Fulbright, 1968, folder “Pueblo Crisis, 1968, Vol. 10, Fulbright Letter,” NSC Historical Archives, Pueblo Crisis, box 31, LBJ.

120 unprecedented level of risk acceptance by North Korea, it may be argued that the DMZ violence and the Blue House raid could not have been predictive of North Korea’s capture of the Pueblo.

But its threatening rhetoric and provocations, particularly against “spy boats,” should have been taken as a warning of North Korean resolve; the United States should have accordingly not engaged in the activity that North Korea proscribed unless it was willing to retaliate, which it was not.

As the Pueblo crisis unfolded, U.S. intelligence collection efforts identified defensive preparations for war and heightened volumes of military activity around the port of Wonsan, where the Pueblo had been brought in, but U.S. analysts did not believe that North Korea was planning an invasion.145 U.S. forces in South Korea and Soviet Bloc diplomats stationed in

Pyongyang alike took special notice of North Korean military actions in the first few days of the crisis. Throughout the crisis but especially in the early days, North Korean raids across the DMZ surged. The day after North Korea seized the Pueblo the Romanian embassy in Pyongyang cabled back to its ministry in Bucharest that since at least the January 21 Blue House raid and continuing as of his writing North Korea began establishing military checkpoints, carrying out

“daytime and nighttime raids,” establishing “armed groups of civilians,” increased police presence at the embassies in Pyongyang, and prohibited foreigners from traveling outside

Pyongyang.146 The same cable message assessed that “North Koreans are currently testing the nerves and feeling out the attitudes of all elements in South Korea, including the US, toward a

145 Richard Mobley summarizes U.S. intelligence views relating to Pyongyang’s war footing during the first week of the crisis in Mobley, Flashpoint North Korea, 80-3. 146 “Telegram from Pyongyang to Bucharest,” January 24, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe (Washington, DC: Woodrow Wilson International Center for Scholars).

121 potential large-scale military confrontation.”147 On January 25 North Korean forces launched six skirmishes against U.S. soldiers along the DMZ, injuring several. In a move that unsettled even

Soviet Bloc diplomats in Pyongyang, North Korea flew two squadrons of MiG fighter jets within five miles of the DMZ; North Korean jets rarely approached closer than 30 miles of the DMZ.148

Public messaging buttressed not only its provocative shows of, and use of, force but also the private messages it was conveying to friendly communist governments and to Rear Admiral

Smith at UNCMAC. Radio Pyongyang played its part by declaring on January 25 that “The criminals who encroach upon others’ sovereignty…must receive deserving punishment,” noting that they were “ready to deal a hundred fold, a thousand fold retaliation.”149 On January 26, the

New York Times quoted the Rodong Shinmun, the newspaper of the North Korean party apparatus, as saying that if “American Imperialists take more high-handed measures,” the North

Koreans would be “ready to adopt decisive countermeasures.”150

Observations from communist diplomats operating in Pyongyang during the Pueblo episode shed some additional, if still limited, light on North Korean thinking throughout the crisis as well. Above all North Korea did not seem to believe that the United States would attack it, though its reasoning is not completely transparent, and it did believe that the United States would ultimately be willing to resolve the matter diplomatically in a manner that would prove humiliating to the United States. Still North Korea was consistent in conveying to all audiences

147 Ibid. 148 Armbrister, A Matter of Accountability, 262; “Information about the Incident with the Ship Pueblo,” January 28, 1968, Archive of the Czech Foreign Ministry, Obtained and translated for the NKIDP by Adolf Kotlik (Washington, DC: Woodrow Wilson International Center for Scholars). 149 As quoted in Ibid. 150 “Imperialists Accused,” New York Times (January 27, 1968): 7.

122 its willingness to risk war in the resolution of the Pueblo crisis, including to Soviet Bloc countries. On January 24, 1968, the day after the Pueblo was captured, North Korean Deputy

Foreign Minister Kim Jaebong convened all of the heads of diplomatic missions accredited to

Pyongyang in order to brief them on the details of the incident. Deputy Foreign Minister Kim’s portrayal of the incident was, as expected, consistent with North Korean claims that the United

States violated their territorial waters with aggressive spying, and that they had a confession from the crew’s captain.151 Kim requested his fellow communist governments to condemn

America’s “pirate-like acts,” and lamented that the situation would likely “become worse because the American imperialists had already deployed an important unit from the Seventh

Fleet.”152

The North Koreans also continually pressed the Soviets at every level of government to determine whether the Soviet Union would support North Korea in the event that all-out war resumed, though on this point it never received an optimistic answer from the Soviet Union.153

While the Soviet Union cautioned North Korea against taking rash action, Soviet diplomats expressed the sense that North Korea did not believe the United States was willing to renew conflict on the Korean Peninsula, which helped explain why North Korea was willing to engage in hostilities with the United States.154 On January 26, 1968, North Korea’s foreign minister,

Park Seongcheol, met with the number two official of the East German embassy in Pyongyang,

151 “Telegram from Pyongyang to Bucharest,” January 24, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 11-12. 152 Ibid. 153 “Telegram from Pyongyang to Bucharest,” January 25, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 14-15. 154 Ibid.

123 and repeated quite coldly and explicitly what North Korea had expressed in other channels: “The

North Korean government will take the same form of action in the future, capturing or even sinking any enemy ship…If the US undertakes any acts of intimidation or uses force, the DPRK will reply through the same means, including force. The resolution of conflict depends entirely upon the Americans.”155

In a February 2, 1968 meeting with the Romanian ambassador to Pyongyang, North

Korean diplomats conveyed their resolve in similarly bombastic terms, stating that U.S. threats solve nothing and they are “determined to fight if we need to fight.”156 The North Korean diplomats also conveyed that “In the given circumstances, we need to strengthen our unity…with all anti-imperialist countries, so as to fight adamantly against imperialists and reactionaries.”157

And in a briefing to friendly communist governments on February 26, weeks after the United

States withdrew its “provocative” military assets from the Sea of Japan, North Korean diplomats underscored that a heightened state of tension persisted, which justified the amplified shows of force that North Korea continued to conduct, which include nation-wide troop movements, air defense exercises, nighttime aircraft exercises, and the proliferation of bunkers and trenches throughout the country.158

155 “Telegram from Pyongyang to Bucharest,” January 26, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 17. 156 “Telegram from Pyongyang to Bucharest,” February 2, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 25-6. 157 Ibid. 158 “Telegram from Pyongyang to Bucharest,” February 27, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 38.

124 Analysis: Explaining Decisions and Outcomes

The final section of this chapter filters the above narrative of the USS Pueblo incident and the period preceding it to assess how well the competing hypotheses on reputation and credibility explain the decision-making and outcomes observed in this case. I discuss the fit of the multiple observations in this case with the expectations of the hypotheses advanced in chapter 2.

The Pueblo crisis was initiated by an act of North Korean violence. The United States believed that if it responded with an act of military compellence, North Korea would respond with escalation to war. Although the United States initiated a gradual military buildup in the region and called up reserve forces, these were non-specific responses to a specific challenge, and had no effect on the crisis. The United States backed down from North Korea’s challenge, meeting North Korea’s attack and seizure with outwardly coolheaded and conciliatory diplomacy, even if internally U.S. decision-makers frustratingly cast about for retaliation options that would meet two conditions: that would get the Pueblo crew back, and that would be sufficiently proportional that it would not lead the crisis to escalate to war. At the first round of diplomatic talks, the United States laid out demands for the return of the crew and an apology.

By the next round of talks and subsequently, not only did the United States abandon its demand for an apology but the bargaining context shifted to being about the terms of a U.S. apology—the

“three As” formulation that North Korea demanded from the opening of the first diplomatic round of talks. The rough sequence of the crisis observed was thus: (1) North Korea decides to initiate a challenge against the United States; (2) North Korea initiates the challenge; (3) the

United States is surprised by the challenge; (4) the United States deliberates about how to

125 respond; (5) the United States backs down, negotiating with North Korea the terms of a U.S. apology and restraining U.S. military forces. So in addition to being able to observe the pattern of interaction that emerged between 1964 and 1968 in this case, we are also able to make multiple observations relating to our hypotheses within the Pueblo crisis itself.

Explanatory Power of Reputations for Honesty

There is some support for the reputations for honesty perspective in this case, but far from total support, and within the case there is not an opportunity to evaluate every claim made by the reputations for honesty literature. As the Pueblo crisis began, the United States had a reputation for honesty and North Korea did not. This case does not offer an opportunity to observe whether U.S. threats became more credible because of honest communication in previous crises (because the United States did not issue threats in this case), but it does offer several observations that fit honesty-based reputational expectations. First, U.S. decision-makers went out of their way to maintain the U.S. reputation for conducting honest diplomacy, as Anne

Sartori and others might expect; indeed, U.S. concerns about diplomatic honesty were the principal reason the incident lasted nearly a year. Second, the United States did not view North

Korea’s threats prior to the Pueblo seizure as credible because of a historical gap between North

Korean threats (word) and actions (deed) wherein innumerable prior threats had gone unfulfilled; the reputations for honesty frame tells us that a failure to match word and deed causes others to interpret threats as lacking credibility. Third, it does appear that the United States found North

Korea’s threats during the crisis credible, but only after the initial challenging action, matching prior threats, occurred; when North Korean word and deed aligned, the United States did find its

126 threats credible. However, reputations for honesty predictions fail to adequately explain at least one crucial aspect of this case. The four years leading up to the Pueblo consisted of repeated interactions in which North Korea initiated challenges against the United States and the United

States backed down. Because backing down from a confrontation is a form of honest communication, the reputations for honesty thesis would expect that North Korea would have been less likely to initiate the Pueblo incident in the first place; from this view, states are supposed to be less likely to challenge states that have a reputation for honesty.

U.S. decision-makers valued the U.S. reputation for honesty throughout the 1964-68 period. Prior to the Pueblo crisis, the United States avoided making empty threats against North

Korea, even though it suffered repeated provocations that gradually escalated in intensity.

During the Pueblo crisis too, the United States refrained from making threats or promises that it was not willing to execute. At several points during the Pueblo crisis, the United States explicitly discussed and rejected the possibility of conceding to North Korea’s demands, securing the release of the Pueblo and crew, and later reneging on the apology. In President

Johnson’s phone call with Amb. Goldberg two days after the Pueblo was captured, President

Johnson referenced with a dismissive laugh Senator Mike Mansfield’s recommendation to offer an instrumental apology and then withdraw the apology after securing the crew.159 Months later, when Major General Park presented the United States with a written apology and requested the

United States sign it to have the crew returned, U.S. officials were divided about whether they should falsely confess, and whether to repudiate any confession made. President Johnson again

159 Telephone Conversation between President Johnson and the Representative to the United Nations (Goldberg), January 28, 1968, Recordings and Transcripts, Recording of Telephone Conversation between President Johnson and Goldberg, Tape F68.01, PNO 2, LBJ.

127 decided to neither offer a false apology, nor to renege on any apology offered.160 On March 4, a member of the NSC staff, Alfred Jenkins, sent a memo to Walt Rostow describing an inauthentic apology to North Korea as “inappropriate” and “not in the American character.”161 Jenkins observed that “absolutely basic to problems of this sort is the integrity, prestige, and credibility of the United States,” and that it would be a mistake to apologize “with tongue quite prominently in cheek” for the sake of saving the crew because the one advantage the United States had was

“not in what we do but in who we are.”162 In September 1968, an NSC memo summarized the evolution of circumstances involving the Pueblo, stating that negotiations had been at an impasse because “We have refused to admit espionage or to apologize for acts we are morally certain we did not commit.”163 And in the negotiations’ end game during the final two months, the United

States focused its effort primarily on convincing North Korea to accept that the U.S. apology would be rescinded after the crew was returned, a position that North Korea eventually accepted.

Another way in which the reputations for honesty hypothesis explains aspects of dyadic behavior in this period particularly well is in U.S. determinations of credibility about North

Korean threats in the period prior to and during the Pueblo crisis. Leading up to North Korea’s seizure of the Pueblo, North Korea established a track record of aggression against the United

States and ROK, but also issued multiple threats specific to the Pueblo that the United States ignored. At the same time, the United States did find North Korea’s intra-crisis threats credible.

160 Armbrister, A Matter of Accountability, 299. 161 U.S. Department of State, March 4, 1968. “Memorandum from Alfred Jenkins of the National Security Council Staff to the President’s Special Assistant (Rostow),” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 289: 649-51. 162 Ibid. 163 U.S. Department of State, September 4, 1968, “Memorandum from Alfred Jenkins of the National Security Council Staff to the President’s Special Assistant (Rostow),” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 312: 697-700.

128 The Pueblo’s intelligence collection missions were approved through a complex process involving the many components of DoD, the intelligence community, and the State Department.

Richard Mobley’s book, Flashpoint North Korea, explains in great detail how and why the

Pueblo mission was approved. The general demand for the collection mission was a determination by the U.S. Intelligence Board in October 1967 to increase intelligence targeting of North Korean capabilities and intentions due to the noticeable increase in provocations in the

Korean theater.164

The proximate decision for the Pueblo mission to move forward as a means of satisfying this demand was ultimately the result of a near consensus among the U.S. interagency that the

Pueblo mission against North Korea was “minimal risk”; North Korea’s general hostility and threatening rhetoric was judged as not representing a credible threat to U.S. intelligence assets.

Mobley contends that given the force demands of Vietnam, it was unlikely that non-Vietnam intelligence missions in the Pacific theater would be approved unless they were assessed as low risk.165 Some parts of the U.S. government assessed risk differently and more systematically than others, yet they all reached the same conclusion: the Pueblo did not face any serious or credible threats by operating off of North Korea’s eastern coast, despite increased hostility and a rise of threats emanating from the North. For example, COMNAVFORJAPAN used a list of criteria for determining an overall minimal risk assessment, including: nature and sensitivity of the operation; possibility of hostile reactions; previous experiences in the proposed area of

164 Mobley, Flashpoint North Korea, 27. 165 Mobley, Flashpoint North Korea, Ch. 3.

129 operation; anticipated intelligence take; and opposing forces.166 The DIA, by contrast, had a four-tiered hierarchy of risk levels based on likelihood of hostile intent, likelihood of intercept actions, and the anticipated degree of hostility.167 At the more senior levels of government, risk assessments hewed to no particular methodology, but rather were the collective judgment of

Deputy Secretary-level officials at the Joint Chiefs of Staff, Office of the Secretary of Defense,

CIA, and State Department. Although specific details and risk considerations were part of the

Deputies’ approval process, the Pueblo mission was one of more than 800 collection missions approved for January 1968, making it impractical to oppose a particular proposed mission by the time it reached their desks unless there were distinctive and noteworthy reasons to do so relative to the many other missions being considered.168 Although one analyst from the National

Security Agency (NSA) did issue a report after the mission was approved warning that the

Pueblo mission may face risks given North Korean sensitivity to U.S. and ROK assets operating in the Sea of Japan, citing specifically the 1965 attempted shoot down of the U.S. RB-47 reconnaissance aircraft,169 decision-makers at every level of the collection mission approval process were so inured to North Korean hostile rhetoric that new threats did not seem any more credible than the steady stream of invectives in the years prior to the Pueblo’s seizure.

DoD and the DIA determined that Pueblo’s mission would face minimal risk because

North Korea did not react to prior naval patrols in the Sea of Japan by an intelligence collection ship, the USS Banner, despite an increase in threatening rhetoric during its mission in 1966 and

166 Mobley, Flashpoint North Korea, 28-9. 167 Mobley, Flashpoint North Korea, 31. 168 U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121, 1646, 1654-5. 169 Armbrister, A Matter of Accountability, 196-7; U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121, 1654-5.

130 1967.170 Put in the language of reputations for honesty, an observed gap between North Korean word and deed led to DoD and DIA’s assessment that North Korea posed a minimal threat to the

Pueblo’s mission. At NAVFORJAPAN, Rear Admiral Johnson approved the Pueblo mission’s

“minimal risk” assessment even though he was aware of increased North Korean provocations and overall hostility in the year prior.171 At CINCPACFLT, the assistant chief of staff for intelligence approved the Pueblo mission without scrutiny: “What we saw and heard [from North

Korea] didn’t seem any different from what we had been seeing and hearing for the past ten years…It took me about as long to approve it as it did to read it.”172 The State Department’s intelligence bureau had similarly assessed that while North Korea’s hostility toward the United

States and ROK intensified in 1966 and 1967, the country’s threatening statements were not useful indicators of what North Korea would actually do.173

As incredulous as the United States found North Korean threats foreshadowing its action against the Pueblo, once the crisis began the United States did deem North Korean threats far more credible than it had before the seizure. During the Pueblo crisis, U.S. intelligence estimated that North Korea was not capable of defeating U.S. and ROK forces on its own, and that in the event of U.S. retaliation against North Korea for the Pueblo China and Russia were

170 Armbrister, A Matter of Accountability, 167. The Banner operated collection missions against North Korea’s eastern coast three times without incident: March 1966, February 1967, and May 1967. See Mobley, Flashpoint North Korea, 24. 171 U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121, 1638. 172 Armbrister, A Matter of Accountability, 189. 173 Memorandum from INR/REA Fred Greene to Claus Ruser, August 1, 1968, “State of Intelligence on North Korea,” Folder Korea 2 of 2, Senior Interdepartmental Group Files 1966-72, box 8, RG 59, NARA; Armbrister, A Matter of Accountability, 28.

131 unlikely to intervene on North Korea’s behalf.174 As Mitchell Lerner has argued, however, these facts about material capabilities and predicted actions did not resonate with U.S.—particularly

Lyndon Johnson’s—views of North Korean threat credibility.175 President Johnson and his advisors believed that North Korea would escalate the conflict and bring on a war if the United

States initiated a military response to the Pueblo seizure, and feared that China or the Soviet

Union would join them in the fight.176 As early as the second meeting between President

Johnson and his advisors after the Pueblo’s seizure, Johnson began distancing himself from the possibility of military force against North Korea on the assumption that it would lead to opening a second war front in Asia.177 As lower level officials worked through and ultimately rejected various forms of military retaliation against North Korea, the reasoning was always that the action would not only fail to secure the release of the crew, but also that the action would incite an escalatory response from North Korea in turn.178 Thus, the prevailing belief among U.S. decision-makers, that North Korea would escalate conflict if the United States took a firm stand

174 The Office of National Estimates of the CIA, “Communist Reaction to Certain US Actions,” January 26, 1968, Pueblo Crisis, 1968, Vol. III, Day-by-Day Documents, Part 3, National Security File NSC History, LBJ; U.S. Department of State, January 26, 1968. “Memorandum from Under Secretary of State (Katzenbach) to President Johnson,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 229: 530-2. 175 Lerner, “A Failure of Perception.” 176 Ostermann and Person, Crisis and Confrontation on the Korean Peninsula; Lerner, “A Failure of Perception”; U.S. Department of State, January 24, 1968. “Summary Minutes of Pueblo Group,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 217: 468-75. 177 U.S. Department of State, January 25, 1968. “Notes on the President’s Thursday Night Meeting on the Pueblo Incident,” drafted by Tom Johnson, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 226: 514-19; Telephone Conversation between President Johnson and the Representative to the United Nations (Goldberg), January 28, 1968, Recordings and Transcripts, Recording of Telephone Conversation between President Johnson and Goldberg, Tape F68.01, PNO 2, LBJ. 178 “Report on Meeting of Advisory Group, January 29, 1968,” February 1, 1968, Folder “Material regarding Vietnam and Pueblo, January-February 1968,” National Security File, Walter Rostow files, box 10, LBJ; Memorandum from the Director of the Korean Task Force (Brown) to the Under Secretary of State (Katzenbach), March 15, 1968, “Possible Actions in the Event the North Koreans Announce that They Intend to Try the Pueblo Crew,” Central Files 1967-69, RG 59, folder POL 33-6 KOR N-US, NARA.

132 against North Korea’s attack against the Pueblo, was a direct cause of the U.S. decision to back down and avoid taking steps to demonstrate resolve; the credibility with which the United States viewed North Korean threats during the crisis was much stronger than before the crisis began.

Although the reputation for honesty logic helps explain the aforementioned aspects of the period prior to and during the Pueblo crisis, it completely fails to explain a crucial aspect of dyadic behavior during this period: the recurrence of North Korea-initiated provocations against the United States. In the years prior to the Pueblo’s seizure, North Korea initiated hundreds of violent challenges against the United States and ROK. Several of these challenges—such as the

1965 attack on a U.S. reconnaissance aircraft, the 1966 raid against U.S. patrol units near the

DMZ, and the 1967 attack on U.S. infantry barracks—specifically targeted the United States.

The U.S. response in each of these cases, as with the hundreds of other attacks across the DMZ and in the Sea of Japan, was to back down, either through conciliatory concessions in a negotiation or by not responding in kind to North Korean provocations.179 According to the logic of reputations for honesty, the repeated U.S. decisions to back down reflected that the attacks did not represent a sufficient challenge to U.S. interests to warrant the risk associated with demonstrating resolve; backing down from each challenge was simply the honest alignment of word and deed. The U.S. reputation for honesty cultivated in this way should have resulted in a reduced likelihood of future challenges from North Korea;180 instead, each instance of backing down was followed by yet another challenge within a year. The seizure of the Pueblo fit this

179 U.S. Department of State, November 8, 1966. Intelligence Memorandum, “Armed Incidents along the Korean DMZ,” FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 98: 209-10; U.S. Department of State, June 23, 1967. Report Prepared by the Office of National Estimates of the Central Intelligence Agency, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 120: 259. 180 Sartori, Deterrence by Diplomacy.

133 pattern as well. So while reputations for honesty as an explanatory frame helps make sense of parts of the narrative of U.S.-North Korea interactions between 1964 and 1968, its predictions of the reputational effects on future conflicts because of decisions to back down are an extremely poor fit for how events unfolded in this case.

Explanatory Power of Reputations for Resolve

The logic of reputations for resolve explains some of the same phenomena in this case as reputations for honesty, but also manages to fill the major explanatory gap left by reputations for honesty. As the Pueblo crisis began, the United States had a reputation for irresoluteness resulting from the word and deed established in the years prior. The interactions throughout this case do not put North Korea in the position of defender at any point because the United States never initiated a challenge against it, nor did the United States ever stand firm in response to a

North Korean challenge. This limits our ability to observe some aspects of reputations for resolve, but gives us powerful explanations of this case in other respects. As with reputations for honesty expectations, reputations for resolve would expect the United States to view North

Korea’s threats, presaging its seizure of the Pueblo, as incredible because so many prior threats of action had gone unfulfilled. The logic of reputations for resolve is such that future threats are not likely to be credible if past threats were not executed. We also find that while North Korea’s seizure of the Pueblo did contribute to a perception of it having the attribute of resolve during the crisis, there did not seem to be a cumulative reputational effect as a result of the many challenges it initiated in the years preceding the Pueblo incident. Unlike reputations for honesty, the consequences for backing down from a confrontation as predicted in the reputations for resolve

134 literature explains precisely the repeated pattern of interaction in this case: when the United

States backed down, it cultivated a reputation for irresoluteness, encouraging future challenges.

This accurately describes the four year period prior to the Pueblo’s seizure, but also offers an explanation for why the Pueblo was seized at all. For reputations for honesty, the recurrence of confrontational interactions in this case is a puzzle; for reputations for resolve, such recurrence is the price the United States paid for repeatedly backing down.

The United States was surprised by the attack on the Pueblo because it did not find North

Korea’s threats against Pueblo-like activity off its eastern coast credible; the United States could not distinguish these threats as having somehow more credibility than the many preceding and ongoing threats that went unfulfilled. Although both the resolve and honesty logics anticipated that the United States would not expect North Korea to fulfill its threats against the Pueblo because so many other threats went unfilled, the resolve logic attributes this outcome (that is, of

North Korean threats lacking credibility) to North Korea’s failure to fulfill past threats, which revealed a lack of resolve; the honesty logic attributes this outcome to North Korea’s failure to communicate honestly in the past. The direct evidence supporting the honesty logic’s explanation for why the United States did not find North Korea’s warnings against the Pueblo credible also supports the resolve logic’s explanation for the same. The near total U.S. consensus about the Pueblo mission being “minimal risk,” for example, explicitly referenced historical precedent—that is, North Korea’s past behavior toward U.S. intelligence collectors in the Sea of Japan—as the basis for not finding North Korean threats credible.181 A reputational explanatory approach wholly anticipated the outcome of U.S. incredulity toward North Korean

181 Armbrister, A Matter of Accountability, 167.

135 threats against Pueblo-like activity prior to January 23, 1968, but does not offer definitive evidence to suggest whether an honesty or resolve logic best explains how this outcome occurred.

The logic of reputations for resolve also points in the same direction as that of honesty to explain the shift in U.S. perceptions of North Korean credibility during the Pueblo crisis, but the evidence does not allow us to identify either the honesty or resolve logic as explanatorily superior. In the shadow of the Pueblo’s seizure, not only President Johnson but also the working groups that reviewed and recommended options for President Johnson and his advisors believed that North Korea would indeed fulfill its threats to escalate if the United States took military action to recover the Pueblo or punish North Korea.182 The day after North Korea captured the

Pueblo, Amb. Portman referenced North Korea’s conduct in the 1964 helicopter incident and

North Korea’s aggression against ROK fishermen in the Sea of Japan as reasons to expect that

North Korea “gives no ground for optimism” that North Korea would respond favorably to U.S. actions to secure the return of the crew.183 Amb. Portman inferred from North Korea’s willingness to seize the Pueblo that it “will not permit any action by us to go unchallenged. They seem confident and sure of themselves…”184 In preparation for a meeting with President

Johnson later in the day, on the morning of January 24 his advisors met to discuss options for

182 U.S. Department of State, January 25, 1968. “Notes on the President’s Thursday Night Meeting on the Pueblo Incident,” drafted by Tom Johnson, FRUS, 1964-68, Vol. XXIX, Part 1, Korea, document 226: 514-19; “Report on Meeting of Advisory Group, January 29, 1968,” February 1, 1968, Folder “Material regarding Vietnam and Pueblo, January-February 1968,” National Security File, Walter Rostow files, box 10, LBJ; Memorandum from the Director of the Korean Task Force (Brown) to the Under Secretary of State (Katzenbach), March 15, 1968, “Possible Actions in the Event the North Koreans Announce that They Intend to Try the Pueblo Crew,” Central Files 1967-69, RG 59, folder POL 33-6 KOR N-US, NARA. 183 U.S. Department of State, January 24, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 219: 481-3. 184 Ibid.

136 responding to the Pueblo. CIA Director Richard Helms offered the meeting’s initial framing observation that the outcome of the Pueblo “is a question of whether we or they have the tougher will.”185 The above indicates that the seizure of the Pueblo did have a near term impact on U.S. perceptions of North Korean resolve that did not exist prior to initiation of the Pueblo crisis.

One aspect of this case that the resolve logic explains uniquely well is the recurrence of

North Korean provocations in the 1964 through 1968 period leading up to and including the seizure of the Pueblo. Questions about intra-crisis credibility assessments are only salient if the crises occur in the first place. The resolve and honesty logics both maintain that how a crisis ends matters, but part ways principally on whether backing down as a means of crisis resolution yields positive or negative consequences. The resolve logic tells us that backing down leads to rivals attributing to you a reputation for irresoluteness, which in turn makes them think their aggression toward you will go unchallenged. It is thus the combination of the rivalry context— constituted by repeated play, incomplete information, and shared enmity due to both historical interaction and incompatible goals—with the dynamics of a reputation for irresoluteness that motivates repeated challenges. The pattern of interaction leading up to and during the Pueblo crisis—with North Korea initiating a challenge, the United States backing down, and North

Korea issuing a new challenge subsequently—fits with the expectations of a reputation of irresoluteness earned by repeatedly backing down. Some evidence from within the Pueblo crisis supports this interpretation as well. Immediately after North Korea seized the Pueblo, for example, Soviet diplomats in Pyongyang assessed that North Korea believed it was conducting

185 U.S. Department of State, January 24, 1968. “Summary Minutes of Meeting,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 217: 468-75.

137 provocations against the ROK and the United States to exploit the unwillingness of the United

States to directly clash with North Korea.186 The day after the Pueblo was seized, the Romanian ambassador to Pyongyang assessed North Korean actions the same way, positing that North

Korea was “testing the nerves and feeling out the attitudes” of the ROK and United States toward renewed conflict in Korea.187 On February 6, the Press Director at North Korea’s foreign ministry told the Romanian embassy in Pyongyang that “the Americans are searching for the best possible pretense to excuse themselves” from the crisis.188 Explicitly taking into account the

U.S. track record of responses to North Korean provocations, the Romanian embassy in

Pyongyang subsequently asserted that it was “the tendency of the Americans to admit their guilt in one way or another in front of the North Koreans,” and that “the prestige and reputation of the

Americans…has been seriously undermined.”189 In the same cable, the Romanian embassy drew a causal link between U.S. irresoluteness and North Korean aggression, observing that since the

United States responded to the crisis with a plea for secret talks rather than a military response, most of North Korea’s infiltrations across the DMZ were specifically targeting areas guarded by the American Second Infantry Division, and that even the ROK lacked trust in the U.S. willingness to ensure its security.190 Richard Helms’ characterization of the Pueblo crisis as

186 “Telegram from Pyongyang to Bucharest,” January 25, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 14-15. 187 “Telegram from Pyongyang to Bucharest,” January 24, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe (Washington, DC: Woodrow Wilson International Center for Scholars). 188 “Telegram from Pyongyang to Bucharest,” February 7, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 28-9. 189 Ibid. 190 Ibid.

138 essentially a battle of wills expresses a similar U.S. understanding of the Pueblo crisis as above, if not the pattern of similarly aggressive behavior that preceded the crisis.191 The aforementioned process tracing evidence, combined with the strong fit between theoretical expectation and empirical outcome, indicates that the reputations for resolve logic explains North

Korea’s repeated challenges to the United States, including the seizure of the Pueblo. As we will see in the next chapter, this pattern would continue beyond the Johnson administration to also plague Nixon and Kissinger.

Alternative Explanations about Credibility

Although Reputation Pessimists have little to say about the role of past word and deed in determining the likelihood of future confrontations, they do offer claims relating to threat credibility that compete with the honesty and resolve logics’ explanations about credibility, above. How do current calculus theory and constant calculus theory’s expectations match the evidence in this case? Even if current calculus theory might be used to explain the pattern of

North Korea initiating challenges and the United States backing down repeatedly from 1964 to

1967, we find that the U.S. decision to back down in response to the Pueblo’s seizure is difficult to explain using current calculus theory. Likewise, U.S. credibility assessments of North Korean threats against Pueblo-like actions are difficult to explain using constant calculus theory. U.S. intra-crisis credibility assessments of North Korean threats do match what constant calculus theory would lead us to expect, but there is scant evidence within the case to suggest that the

191 U.S. Department of State, January 24, 1968. “Summary Minutes of Meeting,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 217: 468-75.

139 reasoning for U.S. credibility assessments of North Korean threats after the Pueblo was seized have to do with preferential biases relating to an in-group/out-group distinction.

Current calculus theory claims that assessing the credibility of another state’s threats is a function of the interaction between the balance of power and the national interests at stake. The more favorable the balance of power and the higher the interests at stake, the more credible are one’s threats and vice versa. 192 Applied to this case, it is unclear why the United States should find North Korea’s threats credible in terms of material capabilities. An extreme power asymmetry between the United States and North Korea is one of the obvious defining characteristics of this relationship. North Korea did possess the capability to start a war with the

United States and ROK, but not to win a war. Even excluding U.S. capabilities, South Korea maintained a standing army of 534,000 to North Korea’s 345,000, a marine force of 31,000 while North Korea lacked a marine force, and a naval force of 18,000 to North Korea’s

10,000.193 South Korea’s air force also outnumbered North Korea’s, with 25, 500 to North

Korea’s 18, 500, but North Korea maintained what were at the time qualitatively superior modern MiG fighters from the Soviet Union that South Korea could not match.194 The point of highlighting these numbers is simply to show that North Korea’s military in the 1960s was robust but not vastly superior to South Korea’s military, let alone the United States. Though the numbers give us a sense of what North Korea could do, they are not inherently meaningful; there is nothing in objective calculations of North Korea’s military capability to suggest that North

Korea’s relative capabilities should be determinative of its threat credibility. Drawing on

192 Press, Calculating Credibility. 193 Mobley, Flashpoint North Korea, 16. 194 Ibid.

140 frequently used data from the Correlates of War project for calculating power using material indicators, we find, unsurprisingly, that from 1964-68 the balance of power between the United

States and North Korea consistently favored the United States by large amounts; U.S. material capabilities in this period ranged from .201 to .208, whereas North Korea’s capabilities ranged from .0048 to .0053, several orders of magnitude smaller than the United States.195

One might argue that including Chinese or Soviet military capabilities with North

Korea’s capabilities might create a more favorable balance of power which would better rationalize, from the current calculus lens, why the United States would find North Korea’s threats credible, but there are two problems with this argument. First, we have evidence from the case that U.S. intelligence assessed that neither China nor the Soviet Union would intervene in a war with the United States on North Korea’s behalf, which recently declassified documents from

Soviet archives corroborates.196 Second, current calculus theory assesses the balance of power in dyadic terms only; it does not specifically accommodate aggregating the capabilities of multiple actors for the summation of capabilities as part of calculating the balance of power. To do so would at any rate drastically complicate the parsimony of current calculus theory given that the challenge of attempting to assess the interests at stake—an already difficult task—would be multiplied several times over. North Korea’s military capabilities certainly affected aspects of

195 Barbieri, Katherine, Omar Keshk, and Brian Pollins. “Correlates of War Project Trade Data Set Codebook,” Codebook Version 2 (2008). Data available at: http://www.correlatesofwar.org/COW2%20Data/Capabilities/nmc4.htm.

196 Sergey S. Radchenko, “The Soviet Union and the North Korean Seizure of the USS Pueblo: Evidence from Russian Archives,” CWIHP Working Paper #47 (Washington, DC: Woodrow Wilson International Center for Scholars); Director of Central Intelligence, February 29, 1968, Update to “North Korean Intentions and Capabilities with Respect to South Korea,” folder NIE 14.2 North Korea,” NSF Country Files, NIEs, box 5, LBJ. See also U.S. Department of State, January 24, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 219: 481-3.

141 this case. Though not specifically relevant to the operationalization of current calculus theory, the local balance of power prevented the United States from being able to immediately respond to the Pueblo’s seizure, though whether the United States would have responded were it able is questionable. Also, the fact that North Korea maintained robust air defenses and that the

Wonsan harbor area was extremely well defended created an environment in which U.S. military operations against North Korea would incur costs, which was part of the U.S. calculation of possible military options. But the role of military capabilities in this case was part of identifying potential costs of U.S. military action, not credibility of North Korea’s threats. Could North

Korea implement its threat to fight a war? Yes, but not to win one. Moreover, its capability to implement its threat to fight a war did not noticeably vary throughout this period, yet we find that the United States did not deem North Korean threats prior to the Pueblo seizure credible, and reversed that credibility assessment during the Pueblo crisis. Material capabilities thus mattered, but are a poor basis for trying to understand U.S. assessments of North Korean threat credibility.

The other relevant variable in current calculus theory, the dyadic balance of interests, is challenged in this case as well. It is unclear on what basis one could argue that the stakes for

North Korea were so high either prior to or during the Pueblo crisis that the United States should find its threats credible on that basis alone. Though it stretches plausibility to the extreme, the argument could be made that from the North Korea perspective, the U.S. presence on the Korean

Peninsula or U.S. reconnaissance missions in international waters against North Korea constituted sufficient violations of North Korean sovereignty that they posed an existential security threat. This was, in essence, what North Korean Deputy Foreign Minister Kim Jaebong

142 told communist diplomats assigned to Pyongyang after North Korea captured the Pueblo: the

“American imperialists” had committed a “new and serious provocative action…the most outrageous one since signing the Armistice [in 1953].”197 Even if one chooses to interpret the

Pueblo’s seizure and North Korea’s rationale for its threats subsequently as meeting the nebulous threshold of North Korean “vital” national interests, this case overall would still pose a puzzle for current calculus theory because both the U.S. presence on the Korean Peninsula and the

Pueblo’s collection mission were nothing new. The U.S. military presence in the period prior to the Pueblo’s seizure remained steady dating back to 1954, and the Pueblo’s mission was hardly the first of its kind against North Korea; the USS Banner, an intelligence collection vessel similar to the Pueblo, conducted three missions in the same general operating area against North

Korea as the Pueblo in 1966 and 1967.198

If North Korea’s vital interests were engaged by the U.S. presence and current calculus theory thus did manage to explain U.S. assessments of North Korean credibility, current calculus theory would still fall short in explaining both the repeated clashes prior to the Pueblo and U.S. assessments of North Korea’s threats presaging its attack on the Pueblo. Further, the insufficiency of current calculus theory to explain this case is exacerbated by the fact that the

Pueblo attack, like the other three cases chosen for this dissertation, represents a dispute of exceptionally high interest to the United States, as I explained in chapter 2; even if North Korean interests were highly engaged, the Pueblo crisis represents the time span in this case within

197 “Telegram from Pyongyang to Bucharest,” January 24, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 11-12. 198 Mobley, Flashpoint North Korea, 24.

143 which the dyadic balance of interests were least in favor of North Korea relative to the United

States. Current calculus theory highlights important, but far from sufficient, explanatory variables in explaining this case.

Constant calculus theory, which makes the social psychology-based assertion that an enemy’s threats will almost always be credible, also fares poorly in explaining the facts of this case. It could be argued that U.S. perceptions of North Korea as an adversary made the latter’s threats during the Pueblo crisis credible, but there is scant process tracing evidence from within the case to support that claim. North Korea was obviously an out-group relative to the United

States, and U.S. decision-makers did express a certain amount of disdain for its adversary, but there is no clear, positive connection between the malign disposition that the United States attributed to North Korea and the U.S. assessment of North Korea’s intra-crisis threats as being credible. More significantly for the constant calculus interpretation of events is the problem that

U.S. in-group/out-group biases toward North Korea did not co-vary with U.S. assessments of

North Korean threat credibility. Given the chronicling of events above, it is safe to assess that the United States was hostile toward North Korea (and vice versa) throughout the 1964-68 period. Yet, the United States dismissed North Korea’s threats presaging the Pueblo attack.

That the United States would deem a rival’s threats incredible is the opposite of what constant calculus theory predicts.

Assessing Hypothesis Performance

We can now revisit the hypotheses I originally laid out in the previous chapter to see how well evidence in this case supports my theoretically based expectations. My first hypothesis—

144 that backing down from a rival’s challenge de-escalates the current crisis, strengthens one’s own future threat credibility, and makes it more likely that a rival will initiate future challenges against it—is mostly supported by this case. The effect of backing down on future threat credibility was not observable in this case because the United States did not issue threats in the period 1964-68, but the other two aspects of H1 fit the circumstances of this case. My second hypothesis—that initiating coercive challenges against rivals does not determinately affect reputation for resolve—finds nuanced support in this case because North Korean decisions to repeatedly initiate challenges did not accrue into a reputation for resolve.

The large gap between North Korea’s incessant, fiery threats and its hostile but incommensurate actions prior to the Pueblo crisis left U.S. decision-makers incredulous when

North Korea issued warnings relating to its subsequent attack on the Pueblo; North Korea’s word-deed mismatch led to the United States dismissing North Korea’s threats, consistent with both the reputations for honesty and resolve logics. When North Korea captured the Pueblo, fulfilling the Pueblo-related threats it made in the months prior, suddenly the United States found

North Korea’s threats credible; once again the reputations for honesty and resolve logics explain the co-variance in word-deed alignment and threat credibility. We also see a recurring phenomenon in this case for which only the reputation for resolve logic offers an answer: North

Korea’s decisions to repeatedly challenge the United States with provocations. Consistent with the reputations for resolve logic and contrary to the reputations for honesty logic, the U.S. decisions not to stand firm when challenged by North Korea in multiple clashes from 1964 through 1967 led to the United States cultivating a reputation for irresoluteness, leaving the

145 opening for North Korea to take action against the Pueblo, thinking the United States would not retaliate, just as it had not done in previous clashes. The Pueblo crisis, as with previous North

Korea-initiated provocations, ended because the United States backed down, either by not responding to an attack or by entering into diplomatic negotiations in which the United States ultimately conceded to North Korean demands. During the Pueblo crisis, North Korean negotiators even directly referenced the 1964 helicopter incident—in which the United States conceded to North Korean demands after protracted diplomatic talks—as the model for ending the Pueblo crisis.199 Drawing on the logic of reputations, this case thus not only illustrates that failing to implement threats hurts credibility while implementing threats strengthens credibility, but also that backing down has a temporally differentiated cost-benefit effect: it de-escalates the current crisis but encourages future challenges as long as rivalry continues to characterize the relationship.

My second hypothesis has to do with discerning whether initiating challenges through bullying or conducting provocations has an effect on accruing a reputation for resolve. I hypothesized that the relationship between the two variables would not be wholly determinative, which appears to bear out in this case. Despite North Korea’s recurring role as challenger in self-initiated conflicts, prior to the Pueblo’s seizure such actions did not have any net effect on the credibility of North Korea’s threats. General Bonesteel, then the USFK commander, seemed to be the only meaningful U.S. actor in this saga who interpreted North Korea’s threats in an actionably serious way, yet even he did not see the threats presaging the Pueblo attack amidst

199 U.S. Department of State, March 4, 1968. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1964-1968, Vol. XXIX, Part 1, Korea, Document 288: 647-9.

146 North Korea’s barrage of general threat making. The reasoning for a reputation for resolve not accruing to North Korea’s as a function of its repeated challenges, in this case at least, seems to be because it had accrued a reputation for bluffing; North Korea’s threats were incredible because so many had gone unfulfilled. Any reputational benefit that North Korea may have gained from its aggressive actions was undermined by the fact that its actions were discordant with its commitments. And yet, after North Korea seized the Pueblo, it did temporarily accrue a reputation for resolve; its threats were credible, but only because its word and deed matched.

There is no way to discern in this case whether it is the resolve or honesty logic that explains whether or when aggressive actions—challenge initiation—leads to threat credibility because both point to the same explanation. But this case clearly shows that threat credibility in a rivalry context is a phenomenon that is best explained in terms of reputation. We will be able to observe this dynamic again in subsequent cases.

Chapter IV: The EC-121 Shoot Down (1969)

This chapter follows the same form as the previous in constructing a historically contextual narrative of events followed by an analysis of how hypotheses performed in this case.

The crisis on which this case focuses is North Korea’s April 15, 1969 shoot down of a U.S. EC-

121 reconnaissance aircraft, an attack that took place less than four months after the United

States secured the return of the USS Pueblo crew in exchange for conceding to North Korean demands for an apology. The EC-121 crisis was the first of the newly inaugurated Nixon administration; Nixon’s National Security Adviser, Henry Kissinger, would look back with regret on the administration’s response as a failure. Although the crisis was short lived and

Vietnam continued as the primary U.S. foreign policy priority, the EC-121 crisis demanded a large amount of attention from Nixon, his White House staff and the U.S. interagency, just as with the other cases in this dissertation. The narrative for this episode is built on the answers to several questions that help us identify the effects of backing down, the sources of threat credibility, and the reasons the incident did not escalate to larger conflict. What was the recent conflict history of the dyad prior to this incident? What was the geopolitical context of the crisis; that is, what role, if any, did China, the USSR, ROK, and Vietnam play in defining the circumstances surrounding the EC-121 crisis? Who initiated coercion and why? How did each view the credibility of the other’s threats during the crisis and why? What actions did each side actually take during the crisis and what can be inferred about the motivations of each side from these actions, and to what extent does process tracing evidence support these inferences?

Finally, how did the crisis end?

147 148 Historical Context

Less than a month after North Korea returned the Pueblo crew in exchange for the United

States conceding the “three As”— admit that the Pueblo violated North Korean territorial space, apologize for the transgression, and assure North Korea that this would not happen again—

Lyndon Johnson left office and Richard Nixon was sworn in as president. But neither the change of presidential administrations nor the long-awaited U.S. apology ameliorated the rivalry dynamics between the United States and North Korea. The active hostility that culminated in the

Pueblo’s seizure continued unabated during Nixon’s first year in office. The EC-121 shoot down was the first foreign policy crisis of the Nixon administration. Though the incident was casually described by at least one scholar as “the first diplomatic event he did not inherit from previous administrations,”1 I contend that historical evidence and insights derived from reputational theories of rivalry behavior reveal that North Korea’s attack on the EC-121 had little to do with the Nixon administration per se; on the contrary, the EC-121 shoot down had much to do with the foreign policy choices of Nixon’s predecessors, which perpetuated active hostility toward North Korea while also making decisions that led North Korea to attribute a lack of resolve to the United States on matters involving the Korean Peninsula.

North Korea v. United States

The U.S.-North Korea rivalry in early 1969 had the recently resolved Pueblo crisis to add to the historical reference points that had accumulated throughout the Johnson administration.

From the Pueblo crisis the leadership in North Korea got what it wanted while U.S. leadership got what it needed. North Korea tussled with the American superpower, attacked its military,

1 Joan Hoff, Nixon Reconsidered (New York: BasicBooks, 1994), 173.

149 received an apology from the United States, and never released the Pueblo ship itself; the United

States secured the safe return of the surviving members of its crew, and prevented a crisis from turning into another war. Within weeks of resolving the Pueblo crisis, however, North Korea began once again initiating raids and ambushes of U.S. and ROK soldiers in the area of the

DMZ. From January 23, 1969 through the April 15 shoot down of the EC-121, North Korea initiated seven such attacks across the DMZ against U.S.-only and U.S.-ROK combined units.

During an ambush on March 15, 1969 of a U.S. 2nd infantry division patrol, North Korea killed one U.S. soldier and wounded three more, and when the medical evacuation helicopter tried to transport the wounded, North Korea shot it down, killing the five crew members and the three wounded.2 North Korea claimed that it shot down the helicopter over North Korean territory, but the United Nations Command (UNC) disputed the claim.3 The ROK attributed this attack, and the other attacks against the United States in the preceding days that did not result in casualties, as North Korea’s reaction to the U.S. announcement to conduct a military exercise called

“Operation Focus Retina” involving several thousand troops from the continental United States, which tested the U.S. ability to quickly mobilize troops for Korea in a contingency.4 The ROK

Ministry of National Defense and subordinate U.S. units requested approval to conduct reprisal raids, which General Bonesteel denied based on the belief that those North Korean provocations were not motivated by a desire to stir up revolution in South Korea but rather as retaliation for

2 See “Appendix 4: Significant U.S.-KPA Firefights, 1966-1969,” in Bolger, An Unfinished War, 137-9. 3 Gabriel Jonsson, Peace-Keeping in the Korean Peninsula: The Role of Commissions (Seoul: Korea Institute for National Unification, 2009), 152. 4 This exercise lasted from March 17-20, but was originally scheduled to begin the day of the North Korean attack, March 15. For a discussion of some of the operational details of Operation Focus Retina, see Bolger, An Unfinished War; Jonsson, Peace-Keeping in the Korean Peninsula.

150 the U.S. decision to conduct Operation Focus Retina.5 In response to this incident, the UNC suspended border demarcation operations, which had previously been routine.

Domestically, the United States and North Korea both experienced political changes in the period from the Pueblo crisis to the shooting down of the EC-121, though neither altered the pattern of interaction described in the preceding chapter—North Korea probes its rivals through repeated challenges, its rival defender chooses not to retaliate or capitulates to North Korean demands, and a crisis ends only to begin anew with a future challenge. In the United States,

Nixon’s ascendance to the presidency brought with him fierce criticisms of Johnson’s handling of the Pueblo crisis. On the campaign trail, Nixon referred to the Pueblo’s seizure and declared that “force must be met by force.”6 Calling President Johnson’s response to the Pueblo seizure

“weak,” Nixon suggested he would have taken a much tougher line if a “fourth-rate military power like North Korea” challenged the United States with a military provocation.7 Though doubtless politicizing the Pueblo crisis for political gain during a bid for the presidency, Nixon came into office with a reputation as a hawk, and frequently used the line, “Remember the

Pueblo” to rile up crowds of supporters while campaigning, implying he would not be as “weak” as he argued the Johnson administration had been.8

Contrary to the prevailing image of North Korea as a one-man dictatorship supported by a nation of loyal supporters, dissenters and oppositionists of varying strengths have existed

5 Bolger, An Unfinished War, 100-1. 6 Len Colodny and Tom Shachtman, The Forty Years War: The Rise and Fall of the Neocons, from Nixon to Obama (New York: Harper, 2009), 43. 7 Carl T. Rowan, “Spy Flights Worth Risks to U.S.,” The Star-Phoenix (May 7, 1969): 3; B. Drummond Ayres, Jr., “Humphrey Asserts Some Foes of War Practice Escapism; Humphrey Calls War Foes Escapists,” New York Times (August 14, 1968). 8 Hoff, Nixon Reconsidered, 174.

151 within North Korea from the time of its founding.9 By 1969, Kim Il Sung managed to silence the most belligerent challengers to his rule, but purges of senior officials who challenged Kim’s policy preferences, or failed to “properly” implement them, continued. In the months between the release of the Pueblo and the shoot down of the EC-121, Kim killed or imprisoned a group of officials whom he accused of disloyalty and failure to implement his policies in both the agricultural and military sectors, including: the defense minister, General Kim Chong-bong; the

KPA’s chief of the general staff, General Choe Gwang; chairman of the KPA political bureau,

Ho Bong-haek; the head of the Reconnaissance General Bureau (long responsible for paramilitary and asymmetric operational planning), General Kim Chong-dae; and several senior members of the Korean Workers’ Party, as well as a handful of other generals, each charged with some aspect of asymmetric operations or forward stationed units along the DMZ.10 Though a continued dearth of historical evidence shrouds North Korean decision-making in this specific event, British diplomats at the time posited that the proximate cause of the decision to shoot down the EC-121 was this internal power struggle, driven by KPA advocates of advanced military technology who were looking for an opportunity to strike a calibrated blow against the

United States.11 Whether the British assessment, Operation Focus Retina, or something else

9 From 1956 through at least 1961, Kim Il Sung used violence, intimidation, and political cooptation to suppress factions within North Korea that opposed his rule, or at least their lack of influence under his rule. It is thought that Kim Il Sung had not fully consolidated his power until the mid-1960s. For a history, see Andrei Lankov, Andrei, From Stalin to Kim Il Sung: The Formation of North Korea 1945 to 1960 (New Brunswick, NJ: Rutgers University Press, 2002); James F. Person, “’We Need Help from Outside’: The North Korean Opposition Movement of 1956,” CWIHP Working Paper No. 52 (Washington, DC: Woodrow Wilson International Center for Scholars, 2006). 10 Dae-Sook Suh, Kim Il Sung: The North Korean Leader (New York: Columbia University Press, 1988), 231; Tai Sung An, North Korea in Transition: From Dictatorship to Dynasty (Westport, CT: Greenwood Press, 1983), 16-18; Suck-ho Lee, “Party-Military Relations in North Korea: A Comparative Analysis,” unpublished doctoral dissertation (Washington, DC: George Washington University, 1983). 11 Mobley, Flashpoint North Korea, 101-3.

152 entirely provided the immediate motivation to shoot down a U.S. aircraft, the proximate cause should not be mistaken for the dyadic structural and historical conditions that made such a proximate cause possible or likely.

Cheap Talk and Credibility before the Shoot Down

Despite the several years of prior low-intensity conflict with North Korea, including the

Blue House raid, Pueblo crisis, and Uljin-Samcheok infiltration all during 1968, the United

States saw no proximate threat of attack from North Korea in the months prior to the EC-121 shoot down. Although the United States believed that North Korea’s policies and general orientation toward the United States and South Korea were unchanged from 1968, it was nevertheless surprised and unprepared for North Korea’s attack, again. USFK’s commander,

General Bonesteel, was aware of several verbal threats and warnings coming from North Korea in the months preceding the EC-121 shoot down. In response to the U.S.-ROK Operation Focus

Retina exercise, Kim Il Sung and North Korea leveled threats against the United States through the limited channels it had. In addition to blaring threats of retaliation against the United States and ROK over DMZ loudspeakers pointed at South Korea—an occasional occurrence even today—for its decision to conduct Operation Focus Retina,12 North Korea communicated threatening criticisms, through direct and indirect channels, that the United States largely ignored. On March 11, the KPA denounced the decision to conduct the military exercise during a meeting of the Military Armistice Commission (MAC) it requested, labeling it “a scheme to

12 Bolger, An Unfinished War, 98-100.

153 provoke a new war in Korea.”13 At another MAC meeting held on March 17, where the UNC condemned the North Korean attack on the U.S. patrol, North Korea took occasion to criticize

Operation Focus Retina as a “…flagrant violation of the armistice” and a “war rehearsal” that would not be tolerated.14 That same month, the KPA submitted a letter to the Neutral Nations

Supervisory Commission (NNSC),15 using the uniquely inflammatory language for which North

Korea has become famous to describe “U.S. imperialist aggressors” and “South Korean puppet army” military exercises as “illegal” and “the most outrageous violation” of the armistice.16 And on April 11, USFK conveyed to CINCPAC that during armistice meetings held in days prior,

“…the North Koreans have been particularly vehement and vicious in warning UN forces about provocative actions.” 17 In an oral history involving General Bonesteel, the USFK commander claimed that he warned CINCPAC that the Air Force in particular should take unspecified precautions when conducting reconnaissance missions against North Korea because they

“…were doing something that was damned suspicious.”18

U.S. unpreparedness for the EC-121 attack, operationally and cognitively, is all the more noteworthy because the repetition of provocations in previous years continued in early 1969—

North Korea initiated seven provocations between January and March 1969 in which U.S. and

13 Jonsson, Peace-Keeping in the Korean Peninsula, 151-2; “7Americans Die As Result of Korean Assault,” Daytona Beach Morning Journal, (March 17, 1969): 6. 14 Jonsson, Peace-Keeping in the Korean Peninsula, 155. 15 The NNSC is a small group of senior officers from impartial nations, two of whom were appointed by the UNC and two by North Korea and the Chinese. The NNSC, along with the UNCMAC, monitor the implementation of the armistice and investigate suspected violations. 16 Jonsson, Peace-Keeping in the Korean Peninsula, 196. 17 Mobley, Flashpoint North Korea, 106. 18 Ibid.

154 ROK soldiers were killed.19 Yet the EC-121 shoot down caught the new Nixon administration completely off guard; it was “…a surprise in every sense of the word,” according to Nixon.20

North Korea’s threats and provocations in the months and years prior had little effect on U.S. actions leading up to the EC-121 shoot down. The United States flew 190 reconnaissance missions in the Sea of Japan off of North Korea’s eastern coast during the first four months of

1969.21 North Korea’s history of initiating provocations, including the Pueblo crisis, did not hinder approval of these missions. According to Kissinger, the ironic purpose of the EC-121 mission was to warn the United States of any surprise North Korean attacks.22 Though lacking direct evidence, Richard Mobley makes the case that DoD and the DIA assigned the EC-121 mission the same category of risk as the Pueblo the year prior: “hostile intent remote, intercept actions unlikely, defensive patrols possible.”23

What we do know from direct evidence is that the EC-121’s reconnaissance mission was approved to operate without a protective escort, and that on July 2, 1968 (before the Pueblo crisis had ended) the State Department approved a JCS request to slacken reconnaissance restrictions against North Korea to be permitted to approach within 40 miles of the North Korean coast.24

General Wheeler, then Chairman of the JCS, assessed the risk associated with reconnaissance

19 See “Appendix 4: Significant U.S.-KPA Firefights, 1966-1969,” in Bolger, An Unfinished War, 137-9. 20 Richard M. Nixon, “The President’s News Conference of April 18, 1969,” in Public Papers of the Presidents, Richard Nixon, 1969 (Washington, DC: U.S. Government Printing Office, 1971), 155-6; Henry Kissinger, The White House Years (Boston, MA: Little, Brown and Co., 1979), 312-14. 21 Department of Defense, “EC-121 Fact Sheet,” unclassified report (May 1969). Available online: http://www.dod.mil/pubs/foi/International_security_affairs/other/500.pdf. 22 Kissinger, White House Years, 313. 23 Mobley, Flashpoint North Korea, 102-3. 24 As soon as the Pueblo was seized, JCS directed that reconnaissance assets give North Korean territory a buffer of at least 80 miles. The July 2 approval to shorten the buffer to 40 miles enabled better intelligence collection. U.S. House Armed Services Committee, Inquiry into USS Pueblo and EC-121, 1677-8.

155 missions in batches, focusing only on anomalies based on whether “… last month this particular program went off without incident.” JCS approval was thus based less on North Korean words

(threats and promises) than on a one-month history of North Korean deeds.25

Additionally, the intelligence community issued a new Special National Intelligence

Estimate (SNIE) on North Korean capabilities and intentions in January 1969, which reflected little learning from the events of the Pueblo crisis (precisely the opposite of what Jonathan

Mercer’s constant calculus theory would predict). Despite a record of numerous challenges to

U.S. aircraft, including at least one attempted shoot down, the shoot down of a U.S. helicopter, the forced landing of another U.S. helicopter, and the seizure of the Pueblo, the SNIE gave no indication that North Korea might seek to attack U.S. reconnaissance aircraft.26 Despite this glaring oversight, the SNIE articulated U.S. assessments of North Korean threat credibility and intentions, making the following judgments: North Korea was engaged in a “process of probing for weaknesses and testing ROK and US resolve”; Kim Il-sung was carefully watching U.S. behavior to “…interpret…US willingness to take a firm stand against further North Korean provocations”; and whether North Korea goes beyond “DMZ and occasional rear-area operations” would depend on how North Korea “judged the probable reactions of the ROK and the US.”27 Still, the SNIE continued, North Korea’s provocations were broadly attributed to a

North Korean strategy to stir up a “revolutionary struggle” in South Korea; Pyongyang did not

25 Armbrister, A Matter of Accountability, 193. 26 U.S. Department of State, January 30, 1969. “Special National Intelligence Estimate,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 1: 1-2. 27 Ibid.

156 intend to provoke a “resumption of major hostilities,”28 suggesting that the totality of North

Korean violent challenges did not alter the general U.S. assessment that North Korea lacked the resolve to provoke a war.29 In aggregate, and despite the cautions urged by General Bonesteel, the above discussion suggests that U.S. policymakers—and the machinery of government with which they were charged—were not acting as one might expect had they taken the threat of

North Korean violence more seriously.

Geopolitical Context

Despite occurring only months after the Pueblo crisis was resolved, the geopolitical context of the U.S.-North Korea rivalry dyad at the time of the EC-121 shoot down had evolved in some respects. North Korea continued to navigate the Sino-Soviet split by charting an independent foreign policy that avoided total alignment with either side. But the cool, critical

Soviet response to North Korea’s seizure of the Pueblo created more distance from the Soviet

Union than the year prior. Relations between North and South Korea were as hostile as ever, and the latter was growing weary of restraint in the face of North Korean violence. In Vietnam, the

Nixon administration began the early phases of a plan for withdrawal, but not before initiating a secret bombing campaign of Viet Cong logistics routes in Cambodia.

Entering 1969, strained relations between China and North Korea were largely a function of the Sino-Soviet split and China’s Cultural Revolution. The Sino-Soviet split, with roots in political and ideological differences that started as early as 1956 with Khrushchev’s de-

28 Ibid. 29 See chapter 3. This was largely a reversal of U.S. thinking in the midst of the Pueblo crisis during which U.S. decision-makers throughout the chain of command believed North Korea might initiate war if the United States used military force against it.

157 Stalinization campaign, had by 1969 ruptured into a string of border conflicts between China and the Soviet Union.30 Other Soviet Bloc countries, dismayed by what they considered petty fratricide by the two giants within the international communist movement, feared that the Soviet

Union would launch a nuclear attack against Chinese facilities.31 As culturally Confucian communist allies who shared a writing system and fought together during the Korean War, China and North Korea shared many common interests. Yet there were real conflicts of interest between the two nations; in China, the Cultural Revolution’s anti-bourgeois, anti-elite character rejected leaders like North Korea’s Kim Il Sung. At the relationship’s worst, in March 1969, only a month before the EC-121 shoot down, Chinese and North Korean ground forces engaged in a firefight with one another across the Sino-North Korean border, resulting in an unknown number of casualties.32 But even without the conflicting interests between China and North

Korea, the latter was strategically motivated to assiduously avoid any appearance of intimacy or solidarity with China while it was still receiving most of its military aid from the Soviet Union; the Soviet Union was North Korea’s primary material benefactor.33 Both before and after the

Sino-North Korean border conflict, however, there were some signs of a Sino-North Korean rapprochement that would become more visible by the end of 1969. In October 1968, the official newspaper of the Korean Workers’ Party celebrated the 18th anniversary of China’s intervention

30 U.S. Department of State, Bureau of Intelligence and Research, March 4, 1969, Intelligence Note, “USSR/China: Soviet and Chinese Forces Clash on the Ussuri River,” Folder POL 32-1 Chicom-USSR, Central Files 1967-69, RG 59, NARA. 31 U.S. Department of State, Bureau of Intelligence and Research, September 18, 1969, Intelligence Note, “Communist China: War Fears and Domestic Politics,” Folder POL Chicom, Files SN 67-69, NARA. 32 Bernd Schaefer, North Korean “Adventurism” and China’s Long Shadow, 1966-72: Working Paper #44 (Washington, DC: Cold War International History Project (CWIHP), 2004), 27. 33 B.C. Koh, The Foreign Policy of North Korea (New York: Praeger, 1969).

158 in the Korean War.34 In February 1969, the diplomatic community in Pyongyang rumored that

North Korea’s foreign minister was planning a trip to Beijing, and by Summer 1969 the two nations did meet in the Sino-North Korean border region.35 By September 1969, five months after the EC-121 shoot down, China and North Korea had rekindled their friendship, and China had, according to some communist bloc diplomatic reporting, recommitted itself to its military alliance with North Korea, culminating in Zhou Enlai leading a diplomatic delegation to

Pyongyang in February 1970 (the first senior Chinese official to travel overseas since the

Cultural Revolution began).36

North Korea’s relationship with the Soviet Union, meanwhile, had been altered by the

Pueblo experience. North Korea still strived to chart an independent course between its giant neighbors, which meant neither total alienation from nor alignment with either the PRC or

USSR, but it relied disproportionately on economic and military aid from the Soviet Union. Half of North Korea’s air force consisted of Soviet fighters and bombers, North Korea’s navy included Soviet submarines and torpedo boats, and its air defense arsenal was continually improved by the Soviet Union before and after the Pueblo crisis.37 And although North Korea rejected the legitimacy of the United Nations, it nevertheless cared a great deal when the UN

Security Council weighed in on Korean Peninsula affairs; the Soviet Union was the only potential source of advocacy for North Korean positions on the Security Council.38 During the

34 Nodong Sinmun editorial, October 25, 1968. 35 Schaefer, North Korean “Adventurism,” 28-9. 36 Schaefer, North Korean “Adventurism,” 29-33. For broad discussions about China’s warming to North Korea see Suh, Kim Il Sung; Shen and Li, After Leaning to One Side. 37 CIA Directorate of Intelligence, Kim Il-Sung’s New Military Adventurism; New York Times, February 1, 1968. 38 B.C. Koh, “North Korea and the Sino-Soviet Schism,” The Western Political Quarterly 22, no. 4 (1969): 940-962.

159 Pueblo crisis, the Soviet Union stood aloof when North Korea requested military backing in case war broke out, and Soviet diplomats chided North Korea for its reckless behavior. If there was any doubt about Moscow’s preference for stability—not military conflict—on the Korean

Peninsula, the Pueblo crisis removed it. Soviet diplomats counseled their North Korean counterparts to take a prudent, restrained course of action against the United States, but grew exasperated by North Korea’s deafness to their advice.39

By 1969, South Korea had grown weary of enduring North Korea’s provocations, and increasingly obsessed with the overall credibility of the U.S. commitment toward the U.S.-ROK alliance. Immediately after the Blue House raid, South Korea’s President Park secretly created his own31-man commando unit charged with assassinating Kim Il Sung as retaliation, which was disbanded after North and South Korea inched toward a short-lived rapprochement in late 1969 and 1970.40 South Korea urged a firm policy against North Korea, explicitly advocating a policy based on the logic of resolve: President Park believed North Korea’s provocations would continue until South Korea demonstrated it was willing to impose significant costs on its enemy.41 But the United States restrained South Korean policy toward North Korea following the Blue House raid, as it had tried to do (mostly successfully) countless times following North

39 “Telegram from Pyongyang to Bucharest,” January 25, 1968, Archive of the Romanian Ministry of Foreign Affairs, Obtained and translated for the NKIDP by Eliza Gheorghe, in The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5: 14-15; Sergey S. Radchenko, “The Soviet Union and the North Korean Seizure of the USS Pueblo: Evidence from Russian Archives,” CWIHP Working Paper #47 (Washington, DC: Woodrow Wilson International Center for Scholars). 40 The story of President Park’s secret assassination squad, its disbandment, and tragic end became an award- winning movie in 2003, at the time the highest grossing Korean movie of all time. The most comprehensive text on the incident to date is Dong-ho Baek, Dong-ho, Silmido (in Korean) (Seoul: Balgeun Saesang, 2004). 41 Choi Kang and Joon-Sung Park, “South Korea: Fears of Abandonment and Entrapment,” in The Long Shadow: Nuclear Weapons and Security in 21st Century Asia, ed. Muthia Alagappa (Palo Alto: Stanford University Press, 2008).

160 Korean provocations from 1965-67.42 For South Korea, the EC-121 crisis was not really a crisis at all; it was the expected consequence of failing to show North Korea that its provocations would not be tolerated, and at any rate qualitatively similar incidents had occurred many times over the prior three years.43 As one official from the ROK government at the time would later recall, “We all initially greeted this as just another big-game provocation by North Korea. There were many other similar instances...We had our own position of course, that is: be firm against these incidents. That was our usual position against any North Korean provocation against the

United States.”44 South Korea had few strategic alternatives to its alliance with the United States at the time, but the combination of U.S. conciliation in response to the Pueblo and Blue House raid, witnessing America’s creeping abandonment of Taiwan, and anticipation of U.S. conciliation toward the North Vietnamese in upcoming Paris peace talks all contributed to South

Korean mistrust of U.S. commitments, giving rise in turn to South Korea’s clandestine nuclear program and secret Kim Il Sung assassination squad.45

In addition to North Korea, South Korea was preoccupied with the future of the U.S. presence in Vietnam, as was the United States. South Korea’s apprehension was on display during meetings with Chung Il-Kwon, South Korea’s Prime Minister, and Nixon during a visit to

Washington on March 30, 1969. Chung recommended that Nixon “be patient and firm with the

Communists” in the Paris negotiations with North Vietnam, and by all means not conciliate or

42 Ibid. 43 Ostermann and Person, Crisis and Confrontation on the Korean Peninsula. 44 Ostermann and Person, Crisis and Confrontation on the Korean Peninsula, 115. 45 Although its analysis exceeds the scope of this dissertation, some evidence suggests that the logic of reputation in an alliance context, for unreliability in this case, may explain South Korean decision-making with regard to its nuclear program. See Kang and Park, “South Korea”; Rebecca C.K. Hersman and Robert Peters, “Nuclear U-Turns: Learning from South Korean and Taiwanese Rollback,” Nonproliferation Review 13, no. 3 (2006): 539-53.

161 show “weakness.”46 Chung also underscored the importance of consulting early and often with

South Korea on the matter of the U.S. presence in Vietnam; South Korea’s large military contribution to South Vietnam would put South Korea in a politically impossible position if its troops were committed to Vietnam while the United States withdrew.47 Given the July 25, 1969 announcement of what came to be called the “Nixon doctrine,” South Korea’s concerns proved well founded.48

In the months prior to the downing of the EC-121, the Nixon administration was already plotting how the United States could honorably extricate itself from the quagmire that Vietnam had become. The key condition, in Nixon’s and Kissinger’s view, was to frame international perceptions of withdrawal in such a way that the United States was not perceived as lacking resolve.49 As early as January 8, 1969, before Nixon was even sworn in, Kissinger was actively discussing the possibility of bombing Viet Cong sanctuaries in Cambodia, thinking it would enable the United States to negotiate with Pham Van Dong from a position of strength.50 On

March 18, 1969 the first secret bombing runs of Cambodian territory took place as the opening salvo of what would become a misplaced campaign lasting throughout Nixon’s first year in office.51

46 U.S. Department of State, March 30, 1969. “Memorandum of Conversation,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 5: 8-10. 47 Ibid. 48 This will be taken up in greater detail as part of the geopolitical context for the next chapter. 49 Hoff, Nixon Reconsidered, 163; Walter Isaacson, Kissinger: A Biography (New York: Touchstone, 1992), 159-62. 50 Isaacson, Kissinger, 171-8; U.S. Department of State, “Memorandum from the President’s Assistant for National Security Affairs (Kissinger) to President Nixon,” January 24, 1969. FRUS, 1969-1976, Vol. VI, Vietnam, January 1969-1970, document 8: 20-2. 51 Isaacson, Kissinger, 171. For a critical overview of the bombing campaign, see William Shawcross, Sideshow: Nixon, Kissinger, and the Destruction of Cambodia (New York: Washington Square Press, 1979).

162 Overview of the Incident

At mid-day on April 15, 1969, a U.S. Navy EC-121 reconnaissance aircraft was operating in international air space over the Sea of Japan, off North Korea’s eastern coast, collecting signals intelligence from North Korean military facilities when a pair of MiG-21 fighter aircraft belonging to the North Korean Air Force pursued and shot down the unarmed EC-121, killing all

31 Americans on board.52 The EC-121 took off from Atsugi Air Base in Okinawa, Japan and was scheduled to fly multiple north-to-south orbits, landing at Osan Air Base in South Korea.

Approximately an hour and a half before the shoot down, U.S. reconnaissance identified two

North Korean MiGs as taking off from the east coast. Shortly thereafter the EC-121 transmitted a routine radio communication; its last. Between nine and eleven minutes before the shoot down, the EC-121 was notified that MiGs have approached to somewhere within 51 and 55 miles of the aircraft; a few minutes later an unidentified U.S. monitoring site begins notifications that

MiGs are responding to a U.S. reconnaissance aircraft. Despite warnings to the EC-121, it drops off the radar screen at 1:50pm on April 15 (local Korea time). The White House was notified less than an hour later (approximately 12:50am U.S. Eastern Standard Time). Kissinger would notify Nixon for the first time around 7:20am on April 15, six and a half hours after the shoot down (U.S. Eastern Standard Time). Immediately following the shoot down, there were some

U.S. military movements as part of standard procedure for crisis reaction and search and rescue.

In the following weeks, the White House would direct some additional military actions, though far too late relative to the actual shoot down, and far short of lethal or demonstrative use of force.

52 The minute by minute account of the EC-121 shoot down can be found in U.S. House Armed Services Committee, Inquiry into USS Pueblo and EC-121; Department of Defense, “EC-121 Fact Sheet,” unclassified report (May 1969). Available online: http://www.dod.mil/pubs/foi/International_security_affairs/other/500.pdf.

163 Some U.S. actions also occurred without being centrally orchestrated, as one might expect of a large and complex foreign policy establishment in a crisis. But the crisis effectively ended with

President Nixon’s April 18 news conference announcing that future reconnaissance mission would be protected, but that the United States would not retaliate for the downing of the EC-

121.53 In a repeat of history, several months after the EC-121 shoot down on August 17, North

Korea would again hold hostage three injured U.S. servicemen after shooting down their helicopter, demanding a Pueblo-like apology formulation in exchange for their release.54

In what follows in the rest of this section, I will highlight the public word and deed of the

United States in reaction to this provocation, dwelling also on the complicated internal deliberations about the U.S. public reaction. I will also consider the limited information we know about this provocation from the North Korean perspective.

Debating Retaliation

The White House Situation Room was notified about an unconfirmed shoot down in the

Sea of Japan at 12:50am on April 15, about an hour after it had occurred. Kissinger’s military assistant, Alexander Haig, was notified 17 minutes after the Situation Room.55 Nixon would not be notified until 7:20am that day, in part because the action took place overnight, and because the shoot down remained unconfirmed, even at 7:20am when he finally was notified.56 Kissinger spoke to Nixon about the shoot down at least four times that day, three of which occurred by

53 The American Presidency Project, “The President’s News Conference: North Korea’s Attack on U.S. Reconnaissance Flights,” (Washington, DC: Public Papers of the Presidents, April 18, 1969). Available online: http://www.presidency.ucsb.edu/ws/index.php?pid=2004. 54 Memorandum, “Helicopter Crew Members Held by North Koreans,” October 4, 1969, NSC Files, Country Files Far East, Korea, Helicopter Incident, box 540, NARA. 55 Kissinger, White House Years, 316. 56 Ibid.

164 phone. Nixon’s initial reaction upon hearing about the incident was to demand retaliation of some kind, focusing specifically on finding out more about a North Korean commercial ship under Dutch flag registry that Nixon believed the United States should seize in international waters, releasing it only when North Korea returned the USS Pueblo.57 Nixon seemed seized with this particular Dutch-flagged ship option, having raised it in each of his discussions with

Kissinger on April 15, even while discussing other kinds of retaliation options. Nixon protested that “we won’t just sit here and do nothing,” and Kissinger agreed.58 He also stated that “the price is too high to pay not to do anything,” recalling an intelligence report claiming that

Egyptian President Nasser told the King of Jordan that, “…After all, it isn’t so risky to defy the

United States—look at North Korea and the Pueblo.”59 Over the course of the first 24 hours of the crisis, however, it became increasingly clear that seizing a North Korean ship under a Dutch flag would be infeasible because of opposition from the Dutch government and the legal requirement to compensate the Dutch for commandeering the ship under international law.60

A range of options was being considered beyond seizing the Dutch-flagged North Korean ship, however, most of which would be considered at an NSC meeting with Nixon the following

57 Memorandum, “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” April 15, 1969 (6:30pm), Nixon Presidential materials, NSC Files, Korea: EC-121 Shootdown, North Korean Reconnaissance Shootdown, 4/9/69-4/16/69, Vol. 1 Haig, box 434, NARA; U.S. Department of State, April 15, 1969. “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 8: 17-20. 58 Ibid. 59 Ibid. 60 Memorandum, “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” April 15, 1969 (10pm), Nixon Presidential materials, NSC Files, Korea: EC- 121 Shootdown, North Korean Reconnaissance Shootdown, 4/9/69-4/16/69, Vol. 1 Haig, box 434, NARA; U.S. Department of State, April 15, 1969. “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 9: 20-21.

165 day on April 16.61 The morning the crisis broke, Alexander Haig asked the Office of the

Secretary of Defense for options that could be presented to Nixon and his cabinet the next day.

Haig specifically posed the following questions for DoD to answer: “1) what options do we have for ‘retaliation’?; 2) what measures can and/or should be taken to prevent the North Koreans from picking up the survivors?; 3) how quickly could we execute an escorted reconnaissance flight like the one that is now missing?; 4) what North Korean assets are available worldwide against which we could take some retaliation?”62

The answers to Haig’s questions were provided in an options paper drafted by the Joint

Chiefs of Staff the same day and conveyed to Kissinger by Melvin Laird, the Secretary of

Defense. DoD divided options into those that were immediately executable, and those that would take a few days to implement, the latter of which potentially required movement of some naval forces committed to Vietnam. According to the Joint Chiefs of Staff paper, the options that could be executed immediately included: diplomatic demands for an apology and potential remuneration; high-speed, high-altitude reconnaissance flights over North Korea; escorted reconnaissance flights in the same area where the EC-121 was shot down; a request to the

Soviets to make demands of North Korea on behalf of the United States; and destruction of

61 The fact that Nixon did not meet with his cabinet until the day after the shoot down has been criticized by some as negligence or detachment, but this ignores the typically secretive decision-making style of the Nixon administration; foreign policy decisions were rarely made using the formal NSC system. For criticisms, see Kissinger, White House Years; Mobley, Flashpoint North Korea; Seymour M. Hersh, The Price of Power: Kissinger in the Nixon White House (New York: Simon and Schuster, 1983). On Nixon’s foreign policy decision-making style, see Hoff, Nixon Reconsidered; Richard M. Nixon, RN: The Memoirs of Richard Nixon (New York: Simon and Schuster, 1978); Alexander Haig, Inner Circles: How America Changed the World (New York: Warner Books, 1992). 62 U.S. Department of State, April 15, 1969. “Papers Prepared by the Joint Chiefs of Staff,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 7: 16-17.

166 North Korean aircraft located on its eastern coast.63 The options that could be executed on short notice but not immediately included: a military demonstration of force; selected air strikes; feints against North Korean air defenses; and a blockade of North Korean ports.64 On the morning of the April 16 NSC meeting, Haig passed a memo to Kissinger with a passionately argued recommendation for a “military retaliatory strike of some type.”65 Haig asserted that the difference between the EC-121 and Pueblo cases was that in the latter case, the hostage component allowed for a “do nothing” pretense, while no such veil existed in the EC-121 case.

He ended the memo by commenting that “If the President is unwilling to trigger all of the signaling devices at his command to convey his/U.S. intent to go to the limit if required, I recommend that we not undertake an action such as an overt attack…” recommending instead a submarine ambush, which would serve as a “punitive action.”66 That same morning, Admiral

John McCain, Commander of the Pacific Fleet, wrote a memo urging General Wheeler,

Chairman of the Joint Chiefs of Staff, to recommend retaliation at the NSC, claiming that North

Korea would perceive the United States as a “paper tiger” if it did not.67 He followed this with a personal note to General Wheeler on April 17, pressing the case for retaliation again, arguing

“…a quick and telling blow should be made against the North Koreans so as to make clear that their aggression will not go unscathed.”68 McCain’s recommendation would prove contrary to

63 Ibid. 64 Ibid. 65 Memorandum, “Shoot-Down Incident,” April 16, 1969, Nixon Presidential materials, NSC Files, Korea: EC-121 Shootdown, North Korean Reconnaissance Shootdown, 4/9/69-4/16/69, Vol. 1 Haig, box 434, NARA. 66Ibid. 67 Memorandum, “Movement of CVA Task Forces,” April 16, 1969, Wheeler Records, Folder 091 Korea (BP) EC- 121 Shootdown,” RG 218, box 31, NARA. 68 Untitled Memorandum from CINCPAC to CJCS, April 17, 1969, Wheeler Records, Folder 091 Korea (BP) EC- 121 Shootdown, RG 218, box 31, NARA.

167 Secretary Laird’s preference and contrary to how General Wheeler would present options during the NSC meeting later that day; it is doubtful that McCain’s advice made its way to Kissinger or

Nixon.

Haig’s minutes from the April 16 NSC meeting on the shoot down show that no decisions were made, and much of the meeting was dedicated to sorting out the facts about what happened, whether any crew might still be alive, and confirming that China and the Soviet Union were unaware of North Korea’s plan for the shoot down, as well as that the EC-121 was always operating at least 38 miles away from North Korean territory.69 In Kissinger’s reflections, this meeting was “unfocused and inconclusive…Military options were reviewed in a haphazard way…There was no discussion of the fundamental issue: whether our failure to respond to the shootdown of an unarmed reconnaissance plane over international waters might not create an impression of such irresolution that it would encourage our enemies in Hanoi and embolden opponents elsewhere.”70 CIA Director Richard Helms—who, along with Secretary of State

Rogers strongly opposed retaliation in this crisis—described the shoot down as falling on a trend line of increasingly provocative behavior. He asserted that provocations from North Korea were on the rise, and that except for the Pueblo, the EC-121 shoot down surpassed most other incidents in aggressiveness.71 Helms added that the United States conducted this same mission

69 U.S. Department of State, April 16, 1969. “Minutes of a National Security Council Meeting,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 13: 28-31; U.S. House Armed Services Committee, Inquiry into USS Pueblo and EC-121. In a conversation with Nixon, Kissinger noted that U.S. tracking of North Korean radar showed that North Korea never saw the EC-121 come closer than 40 miles. U.S. Department of State, April 18, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 17: 34-7. 70 Kissinger, White House Years, 318. 71 U.S. Department of State, April 16, 1969. “Minutes of a National Security Council Meeting,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 13: 28-31.

168 nine times in the prior six months, describing North Korean tracking of these missions as

“seldom done,” which implied that North Korea did not perceive these missions as operationally threatening.72

Nixon principally expressed interest in the history of U.S.-North Korea crisis interactions.

Nixon told the group he wanted information on the 1965 attempted shoot down of the RB-47, as well as on “the history of incidents in this area, especially since1953 through 1955 period.”73 He stated that the fact that this was a routine mission explains why there was no air cover, but he asked, “Why did they do this [provocation] after Pueblo?” Brigadier General Ralph Steakley,

Deputy Director for Reconnaissance on the Joint Staff, responded that “It was not unusual, in the sense of its past pattern—only in sense of its scheduled pattern”; Nixon expressed surprise that this kind of provocation was “normal.”74

After Nixon asked General Wheeler, Chairman of the Joint Chiefs of Staff, to brief the group on military options, it was clear that DoD viewed every potential military option as either unlikely to achieve anything, or too likely to lead to North Korean escalation of the crisis that pursuing them was not worth the risk. Wheeler acknowledged that a show of force could be mobilized within 24-48 hours, but added that the show of force conducted after the Pueblo achieved nothing. Wheeler thought that North Korean retaliation and U.S. casualties would result from striking North Korean air defenses, striking targets near the DMZ with nuclear- capable missiles or conventional artillery, or a ground raid across the DMZ.75 Although Nixon

72 Ibid. 73 Ibid. 74 Ibid. 75 Ibid.

169 initially ordered looking into laying sea-based mines around Wonsan harbor on North Korea’s eastern coast or blockading the port of Wonsan ,76 Wheeler said such actions would be considered an act of war, and would at any rate have little effect since North Korea had no sea- faring ships.77 Wheeler was also pessimistic about the ability to destroy North Korean aircraft off of its eastern coast. Putting the question of North Korean retaliation aside, Wheeler estimated the immediate casualties from any air strikes against North Korean bases or air defenses would range from 2-8%, and would require anywhere from 24 to 250 aircraft for operations.78 The talking points and background documents prepared for Wheeler, which

Secretary Laird passed to Kissinger prior to the NSC meeting, added that DoD opposed all operations against North Korea’s western coast because it contained the majority of North

Korean surface-to-air missile sites, west coast operations posed a high risk of confrontation with

China due to sensitivities in that area, and asset moves from the Sea of Japan to the Yellow Sea would require overflight of South Korea, which DoD deemed undesirable.79 A CIA background paper prepared for the NSC reflected DoD’s concerns while emphasizing potential gains from a retaliatory decision, assessing that the benefits of a “military combat course of action” were that such actions were most likely to achieve the “objectives of deterrence, redress, and maintenance

76 U.S. Department of State, April 15, 1969. “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 8: 17-20. 77 U.S. Department of State, April 16, 1969. “Minutes of a National Security Council Meeting,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 13: 28-31. 78 Ibid. Secretary Laird would tell Nixon two days later that he thought the JCS was underestimating the probability of losses. U.S. Department of State, April 18, 1969. “Memorandum from Secretary of Defense Laird to President Nixon,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 17: 38-42. 79 “Talking Paper for the Chairman, Joint Chiefs of Staff, on an Item to be Discussed at the National Security Council Meeting, 16 April 1969,” April 15, 1969, FOIA request, released on January 9, 2007 (Washington, DC: The National Security Archive).

170 of our rights to use international airspace,” but greatly risked North Korean retaliation.80 The

CIA paper also assessed that the Soviets and Chinese would both avoid any “countermeasures” that might lead to conflict with the United States.81 The CIA also produced a paper concluding that “North Korea’s assessment of future US initiatives, probably are strongly influenced by the

Korean’s interpretation of the US response to the seizure of the Pueblo.”82

Although the State Department’s opposition to any retaliation against North Korea in this crisis is well documented,83 Secretary of State Rogers did not directly weigh in on the decision to retaliate at this meeting, perhaps because the meeting did not yield any vocal advocates in support of retaliation and all the military options discussed came with caveats of loss and retaliation.84 Instead, Rogers focused narrowly on the diplomatic options that were the bureaucratic purview of the State Department, particularly weighing how to respond to North

Korea’s proposal for an April 18 meeting at Panmunjom, a request that the United States received only hours after the shoot down took place.85 Rogers also raised the possibility of communicating indirectly through friends and allies, though without reference to what message might be conveyed, as well as submitting a letter to the United Nations condemning North

Korea’s provocations. The one meaningful point that Rogers made on the matter of retaliation

80 “Alternative Courses of Action in Response to Korean Attack on U.S. Aircraft,” April 16, 1969, FOIA request, released on April 15, 2002 (Washington, DC: The National Security Archive). DoD’s military planners agreed with the CIA’s assessment that Soviet and Chinese intervention in a conflict would be unlikely. See “Talking Paper for the Chairman, Joint Chiefs of Staff, on an Item to be Discussed at the National Security Council Meeting, 16 April 1969,” April 15, 1969, FOIA request, released on January 9, 2007 (Washington, DC: The National Security Archive). 81 Ibid. 82 Memorandum, “Communist Reactions to Certain U.S. Actions,” April 17, 1969, Nixon Presidential Materials, NSC Files, Country Files Far East, Korea, Vol. 1 to 9-69, box 540, NARA. 83 See, for example, Kissinger, White House Years; Hoff, Nixon Reconsidered; Haig, Inner Circles. 84 U.S. Department of State, April 16, 1969. “Minutes of a National Security Council Meeting,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 13: 28-31. 85 Ibid.

171 was to suggest that the United States avoid making threats: “we shouldn’t threaten the other side.

It ties our hands.”86 Kissinger’s only contribution to this meeting was to suggest that whatever the United States conveyed through diplomatic moves, it needed to be consistent with and preferably complement any military actions the United States would take or avoid taking.

Kissinger’s engagement with Nixon on the EC-121 four times the previous day and in the days after the meeting was enough to ensure Nixon understood his views.

But what were U.S. objectives? At the April 16 NSC meeting, there was no discussion about what the United States wanted to achieve; only menus of options responding to requests for as much from Nixon and Kissinger. This would remain the case throughout the crisis, as no explicit consensus ever formed about U.S. objectives in relation to the shoot down, and there was little NSC or presidential-level discussion about what the United States actually wanted to achieve. Still, Kissinger’s NSC staff identified five objectives that were proposed on paper, if not discussed at the meeting: 1) deter future hostile provocations of this kind; 2) seek justice through retaliation or demands of compensation; 3) mobilize international denunciations of

North Korea while decreasing criticisms of the United States; 4) maintain domestic opinion supporting U.S. overseas security commitments in Vietnam and elsewhere; and 5) assert and preserve U.S. freedom in international airspace.87 In a JCS options paper distributed to attendees ahead of the NSC meeting, an overlapping but different set of objectives was proposed: 1)

86 Ibid. 87 “Alternative Courses of Action in Response to Korean Attack on U.S. Aircraft,” April 16, 1969 FOIA request, released on April 15, 2002 (Washington, DC: The National Security Archive); “North Korean Downing of a U.S. Reconnaissance Plane: The EC-121 Incident, April 1969,” October 1969, State Department Historical Studies Division, Office of Executive Secretariat, Folder History Office Research Projects 1969-1974, RG 59, box 6, NARA: 7.

172 receive redress for the attack; 2) “react to the extent required” to prevent further provocations, including actions along the DMZ; 3) prevent escalation, within the limits imposed by needing to achieve the first two objectives; and 4) cause minimum disruption to other military operations worldwide.88 A CIA memo to Director Richard Helms summarized the objectives as being principally about the principle of freedom of movement in international airspace and deterrence of further provocations.89 These three sets of proposed objectives prioritized deterring future provocations and seeking a redress of grievances, but the NSC-staff proposed objectives also prioritized influencing public opinion while the JCS-proposed objectives prioritized not disrupting military missions elsewhere. Neither JCS- or NSC-proposed objectives prioritized among objectives, even though objectives like deterrence logically contradicted others, like leaving unsullied military missions elsewhere.90 Each list of objectives, in other words, identified fundamentally risk-averse and risk-acceptant objectives without prioritizing one over the other. It could be argued either way that the lack of prioritization was a failure of bureaucracy for its inability to propose a logically consistent course of action, or a failure of leadership for an inability to set agendas and priorities that would give the bureaucracy something to respond to consistently. Kissinger’s memoir seemed to do both.91

The day after the NSC meeting, on April 17, Kissinger and Nixon met early in the day to discuss the EC-121, among other issues. On the evening of April 17, Kissinger and Nixon spoke

88 U.S. Department of State, April 15, 1969. “Papers Prepared by the Joint Chiefs of Staff,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 7: 16-17. 89 Memorandum, “Communist Reactions to Certain U.S. Actions,” April 17, 1969, Nixon Presidential Materials, NSC Files, Country Files Far East, Korea, Vol. 1 to 9-69, box 540, NARA. 90 Ibid. 91 Kissinger, White House Years.

173 by phone about how to respond to the EC-121 shoot down and what to say at Nixon’s press conference to occur the following day. The conversation resolved that Nixon would announce the resumption of reconnaissance operations during his press conference the following day, but it did not rule out a retaliatory strike; indeed, the tone of the conversation skewed toward the militant. The many military options from the April 16 NSC meeting that were discussed but not really vetted or deliberated upon had by the time of the April 17 phone call effectively been reduced to two concrete options: either conduct a single air strike against a North Korean air base, preferably the one responsible for the attack, or resume reconnaissance operations in the area where the EC-121 was shot down, but this time under a “protected” status, implying but not committing to armed escorts accompanying reconnaissance flights. Kissinger and Nixon also discussed using tactical nuclear weapons against North Korea in the event that North Korea responded to U.S. retaliation with retaliation of its own: “We might have to go to tactical nuclears and clean it up. All hell will break loose for two months, but at end of road there will be peace in Asia.”92 After the April 16 NSC meeting, Kissinger had a conversation with

Ehrlichman in which Kissinger asked him what he thought the domestic reaction would be to

“Knocking out the base where the Korean planes came from.” When Ehrlichman raised the prospect of North Korean counter-retaliation, Kissinger acknowledged that escalation would result. When Ehrlichman asked how far escalation would go, Kissinger suggested he was prepared for it to “go nuclear.”93 But discussion of nuclear use with Nixon and with Ehrlichman

92 U.S. Department of State, April 17, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 15: 34-7. 93 Isaacson, Kissinger, 181.

174 was limited to the context of thinking two moves ahead of the current decision point, exploring their own willingness to follow through on the downside consequences of a retaliation decision.

The matter at hand remained principally focused on whether to launch an air strike or resume reconnaissance operations in the same area where the EC-121 was shot down. The latter option, resuming reconnaissance operations, was also considered in tandem with or apart from

“Lunch,” the codename assigned to an operation that involved bombing Cambodia in a continuation of the secret bombing that had started the month prior. From Nixon’s and

Kissinger’s perspectives, bombing Cambodia showed firmer resolve to the North Vietnamese ahead of Paris peace talks to end the fighting, which they linked to the North Korea decision on the basis that if they took a conciliatory position toward North Vietnam, the costs would be too high if they also let North Korea’s provocations go unchallenged.94 On Kissinger’s recommendation, Nixon decided not to openly link Cambodia with the North Korea decision, but did decide that he would announce resuming protected reconnaissance operations during his press conference the following day.

Nixon sized up the first military option as “the gains are great and the risks very great.

With option 2, there are no gains and no risks except perhaps down the road.”95 Kissinger argued that the cumulative risk in option 2 would be greater than the risk in option 1. Nixon responded agreeably, venting that “…everytime US fails to react, it encourages some pipsqueak

94 U.S. Department of State, April 17, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 15: 34-7. 95 Ibid.

175 to do something.”96 The decision to go with the latter of the two military options did not seem to satisfy Nixon or Kissinger at the time, as the conversation ended with Kissinger claiming that much could be gained by facing down North Korea, which he characterized as the “toughest character in the Communist camp,” and Nixon suggesting that given the context he could bring any amount of force to bear on North Korea and not take domestic political “heat.”97 Kissinger walked away from his discussions with Nixon on the 17th believing he had decided on launching an air strike against a North Korean air base. Shortly after his deliberations with Nixon,

Kissinger informed H.R. Haldeman that they would bomb a North Korean airfield at noon on

April 21, the following Monday. Although he would later take a more nuanced position, at the time Kissinger expressed the belief that a retaliatory strike against North Korea would be worthwhile, even though it would probably lose liberal-, peace-oriented constituencies and could put at risk congressional funding for the development of an antiballistic missile system on which the Nixon administration placed great importance.98 Though it cannot be firmly validated, it thus appears that Nixon was leaning toward a decision to attack North Korea on April 17, only to change his mind the following day. The next morning, just before his press conference, Nixon again spoke with Kissinger over the phone, doubting that simply announcing a resumption of reconnaissance flights would be enough, describing it as “a piddily thing…it’s not all that significant.”99 After some back and forth about strengthening the language Nixon would use in

96 Ibid. 97 Ibid. 98 H.R. Haldeman, The Haldeman Diaries (New York: Berkeley Books, 1995), 64. 99 U.S. Department of State, April 18, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 19: 44-5.

176 the press conference, he ended the conversation affirming the decision to announce the resumption of reconnaissance flights, but seemed to not yet fully rule out a retaliatory strike subsequently.100

Kissinger,101 some on his NSC staff,102 H.R. Haldeman,103 and the CINCPAC, Admiral

McCain104 all voiced strong support for retaliation against North Korea for the EC-121 shoot down. But Secretary Rogers, the U.S. embassy in Seoul, and CIA Director Helms opposed retaliation in even more strident terms. As the crisis would reveal, Secretary of Defense Laird also made an elaborate entreaty to Nixon on April 18 not to retaliate. Kissinger’s memoirs, as well as some historical accounts, claim that Rogers, Helms, and even Laird, threatened to resign if Nixon retaliated.105 The strength of this claim comes under a shadow of doubt only because some believe that the key decision-makers (Nixon and Kissinger) had an incentive to exaggerate the bureaucratic obstinacy they were facing to help justify backing down from North Korea’s challenge.106 But even if memoirs embellish details at times, we find ample support for the claim that Rogers, the U.S. embassy in Seoul, and Laird were all opposed to a retaliatory strike. On

100 Ibid. 101 Kissinger, White House Years. 102 Kissinger characterized his staff as broadly pushing for retaliation, and military adviser Alexander Haig certainly did, but at least one member of his staff heralded the risks of retaliation in an April 18 memo to Kissinger. See U.S. Department of State, April 18, 1969. “Memorandum from Richard L. Sneider of the National Security Council Staff to the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 16: 37-8. 103 Haldeman, The Haldeman Diaries. 104 Memorandum, “Movement of CVA Task Forces,” April 16, 1969, Wheeler Records, Folder 091 Korea (BP) EC- 121 Shootdown,” RG 218, box 31, NARA. 105 Kissinger, White House Years; Isaacson, Kissinger; Hoff, Nixon Reconsidered. 106 In an oral history, Tom Hughes, Director of the Bureau of Intelligence and Research at the State Department during the Johnson and Nixon administrations, questioned whether anybody threatened to resign, in part because it would have been uncharacteristic for Laird, Helms, and Rogers, but also because Helms’ position at CIA precluded a formal policymaking role and Laird was by that time already deeply involved in orchestrating the secret bombing of Cambodia. For a discussion, see Ostermann and Person, Crisis and Confrontation on the Korean Peninsula, 104- 6.

177 April 17, after Kissinger spoke with Rogers and Laird separately about their views on retaliation,

Nixon spoke with Kissinger about getting staff in line to support a decision for retaliation if it came to it. Nixon stated he already knew Rogers was opposed to doing anything, but he anticipated that Laird would fall in line and publicly support any decision relating to Korea.107

Rogers made a public statement the previous day implying that the United States should not respond, asserting that “The weak can be rash; the powerful must be more restrained.”108

In an April 18 letter to Nixon, Laird commented that among all military options, an air strike was optimal because it was a proportional response, it was a one-time operation with no continuous requirement, and it could be launched by land or sea. However, the same letter argued that retaliation was undesirable for many, mostly dubious, reasons: it was not clear to

Laird that the United States needed the volume of reconnaissance conducted against North

Korea; poor U.S. planning might be at fault for the shoot down in the first place; the risk of

North Korean retaliation exceeded any benefits from “punishing” North Korea with a strike; non-retaliatory means of demonstrating resolve existed, including “never again” deterrent threats and escorted reconnaissance operations; congressional voices calling for retaliation were a small minority; there would be no benefit to linking a strike on North Korea to Paris negotiations with

North Vietnam, and perhaps could lead the United States to appear unreasonably aggressive; and he questioned whether the United States had the capability to fight a major confrontation with

107 U.S. Department of State, April 17, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 15: 34-7. 108 Richard Reeves, President Nixon: Alone in the White House (New York: Simon & Schuster, 2001), 69; Press Telegram (April 17, 1969): 4.

178 North Korea.109 It seems unlikely that Laird actually believed all of these arguments—he knew, for example, about the volume of reconnaissance being conducted and never took action to roll it back, and he excluded South Korean capabilities (which were close to even with North Korea) from his assessment of “capability” to wage a major confrontation against North Korea—but they collectively paint a picture of someone who strongly opposed retaliation but would ultimately conform to presidential prerogative.

In cables back to Washington, the U.S. Ambassador to the ROK, still William Porter, leveraged his frequent engagements with President Park and the ROK government to make a different case against retaliation, assessing it “…would be taken as a signal to the South Koreans to go North,” an outcome that even casual observers knew Washington wanted to avoid.110

President Park’s assessment of the situation was, in hindsight, the most prescient. He asked

Ambassador Porter to convey to Nixon his “firm belief that this incident will happen again as soon as North Koreans have an opportunity.” When Cyrus Vance visited Seoul as a special envoy in the early days of the Pueblo crisis, President Park warned Vance that these kinds of attacks would repeat until the United States stood firm. Now, said Park, the United States was being challenged because it did not make firm threats of retaliation for the Blue House raid and

Pueblo seizure.111

109 U.S. Department of State, April 18, 1969. “Memorandum from Secretary of Defense Laird to President Nixon,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 17: 38-42. 110 Nixon, RN, 474; Hersh, The Price of Power, 731; U.S. Department of State, April 18, 1969. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 18: 43-4. 111 U.S. Department of State, April 18, 1969. “Telegram from the Embassy in Korea to the Department of State,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 18: 43-4.

179 Far from a monolith, the above discussion thus reveals that Nixon’s cabinet and White

House staff expressed a diverse range of views in deliberations over whether to retaliate. The decision to back down in this case had many potential influences, one of which was clearly the threat of North Korean retaliation, which loomed large in the minds of decision-makers. The belief that North Korea would retaliate seemed to be shared by everyone; even Kissinger anticipated North Korea might retaliate, but instead of being a deterrent as it was to some,

Kissinger believed that the United States could out-escalate North Korea if escalation came to pass.

But North Korean threat credibility was far from the only relevant factor. In memoirs by

Kissinger, Nixon, and Haldeman, we see a mishmash of three post hoc rationalizations of the decision not to retaliate: 1) that key cabinet members, Rogers, Helms, and Laird, would resign over the issue; 2) the poor responsiveness of bureaucratic machinery; and 3) the positive public reaction to Nixon’s press conference, which was interpreted as showing restraint. The first of these rationalizations, as discussed, might be an overstatement, particularly in the case of Laird, though opinion on the issue of retaliation was clearly divided. It would be difficult to explain why these men would threaten to resign at the very beginning of an administration over an issue involving North Korea, particularly in the context of the Cambodia bombing campaign and other clandestine acts of aggression that incited no such resignation threats. Also, no archival evidence directly supports the claim that anyone threatened to resign. The second rationalization, dissatisfaction with the bureaucratic process supporting decision-making, elides the fact that

DoD, CIA, and the State Department received no White House guidance about objectives or

180 goals, instead being asked only for “options,” a range of which—military and diplomatic—were presented to Kissinger, on paper, the same day that the shoot down occurred.112 Developing detailed concepts of operation for each option took more time, of course, and some of the military options would require waiting more than 24 hours to execute.

There were many “lessons learned” in bureaucratic best practices upon reflection, particularly within DoD, but none of these seemed to rise to the level of consequence for presidential decision-making. The third rationalization, the positive reaction to Nixon’s press conference, may have been a comparatively greater influence on Nixon’s decision not to retaliate. In Nixon’s conversation with Kissinger just prior to the press conference, it seemed that Nixon was at a minimum still considering retaliation, but after the press conference he ruled it out. The machinery was in action; a working group meeting scheduled for the following day was going to work out the detailed strike plan, only to be informed by Kissinger that no strike would take place. Nixon had to take positive action—telling Kissinger no strike would take place—in order to halt the machinery. Haldeman’s diary entry from April 19, 1969, the day after

Nixon’s press conference, reads, “P well recognizes K’s thesis that a really strong overt act on the part of P is essential to galvanize people into overcoming slothfulness and detachment arising from general moral decay.”113 Haldeman’s entry strongly implies that Nixon believed he had strong incentives to retaliate, yet he did not. Nixon would later tell David Frost that the risks at

112 “Talking Paper for the Chairman, Joint Chiefs of Staff, on an Item to be Discussed at the National Security Council Meeting, 16 April 1969,” April 15, 1969, FOIA request, released on January 9, 2007 (Washington, DC: The National Security Archive). 113 Haldeman, The Haldeman Diaries, 52.

181 that moment outweighed the costs: “I still agree that we had to act boldly; I was just not convinced that this was the time to do it.”114

Orchestrating U.S. Actions

The brief story of what the United States actually did in relation to the EC-121 shoot down spanned a considerably greater length of time than retaliation deliberations themselves.

The types of U.S. actions that took place included military reactions to the shoot down, diplomatic reactions, and post-crisis military contingency planning. The latter of the three types of actions would not be visible to external actors, including North Korea, but the character and extent of the planning implies a certain credibility with which the United States viewed North

Korea in a post-provocation setting, and provides a window into the seriousness with which

Nixon and Kissinger viewed North Korea.

U.S. emergency response measures, focused on crash site identification and search and rescue, began less than 15 minutes after U.S. radars confirmed the disappearance of the EC-121.

The initial reactions included scrambling two F-102 fighter aircraft from Osan Air Base in South

Korea to patrol the Sea of Japan, as well as the dispatch of an HC-130 search and rescue aircraft.

The USS Banner, a Pueblo-like intelligence collection vessel operating in the Sea of Japan, was ordered to curtail its mission and return to port in Japan. Over the course of the first 24 hours, four U.S. naval vessels and a navy P-3 reconnaissance aircraft would also move into the area of the crash site to search the wreckage, located some 90 miles east of North Korea’s coast.115 In contrast with U.S. perceptions of the Soviet Union in the Pueblo crisis, the same day as the EC-

114 Hoff, Nixon Reconsidered, 179. 115 U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121; Department of Defense, “EC- 121 Fact Sheet.”

182 121 shoot down, the State Department sent a cable to the Soviet Union requesting support in the search and rescue effort. In response, three Soviet destroyers moved into the area of the crash site to support the U.S. operation; the Soviet Union and the United States actively cooperated in support of a common goal, reinforcing U.S. government-wide perceptions in this case that North

Korea’s provocation was wholly independent of involvement by other communist nations.116

Beyond the compulsory search and rescue operation in response to the shoot down, two additional U.S. military actions merit highlighting. The first is Nixon’s immediate approval of Secretary Laird’s request to move three aircraft carriers from Vietnam to Northeast Asia, which was translated into a warning order to CINCPACFLT to deploy the carriers the same day as the shoot down.117 The second is DoD’s decision to immediately suspend all aerial reconnaissance operations in response to the shoot down. With the decision to move aircraft carriers into the region—on Nixon’s approval—came additional reinforcement actions, including diverting a U.S. battleship, the USS New Jersey, to CINCPACFLT, moving early warning detection submarines to the region, heightened readiness orders to forces on the Korean

Peninsula and to Fifth Air Force, and deploying F-4 fighters from Japan to the Korean

Peninsula.118 On April 19, JCS also ordered U.S. units in South Korea to be put on alert, and directed additional actions with the caveat that they “…were not to disrupt USFK operations, alarm the ROK, or provoke the North.”119 From April 19, the day after Nixon’s press

116 Department of Defense, May 1969. 117 “Memorandum for the Special Assistant to the President,” April 15, 1969 FOIA request, released on January 9, 2007 (Washington, DC: The National Security Archive). 118 D.L. Strole and W.E. Dutcher, “Naval and Maritime Events, July 1968—December 1969,” United States Naval Institute Proceedings 96 (1970): 14; Mobley, Flashpoint North Korea, 119; Kissinger, White House Years, 320. 119 Colodny and Shachtman, The Forty Years War, 43. Mobley, Flashpoint North Korea, 120.

183 conference, through April 26, U.S. Task Force 71—the three carriers plus supporting vessels and aircraft—conducted a military show of force in the Sea of Japan. Nixon and Kissinger both viewed the action as “weak,” with the latter lamenting that the “show of force” was an empty, even counter-productive gesture: “It was a threat unrelated to a comprehensible demand for action by the other side and therefore likely to be interpreted as empty posturing…we had not even asked for compensation from North Korea; there was no condition for them to satisfy; hence no means of resolution.”120

The other noteworthy military action in response to the EC-121 shoot down was DoD’s order to immediately suspend all global aerial reconnaissance operations. It is not clear from declassified evidence if Secretary Laird was the origin of this idea, but it was a move of such significance that he was required to approve the idea, even if it was not his own, making Laird ultimately responsible for the decision. The decision to stand down on reconnaissance operations had the consequence of un-designating fighter aircraft that would have been assigned for military escort of select reconnaissance missions in the event an adversary attempted to target the reconnaissance aircraft. The curiosity that remains even today is not why Laird chose to suspend aerial intelligence missions against Korea but globally, in Greece, against Cuba, the

Soviet Union, and China too. Laird’s decision, made on the day of the shoot down, would become frustratingly apparent on April 18 following Nixon’s public announcement that the

United States would resume protected reconnaissance missions against North Korea. Nixon and

Kissinger were shocked to find on April 17 that protected reconnaissance missions could not resume for three weeks after Nixon’s announcement that they would resume. The White House

120 Kissinger, White House Years, 320.

184 would only learn about the global scope of the stand down on April 22, by memo.121 Nixon would later write, “…without informing the White House, the Pentagon had also canceled aerial reconnaissance in the Mediterranean and the North Pacific—two of the most sensitive areas of the globe. I was surprised and angered by the situation. The North Koreans would undoubtedly think that they had succeeded in making us back off the reconnaissance flights.”122 Upon finding out about the suspension, Kissinger visited Laird and “ricocheted all over the walls,” saying that he “usurped the president’s authority.”123 Kissinger complained in his memoir that “…halting all reconnaissance in response to a shootdown would convey an impression of insecurity; it hardly suggested that the Administration was determined to defend its rights against the brutal challenge.”124 Kissinger was concerned about the precedent set where “a shootdown of a single plane could put an end to our global reconnaissance system.”125 Despite Kissinger sending multiple memos to Laird over the intervening period trying to resume the reconnaissance flights, and despite Nixon’s public announcement of resumption on April 18, DoD would not resume flights until May 8.

Kissinger recognized the need to harmonize diplomatic and military moves so that each legitimized and supported the other, acknowledging as much in the April 16 NSC meeting.126 As with military moves like DoD’s global suspension of aerial reconnaissance, however, U.S. diplomatic moves fell short of pristine orchestration. The first public statement, by Secretary

121 Kissinger, White House Years, 320. 122 Nixon, RN, 476. 123 Isaacson, Kissinger, 181. 124 Kissinger, White House Years, 317. 125 Kissinger, White House Years, 320-1. 126 U.S. Department of State, April 16, 1969. “Minutes of a National Security Council Meeting,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 13: 28-31.

185 Rogers, preempted Nixon’s April 18 press conference and signaled the United States would not retaliate, even though at the time Nixon seemed to be privately favoring striking a North Korean air base. As I mentioned above, Rogers famously commented on April 17 that “The weak can be rash; the powerful must be more restrained,” which media interpreted as suggesting that the

United States might not retaliate against North Korea’s attack.127

The second public statement, Nixon’s April 18 press conference, basically confirmed

Rogers’ unapproved public signal that the United States would ultimately back down from North

Korea’s challenge, even though at the time of Nixon’s press conference he was still at least considering retaliation. The phrasing of Nixon’s remarks in the press conference was imprecise, but strongly implied that the United States: would not retaliate against North Korea; would resume protected reconnaissance operations; and would retaliate if North Korea attacked again.128 In response to a reporter’s question about how the United States would proceed, Nixon replied: “I have ordered that these [reconnaissance] flights be continued. They will be protected.

This is not a threat; it is simply a statement of fact…Looking to the future, as far as what we do depends on the circumstances. It will depend upon what is done as far as North Korea is concerned, its reaction to the [U.S.] protest [at Panmunjom], and also any other developments that occur as we continue these flights.”129 The same day as Nixon’s press conference, the

United States met with North Korea in a meeting of the UNCMAC at Panmunjom. During the

127 Richard Reeves, President Nixon: Alone in the White House (New York: Simon & Schuster, 2001), 69; Press Telegram (April 17, 1969): 4. 128 William Beecher, “Nixon Declares U.S. Will Protect Planes Off Korea,” The New York Times (April 19, 1969): 1. 129 Richard M Nixon, “The President’s News Conference of April 18, 1969,” in Public Papers of the Presidents, Richard Nixon, 1969 (Washington, DC: U.S. Government Printing Office, 1971), 155-6.

186 deliberations over whether to retaliate, the State Department obsessed about how to respond to

North Korea’s request to hold an UNCMAC meeting on April 18, ultimately deciding to accept the meeting request and use it to lodge a protest.130 Major General Knapp, the U.S. representative, told the North’s KPA representative to the meeting that the shoot down was a

“calculated act of aggression,” and part of an established pattern of armistice violations dating back several years.131 Other than requesting that North Korea “take appropriate measures to prevent similar measures,” the United States made no demands of North Korea, not even an apology.132 The U.S. protest at Panmunjom was thus a delicate one, though, in accordance with orders from Washington, Major General Knapp walked out of the meet when it became clear that

North Korea had no intention of showing contrition or even acknowledgement of the gravity of what it had done.133 Another UNCMAC meeting would not be held for four months, though communication between the United States and North Korea (as well as between North and South

Korea) would remain open informally through the NNSC.134

The final type of actions the United States would take in response to the EC-121 shoot down were of an internal planning nature, not visible to North Korea or any other outside actors.

Following the shoot down, through most of 1969, Nixon would order and DoD would develop and refine contingency plans short of total war for retaliating against North Korean provocations,

130 Kissinger, White House Years, 317. 131 Jonsson, Peace-Keeping in the Korean Peninsula, 179-81. 132 Jonsson, Peace-Keeping in the Korean Peninsula, 180; Chuck Downs, Over the Line: North Korea’s Negotiating Strategy (Washington, DC: AEI Press, 1998), 147-8. 133 Ibid. 134 Ibid.

187 intermittently updating Kissinger on the state of planning.135 When the contingency planning documents were recently declassified much public attention focused on the plans that called for the use of tactical nuclear weapons against North Korea in response to an EC-121-like attack.136

The existence of contingency plans does not mean that the specific actions planned are destined to occur, nor do they mean that the resource requirements identified in a plan will necessarily be available if a contingency plan needs to be executed. What contingency plans do convey, however, is the credibility with which an adversary’s actions are viewed, as well as the threat to

“national interests” that the adversary’s actions pose; if the adversary’s postulated actions were either not believable or not a threat to the nation’s interests, there is little reason to invest the time and resources required to generate plans to address specific contingencies. The hundreds of pages of declassified planning documents are thus significant for the amount of preparation the

United States was willing to make to ensure it possessed a range of tailored options—from airfield strikes, to naval blockades, DMZ ground incursions, and tactical nuclear strikes—for proportional but resolute retaliatory measures.137 The purported objective of the retaliatory options in each case was to deter future similar provocations, though we find that General

135 National Security Study Memorandum 53: Korea Contingency Planning, April 26, 1969, FOIA request, released on March 29, 2006 (Washington, DC: The National Security Archive). 136 Peter Foster, “Richard Nixon Planned Nuclear Strike on North Korea,” The Telegraph (July 8, 2010). Available online: http://www.telegraph.co.uk/news/worldnews/asia/northkorea/7878422/Richard-Nixon-planned-nuclear- strike-on-North-Korea.html. 137 “Review of U.S. Contingency Plans for Washington Special Action Group,” June 25, 1969, FOIA request, released on October 5, 2006 (Washington, DC: The National Security Archive); “Korea Contingency Options,” September 22, 1969, FOIA request, released on March 29, 2006 (Washington DC: The National Security Archive).

188 Wheeler and Secretary Laird both expressed reservations about executing any of the options because of their shared expectation that North Korea would retaliate and escalate in turn.138

The North Korea Perspective

Unlike with the Pueblo crisis, declassified Soviet Bloc documents shed little light on

North Korean thinking in this case because they make little reference to the incident. But we do know certain facts relating to the North Korean shoot down itself, how North Korea communicated publicly and during the UNCMAC meeting at Panmunjom, and North Korea’s reaction to U.S. actions. We also have one declassified conversation that North Korea’s Kim Il

Sung would have with Romania’s Nicolae Ceausescu in which he explained his view of the EC-

121 shoot down. These pieces form a common picture of North Korea as a state that did not view its own provocations as especially risky because each successive provocation reified its image of the United States as lacking resolve. It seems clear from the evidence we are able to muster here that North Korea did not desire war and did not think it could withstand another war, yet it took brazen actions repeatedly because it felt that it could, and the hostile dyadic context with the United States—and with South Korea—gave it little reason to avoid aggressive actions if it thought them unlikely to provoke retaliation.

North Korea’s surge of violence in the month prior to the EC-121 shoot down presaged the level of hostility that made such an act of aggression possible. During March 1969, against the backdrop of the U.S. decision to conduct the Operation Focus Retina exercise and in the shadow of the recently resolved Pueblo crisis, North Korea repeatedly attacked portions of the

138 Untitled Memorandum from the Secretary of Defense (Laird) to the President’s Assistant for National Security Affairs (Kissinger), July 15, 1969, FOIA request, released on January 8, 2007 (Washington, DC: The National Security Archive).

189 DMZ it knew were controlled by the United States, it blared threats and propaganda over loudspeakers pointed toward South Korea, and used UNCMAC meetings as occasions to directly blame the United States for fomenting renewed war hostilities on the Korean Peninsula.139 From

North Korea’s perspective the attack on the EC-121 should not have been perceived as particularly aggressive, since an openly hostile relationship preceded the attack, and a precedent had been set for attacking U.S. forces and U.S. intelligence assets without stimulating U.S. retaliation.

Because the United States was tracking North Korean radars, we know that North Korea believed the EC-121 always remained in international airspace, and never approached closer than

38 miles to North Korean territory.140 North Korea used two MiG 21 aircraft to shoot down the

EC-121, which it had to first deploy to North Korea’s eastern coast in advance of the shoot down.141 Because North Korean radar was able to track the EC-121 in this instance, it was almost certainly able to track previous reconnaissance missions operating in the same area, which the United States had conducted with a predictable rhythm for 15 years, according to

Kissinger.142 Moreover, we know that North Korea’s MiGs chased down the EC-121 before shooting it down.143 The close timing of North Korea’s request for an UNCMAC meeting with the timing of the shoot down additionally suggests North Korean coordination between its

139 Bolger, An Unfinished War; Jonsson, Peace-Keeping in the Korean Peninsula. 140 U.S. Department of State, April 18, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 17: 34-7. 141 U.S. Department of State, April 15, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 8: 17-20. 142 Ibid. 143 U.S. House Armed Services Committee, Inquiry into the USS Pueblo and EC-121.

190 diplomatic and military elements.144 Finally, the shoot down took place on the symbolic occasion of Kim Il Sung’s 57th birthday; imbuing provocations with symbolism by conducting them on meaningful dates has become a trademark of North Korea in subsequent decades.

Taken together, these strands of information strongly imply that North Korea’s shoot down was planned and deliberate, which in the North Korean system would at least need to be condoned by, if not explicitly ordered by, Kim Il Sung. To the extent that the shoot down was a deliberate, centrally orchestrated action, it tells us that North Korea made a decision that it could attack the aircraft without reprisal

Only two hours after the EC-121 was shot down, while the United States was still trying to confirm that a shoot down had even taken place, Radio Pyongyang announced that it attacked the EC-121 as “retaliation against firing by the U.S. Army in the DMZ,” while also claiming that the EC-121 was in violation of North Korean airspace.145 Nearly simultaneous with the radio announcement, North Korea proposed an UNCMAC meeting at Panmunjom for April 18.

During that meeting, North Korea’s attitude was dismissive and calmly accusatory, even though the U.S. representative to the meeting made no formal demands. In response to Major General

Knapp’s complaint that the shoot down was an armistice violation and unwarranted act of aggression, North Korea’s representative, Major General Ri, responded: “You have just referred to a brigandish and aggressive act of the U.S. Government which illegally dispatched a large- sized reconnaissance plane on 15 April last for the purpose of conducting reconnaissance of the interior of our country. I, first of all, ask you: What country owns the EC-121 large

144 North Korea requested the meeting only two hours after the shoot down. Jonsson, Peace-Keeping in the Korean Peninsula, 177; Mobley, Flashpoint North Korea, 114. 145 Quoted in Jonssen, Peace-Keeping in the Korean Peninsula, 179; Downs, Over the Line, 146.

191 reconnaissance airplane you have talked about?”146 Other than North Korea’s opening statement which did not even mention the EC-121, this was the extent of North Korean remarks during the

UNCMAC meeting because General Knapp walked out of the meeting in protest when it became clear from General Ri’s statement that North Korea had no interest in contrition or assuring stability on the Peninsula.

North Korean media offered additional commentary relating to the EC-121 shoot down that reinforced combined themes of celebrating the success of the shoot down, justifying the shoot down as a defensive act, and framing the shoot down as a coercive threat intended to persuade the United States to cease reconnaissance operations. On the day of the shoot down, the Korean Central News Agency supplemented Radio Pyongyang’s announcement by reporting that the United States was trying to provoke a war-like situation by committing “grave provocations” on the Military Demarcation Line and by challenging Korean sovereignty.147 The days prior to and following Nixon’s press conference, on April 17 and 19, North Korean media referenced hundreds of past instances of U.S. “aerial espionage” violating North Korean airspace, explicitly noting multiple types of reconnaissance aircraft, including: the RB-47; the

RB-57; the RC-130; and EC-121.148 Kim Il Sung awarded a commendation to the KPA unit that shot down the EC-121 the day after the attack occurred.149 The commendation was granted for

146 Mun-hang Lee, JSA – Panmunjom (1953-1994) (in Korean) (Seoul: Doseo Choolpan Seohwa, 2001), 49-50. Also quoted in Downs, Over the Line, 147. 147 Korean Central News Agency (KCNA) translation in British Broadcasting Corporation (BBC) Monitoring Service, April 15, 1969, Daily Reports. 148 KCNA translation in BBC Monitoring Service, April 19, 1969, Daily Reports; Pyongyang Home Service, translation in BBC Monitoring Service, April 17, 1969, Daily Reports. 149 Kim Il Sung, “Congratulations to Officers and Men of Unit 447 of the Korean People’s Army on Shooting Down the Heavy Reconnaissance Aircraft of the US Imperialist Aggressive Forces which Illegally Intruded into the

192 combating the “illegal intrusion” of “U.S. imperialist aggressive forces.”150 An April 23, 1969 government statement also described U.S. reconnaissance missions as “hostile acts of espionage,” threatening that “If the reconnaissance planes of the U.S. intrude into the territorial air of our country, we will not sit with folded arms, but will take resolute measures for safeguarding our sovereignty.”151

The final source of the North Korean perspective in this case comes from an extended discussion between Kim Il Sung and Romania’s Nicolae Ceausescu on June 10, 1971, some two years after the shoot down itself. The meeting was a standard and formal affair between heads of state, but Kim Il Sung’s rarely accessible musings and lamentations are helpful to our purpose of understanding thinking about threat credibility and the weight of history on decision-making.

Kim spoke at some length about the lack of hardship that the younger generation in North Korea has had to face, and that they know little of the true nature of Japanese militarism and U.S. imperialism. He lamented that the younger generation filling out the bottom ranks of the North

Korean military was “not well trained for direct confrontation,” adding that “Unless we manage to increase our hatred against exploiters…we will face hardship in the case of a new war.”152 To the extent that Kim actually believed this, it reveals a need for him to maintain a certain level of hostilities with the United States in order to stoke the fires of anti-Americanism, which was

Northern Half of the Republic for Reconnaissance Purposes, April 16, 1969,” Kim Il Sung Works 23 (Pyongyang: Foreign Languages Publishing House, 1985), 423-4. 150 Ibid. 151 Quoted in Downs, Over the Line, 148; Lee, JSA—Panmunjom 50; KCNA translation in BBC Monitoring Service, April 23, 1969, Daily Reports; Pyongyang Home Service, translation in BBC Monitoring Service, April 23, 1969. Daily Reports. 152 “Minutes of Conversation on the Occasion of the Party and Government Delegation on behalf of the Romanian Socialist Republic to the Democratic People’s Republic of Korea,” June 10, 1971, History and Public Policy Program Digital Archive, Archives of the Central Committee of the Romanian Communist Party, Obtained and translated for the NKIDP by Eliza Gheorghe (hereinafter referred to as “Minutes of Conversation”).

193 necessary to ensure that his people could endure a conflict. Because Kim believed the United

States had no stomach for renewed conflict, he could provoke them, leading to a virtuous cycle of morale stiffening in North Korean society that would in turn prepare his people to endure another war.153 Directly addressing his views of American resolve following the Pueblo and EC-

121 crises, Kim assessed, “Americans don’t want to continue this fight. The Americans let us know that it’s not their intention to fight the Koreans again.” Kim expressed a goal vis-à-vis the

United States of “being left alone,” which required the removal of the U.S. presence from the area, which he defined as explicitly including an end to U.S. intelligence collection missions:

“…if they [the Americans] create situations like Pueblo and EC-121, then we are entitled to capturing them or shooting them down…Should the United States withdraw from South Korea, there wouldn’t be any reasons for such incidents, because the South Koreans do not have the material and military basis for such things.”154

During the conversation, Kim clearly took responsibility for North Korean provocations against the United States—he did not deny any North Korean behavior—but he framed them as purely defensive: “We keep our business to our territory…it’s obvious that unless they came into our territorial waters, we couldn’t have captured or sunken their vessels…Regarding the existence of war, the reason is just one: the presence of the Americans in South Korea.”155 We thus see in Kim’s conversation with Ceausescu a North Korean image of the United States as fundamentally lacking resolve on Korean Peninsula issues, rationalizing North Korea’s aggressive behavior as the consequence of the U.S. failure to heed North Korean threats.

153 Ibid. 154 Ibid. 155 Ibid.

194 Analysis: Explaining Decisions and Outcomes

This section assesses the fit of the historical narrative with the competing perspectives on reputation and threat credibility offered at the outset. I find that while some aspects of this case fit with the expectations of current calculus theory, it is the logic of reputation for resolve that not only fits the outcomes in this case, but process tracing evidence directly substantiates the role of reputation here. We also find that North Korea’s provocations, including the recently finished Pueblo crisis, did strengthen U.S. perceptions of North Korean resolve within the crisis, but its challenges did not yield a cumulative reputational effect.

Explanatory Power of Reputations for Honesty

This case offers little support for the reputations for honesty perspective, though available evidence in this case does not allow us to assess one of its key claims. The United States made no threats in the time period examined here, so this case does not help us determine whether backing down (a form of honest signaling) strengthens future threat credibility of the party that backed down. We do find support for the reputations for honesty claim that a word-deed gap reduces threat credibility while word-deed alignment increases it, but reputations for resolve hypotheses expect the same outcomes for different reasons. We also find that each of the key players in U.S. decision-making recommended an honest approach, with specific recommendations differing largely based on different assessments of threat credibility.

But Kim Il Sung’s 1971 admission that North Korea lacked the capability to engage in a war while conducting provocations signaling the opposite suggest that North Korea did not place great value on a reputation for honesty. There are at least three plausible explanations for this, none of which are mutually exclusive. The first is that Kim Il Sung believed that democracies

195 generally lacked resolve, or were at least sufficiently risk averse that they would not seek to provoke a war. The second possibility, which does not necessarily contradict the first, is that the

U.S. history of backing down and meeting North Korean demands with conciliation convinced

Kim Il Sung that the United States would not retaliate if North Korea kept the lethality of its provocations below a certain threshold. The third possibility is simply that as an authoritarian regime, it placed less value on preserving commitments and honest communication than democracies. Taking one of the few direct pieces of evidence reflecting North Korean thinking—Kim Il Sung’s own words in a conversation with another head of state—we find Kim reflecting that “Americans don’t want to continue this fight. The Americans let us know that it’s not their intention to fight the Koreans again.” This statement quite explicitly fits with the reputation for resolve explanation that they could conduct provocations without U.S. retaliation—if Kim Il Sung did not impute U.S. intentions on the basis of past U.S. word and deed, then to what would Kim possibly be referring in this statement? Critically, on the point of greatest divergence between the reputations for honesty and resolve logics—whether backing down encourages or discourages future challenges—we find strong, direct evidence contradicting the expectations of the reputations for honesty hypothesis: U.S. decisions to back down were repeatedly followed by more North Korean challenges, both before the EC-121 shoot down and after, which strongly correlates with the reputations for resolve claim that backing down from a rival’s challenge encourages future such challenges. Combining Kim Il Sung’s statement about his perception of U.S. resolve with the frequency of North Korean provocations following U.S. decisions to back down over time, we can have at least some confidence that

196 North Korea did pay attention to past U.S. word and deed when deciding on challenges to initiate.

North Korea’s history of word-deed misalignment made the prospect of another North

Korean challenge to the United States seem incredible, even though North Korea had in the past challenged the United States directly on many occasions, frequently targeting U.S. patrols along the DMZ, attempting to shoot down a U.S. RB-47 aircraft in 1965, seizing the Pueblo in 1968, and bringing down a U.S. helicopter in March 1969. CIA Director Richard Helms contextualized the EC-121 shoot down as part of a pattern of North Korean provocations going back years, and U.S. representative to UNCMAC, Major General Knapp, said as much to his

North Korean counterpart at Panmunjom.156 Yet, in Nixon’s own words, North Korea’s attack on the EC-121 was “…a surprise in every sense of the word.”157 Kissinger’s memoirs also describe the administration as being taken off guard.158 And although General Bonesteel, the

USFK commander, would later tell that he informed CINCPAC that the North Koreans “were up to something” just prior to the shoot down,159 the U.S. interagency, including OSD, Joint Staff,

DIA, and the State Department, approved the EC-121 with what we can infer was an assessment of minimal risk posed to the aircraft.160 Bonesteel’s purported suspicions stemmed in part from an onslaught of threatening rhetoric that North Korea issued publicly and indirectly through the

NNSC from early March 1969 onward, but others in the U.S. policymaking apparatus showed no

156 U.S. Department of State, April 16, 1969. “Minutes of a National Security Council Meeting,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 13: 28-31. 157 Nixon, “The President’s News Conference of April 18, 1969,” 155-6. 158 Kissinger, White House Years; Isaacson, Kissinger. 159 Mobley, Flashpoint North Korea, 106. 160 Mobley, Flashpoint North Korea, 102-3.

197 sign either that they were aware of these many threats or that they took them seriously if they did.161 Following North Korea’s shoot down of the EC-121, however, we find that most but not all U.S. decision-makers did suddenly find the threat of North Korean retaliation against U.S. actions credible. As a reputation for honesty logic would predict, when North Korea’s word and deed matched—conducting a provocation—it gained credibility. In the shadow of the EC-121 shoot down, Secretary Laird, Director Helms, Secretary Rogers, and General Wheeler all believed that North Korea would be likely to retaliate if the United States targeted a North

Korean air base in a reprisal strike.162 Kissinger and certain members of his staff, as well as

Alexander Haig, had doubts that North Korea would retaliate, but nevertheless believed that if

North Korea retaliated the United States should be able to escalate and win because it possessed superior capability.

This case shows mixed support for the reputations for honesty claim that states value their reputations for honesty. In contrast with what I interpret as North Korea’s frequent bluffing, we see in this case that U.S. decision-makers advocated an honest approach to North

Korea by recommending actions commensurate with perceptions of U.S. interests; nobody recommended bluffing. That key decision-makers could disagree about whether U.S. interests were sufficiently engaged to necessitate retaliation shows the obvious challenge of trying to use interests as an objective or predictive measure in any kind of systematic way. Consistent with the reputations for honesty logic, Laird, Helms, Rogers, and Wheeler opposed retaliation because

161 Bolger, An Unfinished War, 98-100; Jonsson, Peace-Keeping in the Korean Peninsula, 151-2, 155; “7Americans Die As Result of Korean Assault,” Daytona Beach Morning Journal (March 17, 1969): 6. 162 Untitled Memorandum from the Secretary of Defense (Laird) to the President’s Assistant for National Security Affairs (Kissinger), July 15, 1969, FOIA request, released on January 8, 2007 (Washington, DC: The National Security Archive).

198 U.S. interests were not sufficiently engaged to warrant the risk of North Korean escalation, which they viewed as a credible threat.163 Kissinger, Haig, and Nixon, in the early moments of the crisis anyway, argued for a much tougher line against North Korea, but because they were willing to accept the costs of escalation in the event that U.S. retaliation begot North Korean retaliation.164 For them, and though Nixon would change his mind, U.S. interests were sufficiently engaged to warrant a reprisal strike on North Korea.

Each U.S. perspective in this case was thus consistent with a desire to remain honest; they differed not over whether or not to make an empty threat, but rather over the credibility of the threat of North Korean retaliation and whether U.S. interests were sufficiently engaged.

However, to the extent that North Korea believed it lacked the capacity to engage in another war with the United States, its provocations—which signaled willingness, even desire, to go to war— would seem to suggest it had little qualms with bluffing, and did not place great importance on maintaining a reputation for honesty. North Korea’s long history of threatening rhetoric far outpacing its actions, however threatening, indicates the same. Although not necessarily conclusive evidence, Kim Il Sung’s 1971 conversation with Ceausescu revealed that Kim believed North Korea lacked the ability to wage another war, yet North Korea’s provocations seemed to designed to convey the opposite. It is possible that democratic regimes value a reputation for honesty more than authoritarian regimes, but based only on what we see in this

163 See, for example, U.S. Department of State, April 17, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 15: 34-7. 164 Memorandum, “Shoot-Down Incident,” April 16, 1969, Nixon Presidential materials, NSC Files, Korea: EC-121 Shootdown, North Korean Reconnaissance Shootdown, 4/9/69-4/16/69, Vol. 1 Haig, box 434, NARA; U.S. Department of State, April 17, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 15: 34-7.

199 case, it is difficult to conclude anything more than that the United States placed greater emphasis on honest communication than North Korea; empirical evidence thus only partially supports this aspect of the reputations for honesty hypothesis.

On the point of greatest contention within the reputation optimism camp—whether backing down is more or less likely to lead to future challenges—we find that reputations for honesty completely fails to explain the events in this case. Sartori and other reputations for honesty advocates would code the United States going into the shoot down of the EC-121 as having a reputation for honesty. In the four months prior, the Pueblo crisis had been resolved because the United States offered concessions by meeting North Korean demands, and in March

1969 the United States took no reprisal actions against North Korea for its attacks on U.S. units along the DMZ. These instances of honest signaling by choosing to back down were followed not by a deterred North Korea but by the EC-121 shoot down. Several months after the EC-121 shoot down, and the month after Nixon announced the “Nixon doctrine” signaling a partial U.S. retrenchment from Asia, North Korea yet again initiated a clash with the United States, shooting down a helicopter and holding the crew hostage, ultimately demanding the same apology terms it demanded and received for the Pueblo.165 The record on this point is clear: evidence in this case inversely correlates with the expectations of reputations for honesty when it comes to predicting the effect of backing down on the likelihood of future challenges.

Explanatory Power of Reputations for Resolve

165 Memorandum, “Helicopter Crew Members Held by North Koreans,” October 4, 1969, NSC Files, Country Files Far East, Korea, Helicopter Incident, box 540, NARA.

200 Reputations for resolve explain most aspects of this case. It contains strong evidence supporting the idea that states and statesmen value a reputation for resolve. The reputations for resolve logic also offers a fitting explanation for North Korea’s repeated provocations, for U.S. incredulity toward North Korean threats prior to the downing of the EC-121, and the change in

U.S. views of North Korean threat credibility immediately following the shoot down. On the most crucial distinction between reputations for honesty and resolve, the latter outperforms the former, correctly anticipating that U.S. decisions to back down would be followed by subsequent

North Korean challenges. We also find that North Korea’s record of provocations had only a fleeting effect on North Korea’s reputation for resolve, with its impact being limited only to the period immediately following the provocation but not accumulating any discernible effect.

Clear process tracing evidence from within the case shows the value that key U.S. decision-makers placed on a reputation for resolve. Kissinger makes it abundantly clear, both in his memoirs and in declassified records, that a decision to back down would hurt U.S. credibility and invite future aggression. Notwithstanding his decision to the contrary, Nixon echoed the same sentiment as Kissinger. During the phone conversation between Nixon and Kissinger the day after the shoot down, Nixon’s comment that “…everytime US fails to react, it encourages some pipsqueak to do something” reflects an extreme version of the logic of reputations for resolve, implying that if the United States did demonstrate resolve vice backing down, weaker states would be less likely to challenge it.166 In the same conversation, Kissinger added that the

166 U.S. Department of State, April 17, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 15: 34-7.

201 long-term risks of inaction were greater than the short term risks of retaliation.167 Kissinger, moreover, writes in his memoirs that “I believed we paid a price in many intangible ways, in demoralized friends and emboldened enemies.”168 Other important staff beyond Nixon and

Kissinger placed importance on a reputation for resolve as well. All of the options papers—by

NSC staff, JCS, and CIA—identified an objective in the current crisis of deterring future provocations, which was implicitly a statement of belief that actions taken in the present would have consequences for the future.169 More explicitly, Admiral John McCain, Commander of the

Pacific Fleet, argued to General Wheeler that North Korea would perceive the United States as a

“paper tiger” if the United States did not retaliate.170 After the crisis, Fritz Kraemer, a civilian adviser in the Department of Defense, on June 13 wrote to Kissinger decrying the danger of international perceptions of U.S. weakness, citing North Korea’s provocations in particular:

“…the invasion of Czechoslovakia, the capture of the Pueblo, the shooting down of the EC-121, the shelling of our fishing boats by Peru…In each of these cases it may have been statesman-like and prudent not to retaliate by anything but words. In sum total, however, the image of the

United States as a paper tiger is coming dangerously close to being universally accepted.”171

America’s “provocative weakness, Kramer wrote, “constitutes an almost irrepressible

167 Ibid. 168 Kissinger, White House Years, 321. 169 “Alternative Courses of Action in Response to Korean Attack on U.S. Aircraft,” April 16, 1969, FOIA request, released on April 15, 2002 (Washington, DC: The National Security Archive); “Talking Paper for the Chairman, Joint Chiefs of Staff, on an Item to be Discussed at the National Security Council Meeting, 16 April 1969,” April 15, 1969, FOIA request, released on January 9, 2007 (Washington, DC: The National Security Archive). 170 Memorandum, “Movement of CVA Task Forces,” April 16, 1969, Wheeler Records, Folder 091 Korea (BP) EC- 121 Shootdown, RG 218, box 31, NARA. 171 Colodny and Shachtman, The Forty Years War, 43.

202 temptation” for America’s rivals.172 Kramer and other extremely hawkish voices from the Nixon administration ground themselves in a reputation for resolve logic, but it is worth noting that they reflect the most extreme, universal, interpretation of the implications of a reputational logic of resolve.173

But other claims associated with the reputations for resolve literature are potentially more contentious than the point that decision-makers value their own reputations. An important phenomenon observed in this case, involving the strong shift in U.S. perceptions of North

Korean credibility before and immediately after the shoot down, is also explained by reputations for resolve. North Korea’s repeated threats in March and early April were not taken seriously by most U.S. officials other than General Bonesteel. The EC-121 mission was approved to fly without protected escort, as with the 190 previous missions it had flown in the area in the four months prior.174 Thus, based on North Korean actions alone, there were no immediate indications, other than North Korea’s attack on U.S. DMZ patrol units in the months prior, that

North Korean threats would credibly manifest as an attack on a U.S. aircraft. With North Korean rhetoric in the period before the shoot down exceeding its actions, it comes as little surprise from the reputations for resolve perspective that the United States would not find the threat of a North

Korean attack credible. North Korea’s April 15 attack therefore came as “…a surprise in every sense of the word,” as Nixon described it.175 But the explanatory power of reputations for resolve is strengthened by the ability to explain variation in this case, not only U.S. incredulity

172 Colodny and Shachtman, The Forty Years War, 45. 173 To the extent that my findings here support the reputations for resolve logic, my scope is principally limited to the U.S.-North Korea dyad, as I noted in chapter 2. 174 Department of Defense, “EC-121 Fact Sheet.” 175 Nixon, “The President’s News Conference of April 18, 1969,” 155-6.

203 prior to the shoot down. We see that whereas most U.S. officials were shocked that North Korea would attack a U.S. aircraft, once it did, most U.S. officials suddenly found North Korean threats highly credible, inferring North Korean resolve based in part on the downing of the EC-121.

Laird and Wheeler assessed that a military strike posed an undesirable risk of retaliation.176

Nixon commented to Kissinger that the near term risk of a strike was higher than the longer term risk of doing nothing, though Kissinger disagreed.177 And the CIA, who in January 1969 published an SNIE about North Korean capabilities and intentions made nary a reference to the

Pueblo, did not believe North Korea would take any excessively aggressive actions against the

United States, and estimated that future North Korean behavior in general would depend on U.S. demonstrations of resolve: Pyongyang lacked the resolve to start another war, but would continue its “process of probing for weaknesses and testing ROK and US resolve.”178 Despite this, in the midst of the EC-121 crisis, the CIA had changed its assessment of North Korean intentions, arguing that U.S. retaliation would be the response most likely to achieve the goal of future deterrence, but that it also posed the greatest risk of provoking North Korean escalation.179

The CIA added that “North Korea’s assessment of future US initiatives, probably are strongly influenced by the Korean’s interpretation of the US response to the seizure of the Pueblo.”180

176 U.S. Department of State, April 18, 1969. “Memorandum from Secretary of Defense Laird to President Nixon,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 17: 38-42. 177 U.S. Department of State, April 17, 1969. “Record of a Telephone Conversation between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 15: 34-7. 178 U.S. Department of State, January 30, 1969. “Special National Intelligence Estimate,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 1: 1-2. 179 “Alternative Courses of Action in Response to Korean Attack on U.S. Aircraft,” April 16, 1969, FOIA request, released on April 15, 2002 (Washington, DC: The National Security Archive). 180 Memorandum, “Communist Reactions to Certain U.S. Actions,” April 17, 1969, Nixon Presidential Materials, NSC Files, Country Files Far East, Korea, Vol. 1 to 9-69, box 540, NARA.

204 As with the previous Pueblo case, the unique advantage that the reputations for resolve logic offers as an explanatory frame in this case is crisis recurrence—that is, North Korea’s repeated provocations. The combination of a rivalry context with the U.S. reputation for irresoluteness explains the conditions that produced North Korea’s provocations against the

United States multiple times in this case, as in the previous. Reputations for resolve tells us that backing down from a rival’s challenge in the current crisis makes a future challenge from a rival more likely, compared to reputations for honesty’s claim that backing down in the current crisis makes future challenges from a rival less likely. Three months after the United States resolved the Pueblo crisis by conceding to North Korean demands, we see a string of North Korean attacks along the U.S.-controlled portions of the DMZ, ostensibly in response to the U.S. decision to conduct the Operation Focus Retina exercise.

The United States took no action in response to these attacks, which were followed less than a month later by the EC-121 shoot down. By April 18, 1969, Nixon’s press conference made reasonably clear that the United States would not be retaliating against North Korea, but

Nixon publicly hinted that the United States might be in a “never again” mode regarding North

Korean provocations at the level of intensity of the EC-121 shoot down.181 Three months later,

Nixon announced a new foreign policy doctrine that was widely interpreted as a process of U.S. withdrawal from Asia, which North Korea had long sought to achieve.182 Despite the rhetoric of firmness underpinning Nixon’s April 18 statement of restraint, in August, North Korea downed a

U.S. helicopter along the DMZ, holding its crew hostage, creating an incident that resembled a

181 Nixon, “The President’s News Conference of April 18, 1969,” 155-6. 182 For a discussion of the Nixon Doctrine, see Robert S. Litwak, Détente and the Nixon Doctrine: American Foreign Policy and the Pursuit of Stability, 1969-1976 (Cambridge: Cambridge University Press, 1984).

205 mini-Pueblo crisis, including resolution through the same conciliatory formula as the Pueblo.183

From the reputations for resolve perspective, an enduring dyadic hostility with repeated play and imperfect information are conditions under which states are most likely to apply formulas that worked in the past; North Korea paid no direct costs for its provocations, but the rivalry context that motivated them persisted, so it is only logical that the provocations would continue.

Backing down from North Korean challenges incentivized their recurrence.

Alternative Explanations about Credibility

Current calculus theory and constant calculus theory offer competing claims about the source of threat credibility than that offered by either the honesty or resolve types of reputation optimism. Both theories can contribute to limited aspects of our understanding of this case, but both fall short of explaining threat credibility across the case.

Current calculus theory’s negative claim—that decision-makers will not draw on their adversary’s past word and deed when calculating credibility—proves wrong in this case, as discussed above; U.S. officials and intelligence analysts did frequently make reference to North

Korea’s past actions, as did Kim Il Sung when assessing U.S. resolve. But current calculus theory’s positive claims—that the balance of capabilities and balance of interests affect credibility—are factors that were prevalent in U.S. deliberations. And yet, the shift from pre- crisis North Korean threats that U.S. officials found incredible to within-crisis threats that they found credible is not an outcome that fits with what we should logically expect if credibility is based solely on the balance of capabilities and interests, which I discuss below.

183 Memorandum, “Helicopter Crew Members Held by North Koreans,” October 4, 1969, NSC Files, Country Files Far East, Korea, Helicopter Incident, box 540, NARA.

206 Because current calculus theory explains threat credibility as a combination of a favorable dyadic balance of power and high degree of interests, the fact that the United States did not deem North Korea’s threats credible prior to the downing of the EC-121 must be explained by some combination of a balance of power favoring the United States and the U.S. perception that something less than North Korean vital interests were at stake. After the EC-121 shoot down, the dramatic shift in U.S. perceptions of North Korean threat credibility (from incredible to highly credible) would be explained because of either a large shift in the relative dyadic balance of power favoring North Korea or North Korea’s vital interests were at stake. But there was no shift in the balance of power between the previous case and this one; U.S. power vastly exceeded North Korea’s, and this did not change at any point during this case. Drawing again on

Correlates of War data, we find that the U.S. composite index of national capabilities rated at

.194, whereas North Korea’s index rating was the same as the year prior, .005.184

From the current calculus perspective, early U.S. assessments of North Korea as lacking credibility should come as little surprise given the balance of power starkly favoring the United

States, but this makes within-crisis assessments of North Korea as highly credible difficult to explain because the balance of power continued to strongly favor the United States. So what changed before the crisis and after in terms of North Korean interests at stake that the United

States would suddenly view North Korean threats credibly?

Particularly in terms of the relative balance of interests, it was the United States that had

31 servicemen killed versus North Korea’s none; if a shift in the balance of interests meant anything in this case it should have been in favor of U.S. retaliation, and doubting North Korean

184 Katherine, Keshk, and Pollins, “Correlates of War Project Trade Data Set Codebook.”

207 resolve. It is also difficult to make an interest-based case for explaining U.S. views of North

Korean credibility after the shoot down because the United States conducted similar missions in the same general operating area with great frequency in the prior four months, without challenge.

Whether the April 16 NSC meeting, private conversations with Kissinger and his staff, or phone conversations between Nixon and Kissinger, we see no discussion about North Korean interests.

On the contrary, in the April 16 NSC meeting, we see Nixon trying to size up North Korea by asking about North Korea’s history of provocations, particularly in the Sea of Japan; this is explicitly contrary to the expectations of current calculus theory.185

As with the Pueblo crisis, the U.S. decision to back down in this case was strongly informed by U.S. concerns about the ability to fight a war in Korea while the effort in Vietnam persisted, which speaks directly to the relevance of material capabilities in U.S. decision-making.

But U.S. challenges in force management across multiple conflicts had nothing to do with U.S. assessments of the likelihood that North Korea would escalate a conflict should the United States deal a reprisal blow. In summary, prior to the EC-121 shoot down U.S. perceptions of North

Korean credibility are consistent with current calculus expectations in terms of material capabilities, but not perceptions of North Korean credibility within the post-shoot down crisis.

Likewise, U.S. perceptions of North Korean interests did not seem to change from before the shoot down to after; given the balance of casualties, the balance of interests favored the United

States more after the shoot down than before. Thus, neither national interests nor material capabilities can fully explain the variation in U.S. perceptions of North Korean threat credibility,

185 U.S. Department of State, April 16, 1969. “Minutes of a National Security Council Meeting,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 13: 28-31.

208 though both were clearly necessary conditions when U.S. officials calculated North Korea’s within-crisis threat credibility.

Constant calculus theory does not perform quite as well as either current calculus theory or the reputational approaches in explaining either side’s assessments of credibility in this case.

As with the Pueblo crisis, the constant calculus lens cannot account for the U.S. dismissal of

North Korean threats prior to the EC-121 shoot down: if relations were hostile (and they were), then the United States should have believed North Korea’s threats. To be sure, it can be reasonably said that any rivalry context exhibits the in-group/out-group characteristics on which constant calculus theory places its explanatory emphasis, and the U.S.-North Korea rivalry is no exception. Each side maligned the other, be it U.S. frustration with a “fourth rate power” and

“pipsqueak,” or North Korea’s tough talk about the “U.S. imperialist aggressor.”

Where constant calculus errs is in assuming that threat-making by the “other” should be credible because the “self” sees the “other” as hostile or dangerous. Identities matter, but they are not necessarily the sole causal determinant of anything, let alone credibility; one may not trust an out-group, but there is no reason why lack of trust automatically translates into believability. As such, we find that even if one generously applies constant calculus theory to

U.S. within-crisis assessments of North Korean threat credibility—since U.S. officials found

North Korean threats highly credible at that moment—U.S. dismissals of North Korean threats prior to the EC-121 shoot down are contrary to constant calculus’s expectations. By the standards of constant calculus theory, U.S. incredulity toward North Korea is all the more inexplicable because of the evidence that proved the United States “learned” from its interactions

209 (i.e., the Pueblo crisis and other provocations) in a specific way. The CIA’s assessment of North

Korean intentions the month after the Pueblo crisis, which underestimates the probability of a

North Korean provocation against the United States, shows no learning biases that would be expected of an actor operating in a self-other frame.186 And although the United States made no threats in this case, we still find Kim Il Sung’s personal assessment of a U.S. lack of resolve as running contrary to constant calculus theory’s expectations. Indeed, given the extreme power asymmetries involved, one would expect that if ever there were an instance when an actor would automatically attribute threat credibility to its rival it would be when the actor is vastly outgunned by its rival. Yet Kim expressed the belief that the United States had no intention of fighting another war in Korea.187 So at best, we find constant calculus theory’s expectations congruent with U.S. within-crisis credibility assessments, but confounded by U.S. credibility assessments prior to the shoot down.

Assessing Hypothesis Performance

What does the above analysis suggest about the hypotheses on which this dissertation is based? The first hypothesis stated that backing down from a rival’s challenge de-escalates the current crisis, strengthens one’s own future threat credibility, and makes it more likely that a rival will initiate subsequent challenges. This case strongly supports two of the anticipated causal claims about backing down—de-escalating the current situation while making future challenges more likely—but this case does not offer the opportunity to observe any connection between backing down and future threat credibility because the defender (the United States) did

186 U.S. Department of State, January 30, 1969. “Special National Intelligence Estimate,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 1: 1-2. 187 “Minutes of Conversation.”

210 not make threats against the challenger after backing down from the provocation. The co- variance between defender behavior at time 1 and challenger behavior at time 2 revealed the same pattern observed in the previous chapter: North Korea provokes, the United States backs down by either choosing not to respond or conceding to North Korean demands, and North

Korea subsequently provokes again. After the Pueblo crisis ended with U.S. conciliation, within three months North Korea had launched its series of attacks against U.S.-managed areas of the

DMZ. The United States took no action in response other than continuing with the Operation

Focus Retina exercise it already had underway. Less than a month later, North Korea shot down the EC-121. Within three months of Nixon’s April 18 press conference announcing that the

United States would not retaliate, North Korea had launched at least four more raids across the

DMZ, as well as four sea-based infiltration attempts of South Korea.188 Shortly thereafter, North

Korea shot down a U.S. helicopter along the DMZ, capturing its crew and demanding the same terms for resolution that it received in exchange for the Pueblo crew the year prior, which the

Nixon administration eventually gave. This repeated observation confounds the reputations for honesty literature’s claim that backing down decreases the likelihood of future challenges, but is precisely what the reputations for resolve logic would dictate.

Regarding the second hypothesis, we find that North Korea’s repeated provocations did not result in much of an enhanced reputation for resolve (measured as U.S. perceptions of North

Korean threat credibility) beyond the general boundaries of the crisis itself. Although the reputations for resolve literature is largely silent on whether provocative challenger behavior has a different reputational effect than firm defender behavior, we find in this case that North

188 Downs, Over the Line, 148.

211 Korea’s word-deed gap seemed to undermine any reputation for resolve that might have accrued because of North Korea’s provocations. The CIA’s assessment of North Korean intentions, only one month after the Pueblo crew was released, showed no expectation that North Korea might directly attack U.S. forces.189 North Korea’s barrage of threats the month prior to the EC-121 shoot down also fell largely on deaf U.S. ears, except perhaps for General Bonesteel, though

U.S. officials would at least temporarily change their credibility assessments of North Korea after the shoot down. This outcome—North Korean provocations contributing to a very temporary reputation for resolve before dissolving—fits well with a reputational logic, which expects that a word-deed gap undermines one’s credibility. As with the previous case, North

Korea’s reputation for bluffing seemed to prevent North Korea’s track record of provocations from accreting to a reputation for resolve.

189 U.S. Department of State, January 30, 1969. “Special National Intelligence Estimate,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 1: 1-2.

Chapter V: The Panmunjom Crisis (1976)

The previous two chapters showed no variation in the pattern of behavior between the

United States and North Korea, nor in the assigned roles of protagonist and antagonist: North

Korea initiated each provocation, and each time the United States backed down by either not responding to an act of aggression or by responding to aggression with concessions of some kind. In this rivalry dyad, the challenger was always North Korea; the defender, always the

United States. The cases themselves focused on the most intense crisis points between 1966 and

1969, though each case highlighted the added context of numerous other provocation-response interactions with the same outcomes as the two crises. Some seven years after North Korea’s shooting down of the U.S. EC-121 reconnaissance aircraft, we find North Korea and the United

States reprising the same roles of challenger and defender in a crisis as the latest in a series of hundreds of violent attacks across the DMZ took on a particularly gruesome character, with

North Korea ambushing and mutilating two U.S. soldiers in a melee launched in August 1969.

Unlike the hundreds of other challenger-defender interactions over the years, however, the U.S. response to the crisis now called the 1976 “Panmunjom axe murders” was strikingly different than past responses to North Korean provocations, across the DMZ or otherwise.

As with previous chapters, the narrative for this episode is built on the answers to several questions that help us identify the effects of standing firm and the sources of threat credibility.

What was the recent conflict history of the dyad prior to this incident? What was the geopolitical context of the crisis; that is, what role, if any, did China, the USSR, ROK, and Vietnam play in defining the circumstances surrounding the crisis? Who initiated coercion and why? How did each view the credibility of the other’s threats during the crisis and why? What actions did each

212 213 side actually take during the crisis and what can be inferred about the motivations of each side from these actions, and to what extent does process tracing evidence support these inferences?

Finally, how did the crisis end?

Historical Context

Much had changed in the foreign policy landscape between 1969 and 1976. Hostility between North Korean and the United States, and between North and South Korea, would ultimately persist in both dyads in spite of a period of attempted rapprochement during which the pattern of North Korean provocations receded. By 1975, however, rapprochement initiatives had failed, and North Korean provocations against the United States and South Korea resumed in earnest. It is impossible to accurately characterize the U.S.-North Korea rivalry during this period—indeed, it may be argued that between 1970 and 1975 rivalry conditions temporarily abated—without a discussion of the tectonic shifts materializing in U.S. foreign policy, as well as the foreign policy adjustments of North and South Korea during that period.

North Korea v. United States

If the point of North Korea’s repeated provocations from 1966 through 1969 was to fracture the U.S.-ROK alliance and delegitimize the U.S. military presence in South Korea, its tactics did not succeed, though foreign policy trends at the time nevertheless moved in the direction of North Korea’s favor and the U.S.-ROK alliance did experience strain resulting from disagreements about how to respond to North Korea. As early as 1969, North Korea started changing its tactics away from the use of military provocations. As late as October 1969 North

214 Korea was targeting and killing U.S. soldiers in provocations along the DMZ,1 but that same month North Korea announced a desire for peaceful reunification of the Korean Peninsula.2 In

December 1970, a National Intelligence Estimate (NIE) calculated an 80% drop in violent incidents along the DMZ from the year prior.3 The NIE assessed that North Korea’s tactical shift away from the use of provocations and toward international diplomacy and building a subversive political base in South Korea was likely due to the failure of provocations to achieve any geopolitical benefits for North Korea; the United States reinforced its military position in South

Korea, and South Koreans did not rise up to overthrow South Korean President Park Chung-hee.4

The NIE explanation for North Korea’s shift to a less aggressive foreign policy approach in 1970 might be partially correct, but a major 1969 change in U.S. policy that moved the United

States in a direction that North Korea long desired should not be overlooked either. The “Nixon doctrine,” announced in a speech in Guam in 1969, preserved all U.S. treaty commitments and the extension of the U.S. nuclear umbrella to allies, but shifted the primary burden for defense of

U.S. allies to the allies themselves. The doctrine presaged the withdrawal of 20,000 U.S. troops from South Korea the following year, consistent with Nixon’s push for the “Vietnamization” of

1 On October 18, 1969, two months after North Korea downed a U.S. helicopter and held its crew hostage, North Korea killed four U.S. soldiers in an attack targeted against the U.S.-controlled portion of the DMZ. See U.S. Department of State, October 23, 1970. “Memorandum from John H. Holdridge of the National Security Council Staff to the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 43: 115-16. 2 “The Rise and Fall of Détente on the Korean Peninsula, 1970-1974: A Critical Oral History,” (Washington, DC: Woodrow Wilson International Center NKIDP, July 2010). 3 U.S. Department of State, December 2, 1970. “National Intelligence Estimate,” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969-1972, document 80: 200-212. 4 Ibid.

215 the conflict in Vietnam.5 The “mood of détente” that permeated foreign policy in the early 1970s led one member of the North Korean Workers’ Party Politburo to comment to one of his

Romanian counterparts that “Up until now, we used a wide range of methods in South Korea, but we have achieved nothing. In these circumstances, we can’t wage war. What should we do?

Taking the current situation into account, we thought the best thing to do is to launch a peace offensive.”6

North Korean threats and provocations toward the United States did not completely cease between 1970 and 1976—there was, for example, a 1973 showdown with the United States over a South Korean vessel in the Yellow Sea and occasional attacks along the DMZ—but violent provocations became less commonplace as North Korea went on a global diplomatic charm offensive. During this period, which coincided with the Sino-U.S. rapprochement, the United

States and North Korea each made at least superficial attempts in the direction of détente.

Although yielding little, these attempts are notable for the difference in tone compared to the open rivalry behavior that prevailed from 1966 through 1969. In February 1972, while Nixon was in China for the signing of the historic Shanghai Communique, a North Korean delegation was also in town and unsuccessfully scrambled to arrange a meeting between the North Korean and Nixon delegations.7

5 Jeffrey Kimball, “The Nixon Doctrine: A Saga of Misunderstanding,” Presidential Studies Quarterly 36, no. 1 (2006): 59-74. 6 Minutes of Conversation between Nicolae Ceausescu and the economic delegation from the Democratic People’s Republic of Korea, Bucharest, September 22, 1972, NKIDP Document Reader, “New Evidence on Inter-Korean Relations, 1971-1972,” Document #28, Obtained for NKIDP by Mircea Munteanu and translated by Eliza Gheorghe. 7 “The Rise and Fall of Détente on the Korean Peninsula.”

216 In May 1973, North Korea’s foreign ministry requested that the Romanian embassy in

Washington convey a letter from the Supreme People’s Assembly (North Korea’s legislative body) to the U.S. Senate, House of Representatives, and the State Department, urging the withdrawal of U.S. troops from Korea and the signing of a peace treaty.8 In July 1973, diplomatic reporting from the Soviet Bloc recorded that Henry Kissinger had conducted meetings with North Korean officials in Paris, though no confirmation exists about what might have taken place during these meetings.9 In August 1973, the United States decided to try and arrange meetings with North Korean representatives in the U.S. Liaison Office in Beijing, even though South Korea vocally opposed such discussions when the United States informed them about U.S. intentions to do so.10 And again in 1974, Hwang Jang Yop, Chairman of the SPA, sent a letter to the then U.S. Vice President Gerald Ford in his role as President of the Senate, again requesting a peace treaty and a withdrawal of U.S. troops from South Korea.11 But the

U.S. meetings with North Korea, assuming they happened, seem to have affected long-term

Korean Peninsula dynamics little, and neither the Nixon nor Ford administrations took any recorded actions in response to North Korea’s letters to Congress. Notwithstanding a few

8 “Telegram from the First Directorate to Washington, DC, No.01/04493” May 5, 1973, History and Public Policy Program Digital Archive, Romanian Ministry of Foreign Affairs Archives, Matter 210, 1973, Democratic People’s Republic of Korea, Secret, MFA, Folder no. 1495, Obtained and translated for NKIDP by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/114056. 9 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 061.041” February 7, 1973, History and Public Policy Program Digital Archive, Archives of the Romanian Ministry of Foreign Affairs, Matter 220 - Relations with the Democratic People’s Republic of Korea, 1973, Obtained by Izador Urian and translated for NKIDP by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/114580. 10 U.S. Department of State, August 24, 1973. “Backchannel Message 0818 from the Ambassador to the Republic of Korea (Habib) to the President’s Assistant for National Security Affairs (Kissinger), Seoul, August 24, 1973, 0615Z,” FRUS, 1969-76, Vol. E-12, Part 1, Documents on East and Southeast Asia, 1973-1976, document 245: 1. 11 “Letter from Government of North Korea” May 13, 1974, History and Public Policy Program Digital Archive, Gerald R. Ford Presidential Library, Ford Vice Presidential Papers, Office of Assistant for Defense and International Affairs, Files 1973-1974, John O. Marsh, Box 61, North Korea, Obtained for NKIDP by Charles Kraus. Available online: http://digitalarchive.wilsoncenter.org/document/114199.

217 significant exceptions discussed below, U.S. and North Korean failed efforts at dyadic diplomacy during the early 1970s fostered a mostly stable security climate on the Korean

Peninsula, which by 1975 gave way to renewed hostile rivalry dynamics.

Cheap Talk and Credibility before the Crisis

Toward the end of 1973, North Korea raised its military alert level and began sending armed ships into an area known as the Northern Limit Line (NLL), a contested unilaterally imposed maritime boundary established by the United Nations Command some decades prior.12

Although the NLL would become a flashpoint for future conflicts for decades and remains so today, December 1973 was the first time that North Korea directly pulled the United States into a showdown over the disputed area. In a December 1, 1973 meeting of the Military Armistice

Commission at Panmunjom, the North Korean representative condemned South Korea resupplying its positions on South Korean-administered islands within the NLL, claiming the action constituted an armistice violation and a war-inducing action.13 At that meeting, North

Korea explicitly threatened that it would take action to prevent any further such violations. This threat triggered a December 3 meeting of the Washington Special Actions Group (WSAG), a policymaking body that Kissinger established after the EC-121 crisis to take up highly sensitive foreign policy decisions, to decide whether the United States should allow the South Koreans to continue with a planned resupply shipment to leave port on December 5.14

12 It is not clear when the NLL was established; sources range from 1953 to 1965. For a discussion of the NLL as a flashpoint, see Michishita, North Korea’s Military-Diplomatic Campaigns, 52-72. 13 “The Korean Situation and the China Element,” December 3, 1973, NSC East Asian and Pacific Affairs Staff, box 36, Korean Northwest Islands (working file), Gerald R. Ford Presidential Library: Ann Arbor, MI. 14 U.S. Department of State, December 4, 1973. “Minutes of Washington Special Actions Group Meeting, Washington, December 4, 1973, 10:40am to 12:02pm,” FRUS, 1969-76, Vol. E-12, Part 1, Documents on East and Southeast Asia, 1973-1976, document 247: 1-4.

218 In the December 1 WSAG deliberations, the two factors on which the group most fixated were debating the importance of the issue at stake, and whether they should believe North

Korea’s threat. There was a group consensus that the United States did not seek to escalate the situation if North Korea seized or attacked the resupply ship, which implies that if they crossed

North Korea’s “red line” and North Korea implemented its threat of attacking the ship, the

United States would simply back down because the issue was not worth escalation to war. But the group was unable to decide whether it believed North Korea would actually follow through on its threat. North Korea clearly possessed the ability to do what it threatened, and North

Korea’s military movements in the area in the latter half of 1973 were unusual and aggressive enough that it made a North Korean attack seem more plausible, but North Korea had made many empty threats in the past.15 Ultimately the WSAG left it to Kissinger to decide whether to prevent the South Koreans from resupplying the islands, but the U.S. Ambassador to Seoul thought the South Koreans would not listen to a U.S. request anyway; Kissinger took no actions to prevent the December 5 or subsequent shipments, which amounted to the United States defying North Korea’s coercive threat by allowing the South Korean ship to perform its mission.16 North Korea continued to maintain an armed naval presence in the area, but it did not stop the December 5 shipment, waiting instead to attack a South Korean vessel there some months later, in February 1974.17 In the end, the dismissal of North Korea’s “red line” could be

15 Ibid. 16 Ibid. 17 The United States aired its grievances in response to this attack during the 348th Military Armistice Commission meeting. “348th Meeting of the United Nations Command Military Armistice Commission ,” February 28, 1974, History and Public Policy Program Digital Archive, United Nations Command, Military Armistice Commission. Available online: http://digitalarchive.wilsoncenter.org/document/114030.

219 equally well explained as bureaucratic inertia and Nixon administration indecisiveness as it could a judgment that North Korea’s threat was incredible. The result, in any case, was that North

Korea did not immediately implement its threat, as it said it would, but did implement it some months later.

North Korea replaced its charm offensives with the United States and with South Korea through the early 1970s with renewed open hostilities akin to the late 1960s, when it grew frustrated with the continued U.S. military presence in Korea, U.S. unresponsiveness to its proposals for a peace treaty, and its inability to stimulate domestic revolution in South Korea.

On February 24, 1976, the United States introduced F-111 bombers to the Peninsula in order to reinforce the U.S. military presence in Korea as a means of both assuring South Korea of the

U.S. commitment to its defense and optimizing U.S. military capabilities for executing a “short war” strategy if needed.18 Secretary of Defense James Schlesinger had promised the rotation of

F-111s to President Park in a visit to Seoul the August prior.19 Three days later, on February 27,

North Korea’s foreign ministry held a press conference where it declared that the F-111 deployment was proof that “The American Imperialists are the main instigators to the rise of tensions and the unleashing of a war of aggression in Korea…as long as the U.S. troops are stationed in South Korea, it is impossible to get rid of the state of tension and to achieve the

18 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.043,” February 28, 1976, History and Public Policy Program Digital Archive, Archives of the Romanian Ministry of Foreign Affairs, Matter 220 - Relations with the Democratic People's Republic of Korea, 1976, Obtained by Izador Urian and translated for NKIDP by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/114107. Starting in 1975, the USFK commander began working on an updated strategy for a Korea war that could be won in only nine days, though many in the Ford administration were dubious. See NSC Memorandum for General Scowcroft from Thomas J. Barnes, September 29, 1975, “Secretary Schlesinger’s Discussion in Seoul,” FOIA request, released on June 13, 1995, NARA. 19 U.S. Department of State, August 27, 1975, “Memorandum of Conversation, Seoul, August 27, 1975,” FRUS, 1969-76, Vol. E-12, Part 1, Documents on East and Southeast Asia, 1973-1976, document 271: 1-3.

220 unification of the homeland.”20 Despite North Korea’s public and direct threats in February, an

April 1976 CIA assessment of North Korean intentions argued that Kim Il Sung might be favoring renewing limited military attacks because Kim had personally “assessed lessons of the wars in Indochina and the Middle East,” and the North Korean belief that limited military action

“…could be controlled by the North, would not preclude continued Soviet and Chinese assistance, would provoke fear in the U.S. of another Korean War,” and would generate disagreement between the United States and South Korea over how to respond.21

On May 6, 1976, the United States tracked a North Korean merchant ship while it was conducting a military reconnaissance mission in international waters, leading North Korea to threaten that the U.S. reconnaissance flight was “a pretext for invoking a new war in Korea” and a “dangerous military adventure which may lead to an unforeseeable consequence.”22 Only weeks later, on May 25, the North Korean Central News Agency accused South Korea of firing a rifle at a North Korean guard post along the DMZ. Such incidents did occasionally occur, but rarely did they lead to public threats. North Korea called the incident a “grave provocation,” warning that “if the U.S. and South Korea persisted in such provocations, the sentries of the

Korean People’s Army will take a strong retaliatory measure and make the provocateurs pay

20 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.043,” February 28, 1976, History and Public Policy Program Digital Archive, Archives of the Romanian Ministry of Foreign Affairs, Matter 220 - Relations with the Democratic People's Republic of Korea, 1976, Obtained by Izador Urian and translated for NKIDP by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/114107. 21 Intelligence Memorandum, April 22, 1976, “Possible North Korean Initiatives During 1976,” National Security Adviser, Presidential Country Files for East Asia and the Pacific, box 10, Korea (15), Ford. 22 Information Memorandum, May 28, 1976, “North Koreans Charge ROK Provocations and Threaten Possible Retaliation,” National Security Adviser, Presidential Country Files for East Asia and the Pacific, box 10, Korea (16), Ford Library.

221 properly for this.”23 Two days later, a North Korean gun boat came within three nautical miles of the South Korea-controlled Northwest Islands—an extremely rare occurrence—nearly igniting an incident when a South Korean soldier wanted to fire a warning shot at the boat; the South

Korean soldier was prevented from firing by the U.S. commanding officer on the scene.24 This series of threats throughout May is notable because it signifies North Korea’s reversion to open hostility toward the United States and South Korea going into the Panmunjom axe murder incident in August. But the memo (to then National Security Adviser Brent Scowcroft from an

NSC staffer) documenting these threats is also notable because it attempts to estimate the credibility of these threats based principally on variation in North Korean threat-making irrespective of North Korean behavior. The memo mentions that the NSC consulted with the

State Department’s Bureau of Intelligence and Research about the novelty of North Korea’s threats and received affirmation that the specific threat formulations were “a few degrees more threatening than normal.”25 Learning from the previous major crises in the U.S.-North Korea rivalry, the NSC and State Department expected that North Korea’s rhetorical shift suggested that it might launch a military attack against a U.S. reconnaissance aircraft or vessel in the coming months.26 U.S. assessments of North Korean intentions also overlooked a June 1975 incident in the Joint Security Area (JSA)27 at Panmunjom in which, completely unprovoked, a group of North Korean soldiers beat an unarmed U.S. military officer in the middle of a Military

23 Ibid. 24 Ibid. 25 Ibid. 26 Ibid. 27 The JSA is a part of the DMZ in which—prior to the axe murder incident—North Korean and UN (U.S. and ROK) forces could freely intermingle. Small arms were permitted in the JSA, but a strong norm against the introduction of firearms existed, which encouraged personnel to resort to carrying clubs and axe handles.

222 Armistice Commission meeting so badly that he needed to be evacuated to the United States for surgery.

Other, more direct indications of North Korean warnings took place as well. On August

5, 1976, the North Korean Central News Agency broadcast an excerpt from a North Korean government white paper that was unique not only for quoting directly from a North Korean government document, but also in the specific threatening language it used. This broadcast claimed:

“The long-standing tensions in Korea have now reached an acute state as never before…war may break out at any moment…The United States and the South Korean authorities…have now finished war preparations and are going over to the adventurous machination to directly ignite the fuse of war…reminds one of the situation in 1950 when the United States provoked the Korean War. Particularly after it had been ignominiously defeated in and driven out of Indochina, the United States directed the spearhead of its Asian aggression to Korea.”28

The following day, August 6, U.S. and South Korean soldiers entered the JSA to trim a poplar tree that was blocking the line-of-sight of UN observation posts; a routine action. As dictated by precedent, the UNCMAC joint duty officer notified the KPA on July 28 that up to 150 UN personnel would enter the JSA for maintenance work, notified the KPA again on July 29 that the number of UN personnel would be increased, and on August 2 actually conducted a survey of the poplar tree.29 Although North Korea at no point protested this action, on August 6 the KPA guard asked what the United States and South Korea were doing and demanded they leave the

28 Korean Central News Agency radio broadcast, August 5, 1976. 29 Richard G. Head, Frisco W. Short, and Robert McFarlane, Robert, Crisis Resolution: Presidential Decision Making in the Mayaguez and Korean Confrontations (Boulder, CO: Westview Press, 1978), 156.

223 tree alone.30 On August 17, the day before the axe murder incident, North Korean Premier, Park

Song-chul, delivered a strident speech at the Nonaligned Nations Conference in Sri Lanka reiterating the August 5 broadcast, and urging the nonaligned movement to support condemnation of “imperialist maneuvers to provoke a war in Korea.”31 Kim Il Sung was scheduled to attend the conference, but wrote on August 15 that he would be unable to attend because of a “deteriorating situation on the Korean Peninsula.”32

Geopolitical Context

The geopolitical context cast a large shadow on the U.S.-North Korea rivalry leading up to1976. North Korea continued attempting to navigate the Sino-Soviet split with an independent foreign policy, but gradually seemed to move closer to China than the Soviet Union despite repeated assertions that it remained equidistant from both powers. North and South Korea made an attempt at rapprochement starting in 1970, which yielded a hard-negotiated 1972 joint statement announcing a shared desire for peaceful reunification that neither side seemed to value much. But by 1973 it was clear that the North-South rivalry was alive and well, and by 1975 it had become classically hostile. In Southeast Asia, the United States had fully withdrawn from

Vietnam by April 1975, leaving South Vietnam to be overtaken by Ho Chi Minh and the North

Vietnamese nationalists the following year. In May 1975, however, the Khmer Rouge regime in

Cambodia seized a U.S. merchant containership called the Mayaguez in what is considered the last official battle of the Vietnam conflict, despite taking place after the fall of Saigon. The

30 Ibid. 31 Jay Matthews, “North Korea at Summit Seeks Anti-U.S. Move, 3d World Pact,” Washington Post (August 18, 1976), A22. 32 Head, Short, and McFarlane, Crisis Resolution, 158-9.

224 swift, violent U.S. response to the Khmer Rouge’s seizure of the Mayaguez was motivated partly to preserve or restore the U.S. reputation for resolve generally after having walked away from

Vietnam. But the Mayaguez incident is also notable because scholars have identified it as signifying a short-lived period of greater risk acceptance in U.S. foreign policy under President

Gerald Ford.

Sino-U.S. rapprochement in the early 1970s played a significant role in North Korea’s external orientation because it coincided with a rekindling of the Sino-North Korean relationship as well. Following small military clashes along the Sino-North Korean border in early 1969, relations between China and North Korea quickly began to improve. As early as September

1969, North Korean officials traveled to Beijing for meetings to mend ties, with the head of the

North Korean delegation appearing beside China’s Mao Zedong reviewing a parade during the visit.33 A cable from the Romanian embassy at the time reported a North Korean diplomat as saying that it was restoring its historically positive relations with China.34 North Korea’s foreign minister, Park Sung-chol, visited Beijing in February 1970, during which he met with Chinese

Prime Minister Zhou Enlai and convinced him to visit North Korea in April 1970—the first international trip by a senior Chinese official since the Cultural Revolution began.35 Zhou’s trip to Pyongyang yielded a commission to study their common border area as a means of resolving dispute there, and a joint communique announcing communist solidarity.36 North Korea did not immediately render judgment about China’s decision to improve relations with the United States,

33 Bernd Schaefer, North Korean Adventurism and China’s Long Shadow, 1966-1972, Working Paper #44 (Washington, DC: CWIHP, 2004), 29. 34 Ibid. 35 Schaefer, North Korea’s Adventurism and China’s Long Shadow, 30. 36 Schaefer, North Korea’s Adventurism and China’s Long Shadow, 30-31.

225 and when Zhou Enlai traveled to Vietnam and North Korea to inform them, Zhou spent seven hours with Kim Il Sung discussing the matter.37 When North Korea did take a public stance on the Sino-U.S. rapprochement, it took the form of public, full-throated support, choosing to interpret the tectonic geopolitical shift as a U.S. defeat at the hands of communist superiority.38

Despite the improved relations between the North Koreans and Chinese, sources of tension remained. Mao Zedong and Kim Il Sung shared antipathy toward one another; North

Korea feared and loathed China’s cultural revolution; and Chinese officials were disgusted with

North Korea’s philosophy of Juche.39 In conversations with other communist officials from

Eastern Europe, Kim Il Sung took pains to emphasize that it charted an independent foreign policy from the Soviet Union and Beijing, but also that global perceptions of North Korean intimacy with the Chinese was not accurate. In a July 1973 conversation between Kim Il Sung and a visiting Polish delegation, Kim stated that “The DPRK does not allow on its territory any insults against either the USSR or the PR China,” adding that improving North Korean relations with the Eastern Bloc was one way of improving ties with the Soviet Union, and that “these relations can only develop in balanced proportion to relations with the PR China.”40 Eastern

German diplomats reported the same theme in a cable describing Kim Il Sung’s 1975 visit to

Beijing. China’s Deng Xiaoping in particular pressed Kim Il Sung to denounce the Soviet

37 Schaefer, North Korea’s Adventurism and China’s Long Shadow, 33-34;Chen Jian, Mao’s China and the Cold War (Chapel Hill, NC: University of North Carolina Press, 2001). 38 Jian, Mao’s China and the Cold War, 273-5. 39 Juche roughly translates to “self-reliance”; it seemed to permeate many aspects of North Korean politics and foreign policy in the form of a stubborn independence and unwillingness to listen to the entreaties of others. Schaefer, North Korea’s Adventurism and China’s Long Shadow, 30-31. 40 James Person, Limits of the “Lips and Teeth” Alliance: New Evidence on Sino-DPRK Relations, 1955-1984, NKIDP Document Reader (Washington, DC: Woodrow Wilson International Center for Scholars, 2009), 55-6.

226 Union, but Kim refused.41 And although the visit produced a joint communique, it did “not contain any concrete results for bilateral relations,” and failed to address many continuing points of disagreement between the Chinese and North Koreans, chief among them being North Korea’s preference for a hostile policy toward the United States and South Korea.42 But all the latent sources of hostility and disagreement between China and North Korea during this period did remain latent; only a month prior to the Panmunjom crisis, on July 10, 1976, Mao Zedong sent a celebratory message to Kim Il Sung on the 15th anniversary of the Sino-North Korean alliance, which coincided with an editorial in the People’s Daily pledging support for North Korea’s stance against “U.S. imperialist aggression.”43

The five-year period in North and South Korean relations preceding the axe murder incident at Panmunjom was unprecedented, for better and for worse. North Korean provocations against both the United States and South Korea decreased drastically from 1970 through 1975, compared with the five-year period preceding it. In 1970, South Korea’s Park Chung-hee approved his government to secretly reach out to North Korea to improve relations by negotiating reunion visits between Korean families that had been split between North and South

Korea because of either the post-World War II agreement between the United States and the

Soviet Union, or because of familial displacement during the Korean War. These secret talks were far from smooth, with frequent disagreements about everything from meeting agenda to the wording of documents, but by 1972 managed to yield the first joint communique of North and

41 Ria Chae, East German Documents on Kim Il Sung’s Trip to Beijing in 1975, NKIDP Document Reader (Washington, DC: Woodrow Wilson International Center for Scholars, 2012), 4-6. 42 Ibid. 43 Head, Short, and McFarlane, Crisis Resolution, 154.

227 South Korea, agreeing to common principles of autonomy, peaceful unification, and national unity.44 It was a landmark document in the history of North-South relations, and would become a touchstone for future jointly negotiated statements of intent toward one another in subsequent decades, but neither side made any meaningful changes in their posture toward one another as a result of the document at the time.

Indicative of how little had changed in the intentions of each side before and after the joint communique was the series of maritime incidents starting in 1973, most of which (but not all) were initiated by North Korea in the area surrounding the NLL. On October 23, 1973, a

North Korean gunboat and torpedo boat crossed the NLL in the first of more than 40 such assertions between October and December 1973.45 On December 1, North Korea took the occasion of the MAC meeting to convey that it rejected South Korean and UNC claims regarding the NLL and Northwest Islands, issuing an explicit threat that South Korea was not to continue resupplying the Northwest Islands. Some three months later, in February 1974, North Korea attacked two South Korean fishing boats in the vicinity of the NLL, successfully sinking one of them. In June the same year, North Korea would attack and sink a South Korean maritime police patrol boat in the area of the NLL, killing 26. On August 15, North Korea attempted to assassinate Park Chung-hee, killing his wife instead. In November, South Korean DMZ patrol units found the first of several North Korean underground tunnels running from North to South

Korea, some large enough to drive tanks through. Because the tunnels were estimated to be no more than two years old, their mere existence implied that North Korea may have been planning

44 “The Rise and Fall of Détente on the Korean Peninsula.” 45 James Person, After Détente: The Korean Peninsula, 1973-1976, Critical Oral History Conference Briefing Book (Washington, DC: CWIHP, 2011), vi.

228 a renewed invasion of South Korea, even as the two governments made the unprecedented move toward rapprochement in the 1972 joint communique.

Within South Korea too, we find circumstances that inform the strategic context leading up to the Panmunjom axe murders. As student protests and political oppositionists in South

Korea grew louder and more frequent, Park Chung-hee felt that general order and his continued rule were threatened. The world was aware that South Korea was run by an autocrat who took power by military coup; the United States did not like this, but accepted it throughout the 1960s.

In October 1972, however Park dissolved the National Assembly, declared martial law, closed universities across the country, and began arresting political figures who stood in opposition to his regime.46 Through coercive violence and political oppression against his own people, Park intensified his dictatorial tendencies in what became a dark road on South Korea’s path toward democratization. On December 13, 1972, Park created a new constitution called Yushin

(restoration), which ended the martial law he declared in the months prior, but endowed him with effectively total control over South Korean politics and governance.47 Park’s domestic human rights abuses led to congressional investigations in the United States, calls from legislators to suspend U.S. military aid to South Korea until human rights were restored, and a discourse in the

United States debating whether the United States should or needed to have military forces in

South Korea. Against this backdrop, then presidential candidate Jimmy Carter was calling for the removal of troops from South Korea. U.S. discourse, combined with Nixon’s removal of

46 Byung-gook and Ezra Vogel, The Park Chung Hee Era: The Transformation of South Korea (Boston: Harvard University Press, 2011); Bernd Schaefer, Overconfidence Shattered: North Korean Unification Policy, 1971-1975, NKIDP Working Paper #2 (Washington, DC: Woodrow Wilson International Center for Scholars, 2010). 47 Kim and Vogel, The Park Chung Hee Era; Gregg Brazinsky, Nation Building in South Korea: Koreans, Americans, and the Making of a Democracy (Chapel Hill, NC: University of North Carolina Press, 2007).

229 some 20,000 troops from Korea in 1970 and 1971, contributed to a tremendous concern among

South Korean elites that the United States was not a reliable ally. It is in this context that on

December 11, 1974, the U.S. embassy in Seoul cabled Washington for the first time to inform them that South Korea was potentially developing a clandestine nuclear program.

One of the most important factors informing the strategic context, especially for U.S. decision-making, was the evolution of U.S. military activity in Southeast Asia. After a decade of increasingly unpopular war, the United States finally withdrew from Vietnam. On April 17,

Phnom Penh was taken over by the communist Khmer Rouge, and on April 30 Saigon fell to the

North Vietnamese army. For some, particularly U.S. proponents of the war in Vietnam, there was a prevailing feeling of U.S. fecklessness; it set out to do something that it ultimately could not. There were corresponding concerns that the United States might be perceived as an unreliable ally after walking away from Vietnam. According to then President Gerald Ford, this context of battered prestige was relevant in decision-making during the events from May 12 through May 15 surrounding a U.S. ship called the Mayaguez.

On May 12, the U.S.-manned containership, Mayaguez, was seized by Cambodian communist forces in international waters. The communist takeover of Cambodia came with new assertions of sovereignty, including a claimed extension of territorial waters to 90 miles beyond their border. Cambodian forces enforced this internationally unrecognized sovereignty claim accordingly, seizing Vietnamese ships and Thai fishing boats, firing on a South Korean ship, and

230 detaining a Panamanian ship, all in international waters that it claimed for itself.48 When the

Mayaguez transited the area, it too was seized, along with its 40-man crew. Over the next 96 hours, Gerald Ford held four NSC meetings to deliberate on what to do and formulate an action plan. On later reflection, Ford stated, “My feeling was that seizure of a U.S. vessel and crew, especially by a country which had so recently humiliated us, was a very serious matter.”49 Ford would also admit that launching a military assault, while intended to “protect U.S. interests…Perhaps it was also a carryover from the evacuations of Saigon and Phnom Penh.50 In a bloody but quick operation approved by President Ford, the United States launched an amphibious assault on a Cambodian-claimed island, carrier-based air strikes against Cambodian military targets, including a nearby port, and a boarding party of marines that would seize the

Mayaguez and rescue the crew. The outcome was bloody but successful. 10 of 11 U.S. Air

Force helicopters were destroyed or suffered battle damage; 15 U.S. servicemen were killed in the operation; three U.S. servicemen were missing; and 50 U.S. servicemen were wounded by hostile fire. 51 But by May 15, the Mayaguez was successfully secured by U.S. Marines, and the ship’s crew was released unharmed because of U.S. coercion, not purely superior operational force; the Khmer Rouge released the crew in order to seek an end to the U.S. military operation.

During the NSC meetings leading up to the Cambodian operation, “The Pueblo seizure was on everybody’s mind, and so were the most recent North Korean provocations, particularly the

48 The story of the Khmer Rouge’s ascendance to power in Cambodia and its behavior in 1976 can be found in Kenneth M. Quinn, “Cambodia 1976: Internal Consolidation and External Expansion,” Asian Survey 17, no. 1 (1977): 43-54. 49 Interview with President Gerald Ford, as quoted in Head, Short, and McFarlane, Crisis Resolution, 107. 50 Interview with President Gerald Ford, as quoted in Head, Short, and McFarlane, Crisis Resolution, 109. 51 Head, Short, and McFarlane, Crisis Resolution, 139-41.

231 tunneling under the thirty-eighth parallel, the border between the North and South.”52 The analogical learning from the Pueblo experience led Kissinger to argue that any U.S. actions needed “to be read clearly by the North Koreans as well as by the Cambodians,” though some others, notably Secretary of State James Schlesinger, disagreed with the need to set an example for Asia.53

Overview of the Incident

On August 18, 1976, North Korea committed an act of violence in the DMZ area with a lethality not unlike the thousands of attacks it had initiated in the DMZ area over the prior decade. It involved a planned, deliberate melee in which North Korean soldiers bludgeoned to death two U.S. soldiers and injured several more. The proximate trigger for the attack was the

UN need to trim a poplar tree in the JSA of the DMZ because it was shrouding the visibility of key observation posts. The UN work team included five U.S. security forces, four South Korean security forces, a South Korean interpreter, a South Korean foreman, and a four-man U.S. work team. Lt Col Victor Vierra, the man in charge of UN personnel in the JSA, was aware of an increased frequency in North Korean provocations since February 1976, and had received reports of North Korea’s August 5 broadcast full of threatening language.54 Lt Col Vierra had also been briefed on North Korea’s August 6 protest against trimming the poplar tree, but dismissed it as

“no more than usual” harassment, and told the supervising officer—Captain Arthur Bonifas— that if he was harassed he should continue work, but should withdraw if violence seemed

52 Roy Rowen, The Four Days of Mayaguez (New York: Norton, 1975), 141-2. 53 Ibid. 54 Head, Short, and McFarlane, Crisis Resolution, 161.

232 imminent.55 Indicating that the United States was not expecting any trouble during the tree trimming, civilian tours of the JSA continued normal operations on the morning of August 18.56

Firsthand accounts of the incident reported that as soon as the UN work team arrived on the scene, two North Korean officers and nine North Korean guards approached; the leading officer was one of the attackers in the 1975 beating of a U.S. army officer.57 The South Korean interpreter in the UN team informed the North Korean group why they were trimming the tree, to which the lead North Korean officer replied, “Good.” Approximately 20 minutes later, when the poplar trimming was nearly finished, a North Korean officer told the UN detail to stop. After a brief discussion, the North Korean officer said “If you cut more branches, there will be a big problem.” Captain Bonifas heard the interpreter, but told the detail to continue their work, and some 10 minutes later the North Korean side numbered between 28 and 30 guards. The North

Korean officer then repeated himself in more strident terms, stating, “The branches that are cut will be of no use, just as you will be after you die.” The comment was interpreted for Captain

Bonifas, who dismissed it. Repeating himself even more directly, the North Korean officer said that “If you cut more branches, you are going to die.” The North Korean officer removed his watch, and another rolled up his sleeves. One of the officers then yelled “kill,” at which point the North Koreans dealt Captain Bonifas multiple blows as the other North Korean guards descended on the scene with pipes and clubs. The South Korean members of the UN team were

55 Ibid. 56 Head, Short, and McFarlane, Crisis Resolution, 163. 57 Detailed overviews of this incident can be found in Head, Short, and McFarlane, Crisis Resolution; Michishita, North Korea’s Military-Diplomatic Campaigns; Wayne A. Kirkbride, DMZ: A Story of the Panmunjom Axe Murder (Seoul: Hollym, 1984); Hee-do Park, Dora Ojiahneun Dari-ae Seoda (standing at the Bridge of No Return) (Seoul: Saemteo, 1988).

233 mostly left alone, and according to a North Korean defector, Kim Jong-Il, Kim Il Sung’s son and heir apparent at the time, directed that the incident directly target the Americans.58

President Ford would be briefed on the incident at 9am on August 18 (Washington time), and responded with disgust as he directed that a WSAG meet to formulate a response. The most uncertain part of the crisis would last until August 25, though aspects of the crisis would continue through September 6, when the Military Armistice Commission concluded negotiations that would allow both sides to stand down from their crisis military posture. In the remainder of this section, I will review the thinking within the U.S. policy process that led to marshaling a massive and sustained show of force against North Korea in support of U.S. defiance of North

Korea’s demand not to trim the tree by chopping the tree down altogether. Importantly, this case will offer evidence of the North Korean perspective on U.S. credibility more clearly than in the previous cases.

Debating Retaliation

In addition to directing a WSAG meeting, President Ford directed formulation of response options and asked to be kept up to date. Concurrent with attempts to piece together the details of what happened and why, the Joint Staff drew on existing operational plans for Korea, some developed in response to the EC-121 shoot down, as the foundation for response options.

Consensus quickly formed around the need to take firm military action as a general principle, but

Ford considered a range of options, from doing nothing to a fully retaliatory military response.

58 Chang-hyun Jung, Kyeot-aeseo Bon Kim Jeong-Il (“Seeing Kim Jong-Il from beside him”) (Seoul: Kimyeongsa, 200), 202-4.

234 Prior to the WSAG meeting on August 18, Kissinger had a phone conversation with

Phillip Habib, an Undersecretary of State who was U.S. ambassador to Korea during the EC-121 crisis. Kissinger stated unequivocally, “I want retaliation….We cannot have Americans killed. I hope that is clear.”59 When Kissinger asked what could be done, Habib said they could respond with direct retaliation in the DMZ area, or attack a North Korean fishing vessel, but that South

Korean fishing vessels were more numerous and thus more vulnerable than North Korean vessels, and that North Korea’s camps in the DMZ were heavily fortified.60

Kissinger ran the WSAG, and was the source of most ideas and most conclusions during the August 18 meeting. After griping about the failure of CIA, DoD, and the State Department to notify him about the incident in a timely manner, Kissinger immediately weighed the local military balance on the Korean Peninsula, inquiring about how well North Korean forces could perform against South Korea, as well as against the combined alliance.61 The Chairman of the

Joint Chiefs of Staff, Adm. Holloway, informed him that North and South Korea were about even: North Korea had superior numbers, but South Korea had superior quality. Kissinger raised that he spoke with President Ford about the incident and that the President “…feels that some sort of strong action is necessary but does not know precisely what it should be.”62 The group

59 U.S. Department of State, August 18, 1976. “Transcript of Telephone Conversation Between Secretary of state Kissinger and Undersecretary of State for Political Affairs (Habib), Washington, August 18, 1976, 10:05am” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 281. 60 Ibid. 61 U.S. Department of State, August 18, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 18, 1976, 3:47pm,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 282; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford. 62 Ibid.

235 agreed that moving from DEFCON 4 to DEFCON 363 would be a good idea, and Kissinger proposed resuming a B-52 exercise in Korea that he had previously cancelled because he thought it would be too provocative to the Chinese. William Clements noted that they could use the B-

52s to drop live explosives at a bombing range near the North Korean border, which Kissinger approved of on the ground that it would “…be a good lesson for them.”64 When the CIA representative to the meeting asked how the media and U.S. public might react to a change in the

DEFCON level, Kissinger replied, “That has nothing to do with it. The important thing is that they beat two Americans to death and must pay the price.”65 By the end of the meeting,

Kissinger noted “That the President wants to explore the possibility of taking one military step,” urging the group to think about it overnight.66 A WSAG consensus supported not only moving from DEFCON 4 to DEFCON 3 and a live B-52 exercise, but also altogether chopping down the offending poplar tree at issue despite North Korea’s violence-backed threats to leave the tree alone. Kissinger had additionally raised deploying F-111s and F-4 fighter aircraft, and moving an aircraft carrier to the Yellow Sea.

At the WSAG meeting the next day, several decisions were agreed to propose for presidential approval. Over the course of the previous night, acting on the WSAG meeting, F-4s flew into Korea from Japan, and the DEFCON level changed from 4 to 3. The group agreed out of the August 19 WSAG meeting to request presidential approval to launch the B-52 exercise, and to remove the poplar tree, ideally in such a way that it did not precipitate war. The group

63 DEFCON 3 is the most urgent state of alert that U.S. forces can adopt without already being committed to war. At DEFCON 2, war is already inevitable. DEFCON 1 represents an actual state of open war. 64 Ibid. 65 Ibid. 66 Ibid.

236 also agreed to begin moving the carrier task force, led by the USS Midway, to initiate deployment of 18 F-111s from Idaho, and to “develop a contingency plan for hitting the North

Korean barracks near the JSA.”67 The group discussed the balance of capabilities in Korea once again, but abandoned it as a basis for decision-making because, as Kissinger noted, there are many examples from history when the smaller side (South Korea) can win because how forces were used mattered more than how many forces were used.68 Agreeing with CIA assessments,

Kissinger said there was no chance North Korea intended to fight a war, arguing “If they had wanted to launch an attack they would not have beat two of our men to death.”69 Kissinger then asked what North Korea’s aircraft alert response was in the EC-121 shoot down, and George

H.W. Bush—then the CIA Director—said there was no such alert during the EC-121 shoot down, which Kissinger interpreted as further proof that North Korea initiated the incident deliberately, because North Korea’s initiation of a military alert took place quickly after the U.S. men were beaten.70 The group also debated the merits of striking North Korean targets along the

DMZ with either artillery or aircraft, with Kissinger preferring striking the barracks of the North

Korean soldiers who initiated the attack because of its relevance and proportionality to the incident. The Deputy Secretary of Defense, Bill Clements, voiced concern that North Korea would escalate if we attacked, and that any attack by aircraft would make U.S. airbases in South

Korea subject to targeting. Undersecretary Habib noted that if the United States moved in a team

67 U.S. Department of State, August 19, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 19, 1976, 8:12-9:15am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 285; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford. 68 Ibid. 69 Ibid. 70 Ibid.

237 to chop down the tree, the North Koreans would either acquiesce or escalate, but proportionate reaction was not their historical precedent. Kissinger reminded the group that “The purpose of doing something is to show that we are ready to take risks…One always assumes the unlimited willingness of opponents to take risks. The purpose of this exercise is to overawe them. We are

200 million people and they are 16 million.”71

The next day, August 20, Kissinger appeared in a television interview talking about the

U.S. view of the attack: “We cannot permit the principle to be established that Americans can simply be assaulted with impunity any place that some dictator of some aggressive country decides to score some points.”72 He answered a question about whether the public should be alarmed with “It depends on what it indicates about North Korean intentions…whether, we get any satisfactory response to our demands for explanations and reparation.”73 That same day,

General Scowcroft and Kissinger briefed President Ford on recommendations from the WSAG and General Stilwell’s operational plan for chopping down the poplar tree with reinforcements and backups, Operation Paul Bunyan. By this time, F-111s were on their way from Idaho, F-4s had already been flown in from Japan, and the carrier-led naval task force was moving in the direction of the Yellow Sea. Briefing Ford on these details, Kissinger recommended approval of the live B-52 bomber training missions, flying them in from Guam, as well as a retaliatory strike against the North Korean barracks that housed the soldiers who attacked the U.S. soldiers. Ford overruled the recommendation to strike the barracks because he believed North Korea’s military alert was likely defensive in nature, but all the other aspects of the operation were approved,

71 Ibid. 72 “Axe-Murders at Panmunjom,” Korea Herald (August 20, 1976): 23-4. 73 Ibid.

238 including conducting the tree-cutting operation without prior notice to North Korea.74 His hope that the large show of force combined with a specific deterrent threat not to intervene in the tree- cutting operation would strike the right balance between escalation control and demonstrating resolve: “…to gamble with overkill might broaden very quickly into a full military conflict, but responding with an appropriate amount of force would be effective in demonstrating U.S. resolve.”75 But even after Ford vetoed the WSAG recommendation to strike North Korean barracks, during the August 25 WSAG meeting Kissinger and the rest of the group agreed to pursuing a secret plan involving a clandestine infiltration, to be carried out on November 1, along one of North Korea’s coasts to destroy a North Korean target near the DMZ, though for unspecified reasons that plan was never executed (presumably because it became unnecessary).76

With Ford’s approval, Operation Paul Bunyan commenced the followed day, August 21, cutting down the tree with an immense show of force, and without any North Korean direct response or retaliation. Right after the U.S. and ROK forces cut down the tree, Kim Il Sung personally issued an unprecedented but highly equivocal apology for its most recent provocation, leaving the United States in the position of determining what to do next. General Stilwell found

Kim’s message wholly insufficient to prevent further U.S. action, but the State Department recognized Kim’s message as at least somewhat conciliatory. In spite of disagreement about the

74 U.S. Department of State, August 25, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 25, 1976, 10:30am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 286; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford. 75 Interview with President Gerald Ford, as quoted in Head, Short, and McFarlane, Crisis Resolution, 193. 76 U.S. Department of State, August 25, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 25, 1976, 10:30am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 286; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford Library.

239 sufficiency of the apology, everyone on the U.S. side recognized that Kim’s message only met at most two of the three criteria that the United States had conveyed in the August 19 MAC meeting; there was no mention of punishing the men responsible for the attack.77 Because Kim’s message had made the desirable suggestion that in the future UN and North Korean forces in the

JSA should be separated “so that such incidents may not recur in the future,” there was a question about the modalities for establishing such a separation rule.78 So both sides agreed to meet again through a MAC meeting on August 25 to negotiate terms for, in essence, drawing a line within a line—how to separate North Korean and UN forces in the JSA, located within the

DMZ area already separating North and South Korea.

For once U.S. officials felt the crisis atmosphere on the Korean Peninsula seemed to be favoring the United States, and hoped that continuing its military deployment and large-scale series of exercises might help pressure North Korea into conceding terms favorable to the United

States in negotiating the permanent separation of North Korean and UN troops in the JSA:

“None of us wanted to have the media call off the crisis and have the U.S. lose the initiative with

North Korea,” said one U.S. official.79 Having secured an apology from Kim Il Sung, the United

States also needed North Korea to agree to withdraw its presence from the South Korean side of the MDL in Panmunjom, and had also demanded punishment of the North Korean soldiers responsible for the incident. There seemed to be a presumption among U.S. decision-makers that continued coercive signaling would help the United States get what it wanted out of the continuing UNCMAC negotiations.

77 Head, Short, and McFarlane, Crisis Resolution, 200-1. 78 Michishita, North Korea’s Military-Diplomatic Campaigns, 80-1. 79 As quoted in Head, Short, and McFarlane, Crisis Resolution, 199.

240 So the United States and South Korea maintained DEFCON 3, the supplemental F-4s and

F-111s, B-52 training, and the carrier-led naval task force as the UNCMAC convened several more times between August 25 and September 6, when a deal was finally reached that committed each side to guarantee the security of the other within the JSA, and to remaining on their respective sides of the Military Demarcation Line (MDL) within the JSA, including the unilateral removal of four North Korean guard posts on the UN side of the MDL.80 On September 7, having secured all but punishment of the North Korean soldiers responsible for the attack, the

United States and South Korea returned to DEFCON 4, reduced B-52 exercises to a monthly basis, and the supplemental naval vessels and aircraft brought in for deterrent signaling left the area of the Korean Peninsula. An internal NSC memo described the outcome as being on “very favorable terms” for the United States, as well as a “loss of face” and setback for North Korea.81

Despite public solidarity between the United States and South Korea over the incident, however, behind closed doors President Park was dismayed by the U.S. handling of the incident, interpreting the U.S. “failure” to retaliate against North Korea with reprisal strikes of some kind as the United States making “concessions to Pyongyang.”82 There is no evidence to suggest that

North Korea withdrew from the South Korean side of the MDL because of the continued U.S. coercive signaling after it won an apology from Kim Il Sung.

Orchestrating U.S. Actions

80 Michishita, North Korea’s Military-Diplomatic Campaigns, 81; U.S. Department of State, September 5, 1976. “Memorandum from the Deputy Assistant to the President for National Security Affairs (Hyland) to President Ford, Washington, September 5, 1976,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 288; National Security Adviser, Presidential Country Files for East Asia and the Pacific, Korea (19), box 10, Ford. 81 Ibid. 82 U.S. Department of State, August 27, 1976. “Telegram 213541 From the Department of State to the Embassy in the Republic of Korea, August 27, 1976, 2010Z,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969- 76, document 287.

241 The outcome in this case benefited the United States in relative dyadic terms, but it did little to reassure its South Korean ally. Regardless, the orchestration of U.S. word and deed was not only imperfect, but at some points it was externally obvious that the U.S. government was not a unitary actor. Categories of U.S. action in this case break down into on-the-ground reaction to the axe murders as they occurred; military actions authorized by the President or the

NSC; and diplomacy conducted mostly through UNCMAC, but also with some messaging delivered to the UN and to the Chinese, neither of which played a meaningful role in the crisis.

As discussed in the next sub-section, even though these events represent a rare case in which

U.S. signaling actually was perceived by North Korea as intended, U.S. action in this case was nevertheless poorly orchestrated at several points.

During the first President Ford-directed WSAG meeting on August 18, Kissinger expressed disappointment not only about not being notified about the incident until 12 hours after it had occurred, but also about how such an incident was allowed to happen in the first place. The operations centers at each agency were talking to one another, but none passed the information on to principals in a timely manner.83 Reaction teams of security forces were often employed to protect UN personnel in the JSA during potentially vulnerable missions, and reaction forces were available for the tree-trimming operation too. Kissinger could not understand why the reaction team did not act to protect the victims of the attack; in keeping with normal UN practice in the JSA, a duty platoon was posted 660 yards from the poplar tree, and a

83 U.S. Department of State, August 18, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 18, 1976, 3:47pm,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 282; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford Library.

242 quick reaction force was on call approximately 680 yards from the tree.84 Both sets of backup forces were monitoring radio communications of the team trimming the tree. Several North

Korean drop gates dispersed throughout the JSA were blocking road access, which slowed backup forces from arriving at the scene of the attack until after the fighting was over; the backup forces did not arrive on the scene until 11:10am local time, but fighting had ended by

11:07.85

A number of more deliberate military moves took place after the immediate attack ended.

On August 19 in Korea, the day after the first WSAG meeting had approved them, 18 F-4s flew into Korea from Okinawa, Japan, and UN forces moved from DEFCON 4 to DEFCON 3.

Several battalions from the U.S. Third Marine Division also deployed from Okinawa to South

Korea. Mobilization of other military assets that the WSAG had approved would wait until

President Ford’s approval on August 20. With Ford approving all recommendations but reprisal strikes against North Korean barracks, August 20 saw the deployment of the F-111s from Idaho, the carrier-led naval task force (which included a destroyer and four frigates), and three B-52 bombers from Guam for live bombing training in South Korea, relatively close to the DMZ. The first of the B-52 bombers would drop live ordnance for its training shortly before Operation Paul

Bunyan—the tree-chopping operation—began in earnest. As President Ford directed, North

Korea received no advance notice about either the bomber training or Operation Paul Bunyan.

The force mobilized for the operation was fierce.

84 Head, Short, and McFarlane, Crisis Resolution, 159. 85 Ibid.

243 Whereas the original tree-trimming team on August 18 consisted of 15 men, the tree cutting team on August 21 included 110 combined U.S.-ROK forces directly involved in the operation with readied artillery support, 20 utility helicopters, and seven attack helicopters. The report of a U.S. enlisted man who participated in the operation described the tensest moment he ever experienced in Korea, with a shared feeling among UN forces that war actually could break out. As part of the tree-cutting force, several of the ROK Marines acted as if filled with bloodlust, eager to die: “A few of the ROK marines with us unbutton their shirts, showing that they have claymore mines strapped to their chests and they have the clacker (firing mechanism) in their hands. They start yelling and waving at the KPA to come on over.”86 That sentiment permeated the South Korean chain of command; its defense minister, Suh Jeon-chul, delivered a speech from President Park to a group of new minted military officers in Seoul on August 20 stating, “We need a club to deal with a mad dog…Should they dare commit an illegal provocation again, large or small, an immediate punitive action will be taken…There is a limit to our patience.”87 President Park had assigned his most proficient Taekwondo experts to participate in the tree-cutting operation, since 1953 Armistice Agreement limited the number of armed men that could enter the JSA. As the UN team entered the JSA, they removed the drop gates that North Korea had previously emplaced, some of which had hindered the response time of the UN quick reaction force during the August 18 attack. Nothing prohibited the North

Koreans from setting up gates in the JSA, but they became illegal when they prevented UN freedom of movement, as on August 18. North Korea pulled its guards away from the area of the

86 Bill Ferguson, date unknown, Memories of JSA, Sept ’75 – Oct ’76. Available online: http://bill_ferguson.tripod.com/JSA_Memories.htm. 87 “Axe-Murders at Panmunjom,” Korea Herald (August 20, 1976): 24.

244 tree-cutting, watching it take place through binoculars to ensure noninterference. The entire operation, which started at 7am on August 21, was complete by 8:30am. The operation was designed to start before the North Korean guards assigned to the JSA area would take their posts

(normally at 7:30am) but not end until after the North Korean guards should have taken their posts.

U.S. diplomatic initiatives in this case included: condemnatory statements delivered the day after the incident to the KPA from the U.S. head of UNCMAC; public statements by the

State Department, Kissinger, and the White House; the U.S. case made to the UN Security

Council, Kissinger’s discussion with the Chinese, and U.S. consultations with South Korean and

Japanese allies; and U.S. negotiations with North Korea through the MAC. Most of these diplomatic initiatives were discussed in the previous section so I will only cursorily review them here.

As soon as the attack occurred, the U.S. representative to the UN Command responsible for the MAC, Rear Admiral Frudden, met with his KPA counterparts in a MAC meeting on

August 19 at 4pm. Admiral Frudden used this meeting to convey the gravity with which the

United States viewed the situation, and issued the three U.S. demands: North Korean acknowledgment that it was responsible for the attack; assurances that the security of UN forces in the JSA would be guaranteed in the future; and punishment for the offending North Korean soldiers.88 Publicly, the United States took an aggrieved, somewhat belligerent tone at the outset.

As the crisis proceeded, mixed messaging would come out of the United States, and even from within the Executive Branch alone, though mixed messaging seemed to affect North Korea

88 As quoted in Head, Short, and McFarlane, Crisis Resolution, 187.

245 little.89 General Stilwell had also drafted an angry letter directly to Kim Il Sung—the former in his role as commander of UN Command—which Kissinger demanded not be sent: “We are not going to let Stilwell run loose. We are not going to let him act like MacArthur. We could have cut him out completely and insured the whole thing be handled by the DCM [deputy chief of mission].”90 During Kissinger’s television interview on August 20, he publicly reiterated the three demands of North Korea that Admiral Frudden conveyed in the August 19 MAC meeting.91

Shortly after receiving Kim Il Sung’s apology statement the State Department’s press office gave a verbal response rejecting Kim’s statement as a form of contrition: “The United States does not find acceptable a North Korean statement indirectly expressing regret for the killing of two

American officers...We consider this a back-handed acknowledgement…We’re very skeptical about this message. We do not intend to lower our guard any, nor fall for any propaganda ploys.”92 Kissinger did not clear this specific message.93 The following day, however, Kissinger directed a press conference to offer a more balanced interpretation of Kim Il Sung’s statement, after seeing a negative spin on the State Department’s initial comments reflected in U.S. media.94

Other U.S. diplomatic initiatives outside of the U.S.-North Korea dyad, which had little bearing on the outcomes in this case, included circulation of a statement within the United

89 As quoted in Head, Short, and McFarlane, Crisis Resolution, 176-7; White House Statement, August 18, 1976. Weekly Compilation of Presidential Documents 12, no. 34: 1263. 90 U.S. Department of State, August 19, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 19, 1976, 8:12-9:15am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 285; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford. 91 “Axe-Murders at Panmunjom,” Korea Herald (August 20, 1976): 22. 92 Quotation from the Associated Press in “Axe-Murders at Panmunjom”: 27. 93 Head, Short, and McFarlane, Crisis Resolution, 199. 94 Murray Marder, “State Department Reverses Stand on N. Korea’s Regrets,” Washington Post (August 24, 1976): A10.

246 Nations, Kissinger’s consultations with the Chinese, and ally consultations with the Koreans and

Japanese. At the United Nations, the United States chose not to seek a condemnatory resolution of North Korean behavior, but rather to convey a low-key protest and socialize its version of events to the UN Security Council via a written report, in factual language; even Soviet-aligned representatives at the United Nations were surprised the United States did not deliver inflammatory speeches or seek UN resolutions rebuking North Korea.95 With the Chinese,

Kissinger took the approach of speaking to a patron about its client, complaining that the Chinese were not doing enough to rein in North Korean behavior. The Chinese at once attempted to shift blame for the incident by pointing out all the ways in which the United States could have been more prepared for such an incident while at the same time explaining that they hold little sway over North Koran behavior.96 Still, declassified Romanian archives would later reveal that the

Chinese privately believed that North Korea had made a miscalculation from which they gained nothing politically or diplomatically.97 Although Kissinger would continue discussions with the

Chinese intermittently throughout the crisis, it was clear from their initial consultation that they either would not or could not be helpful in arresting North Korean provocations of this type.

Consultations with the South Koreans and Japanese were mostly conducted locally, though some

95 “Telegram from New York to Bucharest, SECRET, Urgent, No. 060.387,” August 20, 1976, Archives of the Romanian Ministry of Foreign Affairs, Matter 220-Relations with the Democratic People’s Republic of Korea, 1976, Obtained by Izador Urian and translated by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/114115. 96 U.S. Department of State, August 19, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 19, 1976, 8:12-9:15am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 285; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford Library. 97 “Telegram from Beijing to Bucharest, SECRET, Urgent, No. 066.252,” August 25, 1976, Archives of the Romanian Ministry of Foreign Affairs, Matter 200-Relations with the Democratic People’s Republic of Korea, 1976, Obtained by Izador Urian and translated by Eliza Gheorghe: http://digitalarchive.wilsoncenter.org/document/114117.

247 consultations with the latter took place between Kissinger and the Japanese ambassador in

Washington, and with Japan’s representative to the United Nations in New York. The State

Department provided official notifications to the Japanese the day of the incident, in particular to preserve Japanese support for the deployment of the USS Midway, U.S. marines, and F-4s from

Okinawa.98

Kissinger wanted the Deputy Chief of Mission at the U.S. embassy in Seoul and General

Stilwell to jointly maintain communication channels with the ROK Ministry of National Defense and President Park, but General Stillwell conveyed U.S. positions without embassy representatives present, much to Kissinger’s rage.99 In Stilwell’s initial cable back to

Washington characterizing his meeting with President Park, he described Park as calm, professional, and generally supportive, though Park also pointed out that a show of force would be insufficient to “teach North Korea a lesson…a show of force, by itself would not impress the

North…deployments after the Pueblo incident did not prevent the downing of the EC-121…”100

Because Stilwell met with Park without embassy representatives, it is difficult to know whether

Stilwell was deliberately manipulating Park’s comments as he conveyed them back to

Washington; in the initial moments of the crisis it was Stilwell who opposed retaliatory measures against North Korea, and it is conceivable that he would misrepresent Park’s views to support his

98 Head, Short, and McFarlane, Crisis Resolution, 181. 99 U.S. Department of State, August 19, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 19, 1976, 8:12-9:15am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 285; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford Library. 100 U.S. Department of State, August 19, 1976. “Telegram 190720Z From the Commander in Chief of the United Nations Command (Stillwell) to the Joint Chiefs of Staff, Secretary of State Kissinger, and Secretary of Defense Rumsfeld, August 19, 1976, 0720Z,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 284.

248 own.101 Park would maintain a public appearance of unity with the United States, but privately believed the United States was not forceful enough with North Korea.102 Privately Park would also contravene U.S. preferences by secretly telling his military special forces to “retaliate thoroughly” if North Korea attacked its forces, smuggling into JSA grenades and various small arms in order to provide it superior fire power for retaliation if needed.103

Some of the most impactful diplomacy in this crisis involved U.S. direct negotiations with North Korea via UNCMAC. The United States used the August 19 MAC meeting the day after the incident to condemn North Korea and to convey its demands—acknowledgement of its guilt, assurances that U.S. and UN personnel would be protected in the JSA, and punishment of the North Korean attackers.104 The U.S. demand for punishment of the attackers in particular was conveyed in dramatic language: “These actions have been exposed as actions of animals.

Not only the UNC but civilized people of the entire world will watch you closely to see that you punish the murderers.”105 As Operation Paul Bunyan commenced on August 21, although North

Korea was not notified of the operation in advance, the MAC did notify North Korea of the purpose of the operation, and conveyed that no additional military actions would occur as long as

101 The August 25 WSAG meeting included discussion about Park’s curious shift from the August 19 view that Stilwell conveyed in which Park supported not retaliating, compared to what they learned privately from Park in the ensuing six days, during which Park repeatedly advocated retaliation. Kissinger made the suspicious quip: “Every time I wanted to hit hard at the North Koreans last week, I was told that Park didn’t want to take military action. Now I gather he wants to do something.” See U.S. Department of State, August 25, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 25, 1976, 10:30am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 286; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford Library. 102 U.S. Department of State, August 27, 1976. “Telegram 213541 From the Department of State to the Embassy in the Republic of Korea, August 27, 1976, 2010Z,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969- 76, document 287. 103 Park, Dora Ojiahneun Dari-ae Seoda, 149, 187. 104 Michishita, North Korea’s Military-Diplomatic Campaigns, 78. 105 As quoted in Jonsson, Peace-Keeping in the Korean Peninsula, 299.

249 its operations team was left alone.106 The subsequent MAC meetings between August 25 and

September 6 would be used to negotiate specific terms for dividing the JSA to provide for better security of UN personnel, per the proposal Kim Il Sung made in his August 21 statement of regret. Throughout this period, the United States would maintain DEFCON 3 and its enhanced military posture; nevertheless, both sides would conduct negotiations relatively calmly, professionally, and largely without the hyperbolic rhetoric that had become so common, especially from North Korea.107 The U.S.-preferred terms of agreement in these negotiations, which North Korea ultimately agreed to, included: “the military of both sides will be restricted to their respective sides of the Military Demarcation Line and will not cross into the territory of the other; that neither side will construct barriers or any other obstacle to the vision or right of way in the territory of the other; the security of each other's personnel will be guaranteed; and that the

North Koreans will remove existing guard posts (4) on the United Nations side in the JSA.”108

The North Korea Perspective

North Korean word and deed in this case reveals more than the previous cases about

North Korean views of U.S. threat credibility. The evidence we can marshal to build the North

Korea perspective includes: 1) its advance warnings and rhetoric relating to both the U.S. presence on the Korean Peninsula generally and U.S. movements of military assets in early 1976, but also specifically relating to the poplar tree in the JSA; 2) North Korean actions in the DMZ

106 Head, Short, and McFarlane, Crisis Resolution, 194-5. 107 Jonsson, Peace-Keeping in the Korean Peninsula, 307. 108 U.S. Department of State, September 5, 1976. “Memorandum from the Deputy Assistant to the President for National Security Affairs (Hyland) to President Ford, Washington, September 5, 1976,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 288; National Security Adviser, Presidential Country Files for East Asia and the Pacific, Korea (19), box 10, Ford.

250 area in the recent past, as well as inferences from the specific way it conducted the attack; 3)

North Korea’s orchestration of public statements immediately following the incident; 4) North

Korea’s messages conveyed during the MAC meetings, including Kim Il Sung’s statement of regret; 5) an important defector account of events during the period of the crisis; 6) North

Korea’s behavior during Operation Paul Bunyan, including North Korea’s public characterization of events; and 7) North Korean violence against the United States after the incident, as well as later North Korean explanations of the events that took place during August and September 1976. The overwhelming impression from this compilation of the North Korean view is that: the August 18 attack was planned and deliberate, though it cannot be conclusively attributed to a decision in Pyongyang despite strong circumstantial evidence in support of such a view; it misjudged U.S. resolve going into this case; and it found the U.S. deterrent threat following the provocation credible, adjusting its behavior accordingly, helped in part by U.S. assurances and restrained U.S. rhetoric.

As discussed in the previous section, North Korea was in a state of renewed hostility with the United States and South Korea in 1975 and 1976. As the United States rotated a unit of F-

111 fighters to South Korea in February 1976 to support its war contingency plans for the

Korean Peninsula, North Korea unfurled a kind of threat not heard in years, asserting that “The

American Imperialists are the main instigators to the rise of tensions and the unleashing of a war of aggression in Korea…”109 Three months later, the State Department’s Bureau of Intelligence

109 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.043,” February 28, 1976, History and Public Policy Program Digital Archive, Archives of the Romanian Ministry of Foreign Affairs, Matter 220 - Relations with the Democratic People's Republic of Korea, 1976, Obtained by Izador Urian and translated for NKIDP by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/114107.

251 and Research took note of what it considered an abnormally intense quality to North Korea’s threatening rhetoric in 1976.110

How did North Korea view the JSA and the DMZ prior to the August 18 attack? In a cable back to the Joint Chiefs of Staff the day after the attack, General Stillwell charged: “What is unique is the death of security personnel (the first caused by a conflict between the two sides in the 23 years since the Armistice Agreement was signed), the severity, and the apparent premeditation with which the KPA attack was carried out.”111 But Stillwell’s characterization was misleading. Violence in the JSA and DMZ area was far from unprecedented. Prior to

August 18, most of North Korea’s violent attacks following the Armistice Agreement had been concentrated in the DMZ area. North Korea had clearly established, moreover, that the JSA was not sacrosanct; at least 25 physical assaults had taken place in the JSA prior to the August 18 attack, including the unprovoked1975 brutal beating of a U.S. officer in the JSA, requiring not only hospitalization but outright evacuation from the Korean Peninsula—an incident that would be raised during the August WSAG meetings to characterize precedent for North Korea’s behavior.112 Less than two months before the attack in Panmunjom, on June 26, more than a dozen North Korean guards ambushed a UN Command vehicle in the JSA and beat two

110 Information Memorandum, May 28, 1976, “North Koreans Charge ROK Provocations and Threaten Possible Retaliation,” National Security Adviser, Presidential Country Files for East Asia and the Pacific, box 10, Korea (16), Ford. 111 As quoted in Michishita, North Korea’s Military-Diplomatic Campaigns, 87. 112 Seuk-ryul Hong, “Wigi sog-ui chongjon hyopchong- Pueblo sakeon-gwa P’anmunjôm tokki salhae sakeon,” (“In Crisis: The Pueblo and Panmunjom Axe Murder Incidents”) Yôksa pip’yông 63, no. 1 (2003): 64-66; U.S. Department of State, August 25, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 25, 1976, 10:30am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 286; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford.

252 personnel with various clubs and boards.113 North Korea continued signaling its intent during the month of the attack. North Korea’s publicly released government white paper on August 5 was unprecedented, and used language even more dramatic than earlier in the year: “The long- standing tensions in Korea have now reached an acute state as never before…war may break out at any moment.”114 This statement was repeated by North Korea’s representative during the

NAM Summit Conference in Sri Lanka the day before the August 18 Panmunjom attack.115

North Korea blamed the “feverish near-war state” on the United States and South Korea, in particular the U.S. military presence on the Korean Peninsula, which it viewed as having malign intentions; especially in the mid-1970s, pushing out the U.S. military was the principal political goal attached to its military provocations.

More direct North Korean signals took place during the month of August, as well as in the way North Korea conducted the actual attack. North Korea’s threatening rhetoric in 1976 was harsh, and was aimed generally at expelling the U.S. military presence from Korea. On

August 6, however, North Korea had drawn a specific line with respect to the poplar tree in the

JSA. Although the tree trimming was a routine action that both sides conducted from time to time, and although the UN Command conducted all the routine notifications to ensure there were no surprises, the North Korean officer on the scene on August 6 warned the UN team to leave the tree alone, which it did at the time.116 On August 17, the day before the attack, Kim Il Sung was scheduled to attend the NAM Summit Conference in Sri Lanka, in hopes of gaining support from

113 Michishita, North Korea’s Military-Diplomatic Campaigns, 85. 114 Korean Central News Agency radio broadcast, August 5, 1976. 115 Matthews, “North Korea at Summit Seeks Anti-U.S. Move, 3d World Pact,” A22. 116 Head, Short, and McFarlane, Crisis Resolution, 156.

253 the Non-Aligned Movement for demanding that the United States military leave Korea, and to support North Korea’s position at the United Nations. In a note from Kim to Yugoslav dictator

Josip Tito, Kim explained that he instead needed to send his a delegate at the last minute to deliver the bombastic speech denouncing the United States and South Korea because circumstances on the Korean Peninsula suddenly required his attention. The unfolding of the actual August 18 attack also revealed tactical indications that North Korea was about to incite an incident. Over the span of about 30 minutes, the KPA officer on the scene at the poplar tree, whom the UN team recognized as one of the instigators of the 1975 beating of a U.S. officer, gave at least four warnings on the spot, each of which was more directly threatening than the previous: the first warning was to simply stop the tree trimming; the second was “If you cut more branches, there will be a big problem”; the third was “The branches that are cut will be of no use, just as you will be after you die”; the fourth was “If you cut more branches, you are going to die.”117 In addition, after the second warning, the North Korean officer signaled for additional

North Korean reinforcements; according to a former UN enlisted soldier stationed on the DMZ at the time, North Korea never initiated an incident unless it first marshaled superior numbers, usually at least 3:1. Only after some 30 KPA soldiers arrived on the scene and North Korea issued two more warnings did the North Korean officer give the order to “kill.”

Only 15 minutes after the attack, North Korea’s foreign minister declared publicly that

“U.S Imperialist Aggression troops with lethal weapons pounced on and beat North Korean guards who had gone to protest that the tree was in their area; that the guards were compelled to

117 Head, Short, and McFarlane, Crisis Resolution; Michishita, North Korea’s Military-Diplomatic Campaigns; Kirkbride, DMZ; Park, Dora Ojiahneun Dari-ae Seoda.

254 take action in self defence and that the incident was a planned provocation and that the actions were to aggravate the prevailing situation. War could break out at any moment.”118 The evening of the attack, on instructions from Pyongyang, North Korea released a statement at the NAM

Summit Conference in Sri Lanka, describing the Panmunjom attack as an “intentional provocative acts against our side,” and continuing to characterize the United States as trying to ignite a war.119 From the time of the attack through the next several days, North Korean media focused almost exclusively on U.S. war preparations, the provocation the United States had perpetrated against North Korean soldiers, a statement from Kim Il Sung as Supreme

Commander of the KPA that all forces should be activated for “battle readiness,” and threatening rhetoric like “The Korean people are not afraid of war and if such a war is imposed upon it, then it will crush the aggressor.”120 The day after the Panmunjom attack, the KPA Supreme

Command issued its version of events, continuing the same messaging as the day prior: “…the

U.S. imperialist aggressors who are making a desperate attempt to start a new war in Korea committed a grave provocative act…our side repeatedly told them that the tree must not be cut…but the enemy side, far from complying with our just demand, collectively pounced, brandishing lethal weapons, upon the security personnel of our side…This reckless provocation of the enemy compelled our security personnel to take a step in self-defence…Such grave

118 Kirkbride, DMZ, 31; Hong, “Wigi sog-ui chongjon hyopchong- Pueblo sakeon-gwa P’anmunjôm tokki salhae sakeon.” Also quoted in Jonsson, Peace-Keeping in the Korean Peninsula, 297. 119 Michishita, North Korea’s Military-Diplomatic Campaigns, 77-8. 120 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.212, August 21, 1976, Archives of the Romanian Ministry of Foreign Affairs, Matter 220-Relations with the Democratic People’s Republic of Korea, 1976, Obtained by Izador Urian and translated by Eliza Gheorghe. http://digitalarchive.wilsoncenter.org/document/114116.

255 provocative act committed by the U.S. side in the joint security area of Panmunjom is a premeditated act to start a war.”121

But North Korean signaling was not only limited to rhetoric immediately following the attack and the subsequent several days. Kim Il Sung released his “battle readiness” order as soon as the UN Command enhanced their alert status to DEFCON 3, and North Korea began mass military mobilization on a national scale as a result. According to Park Pyong-yop, a senior level official in North Korea involved in the Panmunjom crisis that later defected to South Korea, once the UN Command declared DEFCON 3 and started moving in additional military forces on

August 19, North Korea moved to “a complete war-ready condition for three months.”122 On

August 20, North Korea initiated anti-air defense drills.123 “College students were drafted to the armed forces and reserves were called up. Retired officers up to age 50 put on uniforms again.

Production facilities were getting ready to be moved to rear areas…approximately 200,000 people were evacuated from Pyongyang to other areas…men had their military uniforms ready for use when they went to bed. Workers took combat positions for three months leaving their work behind…People were evacuated to mountainous areas to avoid possible heavy bombing.”124 The Romanian embassy in Pyongyang reported on the day of Operation Paul

Bunyan that “…the population is preoccupied, being more worried than on other occasions by

121 Pyongyang Times (August 21, 1976): 1. 122 Jung, Kyeot-aeseo Bon Kim Jeong-Il, 202-3. 123 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.212, August 21, 1976, Archives of the Romanian Ministry of Foreign Affairs, Matter 220-Relations with the Democratic People’s Republic of Korea, 1976, Obtained by Izador Urian and translated by Eliza Gheorghe. http://digitalarchive.wilsoncenter.org/document/114116. 124 Jung, Kyeot-aeseo Bon Kim Jeong-Il, 202-4.

256 the situation…we believe that the current situation is special compared to previous periods when such incidents, more or less similar, took place.”125

North Korean word and deed across the local DMZ area before August 21—the date of

Operation Paul Bunyan—changed noticeably during and immediately after the Operation, even though North Korean public messaging would remain consistently belligerent throughout the crisis. During the August 19 MAC meeting, Major General Han Ju-kyong, the KPA representative, blamed the United States for attacking and severely injuring several KPA soldiers, and ironically demanded that the United States offer assurances that it would not engage in such provocative behavior.126 It was also during the MAC meeting that North Korea nationally broadcasted Kim Il Sung’s decision to move the country to a full “war posture.”127

But on the morning of August 21, 30 minutes before North Korean guards were scheduled to take up their posts in the JSA, the work crew and security personnel entered the JSA to commence Operation Paul Bunyan. North Korea was only notified, by the UN Command, as the operation personnel entered the JSA, at 7:05am. Every reported account of the operation conveys some variation on the theme that the North Koreans in the area at the time were visibly frightened.128 Reinforcing this common account of palpable North Korean fear is the observed fact that not only did North Korea avoid intervening in or retaliating against the UN team, but all the North Korean guards remained outside of the JSA, not taking their positions until after the

125 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.212,” August 21, 1976, Archives of the Romanian Ministry of Foreign Affairs, Matter 220-Relations with the Democratic People’s Republic of Korea, 1976, Obtained by Izador Urian and translated by Eliza Gheorghe. http://digitalarchive.wilsoncenter.org/document/114116. 126 Downs, Over the Line, 153-4; Jonsson, Peace-Keeping in the Korean Peninsula, 299-300. 127 Head, Short, and McFarlane, Crisis Resolution, 188. 128 Oberdorfer, The Two Koreas, 81; Jonsson, Peace-Keeping in the Korean Peninsula; Michishita, North Korea’s Military-Diplomatic Campaigns; Downs, Over the Line; Ferguson, Memories of JSA.

257 UN Command team departed the JSA, an hour after the scheduled time that North Korea’s personnel were supposed to take up their positions. Up to 150 North Korean personnel gathered to watch the operation from a distance outside the JSA, through binoculars.

Only one hour after Operation Paul Bunyan was complete, Major General Han requested a private meeting with Admiral Frudden, which took place at noon the same day as the operation and lasted only 13 minutes. General Han told Admiral Frudden he was instructed to convey a message directly from Kim Il Sung to be delivered to General Stilwell in his role as commander of the United Nations Command. Kim’s statement read:

“It is a good thing that no big incident occurred at Pan Mun Jom for a long period. However, it is regretful that an incident occurred in the Joint Security Area, Pan Mun Jom this time. An effort must be made so that such incidents may not recur in the future. For this purpose both sides should make efforts. We urge your side to prevent the provocation. Our side will never provoke first, but take self defensive measures only when provocation occurs. This is our consistent stand.”129

General Han then added his own commentary rebuking the United States for Operation Paul

Bunyan, which he called a provocation. Though making it clear that it was not part of the message Kim Il Sung instructed him to deliver, General Han threatened that if any such provocation occurs again, North Korea might respond with a repeat of August 18, ending the conversation with “I strongly demand that your side commit no such provocations…the provocation which you committed this morning is a serious one which might cause a serious

[sic] consequences.”130

129 Telegram from CINCUNC to JCS, August 22, 1976. Quoted in Michishita, North Korea’s Military-Diplomatic Campaigns, 80; Head, Short, and McFarlane, Crisis Resolution, 196-7; Downs, Over the Line, 155. 130 Head, Short, and McFarlane, Crisis Resolution, 196-7.

258 Despite North Korea’s avoidance of the JSA during Operation Paul Bunyan, its decision not to retaliate, and Kim Il Sung’s unprecedented statement of regret—which included an embedded proposal to divide the JSA for safety purposes—North Korea’s public diplomacy immediately following the tree-chopping operation was accusatory and threatening, consistent with much of its rhetoric prior to the operation. Radio broadcasts in North Korea immediately after Operation Paul Bunyan accused the United States and South Korea of a “grave provocation” and that the United States was “creating a terrible war atmosphere.”131 On August

25, the day the MAC negotiations to divide the JSA began, North Korea’s foreign ministry gathered the heads of Soviet-aligned and non-aligned diplomatic missions in Pyongyang to officially convey its same version of events it released the day of the August 18 attack, blaming the United States for a premeditated and aggressive attack on its forces. Deputy Foreign

Minister Han Si-hae then cited the massive military show of force that the United States marshaled for Operation Paul Bunyan as proof that the United States desired “to provoke a large- scale military conflict,” adding that “Given the general state of tension,” North Korea had decided not to send or receive foreign delegations for a period of time.132

The same day North Korea’s foreign ministry also invited Korean journalists and foreign press correspondents to receive the same statement it had delivered to Pyongyang’s diplomatic missions, adding that there was “the possibility that at any given moment war breaks out.”133 An

131 “U.S. Sends Jets, Ships to Korea,” Washington Post (August 22, 1976): 1-2. 132 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.219,” August 26, 1976, Archives of the Romanian Ministry of Foreign Affairs, Matter 220-Relations with the Democratic People’s Republic of Korea, 1976, Obtained by Izador Urian and translated by Eliza Gheorghe. http://digitalarchive.wilsoncenter.org/document/114119. 133 “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.220,” August 27, 1976, Archives of the Romanian Ministry of Foreign Affairs, Matter 220/Year/1976/Country: Democratic People’s Republic of Korea,

259 East German diplomat in Pyongyang argued that because North Korea was already in a national state of emergency during Operation Paul Bunyan, there was no chance that North Korea was surprised by U.S. actions during the crisis, and that North Korea “did not desire any further aggravation, which might have caused a hardly controllable escalation….the determination displayed by the U.S. through its spectacular military presence and the combat readiness of its troops.”134

For months after the Panmunjom crisis ended, North Korea—including in personal conversations with Kim Il Sung—continued to proactively raise the attack and insist it was a

U.S. provocation, which the East German defense minister believed suggested not only that

North Korea’s excessive protests belied a certain guilt on North Korea’s part, but also that North

Korea was trying to rationalize the incident in a way that would convince friendly governments that North Korea would “not initiate military actions against the South,” and that “It is willing to join the general course of détente pursued by the socialist states.”135

An October 1976 Kim Il Sung conversation with a small private East German audience sheds further light on the North Korean perspective during the crisis. Definitively clarifying whether North Korea ever punished the instigators of the August 18 attack, Kim said that

“Kissinger demanded that we apologize and punish our soldiers. However, we have no reason to

Telegrams from Pyongyang to the Romanian Ministry of Foreign Affairs, January-December 1976, Obtained and translated by Eliza Gheorghe: http://digitalarchive.wilsoncenter.org/document/114120. 134 “Report on the ‘Axe Murder Incident’ from the GDR Embassy in North Korea,” August 31, 1976, Political Archive of the Federal Foreign Office, Berlin (PolA AA), MfAA. Translated for NKIDP by Bernd Schaefer. Available online: http://digitalarchive.wilsoncenter.org/document/114291. 135 “Report on a Stay of a GDR Military Delegation in the DPRK in October 1976,” October 1976, German Federal Archive, Military Archive (BA-MA), AZN 8283, Translated for NKIDP by Bernd Schaefer. Available online: http://ditigalarchive.wilsoncenter.org/document/114292.

260 sanction our comrades since they just acted as good patriots.”136 On the issue of his statement of regret, Kim clarified “It can always happen during a fight that somebody gets killed. We declared to the Americans that we regret the incident, but we did not concede anything on reparations. Then, the enemies ceased to raise further demands. They withdrew their aircraft and aircraft carriers.”137 This statement reveals the possibility that Operation Paul Bunyan and

U.S. conduct in the MAC meetings could have potentially led to escalation into a larger conflict if the United States had tried to forcibly impose—rather than abandon—its demand that North

Korea punish those responsible; put another way, the United States might have accidentally sparked escalation if it had insisted on North Korea conceding to its initial demands. Kim’s post hoc rationalization of its behavior during the Panmunjom crisis, even if broadly true, overlooks one major concession that North Korea agreed to: the unilateral removal of its guard posts from what would become the UN side of the JSA.

After the Panmunjom crisis, violent incidents across the DMZ became an exception rather than a norm. Since that time, North Korea has occasionally targeted South Korean forces, but has not initiated any attacks on U.S. ground forces. However, the rivalry conditions with

North Korea never abated during the rest of the Cold War, and North Korea’s hostility would subsequently be channeled into violent attempts against the United States on two other occasions, though the latter’s South Korean ally would also become a repeated target of North

Korean provocations for the decades to follow.

136 Ibid. 137 Ibid.

261 On July 14, 1977, a U.S. army helicopter accidentally landed on the North Korean side of the DMZ, and when it realized where it was it attempted to evacuate. As it was attempting to flee, North Korean soldiers shot down the helicopter, killing three U.S. soldiers. Upon learning about the shoot down, the White House issued a statement apologizing: “We are trying to let

[North Korea] know the mistake was made by the crew in going into the DMZ…Our primary interest is in having the incident not escalate into a confrontation and to account for the crew members.”138 North Korea’s media characterization of the incident was sober, factual, and avoided threat-making, to the surprise of the Romanian embassy in Pyongyang, which reported on the incident and how it was concluded.139 The Romanian ambassador attributed the low key, even positive North Korean reaction to the swift, high level apology; it was the first White House apology in the history of U.S.-North Korea relations.140 He also speculated that North Korea was seeking a meeting with Cyrus Vance during his upcoming meetings in Beijing that year, and that the White House’s conciliatory tone gave hope that rapprochement with North Korea was once again possible.141 For most of the Carter administration, North Korea held out the prospect that

Carter’s presidential campaign pledge to withdraw troops from Korea would occur. During the

MAC meeting to address the incident on July 17, Major General Han said that “Taking into

138 As quoted in Downs, Over the Line, 158-9. 139 “Telegram 066678 from the Romanian Embassy in Pyongyang to the Romanian Ministry of Foreign Affairs, US Helicopter Incident on July 14,” July 19, 1977, Archive of the Romanian Ministry of Foreign Affairs, Folder 933/1977, Issue 220/H: Partial US troop withdrawal from South Korea-Discussions regarding the reunification of the two countries, January-October 1977, Obtained and translated for NKIDP by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/114874. 140 Ibid. 141 Ibid.

262 consideration your side’s admission and expression of regret,” North Korea was willing to resolve the incident favorably and repatriate the remains of the soldiers who died in the crash.142

The next aggressive incident took place on August 26, 1981, when North Korea fired an

SA-2 surface-to-air missile at an SR-71 high speed, high altitude reconnaissance aircraft conducting operations off the eastern coast of North Korea in international airspace. Rivalry conditions between the United States and North Korea persisted into the Reagan administration, and had become inflamed by Carter’s reversal of his campaign pledge to withdraw troops from

Korea; still, the last violent incident of note against the United States was the 1977 helicopter shoot down. A U.S. defense official at the time asserted that the SA-2 attempt to shoot down the

SR-71 had to be deliberate because the SR-71 flies at an altitude far out of range of SA-2 missiles, which meant targeting the SA-2 would have required using multiple radar guidance systems in multiple locations to triangulate the SR-71, in advance of the shoot down.143 The SA-

2 was not able to reach the SR-71, detonating out of range but still close enough for the SR-71 pilot to see the missile’s contrails and for his warning systems inside the aircraft to activate.144

It remains unclear based on extant evidence whether North Korea’s attempt to shoot down the SA-2 was a warning signal—an implied coercive threat—or a deliberate attempt at violent provocation that simply failed. During the September 1 MAC meeting held to address

North Korea’s failed attempt on the SR-71, the senior UNC member, Rear Admiral Storms, charged the North Koreans with a “premeditated and unprovoked act of aggression…,” issuing

142 As quoted in Downs, Over the Line, 160. 143 David Binder, “Radar Detector Aboard SR-71 Alerted Pilot to Missile Attack,” New York Times (August 29, 1981). 144 Ibid.

263 the first and most direct verbal threat in the history U.S.-North Korea relations since the signing of the Armistice Agreement: “…the UNC will react against the source of any future such attacks if North Korea chooses to again attack any aircraft under UNC control in airspace in which UNC aircraft are entitled to fly.”145 North Korea charged that the United States “fabricated this absurd incident,” complaining that not only has the United States reversed its decision to withdraw from the Korean Peninsula, but has reinforced its position with additional nuclear and other weapons.146 This incident would prove the last act of violent friction aimed at the United States until the end of the Cold War, though anti-American rhetoric would persist and South Korea would become the victim of numerous additional North Korean attacks over the next decade.

Analysis: Explaining Decisions and Outcomes

How does the historical narrative surrounding the 1976 Panmunjom crisis fit with the expectations of the competing honesty and resolve reputational theories, as well as alternative theories of credibility? On the question of threat credibility, the outcomes in this case are indeterminate themselves, as outcomes here partially fit each of the four hypotheses. On the question of the temporal consequences of backing down or standing firm, the outcomes in this case fit both the reputations for honesty and resolve hypotheses. And on the question of potential connections between initiating challenges against a rival and a reputation for resolve, this case offers support for findings in previous cases: a history of bluffing can impede the accretion of a reputation for resolve for a challenger. Within-case process tracing evidence similarly reveals nuanced multiple causality; reputations, understood as past word and deed,

145 As quoted in Downs, Over the Line, 161. 146 As quoted in Downs, Over the Line, 162.

264 played a causal role in this case, but not in isolation from questions of interests and the relative balance of power. It provides the least support for constant calculus theory—which argues that a state always finds its rival’s threats credible—and offers some support for current calculus theory’s emphasis on the balance of material capabilities as a basis for calculating credibility, but it also somewhat confounds current calculus theory’s similarly weighted emphasis on interests.

Explanatory Power of Reputations for Honesty

This case offers several tests of the reputations for honesty hypotheses. Evidence confirms some aspects of the reputations for honesty logic, and disconfirms others. Going into analysis of these events, the honesty logic anticipated several outcomes: that backing down or standing firm would make future challenges less likely; that a gap between word and deed (that is, threat-making without commensurate follow through) would erode future threat credibility; and that matching word and deed (that is, threat-making with commensurate follow through) would bolster future threat credibility. This case, which allows us to observe all four claims, supports the contention that standing firm makes future challenges less likely, but refutes the argument that backing down also makes future challenges less likely; partially supports and partially undermines the contention that a gap between word and deed erodes future threat credibility; and supports the contention that aligning word and deed bolsters threat credibility.

In May 1976, following months of increased North Korean rhetoric without corresponding provocations, a White House memo reflecting the opinion of the NSC and the

State Department’s Bureau of Intelligence and Research judged that North Korea’s increased

265 threat-making in 1976 was cause to expect future North Korean provocations.147 Although it incorrectly predicted that such attacks would be against U.S. aircraft or naval vessels, the U.S. assessment of North Korean threat credibility was explicitly based on the variation in North

Korean threat-making over the past year, not on the alignment or misalignment of North Korean word and deed.148 This evidence contradicts the reputations for honesty logic. The May 1976

White House memo indicated that some in the U.S. bureaucracy expected North Korean provocations in the coming months, but the only actions directed as a result of this threat credibility were limited to coastal aerial and naval reconnaissance missions, not actions across the DMZ; the May 1976 memo in fact made no mention of historical or future anticipated actions around the DMZ.

As in the previous chapters, even in the rare instances when some in the U.S. government believed North Korean threats, a lack of consensus and a lack of crisis conditions often meant that there would not or could not be any logical action based on the minority who did believe

North Korea’s threats. This dynamic would manifest in a very proximate sense with the action- reaction sequence of events on August 6, August 18, and August 21, providing multiple opportunities to observe the exact same material issue at stake (a poplar tree in the JSA) with the exact same principal actors involved, each with a different outcome. On August 6, the U.S.-led

UN team went into the JSA to trim the poplar tree after following standard procedures for advance notice to the North Korean side. The team backed off when the North Koreans at the

147Information Memorandum, May 28, 1976, “North Koreans Charge ROK Provocations and Threaten Possible Retaliation,” National Security Adviser, Presidential Country Files for East Asia and the Pacific, box 10, Korea (16), Ford. 148 Ibid.

266 site warned them to leave the tree alone; North Korea issued no proximate threat, though the warning came against a backdrop of threatening rhetoric generally, including the August 5 government white paper discussed in the previous section.

Yet on August 18 the United States ignored both North Korea’s threatening rhetoric and its August 6 warning to leave the tree alone, entering the JSA once again to trim the tree; even as

North Korean officers on the scene on August 18 issued multiple ever more direct and ever more threatening warnings, the U.S. work team continued to ignore North Korea’s threats as bluffs.

An additional indicator of U.S. incredulity toward North Korean threats at that time was the fact that civilian tours of the JSA continued normal operations on the morning of August 18.149 And even within the context of North Korea’s enhanced rhetorical flourishes, the commanding U.S. officer on the scene interpreted North Korea’s August 6 warning as “no more than usual harassment.”150 Congruent, then, with the reputations for honesty expectation about threat credibility, North Korea’s word-deed mismatch had once again undermined its own threat credibility and once again left the United States not expecting the next provocation. Operation

Paul Bunyan on August 21 also supports the reputations for honesty claim. The U.S. military show of force accompanying the work team was intended to deter North Korea from intervening as the United States and ROK chopped down the poplar tree that in the previous weeks North

Korea had twice intervened to prevent the United States from even trimming. North Korea not only avoided intervening, but prevented their guards from even entering the JSA during the operation, followed only hours later by a conciliatory statement directly from Kim Il Sung. The

149 Head, Short, and McFarlane, Crisis Resolution, 163. 150 Head, Short, and McFarlane, Crisis Resolution, 161.

267 United States had a long history of backing down from North Korean provocations and avoiding issuing threats; viewed through the reputations for honesty lens, the expectation going into the

August 21 deterrence contest was thus that North Korea should find the U.S. deterrent threat credible, which is precisely what happened.

Also congruent with the reputations for honesty hypothesis, the U.S. decision to stand firm was followed by an extended period with no North Korean provocations. The next incident targeting the United States, the helicopter shoot down in 1977, was not a provocation or challenge but rather a North Korean reaction to a U.S. violation of North Korean airspace, substantiated as such by a swift White House apology. The temporal logic of reputations for honesty expects that demonstrating resolve should be followed by fewer rival challenges, which does bear out in this case; provocations against the United States along the DMZ all but ceased following the Panmunjom crisis.

In sum, the events of August 21 and after support the reputations for honesty logic, as does U.S. incredulity toward North Korean threats leading to the August 18 attack. The two aspects that most challenge the reputations for honesty narrative are North Korea’s repeated challenges against the United States after the latter backed down from previous challenges— culminating in the August 18 challenge—and the May 1976 U.S. assessment that North Korea intended to conduct a violent provocation against the United States based on variations in North

Korean rhetoric that ignored the mismatch in North Korean word and deed. As explained below, however, the evidence that supports the reputations for honesty hypotheses in this case also

268 supports competing hypotheses; that is, to the extent that dependent variable outcomes co-vary with reputations for honesty predictions, they also co-vary with competing hypotheses.

The case narrative supports a multicausal explanation, which makes it difficult to reach a conclusion that the reputations for honesty logic is the dominant explanation, but that is not the threshold for validating the relevance of reputations for honesty; to the extent that reputational theories have explanatory power, it is only in combination with and in a context of other explanatory factors. A reputation for honesty logic proves useful in understanding some aspects of this case, even if it does a poor job explaining others. And it is important to recall that my hypotheses anticipates that one of the claims (effect of backing down on future threat credibility) of a reputation for honesty logic should be confirmed while the other of its claims (effect of backing down on the likelihood of future challenges) should be refuted. As with the previous chapters, we are seeing that the proper scope of a reputation for honesty lens may only be limited to the temporal effects of word-deed alignment on threat credibility. Honesty’s key distinguishing causal claim—that backing down makes future challenges less likely—finds no support in this case, just as H1 expects.

Explanatory Power of Reputations for Resolve

The reputations for resolve perspective is essential to understanding why the events of this case unfolded as they did, though not necessarily in isolation from other explanatory factors.

Evidence in this case, which offer several observations of reputations for resolve hypotheses, is not only congruent with most, but not all, reputations for resolve predictions, but some within- case process tracing evidence also substantiates many of the reputations for resolve claims. The

269 reputations for resolve hypotheses anticipated the following outcomes in this case: that backing down would make future challenges more likely while standing firm would make future challenges less likely; that backing down makes future threats less credible while standing firm makes future threats more credible; and that a gap between word and deed (that is, threat-making without commensurate follow through) would erode future threat credibility. This case tests each of these claims. We find that the evidence here: supports the claim that backing down makes future challenges more likely while standing firm makes future challenges less likely; undermines the claim that backing down makes future threats less credible; and partially supports and partially undermines the claim that a gap between word and deed erodes future threat credibility. Because the case does offer an opportunity to observe the effect of standing firm on the likelihood of future challenges, it does not offer an opportunity to observe how standing firm affects future threat credibility, arguably precisely because the U.S. decision to stand firm in the particular way it did curbed future provocations against the United States.

The logic of a reputation for resolve illuminates why the United States would find North

Korean threats incredible, both in the period prior to the Panmunjom crisis and during August leading up to the proximate, direct threats North Korea issued on August 18. North Korea had issued countless threats before without action, so the United States saw little reason to treat the threatening rhetoric in North Korea’s August 5 government white paper, or the August 6 warning not to touch the poplar tree, seriously. Even on August 18 as the KPA soldiers at the scene were uttering direct threats like “If you cut more branches, you are going to die,” U.S. soldiers on the

270 scene found nothing particularly credible or alarming about such threats.151 It is difficult to comprehend such a dismissive reaction to North Korean threats except in the context of reputations—past word and deed.

In addition, there is strong within-case evidence from the August 19 WSAG meeting that supports a resolve-based interpretation of North Korean credibility the day after the North

Korean attack. Reflective of a reputation for resolve interpretation, Kissinger and the CIA both agreed that North Korea had no intention of starting a larger conflagration or fighting a war; “If they had wanted to…they would not have beat two of our men to death” because it would sacrifice the element of surprise.152 Kissinger also took North Korea’s behavioral history in relation to other provocations into account in assessing its credibility, inquiring specifically about what North Korea’s alert posture was after the EC-121 shoot down.153 Of note, Deputy

Secretary of Defense, Bill Clements, voiced concern that North Korea would escalate in retaliation to a U.S. strike of any kind, without specifying the basis for his assessment. Like

Kissinger and the CIA though, Undersecretary of State Habib also drew on North Korea’s history of word and deed, arguing that historical precedent did not support expecting North

Korean proportional responses, but rather either acquiescence or escalation.154 Arguing that U.S. credibility would require an actual show of U.S. resolve, Kissinger reminded the group that “The

151 Head, Short, and McFarlane, Crisis Resolution, 164. 152 U.S. Department of State, August 19, 1976, “Minutes of Washington Special Actions Group Meeting, Washington, August 19, 1976, 8:12-9:15am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 285; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford. 153 Ibid. 154 Ibid.

271 purpose of doing something is to show that we are ready to take risks…The purpose of this exercise is to overawe them.”155

Prior to the Panmunjom crisis, the United States had a long-running history of backing down from North Korean challenges. The concessions made to North Korea during the 1968

Pueblo crisis and the U.S. non-reaction to North Korea’s 1969 shoot down of a U.S. reconnaissance aircraft in international airspace were the most high-profile examples of hundreds of data points in the U.S.-North Korea rivalry in which North Korea initiated a challenge and the

United States backed down, preventing crisis escalation but making future confrontations more likely. The reputations for resolve logic explains this repeating dynamic as a function of North

Korean assessments of U.S. irresoluteness, which were derived at least in part from the U.S. history of backing down. From this perspective, it comes as little surprise that North Korea would attack U.S. troops on August 18 after warning them on August 6 to leave the JSA poplar tree alone; nothing about U.S. word or deed incentivized North Korean restraint while hostile rivalry conditions persisted.

Strengthening confidence in a reputation for resolve interpretation, the U.S. decision to stand firm—which took the form of Operation Paul Bunyan on August 21—introduces the first opportunity in this dyad to observe the temporal effects of not backing down from a challenge.

North Korea found the U.S. deterrent threat credible because it demonstrated resolve in a way that aligned word and deed; to this end, Kissinger noted in the August 25 WSAG meeting that

“Last week I was in favor of firm action but it was overruled at Vail, not by this group [but by

155 Ibid.

272 President Ford]. It was a tragedy. I have never seen the North Koreans so scared.”156 Of course, the U.S. deterrent threat combined a somewhat restrained verbal threat with a large, implicit threat attached to the large military force marshaled as part of the August 21 operation. As well, often missed in the retelling of the Panmunjom crisis is the U.S. decision to abandon its compellent demand for North Korea to punish the KPA soldiers responsible for the August 18 attack. Kim Il Sung would later tell a communist bloc ally that complying with such a demand was unacceptable and the crisis was only resolved because the United States walked away from that demand.157 It is important to acknowledge the nuanced way this crisis ended, with U.S. resolve, but also with U.S. conditional assurances of non-action and ultimately of the United

States compromising its initial objectives for the sake of a positive outcome. It is possible that marshaling raw coercive force is sufficient to achieving a reputation for resolve—conveying word and deed that a rival deems credible—but a reputation for resolve is not good for much if it does not contribute to a desirable outcome. In this case, a reputation for resolve only contributed to a desirable outcome because of more nuanced assurances and compromises that the United

States took in tandem with its show of force.

Whereas all other aspects of this case support a reputation for resolve interpretation of events, the May 1976 White House memo judging that North Korea’s increased threat-making in

1976 was cause to expect future North Korean provocations is inexplicable when viewed through

156 U.S. Department of State, August 25, 1976, “Minutes of Washington Special Actions Group Meeting, Washington, August 25, 1976, 10:30am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 286; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford. 157 “Report on a Stay of a GDR Military Delegation in the DPRK in October 1976,” October 1976, German Federal Archive, Military Archive (BA-MA), AZN 8283. Translated for NKIDP by Bernd Schaefer. Available online: http://ditigalarchive.wilsoncenter.org/document/114292.

273 a reputation for resolve lens.158 As with a reputation for honesty approach, the reputation for resolve hypotheses cannot account for believing threats on the basis of the threat alone, particularly threats from a rival with a recent history of bluffing.

Alternative Explanations about Credibility

Although current calculus theory and constant calculus theory make no testable claims about the temporal consequences of word and deed, each offers useful perspective on calculations relating to credibility in the events in this case, though both also face serious empirical challenges in explaining certain aspects of the case.

Current calculus theory urges us to enter analysis of credibility assessments in this case by ignoring all past behavior and statements, and to instead focus on two things: the current balance of power within the rivalry dyad, and the current balance of interests. One side will judge the other’s threat credibility as high when the balance of power and the balance of interests favors the state being judged (and vice versa). Current calculus theory also makes allowances for either the balance of power or the balance of interests to be unfavorable while still assessing high threat credibility if one of the variables identified takes on an extreme value in the supporting direction, meaning, for example, that a state can have relatively weak capabilities but still make credible threats if its vital interests are sufficiently engaged. Drawing again on data from the Correlates of War project for calculating power using material indicators, we find, as in previous chapters, that the balance of capabilities vastly favored the United States in 1976 and the preceding period: a .141 score for the United States, compared to a .007 score for North

158Information Memorandum, May 28, 1976, “North Koreans Charge ROK Provocations and Threaten Possible Retaliation,” National Security Adviser, Presidential Country Files for East Asia and the Pacific, box 10, Korea (16), Ford.

274 Korea.159 Given the vast power asymmetry, a case can be made from the current calculus theory perspective that the United States found North Korea’s threats in the period prior to the

Panmunjom crisis incredible as long as North Korea’s vital interests were not engaged.

During Operation Paul Bunyan, North Korea did clearly find the U.S. deterrent threat credible: even though only three days prior North Korea had attacked U.S. soldiers for trimming the JSA tree, on August 21 they restrained themselves from even entering the JSA, and multiple sources have identified that the North Korean soldiers on the DMZ were scared, an observation supported by diplomatic cables in Pyongyang indicated that the North Korean public was fearful of a U.S. attack.160 Current calculus theory would predict as much given the dramatically superior military capabilities of the United States, and the large U.S. show of military force immediately preceded the deterrent outcome, Kim Il Sung’s apology, and North Korea’s unilateral withdrawal of four of its observation posts from the UN side of the JSA. This case also validates current calculus theory insofar as the U.S. deliberations during the Panmunjom crisis drew on the balance of military capabilities on the Korean Peninsula. During the August

18 and 19 WSAG meetings, Kissinger and others inquired about the balance of military capabilities as they were determining North Korean threat credibility and how the United States should respond.161

159 Katherine, Keshk, and Pollins, “Correlates of War Project Trade Data Set Codebook.” 160 U.S. Department of State, August 25, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 25, 1976, 10:30am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 286; Oberdorfer, The Two Koreas, 81; Jonsson, Peace-Keeping in the Korean Peninsula; Michishita, North Korea’s Military-Diplomatic Campaigns; Downs, Over the Line; Ferguson, Memories of JSA. 161 U.S. Department of State, August 18, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 18, 1976, 3:47pm,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 282; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford Library; U.S. Department of State, August 19, 1976. “Minutes of

275 Current calculus theory falls short as an explanation in four aspects of this case though:

North Korean incredulity toward the United States leading up to its August 18 attack in the JSA, followed by North Korea’s view that the August 21 deterrent threat was credible; the U.S. perception that its interests were sufficiently engaged that threats and potentially an attack were warranted; Kissinger’s, Undersecretary Habib’s, and the CIA’s assessment of North Korean threat credibility during the Panmunjom crisis explicitly, in part, based on North Korean behavior in past crises; and the May 1976 White House memo assessing North Korean threat credibility specifically in terms of shifts in threat-making over the prior year.

North Korea’s repeated challenges against the United States prior to the Panmunjom crisis indicate that it did not expect the militarily superior United States to retaliate, an inference supported by North Korean statements during the early 1970s in which North Korea indicated it was not in a position to wage a war, even while it was conducting provocations. Current calculus theory could explain this in terms of an asymmetry of interests strongly favoring North

Korea, but that would pose a puzzle itself because on August 21 North Korea did find the U.S. deterrent threat credible. The dyadic capability asymmetry before and during the Panmunjom crisis were essentially constant, and the issue at stake—whether one defines the issue at stake as generalized DMZ competition, the specific issue of the poplar tree, or the U.S. military presence on the Korean Peninsula—was also constant over that period. What, then, explains both North

Korean incredulity toward the United States leading up to the Panmunjom crisis but also North

Washington Special Actions Group Meeting, Washington, August 19, 1976, 8:12-9:15am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 285; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford.

276 Korea’s view that the U.S. threat accompanying Operation Paul Bunyan as highly credible?

Current calculus theory is strained to explain this.

That U.S. decision-makers believed vital or important U.S. interests were at stake in this crisis is wholly inexplicable from the vantage point of current calculus theory. North Korea had attacked U.S. and ROK forces thousands of times across the DMZ over the decades, and North

Korea had specifically targeted U.S. soldiers in assaults several times in the years prior to the

Panmunjom crisis. Even the JSA, where North Korean, U.S., and South Korean forces were historically authorized to intermingle, was not sacrosanct, proven by multiple beatings of U.S. troops in the year before the crisis. It could be, and was, argued that there was a principle at stake with temporal consequences—that allowing North Korean aggression without consequences might beget more aggression162—but this logic is precisely what current calculus theory tells us is irrelevant in assessing the credibility of others and in determining the balance of interests. According to current calculus theory, interests are objectively knowable and static, and the criteria for greater interests are based on how grave the threat is; a threat cannot be a vital interest unless the issue at stake directly threatens the existence of the state. The current calculus view of interests could potentially explain why U.S. interests over the previous decades were not sufficiently engaged to necessitate a response to North Korean challenges, but the JSA attack on

August 18 was smaller in scale than in numerous prior incidents, even along the DMZ. Current calculus theory is also somewhat repudiated by direct statements made during the August 19

WSAG meeting. As the WSAG deliberated about North Korean intentions, Kissinger,

Undersecretary Habib, and the CIA each raised past North Korean behavior—specifically during

162 Michishita, North Korea’s Military-Diplomatic Campaigns, 87.

277 the EC-121 shoot down—contrasting North Korean posturing in past crises with its behavior immediately following the August 18 JSA attack.163 During the WSAG meetings, the

Kissinger-led group drew on a number of factors in weighing how to handle the situation they faced, including the balance of capabilities on-Peninsula, but current calculus theory tells us explicitly that one of those factors should not be the adversary’s past behavior, which in this case is flatly wrong.

The final aspect of this case that poses a challenge to current calculus theory is that which also vexed the reputational approaches to explaining credibility in this case: the May 1976 White

House memo. Although the memo anticipated the wrong type of provocation as a result of North

Korean threat-making—expecting an attack on a reconnaissance asset—it nevertheless assessed

North Korean threat credibility without reference to either U.S. or North Korean capabilities.164

In aggregate, although current calculus theory’s positive claim that we should look to capabilities and interests as determinants of threat credibility proves accurate here, the fact that there is direct, explicit process-tracing evidence that decision-makers also took into account adversary behavior while making credibility assessments is sufficient to disconfirm current calculus theory’s claim that past word and deed are causally irrelevant.

Constant calculus theory quite simply predicts a constant value over time for something that proves empirically to vary over time. In dyadic rivalry conditions, constant calculus theory

163 U.S. Department of State, August 19, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 19, 1976, 8:12-9:15am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 285; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford. 164 Information Memorandum, May 28, 1976, “North Koreans Charge ROK Provocations and Threaten Possible Retaliation,” National Security Adviser, Presidential Country Files for East Asia and the Pacific, box 10, Korea (16), Ford.

278 tells us that each side will view the other’s threats as always credible, which does not prove correct here. There are several instances in this case where one side (or at least individual decision-makers on one side) viewed the other side’s threats credibly, but it is a stretch to explain these instances in terms of constant calculus theory when there were key instances within the same case when one side did not view the other’s threats credibly. In particular, North Korea saw no credible threat in the United States during the sequences from August 6 through August

18; the United States saw no credible DMZ-area threat leading up to the August 18 attack; and some U.S. decision-makers, particularly Kissinger and the CIA, saw North Korean threats as incredible during crisis deliberations following the August 18 attack. Notably, however, constant calculus theory is the only of the four competing hypotheses on threat credibility that can actually account for the May 1976 White House memo. In that memo, North Korea’s “cheap talk” was assessed as credible in and of itself, simply because it occurred within a rivalry context, without explicitly considering the balance of capabilities, expressions of relative interests, or past word and deed. This aspect of the case suggests that, whatever constant calculus theory’s spotty record of predicting when an adversary’s threats are likely to be credible, it is the only of four perspectives that offers an explanation for the May 1976 White

House memo.

Assessing Hypothesis Performance

Recalling that this dissertation is principally concerned with the temporal consequences of rivalry behavior, I advanced two hypotheses in chapter II. H1 expects that backing down from a confrontation initiated by a rival challenger strengthens the future threat credibility of the state

279 that backed down, makes future threats more likely, and de-escalates the crisis in which the defender backed down. H2 expects that initiating coercive challenges against rivals does not strengthen the challenger’s future threat credibility. H1 and H2 both make predictions wholly congruent with the outcomes observed in this case. These hypotheses, which are a selective composite of insights from the literature on reputations and rivalries, shed light on two puzzles that manifest in the Panmunjom crisis: Why can some rivalries experience recurring crises without escalating to war, and why does the weaker power in an asymmetric rivalry so frequently challenge its stronger rival?

In the context of this chapter, H1tells us that the U.S. decision to back down repeatedly from North Korean challenges—as in the Pueblo crisis, EC-121 shoot down, October1969 helicopter incident, and every other incident prior to the 1976 crisis—made it more likely that the

United States would be challenged subsequently. Only the reputation for resolve (irresoluteness) logic offers an explanation for this phenomenon, which interprets the August 18 attack in the

JSA as at least partially a function of the U.S. decision to repeatedly back down from past North

Korean challenges while allowing rivalry conditions to persist. The reputational logics of honesty and resolve, meanwhile, explain the dramatic reduction in North Korean provocations after the U.S. decision to stand firm (in the form of Operation Paul Bunyan) as a demonstration of U.S. resolve that aligned U.S. word and deed; the United States made a subtle threat and marshaled a large, costly force to signal its seriousness. From the resolve perspective, the decrease in post-Panmunjom crisis provocations resulted from the United States taking a firm stand, which is an honest form of communication. From the honesty perspective, it did not

280 necessarily matter that the United States was demonstrating resolve by standing firm because it was the honesty of that act that explains the decrease in post-Panmunjom crisis provocations.

As in previous cases, the decision to back down—this time North Korea’s decision—did effectively end the crisis, as expected. But unlike the previous cases, for the first time in this rivalry relationship, we are able to bring evidence to bear on the second observation in H1—the effect of backing down on future threat credibility. The U.S. decisions to repeatedly back down over the decade prior to the Panmunjom crisis seemed to only encourage North Korean challenges as long as the hostile character of the rivalry remained unchanged. Even though the

United States had a history of backing down from North Korea, the latter nevertheless found

U.S. coercive threats during the Panmunjom crisis highly credible. It is not possible to conclusively attribute this outcome to a U.S. reputation for honesty because U.S. material superiority also played a role in making U.S. threats credible, as did the fact that the type of word and deed alignment was matching a threat with the marshaling of sufficient force to execute the threat. Nevertheless, it is notable that a history of honestly backing down—that is, not backing down after issuing a threat but rather backing down without issuing a threat—did not negatively impact U.S. credibility when it did later decide to issue threats.

H2, which posits that standing firm in response to a challenge can have an observably different effect on an adversary’s perceptions of your resolve than demonstrating resolve by initiating challenges, finds support in this case as well. It seems that the explanation for why

North Korean provocations did not have a systematic impact on U.S. perceptions of North

Korean resolve has to do with a different type of reputation—a reputation for bluffing. North

281 Korea’s history of grandiose threat-making without consistent follow through is the most plausible explanation for why Captain Bonifas dismissed North Korea’s very direct in-person threats on August 18; the North Korean guards directed him to cease trimming the poplar tree under threat of death, multiple times, yet he ignored their threats even though he was told that if he believed a violent situation would occur he should immediately withdraw.165 LtCol Vierra,

Captain Bonifas’ supervisor, had ordered the tree trimming on August 18 with full cognizance of

North Korea’s heightened threat-making and North Korea’s August 6 warning against trimming the poplar tree, but considered these threats as bluffs with “no more than usual” intentions behind them.166 Even in this case, however, the correlation between a history of bluffing and a lack of future threat credibility is not perfect. During the WSAG deliberations over whether and how to retaliate for the Panmunjom attack, despite North Korea’s history of bluffing, Deputy Secretary of Defense Bill Clements was likely not alone in believing that North Korea would attack U.S. bases in South Korea if the United States executed even a limited strike against North Korea.167

So even though we do observe direct evidence in this case that a history of bluffing can undermine threat credibility and prevent a reputation for resolve from accruing, such a claim is clearly not axiomatically true. It would be accurate to characterize this case as supporting the somewhat modest H2 claim that initiating challenges does not have a determinate effect on the

165 Jung, Kyeot-aeseo Bon Kim Jeong-Il, 202-4; Head, Short, and McFarlane, Crisis Resolution, 160-5. 166 Head, Short, and McFarlane, Crisis Resolution, 161. 167 U.S. Department of State, August 19, 1976. “Minutes of Washington Special Actions Group Meeting, Washington, August 19, 1976, 8:12-9:15am,” FRUS, Vol. E-12, Documents on East and Southeast Asia, 1969-76, document 285; National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27, Ford.

282 challenger’s reputation for resolve—implying that standing firm may do more for future threat credibility than being a provocateur.

There may be two reasons for this. The first could be that reputations are not earned

“cheaply,” meaning that provoking through small-scale attacks near the bottom of a metaphorical

“escalation ladder” is not a costly enough signal to earn a reputation for resolve. This seemed to be how Kissinger sized up North Korean intentions during the WSAG meetings. This raises a valid question about whether a reputation for resolve needs to be earned, or whether it is assumed until proven otherwise (by bluffing). The second reason could be that, as we have seen in the perceptions of LtCol Vierra and Captain Bonifas, North Korea’s history of making empty threats really did serve as a basis for not believing threats in the current context. Nothing about these two explanations makes one necessarily incompatible with the other. However, when U.S. decision-makers in previous crises found North Korea’s threats credible only in the immediate aftermath of a provocation but not before, the reasoning I posited was that the actual challenge represented word-deed alignment—that is, North Korean honesty. The effect on North Korea’s reputation was not cumulative. Because North Korea’s provocations did seem to bolster its credibility in the context of the crisis of the day, that there is likely some connection between challenge initiation and threat credibility, but it is a tenuous connection with little staying power.

Chapter VI: The North Korean Nuclear Crisis (1993-4)

The crises of previous chapters shared a common rhythm: a hostile dyadic context gives rise to North Korean empty threat-making, followed sometime later by a North Korean military provocation, followed in turn by a U.S. response that ends the crisis. In all but the Panmunjom crisis, the United States responded by backing down—that is, by not retaliating and/or offering concessions of some kind to North Korean demands. In the 1976 Panmunjom crisis, the United

States stood firm—using coercive signaling to both send a message of resolve and to deter another North Korean attack as the UN Command cut down a contentious poplar tree—but also made conditional reassurances to North Korea and modified its demands to accommodate its rival. During the North Korean nuclear crisis, the focus of this chapter, the issue at stake was very different than in previous crises, and the role of challenger and defender difficult to identify at times. The proximate issue at stake was North Korea’s march toward developing nuclear weapons and its unprecedented attempt to withdraw from the Nuclear Nonproliferation Treaty

(NPT). Much of the coercive signaling that took place in this case can only be appreciated in its historical context. Still, the reputational framework that guided the inquiry of previous chapters is, as will be shown, a useful way of understanding what happened and why.

As with previous chapters, the narrative for this episode is built on the answers to several questions that help us identify the temporal effects of backing down or standing firm and the sources of threat credibility. What was the recent conflict history of the dyad prior to this incident? What was the geopolitical context of the crisis? Who initiated coercion and why?

How did each view the credibility of the other’s threats during the crisis and why? What actions did each side actually take during the crisis and what can be inferred about the motivations of

283 284 each side from these actions, and to what extent does process tracing evidence support these inferences? Finally, how did the crisis end?

Historical Context

Following the resolution of the Panmunjom crisis in 1976, North Korea largely curbed its use of military violence against the United States, focusing it instead on South Korea. Despite remaining on a very hostile footing throughout the 1980s, the fall of the Berlin Wall in 1989 and the definitive end of the Cold War by 1991 had ushered in the beginnings of a short-lived sea change in U.S. and South Korean relations with North Korea. But North Korea’s longstanding clandestine pursuit of nuclear weapons put it in direct conflict with U.S. interests in preventing the creation of new nuclear states.

North Korea v. United States

With the end of the Cold War and the cessation of military provocations aimed at the

United States, the contemporary basis for hostility between the United States and North Korea stemmed from the latter’s persistent acts of violence against South Korea—a U.S. treaty ally—as well as its use of and support for acts of political terrorism in the 1980s. North Korea’s nuclear ambitions were a U.S. concern during the Cold War, but its nuclear program was sufficiently secretive and nascent that it did not engender feelings of crisis on the Korean Peninsula. By the end of the Cold War, North Korea’s desire for nuclear weapons had become a symptom of endemic feelings of insecurity and persistent rivalry conditions with the United States and South

Korea. Conveying this sentiment of extreme vulnerability in 1980, one North Korean official pointed out to an Italian communist party delegation that North Korea was at the center of the

285 Sino-Soviet conflict and the Sino-U.S. rapprochement, and that all the while “Below the 38th

Parallel a million soldiers have arms turned against us…If we are not careful, if we do not make the correct moves, we run the risk of being crushed or sold.”1 Xenophobia and alienation from the outside world is as much a part of North Korean identity as any of its indigenous traditions.2

North Korea’s interest in obtaining nuclear weapons dates back to the Korean War, when

Eisenhower employed a dubious threat of nuclear coercion in hopes of resolving the long- running standoff.3 Although the threat of nuclear use has been shown not to have played any meaningful role in resolving the Korean War, some in the United States, including Nixon, believed nuclear threats helped bring the Korean War to an end; North Korea has ever since felt that its vulnerability to nuclear attack could only be mitigated by its possession of nuclear arms.4

So after the Korean War ended in armistice, North Korea started a gradual, long-term process to establish a nuclear weapons program, documented as early as 1956, when the Soviet Union had agreed to nuclear cooperation with North Korea, at the latter’s request.5 In 1964, Kim Jong-il had a letter delivered to China’s Mao Zedong requesting cooperation on nuclear development for

North Korea.6 Though China refused that specific request, it would not be North Korea’s last,

1 As quoted in Thomas P. Bernstein and Andrew J. Nathan, “The Soviet Union, China, and Korea,” in The U.S.- South Korean Alliance, eds. Gerald L. Curtis and Sung-joo Han (Lexington, MA: Lexington Books, 1983), 119. 2 Brian R. Myers, The Cleanest Race: How North Koreans See Themselves and Why It Matters (Brooklyn, N.Y.: Melville House Publishing, 2011). 3 Roger Dingman, “Atomic Diplomacy during the Korean War,” International Security 13, no. 3 (1988/89): 60-86. 4 For a discussion of this history, see Jonathan D. Pollack, No Exit: North Korea, Nuclear Weapons and International Security (London: International Institute for Strategic Studies, 2011); Michael J. Mazarr, North Korea and the Bomb: A Case Study in Nonproliferation (New York, NY: St. Martin’s Press, 1995), 15-34. 5 Alexandre Y. Mansourov, “The Origins, Evolution, and Current Politics of the North Korean Nuclear Program,” Nonproliferation Review 2, no. 3 (1995): 25-6. 6 Oberdorfer, The Two Koreas, 252-3.

286 and its interest in nuclear weapons persisted throughout the Cold War.7 North Korea’s efforts at nuclear outreach, combined with its secret indigenous research, eventually yielded what U.S. reconnaissance satellites identified in 1980 as a 5-megawatt nuclear reactor facility in the city of

Yongbyon, an observation that put North Korea’s security goals directly at odds with U.S. stated security interests.8

Without the use of coercion but with heavy reliance on the Soviet Union to pressure

North Korea, the United States managed to gain North Korea’s accession to the NPT in 1985, in exchange for nuclear power plants from the Soviet Union. One of the requirements of NPT membership is the conclusion of a comprehensive safeguards agreement with the IAEA within

18 months of accession, in order to ensure that all nuclear material in a country is not diverted for weaponization purposes.9 Because of a clerical error, North Korea did not receive the correct paperwork until the end of the 18 month period, followed by a further 18 months with no progress toward concluding a comprehensive safeguards agreement.10 As a result, North Korea was by default a noncompliant NPT member that was increasingly suspected of reprocessing spent fuel rods from its nuclear facilities, a key step in the development of nuclear weapons. By the end of 1989, the U.S. intelligence community had assessed that North Korea had reprocessed enough fuel rods to create one or two plutonium-based nuclear weapons.11

7 Ibid. 8 Joel S. Wit, Daniel B. Poneman, and Robert L. Gallucci, Going Critical: The First North Korean Nuclear Crisis (Washington, DC: Brookings Institution, 2004), 1. 9 Federation of American Scientists, June 7, 1994. “Director General’s Statement to the Board of Governors on the Democratic Peoples’ Republic of Korea,” Available online: www.fas.org/news/un/dprk/dgbg1994n01.html. 10 Wit, Poneman, and Gallucci, Going Critical, 4. 11 Jeffrey R. Smith, “N. Korea and the Bomb: High-Tech Hide and Seek,” Washington Post (April 27, 1993), A1.

287 In an effort to curb North Korea’s pursuit of nuclear weapons, the George H.W. Bush administration sought “comprehensive engagement” with North Korea, which allowed for improved bilateral relations if North Korea abandoned its nuclear weapons acquisition efforts.12

National Security Review-28 (NSR-28), an interagency study proposing the U.S. government way ahead on North Korea policy, provided a path to North Korea’s nuclear abandonment, which included maintaining a robust military presence for the purpose of general deterrence against North Korean aggression, promoting dialogue between North and South Korea, gaining

North Korean accession to nonproliferation regimes, and preventing North Korea from accessing certain enrichment and reprocessing technologies.13 In essence, NSR-28 recommended adding diplomatic engagement with North Korea to the existing status quo ante on the Korean

Peninsula. NSR-28 also identified other U.S. goals vis-à-vis North Korea, including preventing

North Korea-sponsored terrorism and constraining its sales of nuclear, chemical, and ballistic missile materials.14 The “comprehensive engagement” approach was significant because it conditioned U.S. policy toward North Korea principally on the basis of the latter’s pursuit of nuclear weapons; other sins may have been more or less forgivable, but dismantlement of North

Korea’s nuclear program became the priority as the U.S. presidency transitioned from the Bush administration to the Clinton administration in 1993. NSR-28 suggested that the presence of

U.S. nuclear weapons were undermining U.S. goals relative to the Korean Peninsula, a key

12 Joel S. Wit, “North Korea: Leader of the Pack,” Washington Quarterly 24, no. 1 (2001): 1. 13Joel S. Wit, “Clinton and North Korea: Past, Present, and Future,” in Solving the North Korean Nuclear Puzzle, eds. David Albright and Kevin O’Neill (Washington, DC: Institute for Science and International Security, 2000), 203. 14 Ibid.

288 judgment that ensured the Korean Peninsula would not be exempted from the U.S. global withdrawal of non-strategic nuclear weapons announced on September 28, 1991.

Geopolitical Context

The three most significant geopolitical trends relating to the Korean Peninsula that framed the nuclear crisis were the U.S.-Russian rapprochement that accompanied the withering of the Soviet empire and the ascendance of proliferation as a U.S. national security priority;

North Korea’s growing isolation from the international community; and renewed but ultimately failed attempts at improving relations between North and South Korea after more than a decade of occasionally intense dyadic violence as North Korea diverted its military focus away from the

United States and toward South Korea after the 1976 Panmunjom crisis.

The thaw in U.S.-Soviet relations that came with the latter’s proclamation of glasnost and perestroika at the end of the Cold War eliminated the only peer competitor of the United States in terms of nuclear arms and power projection; the United States had no equal in military terms.

Although the unprecedented era of U.S. unipolarity in the 1990s gave birth to a range of opinions about the durability of a unipolar order and the implications of liberal hegemony, one consequence was that U.S. “grand” strategy no longer had a grand, simplifying focus.15 Without a singular principal adversary to plan against, proliferation of weapons of mass destruction

(WMD) had become the U.S. foreign policy establishment’s primary security concern. During the George H.W. Bush administration, this sentiment manifested in President Bush’s September

27, 1991 announcement that it was withdrawing all ground-launched non-strategic nuclear

15 For the range of debates about the durability and desirability of U.S. unipolarity, see Michael E. Brown, Owen R. Cote Jr., Sean M. Lynn-Jones, and Steven E. Miller, Primacy and Its Discontents: American Power and International Stability (Cambridge, MA: The MIT Press, 2008).

289 weapons deployments globally, which served the U.S. nonproliferation strategy vis-à-vis North

Korea at the same time that it served as a confidence-building measure in U.S.-Soviet relations.

16 The decision to withdraw nuclear weapons from the Korean Peninsula was reached at least a year earlier than the announcement of global withdrawal, but implementation of the decision was delayed because of the Bush administration’s desire to preserve the status quo ante in terms of regional deterrence in Asia as it focused its attention on repelling Saddam Hussein’s invasion of

Kuwait in 1990.17

The Clinton administration inherited its predecessor’s emphasis on nonproliferation as a national security priority, embracing it as the principal focus of its foreign policy during the first term. The Pentagon’s 1993 Bottom-Up Review put proliferation at the forefront of security concerns for DoD over the course of the next decade,18 and the companion National Military

Strategy issued by the U.S. Chairman of the Joint Chiefs of Staff identified proliferation as the primary threat to U.S. security.19 So while it may be debatable among academics what a nation’s most vital security interests should be, there is no debate that the events of the North Korean nuclear crisis coincided with an increasing U.S. prioritization of nonproliferation above other security objectives, offering a historically contextual explanation for why the United States would be willing to seek coercive options against North Korea for the latter’s pursuit of a nuclear program in the first place.

16 George H.W. Bush, September 27, 1991, “Address to the Nation on Reducing the United States and Soviet Nuclear Weapons,” in the UC Santa Barbara American Presidency Project, Available online: http://www.presidency.ucsb.edu/mediaplay.php?id=20035&admin=41. 17 Mazarr, North Korea and the Bomb, 60-1. 18 Les Aspin, Report on the Bottom Up Review (Washington, DC: U.S. Department of Defense, October 1993), 5-12. 19 “Joint Chiefs See Proliferation as Top Threat,” Defense Daily (January 27, 1994): 133.

290 The second issue giving important color to the onset of the nuclear crisis was North

Korea’s growing isolation from the international community. As discussed in previous chapters, during the Cold War North Korea sought to navigate the Sino-Soviet split by maintaining equidistant relations from both China and the Soviet Union. Although much of the international communist movement did embrace North Korea through much of the 1950s and 1960s, its foreign policy tactics—relying on military provocations—contributed to its own international political isolation well before the end of the Cold War. The particularly gruesome beating of

U.S. soldiers that sparked the Panmunjom crisis in 1976 laid bare a kind of hostility that even the communist bloc at the time found distasteful.20 North Korea’s use of terrorist tactics against

South Korea in the 1980s, most notably in a 1983 presidential assassination attempt in Burma, created further distance between North Korea and once friendly governments that rejected such behavior.21 By the late 1980s, North Korea had also developed an international reputation as a deadbeat debtor, defaulting on most of the loans and credits it accrued during the 1960s and

1970s. After the 1976 Panmunjom crisis, for several years Kim Il Sung would meet with foreign leaders from the Soviet Bloc and Non-Aligned Movement and insist that its insecurity drove its military spending, which in turn made it difficult to repay debts.22 The end of the Cold War only

20 “Telegram from Beijing to Bucharest, SECRET, Urgent, No. 066.252,” August 25, 1976, Archives of the Romanian Ministry of Foreign Affairs, Matter 220 - Relations with the Democratic People's Republic of Korea, Obtained by Izador Urian and translated by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/114117. 21 For a discussion of how North Korean provocations contributed to international isolation, see Yong-ho Kim, “North Korea’s Use of Terror and Coercive Diplomacy: Looking for Their Circumstantial Variants,” Korean Journal of Defense Analysis 14, no. 1 (2002): 45-67. 22 Brian C. Koh, “North Korea in 1987: Launching a New Seven-Year Plan,” Asian Survey 28, no. 1 (1988): 62-70; “TELEGRAM 066.569 from the Romanian Embassy in Pyongyang to the Romanian Ministry of Foreign Affairs,” February 28, 1978, AMAE, Folder 784/1978, Issue 220: Features of political-diplomatic relations between the Democratic People's Republic of Korea and some countries in Europe, Asia, Africa, America (Cyprus, Spain, USA,

291 brought further isolation of North Korea; not only did its one-time allies East Germany and

Romania transform politically, but Russia began cooperating with the United States on nonproliferation and arms control at the same time that it normalized its long frozen relations with South Korea in 1990.23 By 1992, even the Chinese had normalized relations with South

Korea through a process that started in 1988, and in 1993, the month before North Korea declared its intention to withdraw from the NPT, Russia informed North Korea it was no longer willing to fulfill its mutual defense treaty with North Korea, formally leaving it with no clear allies in Northeast Asia.24 In response to the news that Russia and South Korea had normalized relations, Kim Il Sung sent a letter to Moscow alluding to its need to develop nuclear weapons:

“In such circumstances, the DPRK-Soviet alliance would become meaningless and we would have to prepare countermeasures to produce our own nuclear weapons to replace the weapons we had depended on for our alliance.”25

The third dimension informing the geopolitical context of the nuclear crisis was the state of relations between North and South Korea, which directly contributed to the onset of the crisis, if not to its resolution. North-South relations were on a historically positive track before being subsumed by contestation over North Korea’s nuclear program. In 1987, South Korea transitioned from autocracy to fledging democracy, electing President Roh Tae-woo to office. In

Bangladesh, Philippines, India, Indonesia, Japan, Pakistan, Sri Lanka, Central African Republic, Egypt, Gabon, Iraq, Iran, Libya, Nigeria, Mozambique, Syria) January 7, 1978 – September 23, 1978, Obtained and translated for NKIDP by Eliza Gheorghe. Available online: http://digitalarchive.wilsoncenter.org/document/116417. 23 Jae-nam Ko and Sung-whan Choi, “Russia’s South Korea Policy: A Comparison of Soviet and Post-Soviet Dynamics,” East Asian Review 14, no. 2 (2002): 39-60. 24 Hak-joon Kim, “The Establishment of Chinese-South Korean Diplomatic Relations: 1988-1992,” Journal of Northeast Asian Studies 13, no. 2 (1994): 31-8; Mazarr, North Korea and the Bomb, 96. 25 As quoted in Dong-won Lim, Peacemaker: Twenty Years of Inter-Korean Relations and the North Korean Nuclear Issue (Stanford, CA: Stanford Asia-Pacific Research Center, 2012), 94.

292 1988, President Roh issued the July 7 Special Declaration, indicating his vision for transitioning

North Korea from an enemy into a partner. According to Lim Dong-won, a South Korean official who served in many roles in the 1980s and 1990s, including as a principal in inter-

Korean dialogues, President Roh seized on the 1989 U.S.-Soviet agreement in Malta to end the

Cold War by announcing that he sought to bring an end to the Cold War on the Korean

Peninsula. To capitalize on what he saw as structural momentum incentivizing rapprochement,

President Roh invited North Korea to begin what came to be called “the North-South High-Level

Talks,” starting with meetings as early as February 1989 and continuing for a year and a half before yielding two landmark documents in the history of inter-Korean relations: a general rapprochement vision document called the “Basic Agreement,” and the Joint Declaration for the

Denuclearization of the Korean Peninsula (JDD).26 The administration of South Korean

President Kim Young Sam, which succeeded Roh Tae-woo in 1992, did its part to try and continue implementation of the 1991 agreements negotiated with North Korea, but like his predecessor, Kim Young Sam’s administration was divided into those who believed the nuclear issue with North Korea could either be resolved in parallel with or subsequent to an improvement in inter-Korean relations and those who believed relations could not and should not improve if North Korea continued its march toward nuclear armament.

Despite its lack of consensus about how to approach North Korea policy, Kim Young

Sam’s administration attempted to follow through on the terms of the Basic Agreement and JDD

26 As revealed below, South Korean preferences during the nuclear crisis were somewhat erratic, reflecting what several participants in the United States and South Korea observed as several constituencies within the Roh administration—and his successor’s administration—who either believed North Korea could not be trusted, or who believed that nuclear negotiations needed to be resolved before proceeding with a broader rapprochement. See Lim, Peacemaker, 108-13.

293 before being overtaken by the coercive diplomacy that started in 1993 between the United States and North Korea. The Basic Agreement identified a goal of improving relations across every dimension of social life, including the economic, cultural, and eventually political and military dimensions too. North Korea proposed the JDD as a favorable response to the Bush administration’s decision to remove nuclear weapons from the Korean Peninsula.27 Agreement on the Basic Agreement and JDD in 1991 were contingent on satisfying two major North Korean demands that required coordination with, and to some degree concurrence from, the United

States: canceling the U.S.-ROK Team Spirit annual military exercise in 1992, and withdrawing

U.S. non-strategic nuclear weapons from the Korean Peninsula.

As mentioned previously, the U.S. announcement that it was withdrawing its non- strategic nuclear weapons from global deployment in September 1991 was a major enabling condition for progress in North-South relations. The cancellation of Team Spirit in 1992 was a longstanding North Korean demand whose agreement was floated to the North Koreans in late

1991, and formally announced during President Bush’s summit with Roh Tae-woo in January

1992. As a ritualistic symbol of reciprocity and a practical security precaution, the North Korean military would mobilize and exercise its forces annually in response to Team Spirit, which was not only financially costly for North Korea over time, but also conditioned the North Korean military and populace to view Team Spirit as a meaningful representation of U.S. aggressive intentions.28 Even though the Basic Agreement and JDD would become important milestones

27 Lim, Peacemaker, 104-7. 28 Chang-hui Kim, “Pukhan, 80 Nyun-dae Jungban Seodok-aeseo Haekmuljil Goo-eep” (North Korea, Purchased Nuclear Material from West Germany in the Mid-1980s), Sindong-a (1995): 134.

294 for dialogue with North Korea in subsequent years and decades, both fell apart almost as soon as the terms of the agreements required implementation. The collapse of the JDD was germane to the nuclear crisis because North Korea’s unwillingness to allow South Korean inspections of

North Korean nuclear facilities—which were required under the terms of the JDD—introduced further tensions and suspicions about North Korea’s increasingly non-secretive nuclear weapons program.

Although 1992 started with the promise of progress in both North-South relations and efforts to halt North Korea’s nuclear ambitions, the IAEA’s discovery of numerous violations combined with North Korea’s refusal to accept on-demand inspections by either the IAEA or

South Korea was sufficient to bring on North Korean threats of war by the end of the year, and a crisis of nonproliferation in 1993, the resolution of which all but ensured its recurrence the following year. Starting with Hans Blix’s visit to Yongbyon, North Korea in May 1992, the

IAEA was able to inspect specific facilities and areas that North Korea designated. In parallel with IAEA inspections throughout 1992, however, negotiations between North and South Korea over implementation of the JDD through the Joint Nuclear Control Committee (JNCC)29 were at an impasse over South Korea’s demands to conduct “challenge inspections” of any suspect facility on 48 hours’ notice, which nonproliferation advocates believed was crucial both to avoid

North Korean deception and to trace the amount of past plutonium reprocessing that may have taken place. This stalemate might not have affected the continuation of North Korea-controlled

IAEA inspections were it not for the fact that, during the JNCC talks, South Korea wielded the

1993 resumption of Team Spirit as a threat that would only be lifted if North Korea accepted

29 The JNCC was established in December 1991 to implement the JDD.

295 challenge inspections. During the JNCC talks, North Korea would not agree to challenge inspections, and in addition to the cancellation of all future Team Spirit exercises, North Korea demanded that it be allowed to inspect U.S. bases in South Korea to verify that it no longer deployed nuclear weapons in South Korea.

In sum, the saga of on-again, off-again North-South talks became intertwined with nuclear diplomacy between the United States and North Korea, leading to an exchange of threats toward the end of 1992 and the onset of crisis in 1993.

Overview of the Incident

The crisis dimension of this case consists of two distinct phases, separated in time by a brief interlude during which U.S. policymakers tried but failed to prevent the second phase from ever occurring: from March 8, 1993 (with roots in threats from October 1992) to June 11, 1993, and from March 19 to June 16, 1994. The way in which the second part of the crisis was very much a repeat of the first reminds us that how a crisis ends can matter a great deal for the likelihood of something similar recurring. In both parts of the crisis, North Korea employed a brinkmanship strategy of threat-making that would escalate in response to pressure. The United

States entered the crisis with an approach of “gradual escalation” and expressed a preference for coercion for most of the crisis, but in the face of new or escalating North Korean threats, often resorted to short-term compromises and conciliation to meet North Korean demands.

From the beginning of the crisis, North Korea declared to U.S. negotiators it would only abandon what it claimed was its peaceful nuclear program in exchange for the construction of light-water reactors (LWRs), and would only agree to continued IAEA inspections of its facilities if the United States and South Korea agreed to permanently cancel its Team Spirit

296 military exercise; the United States agreed to offer security assurances and a pledge of non- aggression against the North; and the United States agreed to allow North Korean inspections of is bases in South Korea to verify that U.S. nuclear weapons had been withdrawn from the Korean

Peninsula. North Korea would also insist that all negotiations take place directly with the United

States, something that some in the Clinton administration, as well as in South Korea, viewed as a concession or “reward” in itself. At no point during the crisis would North Korea allow the original U.S. and South Korean demand for some form of challenge inspections to apply to any of North Korea’s suspected nuclear facilities. Faced with coercive pressure from North Korea, however, the United States would occasionally alter its objectives to also include preventing

North Korea’s withdrawal from the NPT, preventing a break in the continuity of IAEA nuclear safeguards (which ensured nuclear enrichment did not become weaponized), and restoring nuclear safeguards after North Korea temporarily expelled the IAEA.

The first part of the crisis started with the combined breakdown in North-South talks and the IAEA’s growing discoveries of inconsistencies between North Korea’s plutonium reprocessing claims and evidence it found during its limited inspections of North Korean facilities that the latter very much controlled. In October 1992, North Korea threatened that it would end “peace talks” with South Korea and take other unspecified measures if Team Spirit went forward in March 1993, threats which U.S. officials heard but ignored. Throughout

February 1993, North Korea began making coercive threats designed to prevent the application of sanctions and to get the triumvirate of United States, IAEA, and South Korea to cease its pressure for the North to accept special inspections. On March 8, 1993, Kim Il-Sung mobilized

297 the nation’s military and declared a state of “semi-war,” a term that had not been used since 1983 when North Korea attempted to assassinate South Korea’s president and had only been used three other times in North Korea’s history.

On March 12, North Korea’s First Vice Foreign Minister Kang Sok Ju announced in a press conference that if North Korean non-compliance with the IAEA’s request for special inspections was referred to the UN Security Council for sanctions, there would be “grave consequences.”30 Also on March 12, 1993, North Korea formally began the 90-day process to withdraw from the NPT. The day before the UN Security Council issued a presidential statement urging North Korea to end its NPT withdrawal process, on April 7, 1993, North Korean media argued that concerns about its nuclear program “should be resolved through negotiation between the DPRK and the United States,” and that if the UN Security Council pursued sanctions it would lead to “unpredictable consequences.”31 The first part of the crisis would end only hours before North Korea’s withdraw from the NPT would be formally effectuated, late into the night on June 11, 1993. The crisis temporarily abated at that point because of a satisficing agreement that required little of the United States and even less of North Korea; the former would forswear invasion and declare no hostile intent toward the North (and one month later agree to exploring furnishing North Korea with LWRs), and the latter would agree to suspend its withdrawal process from the NPT (but not to either rejoin the NPT nor to allow any meaningful nuclear inspections).

30 “Statement Notes Withdrawal,” Korean Central Broadcast System (March 12, 1993); Wit, Poneman, and Gallucci, Going Critical, 37. 31 “Pyongyang Berates IAEA, UN over Possible Scrutiny,” Foreign Broadcast Information Service (FBIS) Trends (April 12, 1993).

298 Only three months later, it started becoming clear that negotiations over North Korea’s nuclear program may not resolve the issue at stake, and North Korea’s participation in North-

South JNCC talks had a largely perfunctory quality. As the United States wielded veiled threats and crept toward a decision to send military reinforcements and pursue UN sanctions against the

North, the North continued with war preparations, blocking IAEA inspections, and advancing its nuclear programs. When the United States announced it would be sending Patriot missiles and threatened the resumption of Team Spirit in 1994, North Korean threats resumed in earnest, spiking with a panic-inducing threat to destroy Seoul on March 19, 1994. The crisis came to an end three months later because of a negotiated agreement reached by former President Jimmy

Carter, who intervened in spite of White House preferences and successfully struck a deal that prevented the crisis from escalating to violent conflict.

Unlike in previous cases, coding the initial challenger in this case is debatable, though I make the case that the United States was actually the initial challenger. On the one hand, the

United States and South Korea wielded the threat of resuming Team Spirit in 1993, having made clear at the time that the cancellation of Team Spirit in 1992 was only a one-year hiatus. In fairness to the conventional view casting North Korea as challenger, the threat of a highly scripted military exercise, no matter how large, can really only be considered a coercive threat if the defender sees it as a threat and its employment is intended to achieve something diplomatically or politically. But on the other hand, as I discuss subsequently in the section addressing North Korea’s perspective, North Korea did frequently characterize Team Spirit as a threat and the threat of its resumption was intended to compel North Korea to grant the IAEA—

299 or at a minimum South Korea—the ability to conduct challenge inspections of suspect nuclear facilities.

Debating Coercion

The resumption of Team Spirit in 1993 was part of Clinton’s “gradual escalation” approach to compel North Korea to cooperate with IAEA inspection demands, but was also to some extent coercion by path dependence. The Director of the Defense Intelligence Agency at the time admitted that “Anything that we do, however defensive it looks, is offensive to them.”32

Mike Mazarr, a scholar who interviewed dozens of Pentagon officials about the crisis at the time, reported that Team Spirit’s resumption at the 1992 SCM was deliberately announced with the intention of using it as a lever of influence in negotiations with the North.33 Charles Kartman, then Country Director for Korea at the State Department, described the thinking of most U.S. officials at the time: “The advantages of using Team Spirit as a club appealed to many people…that was the mind-set here.”34 But the decision to conduct the Team Spirit exercise in

1993 was made by the outgoing Bush and Roh administrations in 1992, and inherited by the

Clinton and Kim Young Sam administrations. Clinton and Kim did not seek to undo a decision of the past administration, in part because nonproliferation was the cornerstone of Clinton’s foreign policy and his staff felt the need to appear resolved on the issue, and in part because of

32 Interview with General James Clapper, as quoted in Leon V. Sigal, Disarming Strangers: Nuclear Diplomacy with North Korea (Princeton, NJ: Princeton University Press, 1998), 75. 33 Mazarr, North Korea and the Bomb, 91. 34 Interview with Charles Kartman, as quoted in Sigal, Disarming Strangers, 46.

300 “inertia…Events already set in motion by the time President Clinton assumed office appeared to follow an inexorable course.”35

During working level meetings between the United States and North Korea in January in

New York in January 1993, North Korea warned the United States directly about moving forward with Team Spirit: “…the whole focus was, if you guys start Team Spirit, we will respond with drastic measures,” which U.S. officials dismissed as the kind of empty bluster that often accompanied Team Spirit exercises.36 Adding to the gradual “screw tightening” on North

Korea, on February 22, 1993, U.S. intelligence officials took the historically assertive step of presenting to the IAEA Board of Governors what it considered irrefutable evidence that there were at least two undeclared, uninspected sites in North Korea wherein the North was engaged in illegal nuclear activity that had no explanation other than weaponization. Faced with North

Korean intransigence on the issue of IAEA inspections and the breakdown of North-South talks,

U.S. officials knew offering this intelligence to the IAEA would result in North Korea’s case being referred to the UN Security Council, a necessary first step in the process of establishing global sanctions on North Korea’s proliferation activity.37 Only the day prior to the IAEA meeting, on February 21, 1993, North Korea made its starkest threat yet: “If any special inspections and sanctions are enforced on us, and if the sacred lands of our fatherland are

35 Wit, Poneman, and Gallucci, Going Critical, 18-19. 36 Interview with anonymous U.S. official as quoted in Sigal, Disarming Strangers, 47. 37 Jeffrey R. Smith, “North Korea and the Bomb: High-Tech Hide and Seek,” The Washington Post (April 27, 1993): A1, A11.

301 trampled underfoot by big countries, this will be a dangerous fuse that will drive all lands, including the North and the South, into the crushing calamities of war.”38

The Clinton administration was shocked by North Korea’s March 12 announcement of withdrawal from the NPT, despite explicit verbal and written warnings that it might do so on at least four occasions between February 15 and February 22.39 In the words of a senior administration official commenting about North Korean threats generally, it was difficult to separate “the nuggets of reason” from “the ocean of vitriol.”40 The administration debated whether it was prudent for the IAEA’s referral of North Korean non-compliance to coincide with the Team Spirit exercise, “But the standard (albeit erroneous) wisdom was the situation would cool down once Team Spirit was over.”41 U.S. Ambassador to Seoul, William Clark, sent a memo to the State Department after the “semi-war” declaration but before the NPT withdrawal announcement that tried to downplay North Korea’s threats, arguing that there was “very little reason for serious concern” about North Korea’s escalation of words.42 In spite of North Korea’s threats of war, in other words, the general consensus prior to March 12, 1993 was that North

Korea’s rhetoric did not indicate an imminent attack of any kind because the conduct of past

Team Spirit exercises also incited North Korean hyperbole without incident. On the issue of sanctions, however, the U.S. intelligence community did assess North Korea’s threat credibly— that UN sanctions would result in, if not war then at least provocations—presumably based on

38 Cited in FBIS, East Asia Section (February 22, 1993): 11-13. 39 Mazarr, North Korea and the Bomb, 96-7. 40 Interview with anonymous U.S. official as quoted in Sigal, Disarming Strangers, 59. 41 Wit, Poneman, and Gallucci, Going Critical, 19. 42 Wit, Poneman, and Gallucci, Going Critical, 25.

302 the dramatic changes in the North Korean military that had been happening over the course of the several months prior.43

Despite North Korea’s withdrawal announcement in response to pressure for inspections and the resumption of Team Spirit, the United States refused to abandon the principle of “gradual escalation” while finding itself needing to respond to North Korea’s credible escalation of the situation with compromises and conciliation once it realized North Korean threats were more than propaganda. A White House official commented at the time that even though the U.S. preference was coercion, “To some extent the diplomatic effort was forced on us by tactical considerations.”44 The United States took a conciliatory tone in response to North Korea’s enhanced war posture and process of NPT withdrawal, being careful to offer no public threats but instead to encourage North Korea to return to the negotiating table.45 The U.S. commitment to gradual escalation foreclosed on the immediate possibility of a preventive strike on North

Korean nuclear facilities or similarly extreme military actions—though such options were being discussed at DoD and within the interagency—and forced the United States to first build the multilateral coalition necessary to eventually be able to apply UN sanctions. The escalatory logic framing U.S. choices was that there could be no sanctions without diplomacy, and no military action without sanctions.46

43 Michishita, North Korea’s Military-Diplomatic Campaigns, 106; Wit, Poneman, and Gallucci, Going Critical, 32. 44 Interview with anonymous U.S. official as quoted in Sigal, Disarming Strangers, 59. 45 Alan Elsner, “U.S. Takes Mild Approach to Korean Nuclear Threat,” Reuters (March 16, 1993); George Gedda, Associated Press (March 12, 1993). 46 This logic would later be called “hawk engagement,” the idea being that attempted diplomacy was a necessary prior condition to use force in a way that the international community would condone. See Victor D. Cha, “Hawk Engagement and Preventive Defense on the Korean Peninsula,” International Security 27, no. 1 (2002): 40-78.

303 And so it was that when North Korea announced that their nuclear program was “not a matter to be discussed in the UN arena but a problem that should be resolved through negotiation between the DPRK and the U.S.,” the United States complied.47 With the conclusion of Team

Spirit on March 19 and North Korea’s subsequent downgrade of its “semiwar” state, on April 22,

1993, the State Department announced it was willing to negotiate North Korea’s return to the

NPT, appointing Assistant Secretary Robert Gallucci as the lead U.S. negotiator.48 Whether or not diplomatic engagement should have been imbued with the meaning of “reward” or

“punishment” is a matter of debate, but the United States did imbue it with such meaning, deliberately withholding senior level negotiations with North Korea before the NPT withdrawal announcement only to agree to them after the announcement as a means of preventing North

Korea’s total withdrawal.49 On June 2, U.S.-North Korea high-level talks with Gallucci began, even though the U.S. preference remained coercion: “Everyone on the American delegation was gung-ho to sock it to these sons-of-bitches.”50 Preferences notwithstanding, the United States resorted to minimal concessions in exchange for minimal promises from North Korea, gaining the latter’s agreement to suspend its process of withdrawal from the NPT—though not explicitly opting back in to the NPT—late on June 11, only hours before the formal process would have been complete. In exchange for this suspension of withdrawal, the United States reiterated the security assurances that Undersecretary Kanter offered North Korea during his “political level”

47 “North Korea Calls for Direct U.S. Talks on Nuclear Inspections,” United Press International (April 5, 1993). 48 Douglas Jehl, “U.S. Agrees to Discuss Arms Directly with North Korea,” New York Times (April 23, 1993): A10. 49 The initial U.S. decision to withhold “high-level talks” is historically curious given that the United States had initiated regular lower level talks with North Korea starting as early as 1989. Daniel Russel,“U.S.-North Korean Relations,” in Current Issues in Korean-U.S. Relations: Korean American Dialogue, Man-woo Lee (Seoul: Institute for Far Eastern Studies, 1993), 45. 50 Sigal, Disarming Strangers, 63.

304 meeting with North Korea in 1992 and agreed to continue bilateral negotiations with North

Korea.

Having staved off North Korea’s NPT withdrawal by mere hours, negotiations continued off and on throughout the rest of 1993, with the high point being a July 19, 1993 agreement in which North Korea agreed in principle to continued IAEA inspections—but not “special” or

“challenge” inspections—in exchange for the United States’ agreement to explore the provision of light-water reactors to North Korea for energy purposes. Yet even this promising moment took place against a backdrop of generalized threats. Despite sporadic negotiations throughout the year, the United States at no point abandoned its gradual escalation strategy. In a presidential visit to Seoul on July 14, 1993, President Clinton offered what came to be interpreted as a thinly veiled nuclear threat: “it is pointless for them to try to develop nuclear weapons, because if they ever use them it would be the end of their country.”51 At one point the United States even reverted to again using diplomatic engagement as a symbolic tool, with Gallucci announcing after the July 19 agreement that another round of high-level talks and economic assistance would not begin until North Korea was in compliance with its NPT obligations.52 By October, North

Korea had publicly proposed a sort of grand bargain that granted controlled IAEA inspections in exchange for the permanent cancellation of Team Spirit, the removal of U.S. economic sanctions, and diplomatic recognition.53 Only four days prior, on November 7, President Clinton appeared on Meet the Press and declared that “North Korea cannot be allowed to develop a

51 Gwen Ifill, “In Korea, Chilling Reminders of Cold War,” New York Times (July 18, 1993): IV-1. 52 U.S. Department of State, Dispatch IV, no. 30 (July 26, 1993): 535. 53 David E. Sanger,. “Seoul’s Big Fear: Pushing North Koreans Too Far,” New York Times (November 7, 1993): A16; R. Jeffrey Smith, “North Korea Deal Urged by State Dept.,” Washington Post (November 15, 1993): A31.

305 nuclear bomb. We have to be very firm about it.”54 Minutes later, when asked about the prospect of a preemptive strike, Clinton demurred instead of ruling it out.55

Outgoing Secretary of Defense Les Aspin, in a television interview with Bob Woodward, described Clinton’s remarks as an ultimatum not unlike that posed to Saddam Hussein over his invasion of Kuwait: “We will not let the North Koreans become a nuclear power.”56 Two months prior Aspin had repeated the unofficial line of many U.S. officials in socializing the public—and North Korea—to the expectation that negotiations would not end the nuclear program and sanctions would likely be required.57 Two months later, on December 3, 1993, a leak of an intelligence community assessment that North Korea may already possess a nuclear device raised public alarm. There was no consensus on North Korea policy relevant issues within the intelligence community, and DIA and INR were most at odds with one another. INR believed that North Korea’s nuclear progress was still reversible and that it was looking for a face-saving negotiated solution, but the DIA assessed that “Based on North Korea’s actions to date…Pyongyang will continue its nuclear weapons program despite any agreement it signs to the contrary.”58

Several subsequent agreements were reached between the United States and North Korea in late February 1994, including IAEA limited access—not special inspections—to Yongbyon;

54 “Interview With Timothy Russert and Tom Brokaw on Meet the Press: 7 November 1993,” Public Papers of the President of the United States: William J. Clinton, Book II (Washington, DC: U.S. Government Printing Office, 1993), 1921. 55 “Interview With Timothy Russert and Tom Brokaw,” Public Papers of the President, 1921. 56 As quoted in Mark Dean Millot, “Facing the Emerging Reality of Regional Nuclear Adversaries,” The Washington Quarterly 17, no. 3 (1994): 47. 57 Cited in FBIS, East Asia Section (September 29, 1993): 25-6. 58 R. Jeffrey Smith, “U.S. Analysts are Pessimistic on Korean Nuclear Inspection,” Washington Post (December 3, 1993): A1; Bill Gertz, “N. Korea Unlikely to Give Up Nukes,” Washington Times (August 24, 1994): 4.

306 continued IAEA access to declared facilities; the restart of North-South dialogue; the cancellation of the 1994 Team Spirit; and a third round of principal-level talks between the

United States and North Korea on March 21.59 These agreements ended up having a very short shelf life, with the nuclear crisis renewing in earnest when North Korea leveled one of its most bombastic threats yet during the quickly collapsing North-South talks on March 19, 1994: “We are ready to respond with an eye for an eye and a war for a war. Seoul is not very far from here.

If war breaks out, we will turn Seoul into a sea of fire.”60 This threat quickly ended the North-

South talks and led to the rescheduling of Team Spirit for 1994, though it would never actually take place.61 The threats came a week after the IAEA had revealed that over the last year, in parallel with the on and off negotiations, North Korea had substantially advanced its nuclear weapons program and it was effectively unable to monitor what was happening.62

A White House Principals Committee meeting held the following day reaffirmed an approach of gradual escalation to compel North Korea to comply with the agreements just reached.63 Within two days of that meeting, the United States initiated a compellent effort greater than at any point during the crisis previously, announcing that Patriot missile batteries would be shipped to South Korea64 at the same time that it started trying to shore up support for

59 Sigal, Disarming Strangers, 104-7. 60 John Burton, “N. Korea ‘Sea of Fire’ Threat Shakes Seoul,” Financial Times (March 22, 1994): 6. 61 R. Jefrrey Smith, “Perry Sharply Warns North Korea,” Washington Post (March 31, 1994): A1. 62 Drennan, “Nuclear Weapons and North Korea,” 172-3. 63 Wit, Poneman, and Gallucci, Going Critical, 152. 64 The decision to send Patriot batteries occurred during a Principals Committee meeting at the White House on 8 February; nevertheless, the announcement that took place in the context of ongoing events in March drew attention to the U.S. commitment to match threats with actions. Wit, Poneman, and Gallucci, Going Critical, 134.

307 UN Security Council sanctions.65 On March 31, U.S. efforts to secure a UN Security Council

Presidential Statement calling on North Korea to allow IAEA inspectors to conduct inspections, as well as to implement the terms of the North-South Joint Declaration, culminated in a success.66 Also at the end of March, Secretary of Defense William Perry stated publicly and unequivocally that the United States would stop North Korea from building nuclear weapons.67

U.S. coercive signaling only led to greater North Korean intransigence and an escalation of threats, with North Korea’s foreign ministry announcing on March 21, 1994 that U.S. actions

“may bring the Korean nation back to the phase of confrontation and war.”68 On March 27, the

North warned that U.S. military reinforcement plans might trigger war: “The United States must not forget…it was Korea where it drank the bitter cup of its first defeat.”69 On April 28, 10 days after the announced Patriot batteries arrived in South Korea, the North declared the 1953

Armistice Agreement invalid, expressing its intent to withdraw from the Military Armistice

Commission.70

North Korea’s threatening statements were coupled with corresponding deed. Over the course of the prior year, according to the U.S. National Intelligence Council, as much as 70% of

North Korea’s military forces deployed forward into the area near its side of the DMZ.71 The

North Korean military was also on a heightened state of alert, which was normal during its

65R. Jeffrey Smith and Ann Devroy, “Clinton Orders Patriot Missiles to South Korea,” Washington Post (March 22, 1994): A1. 66 “UN Security Council Urges North Korea to Allow Inspections,” April 1, 1994, Statement by UN Security Council President Jean-Bernard Merimee. Available online: http://www.fas.org/news/dprk/1994/940404-dprk- usia.htm. 67 R. Jeffrey Smith, “Perry Sharply Warns North Korea,” Washington Post (March 31, 1994): A1. 68 As quoted in Sigal, Disarming Strangers, 107. 69 Cited in FBIS, East Asia Section (March 28, 1994): 24. 70 Wit, Poneman, and Gallucci, Going Critical, 413. 71 Sigal, Disarming Strangers, 74.

308 Winter military training cycle (which usually lasted through March), but analysts observed that some of the military activity was inconsistent with normal Winter training cycle activity— including testing emergency communications networks and holding mass protest rallies to bolster national unity—which heightened concern that North Korea was indeed preparing to implement its threat of war.72

The White House, State Department, and DoD operated throughout the 1994 part of the crisis largely on the assumption that North Korea’s threats that sanctions meant war were credible, which was why they believed defensive reinforcements such as Apache helicopters,

Patriot batteries, and possibly additional troops seemed prudent, even if those defensive actions would become self-fulfilling. Secretary Perry described these moves as necessary prior conditions to a sanctions declaration.73 He would go on to say that “Acquiescing now to an active North Korean nuclear program would invite a future crisis. Taking military action now would invite an immediate crisis…It is possible that North Korea could misperceive these efforts

[defensive military preparations] as provocations. We must face that possibility.”74 In a defense briefing to President Clinton on May 19, Secretary Perry, the Chairman of the Joint Chiefs of

Staff, and the USFK Commander focused on two major elements in their assessments: North

Korea’s rhetoric, and the balance of military capabilities. Secretary Perry described the military balance on both sides of the DMZ and highlighted the heightened threats, which had grown

72 Wit, Poneman, and Gallucci, Going Critical, 127,160; “North Sharpens Rhetoric on Patriot Missiles,” March 30, 1994. FBIS Trends. 73 “Remarks by Secretary of Defense William Perry to the Asia Society,” Federal News Service (May 3, 1994). 74 Ibid.

309 “increasingly obnoxious.”75 Beyond these assessments, all three military leaders tried to convey the upper end costs of a North Korean miscalculation, which they estimated at one million lives and one trillion dollars.76

Much of the intelligence community substantiated the policymakers’ presumption-of- credibility approach as the crisis went from simmer to boil in 1994. The National Intelligence

Officer for Indications and Warning, responsible for monitoring sets of observable indicators for whether and when attacks were probable, grew alarmed by “the combination of Pyongyang’s escalating rhetoric and military moves.”77 But issuing an actual “warning of war” was difficult because, as some dissenting voices within the intelligence community would highlight, even if

“almost every indicator is lit up…They could stay that way for years or they could attack tomorrow.”78 Korea specialists in the policymaking and intelligence communities explicitly looked to past North Korean word and deed as they informed present day decision-making about whether North Korea’s threats of war were credible; they were disturbed by the fact that unlike during previous provocations like the Pueblo seizure or the 1976 Panmunjom crisis, North Korea

“backed them up with military moves and mobilization of its population.”79 All of this was sufficient to justify policymaker assumptions that decisions to move forward with sanctions, or even just defensive preparations, should account for the potential of North Korean retaliation.

North Korea’s threat escalation, combined with the pressure resulting from its rapid and difficult to observe progress in its nuclear program, led the United States to again make

75 Wit, Poneman, and Gallucci, Going Critical, 179-80. 76 Ibid: 180-1. 77 Ibid: 127. 78 Ibid: 128-9. 79 Wit, Poneman, and Gallucci, Going Critical, 160.

310 contradicting compromises of its gradual escalation strategy, offering to cancel the Team Spirit exercise it had just rescheduled, though North Korea rejected the offer.80 On June 16, 1994,

Secretary Perry met with President Clinton, minutes away from presenting him with three options for how to respond to North Korea should the next and final diplomatic overture fail.

Each of the three options were what the military calls “flexible deterrent options,” which involved military reinforcements to the Korean Peninsula that could be used for readiness, coercion, or defense in war, varying only in size and timeline for the flow of forces.81 Before

Clinton could choose, he was interrupted by a phone call indicating that former President Jimmy

Carter was in Pyongyang and had just struck a deal with Kim Il Sung to freeze North Korea’s nuclear program and allow the return of IAEA inspectors. At that moment Jimmy Carter embodied the meaning of the phrase deus ex machina, and it was on that basis that the crisis ended, and coercive bargaining would shift to diplomatic bargaining.

Orchestrating U.S. Actions

If coercion is difficult, then orchestrating a two-track policy of coercion and conciliation seems to be all the more difficult. The preceding section described a Clinton administration bent on taking a coercive approach to North Korean proliferation, at first not finding North Korean threats credible and later reversing that credibility assessment as North Korea began marrying word and deed. At several points, particularly during the two phases of the nuclear crisis, the

Clinton administration offered concessions to North Korea, departing from its preference for

80 Reiss, Bridled Ambition, 268. 81 Oberdorfer, 2000: 324-5; William J. Perry, “Standing at the Brink in North Korea: The Counterproliferation Imperative,” in Preventive Defense: A New Security for America, eds. Ashton B. Carter and William J. Perry (Washington, DC: Brookings Institution Press, 1999), 131.

311 gradual escalation even as it continued to pursue it in other ways. If one looks beyond rhetoric and internal deliberations to account for the actions the United States took, multilateral diplomacy was the most time-consuming and arguably most important task. But we also see, to some extent, the “illusion of orchestration”; U.S. officials invested most of their energy into trying to manipulate North Korea through subtle signals and bargaining. While it is unclear and somewhat doubtful that much of the signaling was meaningful in the final analysis, to the extent we can see that signals were received the response from North Korea was escalation. The two most significant U.S. actions beyond seeking multilateral sanctions—Team Spirit and the deployment of Patriot batteries—were not clear functions of executing a strategy, but rather were thrust on an administration with few options and little ability to control who spoke for the U.S. government and when. Each of these actions, moreover, produced negative short-term reactions that left the United States in a position where it was forced to both back down and coerce. This section illustrates the extent to which the Clinton administration lacked the ability to orchestrate action and shape events, instead being very much shaped by them. The strongest and most visible coercive U.S. signals—threats of sanctions, the Team Spirit military exercise, and the deployment of Patriot batteries—were signals controlled less by the White House than by bureaucrats and foreign partners’ preferences.

When the first phase of the nuclear crisis started in 1993, the United States immediately began coordination with the Chinese and South Korean governments, motivated by the importance of each in gaining approval for and implementation of UN sanctions. Consultations happened early and often with both partners, both of which were opposed to sanctions initially.

312 Nevertheless, through a combination of administration leaks and carefully worded statements, the United States made clear that sanctions was the direction of U.S. policy if North Korea did not comply with the NPT.82 On March 12, 1993, the Chinese foreign ministry told Stapleton

Roy, the U.S. Ambassador to Beijing at the time, that “pressure or coercion” against North Korea was unacceptable.83 On March 24, Chinese Foreign Minister Qian Qichen stated flatly that

“China opposes sanctions against the DPRK for its decision to withdraw from the NPT.”84 On

March 30, an anonymous U.S. official was quoted as saying that “Sanctions require, absolutely require, that the Chinese participate.”85 Beijing did offer to facilitate cooperation to the extent it was possible, hosting U.S. and North Korean delegations during the May and June 1993 talks to resolve the first part of crisis.86 When the IAEA Board of Governors declared North Korean non-compliance with the NPT on March 31, 1993, China refuted the charge as “improvised merely as a justification for referral of the matter to the Security Council.”87

But persistent U.S. diplomacy coupled with persistent North Korean intransigence gradually gained China’s cooperation, resulting in UN support in the form of a Security Council resolution on May 11, 1993, a General Assembly resolution in November 1, 1993, and UN

Presidential Statements on March 31 and May 1, 1994; with China’s reluctant backing, each UN statement grew progressively firmer as North Korea’s nuclear program and posturing persisted

82 See, for example, Douglas Jehl, “U.S. Pressing Plan on Arms Pact to Force North Korea to Comply,” New York Times (March 13, 1993): A3. 83 Wit, Poneman, and Gallucci, Going Critical, 31. 84 As quoted in Mazarr, North Korea and the Bomb, 114. 85 Warren Strobel, “Defiant N. Korea Tests U.S. Effort to Control Nukes,” Washington Times (March 30, 1993): 1. 86 For a discussion, see Mazarr, North Korea and the Bomb, 116-8. 87 R. Jeffrey Smith, “N. Korea Censure Seen as Turning Point in Arms Control,” Washington Post (April 7, 1993): A23.

313 throughout both phases of the crisis.88 The United States was the leading voice on sanctions, but the state of progress toward sanctions never moved further than China was willing to go.

Although the South Korean government was divided over what to do about North Korea, the one constant in the South Korean approach was the Kim Young Sam administration’s need to appear in control and integral to resolving the evolving nuclear challenges that North Korea posed. South Korea’s Deputy Prime Minister reflected as much when he said in 1993, “There is no possibility that we will be estranged from resolving the issue of the Korean Peninsula even after the high-level talks between the United States and North Korea.”89 Although South Korea was paranoid about perceptions of being sidelined as the United States and North Korea engaged in direct crisis bargaining, it was South Korean decisions that, for better or worse, accelerated the crisis at certain points.

South Korea was a major force behind decisions to threaten, cancel, and plan for the

Team Spirit military exercise, sometimes irrespective of U.S. preferences at any given time. In

March 1993, when the first phase of the crisis culminated in North Korea’s declaration of “semi- war,” a major impetus was the decision to resume the 1993 Team Spirit exercise. During the

U.S.-ROK SCM in October 1992, the South Korean defense minister’s staff insisted that the

United States jointly announce that planning for Team Spirit would resume in 1993, reportedly because South Korea was frustrated by North Korea’s rejection of South Korean negotiating positions during their JNCC talks in the months prior.90 Through the ups and downs of the

88 Chronologies of key events in the crisis can be found in many places, including Wit, Poneman, and Gallucci, Going Critical, Appendix A; Sigal, Disarming Strangers, Appendix I. 89 As quoted in Wit, Poneman, and Gallucci, Going Critical, 84. 90 For an overview of South Korea’s role in the Team Spirit decision, see Mazarr, North Korea and the Bomb, 90-1.

314 North-South dialogue in 1993, South Korea would often use the threat of Team Spirit resumption and the offer of its cancellation as a point of leverage in its negotiations.91 In mid-November

1993, the United States was prepared to agree to a North Korean proposal for a third round of principal-level talks and the 1994 cancellation of Team Spirit in exchange for North-South dialogue and inspections of facilities already under safeguard, which would have helped ensure the continuity of existing safeguards preventing North Korea’s declared facilities from weaponizing its nuclear material.92 The approach fell apart, however, on November 23, when

South Korean president, Kim Young-sam, refused to agree to the new U.S. approach, or any approach, that did not first require North Korea to comply with its IAEA obligations and engage in satisfactory North-South Dialogue.93

While the South Korean government never opposed UN sanctions as vehemently as did

China during the first phase of the nuclear crisis, South Korean officials in meetings with U.S. officials expressed indirect forms of reluctance to move forward with sanctions. At the end of

March 1993, Secretary of State Warren Christopher met with his South Korean counterpart to gain South Korean agreement to a carrot-and-stick approach to North Korea. The latter agreed in principle but noted that “pressure alone will not work” against North Korea, and emphasized that concessions were crucial to North Korean change.94 During the U.S.-ROK SCM in October

1993, the South Korean defense minister argued that “in the event of sanctions, North Korea

91 See Wit, Poneman, and Gallucci, Going Critical, 78-116. 92 R. Jeffrey Smith, “U.S. Weighs N. Korean Incentives,” Washington Post (November 7, 1993): A31; David E. Sanger, “U.S. Revising North Korea Strategy,” New York Times (November 22, 1993): A5; Sigal, Disarming Strangers, 82-4. 93 Ruth Marcus and R. Jeffrey Smith, “U.S., South Korea Shift Strategy on North,” Washington Post (November 24, 1993): A12; Thomas L. Friedman, “U.S. and Seoul Differ on Appeal to North Korea on Nuclear Sites,” New York Times (November 23, 1993): A16; Reiss, Bridled Ambition, 262; Sigal, Disarming Strangers, 88. 94 David E. Sanger, “Neighbors Differ on How to Chasten North Korea,” New York Times (March 31, 1993): A9.

315 could perpetrate military provocations.”95 After a final breakdown of North-South dialogue in

1994, however, the South Korean government would support the U.S. coercive effort unequivocally, with Kim Young Sam finally approving “all necessary steps” for sanctions on

June 3.96

The deployment of Patriot batteries to South Korea, the signal that one would have thought the White House would exercise the most control over, was actually initiated by the commander of U.S. Pacific Command , Admiral Charles Larson, who had “gone a long way before the civilian side knew it.”97 Bringing it closer to reality and certainly making the decision known publicly, on January 25, 1994 the U.S. discussions about deploying Patriot missiles leaked to the press.98 The push toward Patriot deployments was the most visible discussion, but threats of additional deployments were leaked at the same time, including an aircraft carrier,

1,000 troops for Team Spirit, and a U.S.-ROK alliance objective of seizing Pyongyang in the event of war.99 Even with the leaks, however, the South Koreans had not yet agreed to allow the

Patriot deployment, to say nothing of other rumors. When the final phase of the nuclear crisis broke on March 19, 1994, the proximate cause was the collapse of North-South talks, which fell apart with mutual recrimination and the North’s infamous “sea of fire” threat. Prior to this moment, South Korea had not agreed to the U.S. deployment of Patriot batteries. On March 23,

95 “North Korea Backs Out of Inter-Korean Talks for Envoy Exchange,” North Korea News (November 15, 1993): 709; “North-South Contact Cancelled after Interview,” Korean Central Broadcasting Network (November 3, 1993). 96 “President Phones Advisers about Nuclear Issue,” Yonhap (June 3, 1994); Wit, Poneman, and Gallucci, Going Critical, 196. 97 Interview with anonymous U.S. official, as quoted in Sigal, Disarming Strangers, 101. 98 Michael R. Gordon, “U.S. Said to Plan Patriot Missiles for South Korea,” New York Times (January 26, 1994): A1. 99 “Daily Reports ROK, U.S. Prepare for Team Spirit,” Joongang Ilbo (February 7, 1994): 1; Michael Gordon and David Sanger, “North Korea’s Huge Military,” New York Times (February 6, 1994): 1; R. Jeffrey Smith, “South Korea Warns North of Plan to Hold War Game,” Washington Post (February 1, 1994): A7.

316 however, seeing no prospect of gain in North-South relations, President Kim sent a letter to

President Clinton agreeing to a gradual escalation approach and the deployment of Patriot batteries.100

In sum, the most visible coercive signals rarely came directly from the White House, though each signal would eventually gain White House imprimatur. The Team Spirit military exercise was decided by the outgoing U.S. and ROK administrations, and it was South Korea’s

Ministry of National Defense that pushed for its resumption for 1993. The idea to deploy Patriot batteries was initially advanced by the commander of U.S. Pacific Command, outside of the normal interagency process, and leaked to the press well before a decision had been finalized.

Once the Patriot battery deployment was approved by the White House, it was unable to do anything until the South Korean government approved, which did not occur until months later when North-South talks broke down. And although the United States was the strongest voice pushing the threat of sanctions, the multilateral nature of a sanctions threat inherently meant that the U.S. threat could extend no further than key partners (China and South Korea) were willing to go. Although I have focused on the coercive aspect of U.S. signaling, even the most conciliatory action—former President Jimmy Carter’s backchannel negotiations with Kim Il

Sung—was done without either the approval or direction of the White House.

The North Korea Perspective

Did North Korea find U.S. coercive signals credible during the nuclear crisis? When

North Korea made threats, how often did they follow periods of U.S. concessions? How did

100 Wit, Poneman, and Gallucci, Going Critical, 150-1; “First Shipment of Patriot Missiles Arrive,” Yonhap (April 18, 1994).

317 North Korea react to U.S. threats? To the extent that evidence allows, here we must consider how North Korea perceived U.S. signals. From the collection of news reports and firsthand accounts of North Korean words and deeds in the nuclear crisis, we find at least some evidence to suggest that North Korea: perceived Team Spirit as a military threat; was the first to issue verbal threats but in all likelihood did not initially perceive itself as a challenger (unlike in previous crises); found U.S. threats credible, basing that credibility assessment on a combination of U.S. capabilities and U.S. behavior in the Gulf War; and, despite finding U.S. threats credible, took coercive measures when the United States stood firm and when it backed down, not only when it backed down.

In the conventional narrative of the nuclear crisis, North Korea is framed as the challenger, having increased tensions with extreme rhetoric in October 1992 and having created real pressure with its February 1993 announcement that it would withdraw from the NPT, combined with verbal threats of war.101 North Korea attributed significant meaning to Team

Spirit, calling it “a large-scale thermonuclear war rehearsal…aimed at waging an attack.”102

These complaints go back as early as 1984, when Kim Il Sung complained to his East German counterpart that: “When the Team Spirit exercise occurs, we go on nationwide alert, mobilizing workers to the military and, therefore, stop production activities for over a month.”103 The

United States and South Korea knew North Korea viewed Team Spirit as a coercive threat, which is why throughout talks in 1992 they threatened it would resume in March 1993. During

101 See, for example, Drennan, “Nuclear Weapons and North Korea.”; Down, Over the Line. 102 As quoted in Mazarr, North Korea and the Bomb, 92. 103 Chang-hui Kim, “Pukhan, 80 Nyun-dae Jungban Seodok-aeseo Haekmuljil Goo-eep” (North Korea, Purchased Nuclear Material from West Germany in the Mid-1980s), Sindong-a (1995): 134.

318 the U.S.-ROK SCM in October 1992, it was announced that planning for Team Spirit 1993 would commence, to which North Korea responded that the United States was “…running amok with large-scale joint military exercises…the United States also staged the Focus Lens exercise some time ago…thus aggravating the military tension.”104 On October 13, 1992, North Korea referred to Team Spirit as a “criminal act” trying to “drive the North-South dialogue to a crisis.”105 Also referring to Team Spirit, North Korea’s March 12 announcement declared that it would proceed with the process of withdrawal from the NPT “until the United States stops its nuclear threats against our country.”106 That North Korea issued subtle threats at the end of 1992 and direct threats in February 1993 is thus unambiguous. While debatable whether the U.S. and

South Korean threats to resume Team Spirit counts as a prior coercive signal, I argue based on the above that it was both intended as, and perceived as, a coercive signal.

In earlier chapters and with some exceptions, much of the United States’ orchestration of word and deed for signaling purposes during crisis was imperceptible to North Korea, to the extent we were able to observe North Korean perceptions in each instance. By contrast, the previous sections of this chapter reveal that, during the nuclear crisis, most U.S. threats tended to be oblique—occurring largely through press leaks, military exercises, and military deployments—but that North Korea nevertheless did believe them.

Presaging the drama that would proceed over the next year and a half, on February 21,

1993, North Korea threatened that “If any special inspections and sanctions are enforced on

104 Cited in FBIS, East Asia Section (October 5, 1992): 9. 105 Cited in FBIS, East Asia Section (October 13, 1992): 7; Nodong Sinmun (October 13, 1992). 106 “Statement Notes Withdrawal,” Korean Central Broadcast System (March 12, 1993); Wit, Poneman, and Gallucci, Going Critical, 37.

319 us…this will be a dangerous fuse that will drive all lands, including the North and the South, into the crushing calamities of war.”107 In March 1993, as the United States and South Korea were initiating Team Spirit, North Korea reacted militarily and rhetorically in a more dramatic way than in reaction to Team Spirit exercises from previous years. The day before the start of Team

Spirit, Kim Jong Il issued a national military order to enter a state of “semi-war,” castigating

Team Spirit as a “nuclear war game preliminary to invasion of North Korea.”108 The alert status of “semi-war,” which North Korea had not invoked since 1983, brought with it the cancellation of foreign visas, mass unity rallies, and the unprecedented jamming of Korean and Japanese radio broadcasts.109 Then, on March 12, came the North’s announcement of withdrawal from the

NPT, warning as well of “grave consequences” if sanctions were pursued.110 Added to this is defector testimony claiming that Kim Jong Il, who ran North Korea’s military prior to succeeding his father, spent much of March 1993 in a military bunker, issuing commands to field units, a curious action if North Korea did not anticipate the possibility of conflict.111

All of this suggests that North Korea took seriously both U.S. threats of sanctions, and more importantly, the possibility of Team Spirit as a prelude to invasion akin to the Gulf War.

North Korea feared that it would become “another Iraq,” and had become a careful study of the recent U.S. way of war, which in the Gulf War involved amassing forces prior to conflict using

107 Cited in FBIS, East Asia Section (February 22, 1993): 11-13. 108 “Order No. 0034 of the KPA Supreme Commander,” Pyongyang Times (March 8, 1993): 1; “Kim Chong Il Declares ‘Semi-State of War,’” BBC Summary of World Broadcasts (March 9, 1993); “North Korea Proclaims Semi-War State as Countermeasure against Team Spirit,” North Korea News 674 (March 15, 1993): 2. 109 “North Korea Escalates Warlike Mood among Populace,” North Korea News 675 (March 22, 1993: 1-2. 110 “Statement Notes Withdrawal,” Korean Central Broadcast System (March 12, 1993); Wit, Poneman, and Gallucci, Going Critical, 37. 111 Testimony of North Korean defectors, as referenced in Wit, Poneman, and Gallucci, Going Critical, 37.

320 the FDO model,112 rather than flowing masses of forces after conflict had started.113 It seems reasonable to infer from North Korea’s attention to U.S. behavior in Iraq that North Korea assessed U.S. threat credibility—leading to an escalation of word and deed—in terms of how its military moves matched those in the run up to the Iraq War in 1991. During an April 1994 meeting in Panmunjom, one of the North Korean negotiators informed the United States that

“This will not be a situation like the Iraq war. We will not give you time to collect troops around

Korea to attack us…if it is clear you are going to attack, then we will attack.”114 North Korea’s anomalous military movements through the end of 1993 and early 1994 were consistent with the type of maneuvering and deployments necessary to try and operationally counter the way the

United States waged the first Gulf War.

Even though we see some evidence that North Korea did receive U.S. coercive signals and found them at least as credible as the United States intended—arguably more so given the

Gulf War lens through which North Korea seemed to view U.S. word and deed—neither U.S. threats nor the scattered offers of concessions at various points of the crisis led to North Korea backing down. On the contrary, the most significant U.S. threats were met with escalation in

North Korean word and to some extent deed as well, while U.S. concessions in the face of North

Korean resolve all but ensured a replay of Spring 1993 one year later.

112 “Flexible deterrent options” or “FDOs” are options built into contingency plans that allow U.S. forces to provide early response to an emerging crisis. FDOs are designed to give decision-makers deterrent-oriented options to prevent conflict, but also to quickly mass forces that could be used to execute a contingency plan should deterrence fail. 113 Wit, Poneman, and Gallucci, Going Critical, 34-9. 114 Wit, Poneman, and Gallucci, Going Critical, 169; Marion Creekmore, Jr., A Moment of Crisis: Jimmy Carter, The Power of a Peacemaker, and North Korea’s Nuclear Ambitions (Cambridge, MA: PublicAffairs, 2006), 109.

321 As the United States started generating pressure on North Korea by initiating Team

Spirit, nudging the IAEA toward referring North Korea’s NPT non-compliance to the UN

Security Council, and intimating that sanctions against North Korea would be necessary if its nuclear program were not curbed, North Korea responded venomously. Instead of seeking relief at the bargaining table, North Korea responded with pressure in kind, making clear that the imposition of sanctions would lead to war.115 Although a verbal North Korean threat of war was not unprecedented, the fact that such threats occurred in tandem with a state of “semi-war” and the vast majority of North Korea’s military capabilities forward deployed near the DMZ

(increasing its ability to launch an attack with little to no warning) made clear that it did not intend to bend to pressure. Underscoring North Korea’s approach to pressure in this case, a

North Korean diplomat noted during a round of North-South talks that the U.S. military buildup and sanctions would be met in kind: “We are ready to respond with an eye for an eye and a war for a war.”116 Its “eye for an eye” mentality, a common sentiment in dyadic rivalries, helps us understand why, after news of the U.S. Patriot missile deployment leaked in January 1994, North

Korea responded by condemning U.S. “hardliners” for trying to inflame tensions with Patriot missiles, warning that “catastrophic consequences” would result.117 And as the nuclear crisis was reaching its boiling point and U.S. commentators were increasingly talking of war, North

Korea launched short-range anti-ship missiles into the Sea of Japan on May 31 and again on June

115 Cited in FBIS, East Asia Section (February 22, 1994): 11-13; Downs, Over the Line, 228-30. 116 John Burton, “N. Korea’s ‘Sea of Fire’ Threat Shakes Seoul,” Financial Times (March 22, 1994): 6. 117 Pyongyang Times (February 5, 1994): 3.

322 2, followed by a visible June 14 test of an engine that could be used for the Taepodong, North

Korea’s intercontinental ballistic missile.118

Leon Sigal, a scholar who interviewed dozens of officials at the time of the nuclear crisis, described North Korea’s approach as “tit-for-tat,” but that description implies that while force would be met with force, concessions would be met with concessions too.119 Yet we see that the first phase of the nuclear crisis ended because the United States backed down while facing North

Korea’s threats, satisfying North Korea’s demand for a declaration of non-aggression, agreeing to the direct high-level talks that North Korea had long demanded, and the offer to “explore” providing North Korea with LWRs for nuclear energy. In exchange, North Korea agreed to

“temporarily” suspend its withdrawal from the NPT, but offered no compromises relating to

North-South reconciliation, IAEA special inspections, South Korean inspections, or a freeze of its nuclear activity.

Following the conclusion of the first phase of the nuclear crisis in summer 1993, we find that North Korea did not behave conciliatorily, but rather levied increased demands, continuing all the while its preparations for war. On October 12, 1993, North Korea formally conveyed to the United States that they must: abolish the Armistice Agreement and sign a peace treaty; provide, at no cost, LWRs for civilian energy; normalize diplomatic relations and agree to statements of respect for sovereignty and non-interference in internal affairs; and promise to take

118 Hanguk Ilbo (June 4, 1994): 2; Chosun Ilbo (July 2, 1994): 2. 119Sigal, Disarming Strangers.

323 balanced positions between North and South Korea in future policy matters.120 From North

Korea’s perspective, the United States was less than cooperative following the first phase of the crisis—given that the United States refused (initially) to proceed with high level talks until North

Korea made progress with the IAEA and South Korea—which was the seed corn that made the second phase of the crisis probable: the United States did not take steps to ameliorate dyadic hostilities, yet nevertheless offered concessions in that hostile context, with the result that North

Korea simply increased its demands and demonstration of resolve later. In the epilogue to the nuclear crisis, which, like the first phase of the crisis also ended with reluctantly proffered U.S. concessions, we see that although after 1994 U.S.-North Korea hostilities never quite reached the crisis level of 1993-’94, much of the two decades that followed did consist of heightened North

Korean demands, inflammatory rhetoric, and the exchange of coercive signals, principally over its nuclear program.

Particularly in the context of the history of the U.S.-North Korean rivalry and the latter’s conduct in previous dyadic conflicts, the fact that in the nuclear crisis North Korea was willing to challenge the United States irrespective of whether the United States initiated a challenge, stood firm, or backed down says something about North Korea’s interest in the issue at stake—its nuclear program—compared to its interests in the previous crises. One scholar has described

North Korea’s perceptions of its own provocations starting in the 1960s as “nwae-jeon” (literally

“mind combat” or “strategic battle”), which suggests North Korea perceived its military actions

120 These demands are scattered piecemeal throughout the many firsthand accounts of the nuclear crisis, but are identified as a specific list of demands conveyed on a specific date in Michishita, North Korea’s Military- Diplomatic Campaigns, 110.

324 as part of a kind of complex strategic bargaining at the political level.121 While nwae-jeon may have been present in some aspects of the nuclear crisis, North Korea’s willingness to flout U.S. demands even in the face of military threats implies that North Korea perceived its nuclear program as being of existential importance, and therefore less susceptible to coercive pressure.

Analysis: Explaining Decisions and Outcomes

As the above narrative conveys, the cast of relevant actors and factors involved in explaining this case is dizzyingly complex. The firsthand retellings of this story across many scattered memoirs reflect as much, making a simplified rendering of reality—that focuses on the factors most relevant to reputation and credibility—all the more valuable. Not only is the story of the nuclear crisis more comprehensible this way, but the nuclear crisis in turn helps us discern which perspectives of credibility and reputational effects have the greatest explanatory power.

On the question of threat credibility, we find the least support for constant calculus theory and the greatest support for the reputations for resolve hypothesis, while finding mixed support for current calculus and reputations for honesty expectations. On the question of the temporal consequences of backing down or standing firm, the outcomes in this case seem to fit the resolve hypothesis better than the honesty hypothesis, but it is difficult to judge definitively because backing down often occurred against a backdrop of coercion. And on the question of potential connections between initiating challenges against a rival and a reputation for resolve, this case offers some support for findings in previous cases suggesting that a history of bluffing can impede the accretion of a reputation for resolve for a challenger, but again, it is difficult to reliably assign the role of challenger and defender in this case.

121 Michishita, North Korea’s Military-Diplomatic Campaigns.

325 Explanatory Power of Reputations for Honesty

This case offers several tests of the reputations for honesty hypotheses. Evidence confirms some aspects of the reputations for honesty logic, and disconfirms others. Going into analysis of these events, the honesty logic anticipated several outcomes: that backing down or standing firm would make future challenges less likely; that a gap between word and deed (that is, threat-making without commensurate follow through) would erode future threat credibility; and that matching word and deed (that is, threat-making with commensurate follow through with military signaling) would bolster future threat credibility. This case, which allows us to observe all four claims, supports the contention that aligning word and deed by standing firm bolsters threat credibility; supports the contention that a gap between word and deed undermines threat credibility; partially supports the contention that standing firm makes future challenges less likely; and refutes the claim that backing down makes future challenges less likely.

The logic of reputations for honesty performed best when it comes to anticipating the effect that a gap between word and deed—that is, a history of bluffing—has on threat credibility.

The only clear opportunity to observe this dynamic was at the beginning of the crisis, when U.S. officials dismissed North Korea’s threats that it was willing to risk war over sanctions, as well as the threat that it would withdraw from the NPT if pressure for IAEA inspections continued. The evidence in support of the reputations for honesty logic here is strong and direct. The U.S. ambassador in South Korea recommended in a memo that the White House not pay too much attention to North Korea’s threats of war because it had often made threats associated with Team

326 Spirit that went nowhere.122 When North Korea directly warned a U.S. diplomatic delegation of dire consequences if Team Spirit resumed, the U.S. officials involved also dismissed the threat because of North Korea’s poor track record of matching word and deed, particularly when it came to bluster in response to Team Spirit.123 And an anonymous U.S. official summarized the problem in a manner similar to how it has been described by U.S. officials in previous chapters: there was no way to separate “the nuggets of reason” from “the ocean of vitriol.”124 Focusing specifically on commitment credibility generally rather than threat credibility specifically, the

DIA assessed that “Based on North Korea’s actions to date…Pyongyang will continue its nuclear weapons program despite any agreement it signs to the contrary.”125 Of course, further strengthening the reputations for honesty claim about the effects of a reputation for bluffing is variation in threat credibility assessment when North Korea’s gap between word and deed shifted to alignment of word and deed. Discussed in the next two paragraphs, this variation in outcome is precisely what happened when North Korea shifted from its track record of empty threat- making to mobilizing its military commensurate with its threats of war.

Another honesty-based claim that performed well in this case is that when aligning word and deed comes in the form of standing firm, it bolsters threat credibility, though a reputation for resolve would expect the same outcome under the same conditions. If the United States was the initial challenger in the first phase of this case because of its Team Spirit threats, then North

Korea’s decision to stand firm and threaten war in kind fits the reputations for honesty

122 Wit, Poneman, and Gallucci, Going Critical, 25. 123 Interview with anonymous U.S. official as quoted in Sigal, Disarming Strangers, 47. 124 Interview with anonymous U.S. official as quoted in Sigal, Disarming Strangers, 59. 125 R. Jeffrey Smith, “U.S. Analysts are Pessimistic on Korean Nuclear Inspection,” Washington Post (December 3, 1993): A1; Bill Gertz, “N. Korea Unlikely to Give Up Nukes,” Washington Times (August 24, 1994): 4.

327 explanation because the U.S. response to North Korea’s decision to stand firm was to back down, implying that it found North Korea’s threats credible. North Korea’s decision to stand firm represented the alignment of word and deed—it did not simply threaten war in February and

March 1993, but put its forces on alert, conducted joint military exercises, and moved the majority of its forces forward near the DMZ, as one might expect from a state of “semi-war.”

From October 1992 to July 1993, in fact, North Korea declared a period of war preparation, siphoning agricultural and industrial labor for military mobilization, stockpiling rations and weapons inventory, and establishing an operational command focused on war preparations.126

The United States found North Korea’s threats credible once it initiated the process of

NPT withdrawal combined with its military signaling and war footing, which is why despite harboring preferences for coercion, the United States instead opted to respond to North Korea’s threats by scrupulously avoiding making threats and instead meeting several of North Korea’s demands.127 North Korea also found the threat represented by Team Spirit credible, which is arguably why Kim Jong Il was in a command bunker for most of March 1993,128 and why North

Korea began its war preparations concomitant with the decision to resume Team Spirit.129

During the interregnum between the two phases of the nuclear crisis, the U.S. National

Intelligence Officer for Indicators and Warning found North Korea’s general threats credible, raising alarm at “the combination of Pyongyang’s escalating rhetoric and military moves.”130

126 Republic of Korea Ministry of National Defense, Defense White Paper, 1994-1995 (Seoul: Korea Institute for Defense Analyses, 1995), 72. 127 Alan Elsner, “U.S. Takes Mild Approach to Korean Nuclear Threat,” Reuters (March 16, 1993); George Gedda, Associated Press (March 12, 1993). 128 Testimony of North Korean defectors, as referenced in Wit, Poneman, and Gallucci, Going Critical, 37. 129 Republic of Korea Ministry of National Defense, Defense White Paper, 72 130 Wit, Poneman, and Gallucci, Going Critical, 127.

328 And with the onset of the second phase of the crisis in 1994, subtle U.S. threats of Patriot missile deployments, the potential resumption of Team Spirit, and the continued pursuit of sanctions were viewed credibly by North Korea, who interpreted these signals as presaging a military buildup that would lead to an attack.131

In addition to the previous paragraphs’ discussion about the effects on threat credibility of aligning word and deed, the reputations for honesty logic also makes claims about the likelihood of future challenges as a result of word-deed alignment. The effect of standing firm as a form of word-deed alignment produced mixed results in this case. On the one hand, North

Korea stood firm against what it perceived as U.S. coercive threats in the form of the 1993 Team

Spirit exercise, yet North Korea’s demonstration of resolve only compelled the United States to resolve the immediate crisis situation with conciliation; it did not compel the United States to abandon its overall commitment to gradual escalation. Neither did U.S. demonstrations of resolve in early 1994 lead to a cessation of North Korean threats and challenges. For either side, standing firm seemed to affect the threat credibility assessments of the other, but not the likelihood of future challenges. Even the end of the crisis in June 1994 did not mean an end to challenge initiation, given that the exchange of threats and coercive signals became a continued part of U.S.-North Korea interaction after the crisis ended, accelerating in frequency and intensity after the George W. Bush administration entered office in 2001. So on balance, we find little evidence to support the honesty-based claim that aligning word and deed by standing firm makes future challenges less likely.

131 Wit, Poneman, and Gallucci, Going Critical, 169.

329 Where reputations for honesty most fail to deliver is in its prediction of the effect of backing down on crisis recurrence. This is the claim that most distinguishes a reputations for honesty logic from one of resolve, and, as with previous chapters, it is simply incongruent with how events played out. Backing down is a form of honest communication; an alignment of word and deed. Such alignment is supposed to reduce the likelihood of being challenged again in the future. Yet the United States backed down to end the first phase of the crisis and prevent escalation in summer 1993, only to face a challenger levying increased demands by Fall 1993.

While it is difficult to make a strong argument for coding either the United States or North Korea as backing down to end the crisis in June 1994 given the unique way it ended with Jimmy

Carter’s entry, it is notable that North Korea would repeatedly challenge the United States following the 1994 Agreed Framework, not only over ballistic missiles in 1998, but over the unresolved nuclear issue that drove the 1994 crisis starting in 2002.

Explanatory Power of Reputations for Resolve

Reputations for resolve hypotheses anticipated the following outcomes in this case: that backing down would make future challenges more likely while standing firm would make future challenges less likely; that backing down makes future threats less credible while standing firm makes future threats more credible; and that a gap between word and deed (that is, threat-making without commensurate follow through) would erode future threat credibility. These claims are only partially congruent with how events played out. The reputations for resolve logic performs best where an honesty-based logic performs worst: the effect of backing down on the likelihood of future challenges. A resolve-based logic also correctly anticipated, similar to an honesty-

330 based logic, that a gap between word and deed erodes future threat credibility. The reputations for resolve claim that standing firm makes future threats more credible finds support in the nuclear crisis, but the complementary claim that standing firm makes future challenges less likely finds little support here.

As with previous chapters, we see that backing down from a rival’s challenge encourages future challenges, even though it often ameliorates the immediate crisis circumstances brought on by the challenge in the first place. When the United States backed down from North Korea’s threats and instead offered to satisfy some of its demands in order to prevent the latter’s withdrawal from the NPT in summer 1993, it took less than three months before North Korea increased its demands on Clinton administration negotiators and began again issuing threats proscribing sanctions and special inspections.132

A resolve-based logic illuminates another important aspect of the nuclear crisis as well— why U.S. officials would be incredulous about North Korean threats of war and NPT withdrawal prior to March 1993, only to reverse course for the remainder of the crisis and move to a presumption-of-credibility approach when it came to North Korea’s threats. Prior to North

Korea’s March 12 declaration, U.S. officials interpreted North Korean threats as the bluster that typically accompanied Team Spirit exercises.133 North Korea had a long history of issuing empty threats. Yet the NPT withdrawal process was unprecedented, a declaration of “semi-war” had only occurred three other times in North Korean history, and its irregular military maneuvers and nationwide mobilizations accompanied its warnings of war. It was only after these signals of

132 Elsner, “U.S. Takes Mild Approach to Korean Nuclear Threat”; Gedda (March 12, 1993). 133 Wit, Poneman, and Gallucci, Going Critical, 25.

331 resolve, which deviated from prior North Korean behavior and posture, that the United States changed its view of North Korean threat credibility. Korea experts in the Clinton administration looked to North Korea’s past word and deed, noting that in contrast to previous crises in U.S.-

North Korea relations, North Korea was complementing its threats in the nuclear crisis “…with military moves and mobilization of its population,” which they interpreted as a worrying sign.134

As a resolve-based reputational logic would expect, standing firm also exercised a positive effect on North Korean perceptions of U.S. threat credibility during the nuclear crisis, specifically looking to the latter’s behavior in the Gulf War. Up to this point I have limited my inquiry about credibility to the dyadic rivalry context, but direct process tracing evidence from the dyadic history tells us North Korea referred to U.S. behavior during the Gulf War as at least a partial basis for interpreting U.S. word and deed during the nuclear crisis: “This will not be a situation like the Iraq war. We will not give you time to collect troops around Korea to attack us.”135 North Korea’s unprecedented military response to Team Spirit in 1993, including the claim that Kim Jong Il spent most of that month in a command bunker, suggests that it viewed the 1993 Team Spirit as a greater threat than past Team Spirit exercises, possibly because it involved an amassing of troops akin to how the United States fought Iraq during the first Gulf

War. North Korea’s military movements, and its growing emphasis on multiple rocket launchers, artillery that could range Seoul, and chemical weapons support this interpretation.

The final claim associated with a resolve-based reputational logic—that standing firm makes future challenges less likely—does not fare well in the nuclear crisis. North Korea’s

134 Ibid: 160. 135 Ibid: 169.

332 decision to stand firm in the first phase of the crisis did lead to the United States backing down, which produced the summer 1993 joint statement. But standing firm in all other instances by each side during the crisis only seemed to rigidify the other. North Korea initiated its “sea of fire” threats in early 1993 and the United States in turn threatened to resume Team Spirit and approved Patriot missile deployments; the United States threatened sanctions and approved the

Patriot missile deployments only to have North Korea respond with alarming military movements, impromptu large-scale exercises, and increased war rhetoric. After summer 1993, standing firm did not reduce the probability of future challenges but seemed to increase them, leading to counter-threats and escalation by either side.

Alternative Explanations about Credibility

Like the two reputation-based explanations of this case, current calculus theory and constant calculus theory have ambiguous results here. Both make claims that do not align with outcomes in the nuclear crisis. While neither makes testable claims about the temporal consequences of word and deed, each adds perspective on how calculations of credibility were made.

Current calculus theory tells us that decision-makers will not look to past word and deed when assessing the credibility of another’s threats. This negative claim simply proves untrue in this case. The United States judged North Korea’s threats prior to March 12, 1993 as incredible based explicitly on North Korea’s history of bluffing. The U.S. NIO for Indicators and Warning, as well as Korea specialists in the Clinton administration, referred specifically to North Korea’s past behavior when claiming that its threats were credible. The U.S. presumption-of-credibility

333 approach only followed North Korea’s unprecedented military posturing. And North Korea’s judgment of U.S. threat credibility explicitly referenced U.S. behavior in the Gulf War. This is not to say that every rendering of judgment about credibility draws on observations of past behavior, but there is no question that past behavior was used to inform many of the contemporary credibility assessments relating to the nuclear crisis.

Current calculus theory’s positive claim—that we should look to the balance of capabilities and interests as a basis for credibility assessments—fares much better than its negative claim in explaining the nuclear crisis, though it too has mixed results. One of the factors that made North Korea’s threats of war during the crisis credible was that it was acquiring and deploying the capabilities necessary to optimize its performance in a war: it was deploying increasing numbers of long-range artillery and multiple-rocket launchers, it enhanced its military readiness with dramatically increased frequency of military exercises, and it was stockpiling chemical weapons.136 It was this activity, coupled with North Korea’s threatening rhetoric, the compelled Washington to begin presuming that North Korea’s threats were credible once the nuclear crisis was underway. There is also some direct process tracing evidence that supports the current calculus claim that existing capabilities play an important role in credibility assessments: when DoD leadership briefed President Clinton on May 19 about the likelihood of a North

Korean attack under various conditions, they focused not only on North Korea’s increased threat- making compared to past crises, but primarily the balance of military capabilities in North and

136 Republic of Korea Ministry of National Defense, Gukbang Baekseo, 1995-96 (Defense White Paper, 1995-96) (Seoul: Ministry of National Defense, 1996), 64; Joseph S. Bermudez, Jr., The Armed Forces of North Korea (London: I.B. Tauris, 2001).

334 South Korea.137 They also encouraged Clinton to take North Korea’s threats seriously, focusing on the costs North Korea had the capability to impose rather than on whether it would impose them.138

For all its aforementioned virtues in helping us understand what happened and why in the nuclear crisis, current calculus theory falters in multiple important respects. First, as with previous crises, the balance of capabilities vastly favored the United States, before, during, and after the nuclear crisis, yet U.S. assessments of North Korean threat credibility varied over time, even though its assessment of North Korean capabilities did not vary over time. Second, current calculus theory tells us to look to the balance of interests when capabilities alone do not determine threat credibility. Yet the issue at stake—North Korea’s nuclear program—was a

“vital” interest to both the United States and North Korea.139 For the former, the Bottom-Up

Review and National Military Strategy both defined nuclear proliferation as the greatest security threat the United States faced, which is arguably why it was willing to risk the use of coercion in the first place. For the latter, the nuclear program was, for historical and geopolitical reasons, of existential importance, even though some in the State Department have historically argued that it was primarily a bargaining chip.140 For both sides, the interests at stake were essentially unchanging throughout the crisis, as did perceptions of capabilities by both sides, leaving as a

137 Wit, Poneman, and Gallucci, Going Critical, 179-80. 138 Ibid: 180-1. 139 It would be difficult to argue that North Korea’s nuclear program was an existential threat to the United States, but it was defined as the top security issue and was a threat to U.S. values, which Charles Hermann defines as an important attribute of a crisis. Hermann, Crises in Foreign Policy. 140 Pollack, No Exit; Mazarr, North Korea and the Bomb, 15-34. For the authoritative debate on how to interpret North Korea’s nuclear program, see David C. Kang and Victor D. Cha, Nuclear North Korea: A Debate on Engagement Strategies (New York: Columbia University Press, 2003).

335 puzzle for current calculus theory why U.S. threat credibility assessments would vary. Finally, if we gave current calculus theory the benefit of the doubt in explaining North Korea’s perception of U.S. threat credibility as constant because the balance of interests somehow favored the

United States, we would not be able to then also explain why the United States would find North

Korea’s threats credible during the nuclear crisis.

Constant calculus theory offers a plausible contribution to our understanding of the U.S. presumption-of-credibility approach once the nuclear crisis started, but, as with previous chapters, effectively fails to explain any instances of variation in the dependent variable—threat credibility—because constant calculus theory’s a priori prediction is that states always view their rival’s threats as credible due to their identity-based distinction as a hostile “other.” While in reality threat credibility is not a simple binary matter and there are many shades of grey between threats being highly credible and threats being completely incredible, a graduated approach to describing threat credibility does not change the overall ability of constant calculus theory to explain credibility in this case relative to explanations offered by reputational or current calculus perspectives. Once North Korea started issuing threats in 1993, the current calculus perspective tells us that the United States should have found them all credible. After March 12, 1993, when the United States began discussing North Korean threats and potential U.S. responses, it did so with the premise that North Korea would act according to the threats it was making; this observation is congruent with constant calculus’ expectations. But constant calculus theory falls short in being able to account for why the United States would have viewed North Korean threats prior to March 12 with incredulity.

336 The logic of constant calculus also faces difficulty explaining why North Korea’s reaction to Team Spirit 1993 would be so credible as to warrant declarations of “semi-war,” and all the military preparations that accompany such as state of alert, when Team Spirit had been recurring annually without inciting the same degree of North Korean threat and, seemingly, fear.

Constant calculus theory offers us a perspective on this case not offered by the alternative theories of credibility. Moreover, it helps us understand why we see a genuine hostility in the way some U.S. officials described North Korea: “Everyone on the American delegation was gung-ho to sock it to these sons-of-bitches.”141 But the value of the social-psychological insights that current calculus offers must be measured relative to the insights generated by the alternative theories of credibility considered here, since it is credibility and the related concept of reputation that is at issue in this study. Assessed in these terms, constant calculus theory struggles once again because it makes a constant prediction about a value that continually seems to vary over time.

Assessing Hypothesis Performance

The two hypotheses I advanced at the beginning of this dissertation drew on aspects of the reputational explanations described above:

H1: Backing down from a confrontation initiated by a rival challenger increases future threat credibility, increases the likelihood of future challenges, and de-escalates the current crisis.

H2: Initiating coercive challenges against rivals has an indeterminate effect on a reputation for resolve.

141 Sigal, Disarming Strangers, 63.

337 Of the four cases in this dissertation, H1 and H2 perform worst in explaining the nuclear crisis.

We find that H1 is congruent with the dependent variable outcomes it would anticipate in this chapter, but we do not have high confidence that the causal logic of H1 is the best, or even primary, explanation for events in this case. To the extent that we can actually observe H2 in the nuclear crisis, we actually find disconfirming evidence, but the reality of the nuclear crisis is too complex to be able to reliably code challenger and defender roles throughout.

The nuclear crisis included one clear instance of backing down, and at least one additional instance that might be coded as backing down, though it is less clear. In the first instance, the United States backed down in summer 1993 facing North Korean threats of war and

NPT withdrawal, accommodating some of North Korea’s demands in order to forestall withdrawal. The other, more ambiguous instance of backing down was the apex of the crisis in

June 1994, where U.S. agreement to Jimmy Carter-brokered concessions defused the crisis. I describe this second instance of backing down as ambiguous because North Korea also partially accommodated U.S. demands by agreeing to a package deal that would freeze its nuclear reprocessing and, over time, end its nuclear program; the Agreed Framework that resulted from

Jimmy Carter’s meetings with Kim Il Sung was a negotiated outcome between two parties.

Nevertheless, the United States agreed to tangible and intangible concessions, which has been coded elsewhere as U.S. accommodation of North Korean demands.142

After the United States backed down in summer 1993, we see all three of the outcomes

H1 anticipated occurring. As expected, the first phase of the crisis came to an end because the

United States was willing to satisfy key North Korean demands, as was the end of the crisis in

142 Rock, Appeasement in International Politics, 150-1.

338 June 1994. Backing down proved effective in de-escalating the crisis. But H1also posits that backing down makes future challenges more likely, which also seems to bear out in this case.

Less than three months after the United States agreed to explore providing LWRs to

North Korea in exchange for a suspension of its NPT withdrawal threat, North Korea increased its list of demands, and began issuing threats anew. After the United States reluctantly agreed to the concessions package in the Agreed Framework to end the second phase of the crisis, North

Korean challenges—in the context of threats made relating to its nuclear program, as well as apart from the nuclear program—continued, consistent with H1 claims. After North Korea test- launched an ICBM that traveled over Japan in 1998, it began two years of missile negotiations to try and curb North Korea’s missile production and proliferation.143 Not to be outflanked in terms of aggression, when the George W. Bush administration came to office talking tough and making clear that it had no intention to “appease” North Korea, the latter began making classically fiery threats concomitant with its actual withdrawal from the NPT.144 Since the end of the nuclear crisis in 1994, North Korean threats and military provocations have occurred directly against either the United States or South Korea during every period during which the United States was not in diplomatic negotiations with North Korea.145

The final aspect of the nuclear crisis that appears congruent with the claims of H1 is the positive correlation between backing down and future threat credibility. North Korea did seem

143 For a history, see Michishita, North Korea’s Military-Diplomatic Campaigns, 117-37. 144 The history of U.S.-North Korea clashes during the Bush administration is still being written, but for an early firsthand account see Charles L. Pritchard, Failed Diplomacy: The Tragic Story of How North Korea Got the Bomb (Washington, DC: Brookings Institution Press, 2007). 145 Office of the Korea Chair, Record of North Korea’s Major Conventional Provocations since the 1960s (Washington, DC: Center for Strategic and International Studies, 2010).

339 to find U.S. coercive signaling credible after the latter accommodated North Korea’s demands in summer 1993, hence the North Korean comment to U.S. officials in April 1994 (immediately following the U.S. decision to deploy Patriot missiles) that “This will not be a situation like the

Iraq war. We will not give you time to collect troops around Korea to attack us .”146 But North

Korea believed U.S. threats before the United States had begun accommodating North Korean demands in summer 1993, shown in North Korea’s national mobilization and war footing in response to U.S. threats leading up to and including the Team Spirit exercise in March 1993.

The continuity of North Korean views of U.S. credibility across the nuclear crisis regardless whether the United States backed down or not suggests that factors other than honest

U.S. signaling (in the form of backing down) informed North Korea’s credibility judgment about the United States. The most compelling explanations for North Korean views of U.S. credibility, then, are some combination of the vast power asymmetry between rivals, which obviously preceded the crisis and continued throughout, and U.S. behavior during the Gulf War, which also preceded the nuclear crisis and which North Korea referred to subsequently as cause for its concern. So even though we can explain North Korean views of U.S. credibility better with alternatives to the honesty thesis, and even though we lack process tracing evidence to attribute the cause of North Korean views of U.S. credibility to backing down per se, it is nevertheless noteworthy that U.S. decisions to back down did not erode North Korean perceptions of U.S. threat credibility. Based on this case alone, we cannot conclude that the U.S. decision to back down made its subsequent threats more credible, but we can conclude that North Korea perceived its subsequent threats credible even though the United States had just backed down.

146 Wit, Poneman, and Gallucci, Going Critical, 169; Creekmore, A Moment of Crisis, 109.

340 It is quite difficult to reliably interpret the explanatory power of H2 in the nuclear crisis.

The assessment problem stems from coding challenges; in the multiple observations in this case, who was the challenger and who was the defender? I have made the case that the United States initiated the challenge against North Korea, primarily by wielding its threat of the Team Spirit exercise but also with allusions to the possibility of sanctions, which could have devastated the internationally isolated North Korea had they been implemented. Beyond this opening salvo of coercive signals, however, the case can be argued either way that North Korea or the United

States was the initiating challenger in the second phase of the crisis. After the end of the first phase of the nuclear crisis in summer 1993, North Korea continued the progressive forward movement of its military, and allowed the work of its newly established military command— responsible specifically for war preparations—to continue.147 It was in this context, and its rhetorical persistence that sanctions meant war, that North Korea levied its increased demands in

October 1993. In this way it might be said that the second phase of the crisis was initiated by

North Korea.

But the United States, despite accommodating North Korea in June and July 1993, never abandoned its general commitment to coercion, and conveyed multiple oblique threats in the latter half of 1993. In September 1993, Secretary Aspin stated publicly that sanctions would likely be required.148 President Clinton’s November 7 appearance on Meet the Press saw Clinton draw a red line with ambiguity about the consequences, indirectly hinting at the possibility of a

147 Republic of Korea Ministry of National Defense, Defense White Paper, 72. 148 Cited in FBIS, East Asia Section (September 29, 1993): 25-6.

341 preemptive strike declaring that.149 Following Clinton’s interview, Secretary Aspin told Bob

Woodward that Clinton’s comment about North Korea was a sort of ultimatum, reiterating that

“We will not let the North Koreans become a nuclear power.” 150 A reasonable person might point to U.S. comments such as these as proof that the United States was the challenger initiating the second phase of the crisis.

Proceeding, then, with the one clear observation of challenger identification, we find an outcome that seems to repudiate H2. Specifically, H2 anticipates that initiating coercive challenges does not lead to any clear strengthening of future threat credibility. In this case, however, we see that North Korea believed the initial threats the United States issued, leading to

North Korea reacting to Team Spirit with an unprecedented level of military preparations, seeming paranoia, and ultimately counter-coercive signaling. Given the historical context of

U.S. behavior during the Gulf War and the vast capability asymmetry between the United States and North Korea, it seems that initiating a coercive challenge—though somewhat counterproductive when assessed in terms of achieving its goals—did contribute to the U.S. reputation for resolve in North Korea’s eyes. It is worth noting as well that the U.S. decision to coerce North Korea at the outset was aligned with its own perceptions of its interests. The

United States had made nonproliferation its highest security priority; whether such a priority is as

“vital” to the security of the United States as North Korea’s nuclear program was to its continued existence is doubtful, but nevertheless a matter of debate, underscoring what I have repeated in

149 “Interview With Timothy Russert and Tom Brokaw on Meet the Press: 7 November 1993,” Public Papers of the President of the United States: William J. Clinton, Book II (Washington, DC: U.S. Government Printing Office, 1993), 1921. 150 As quoted in Millot, “Facing the Emerging Reality of Regional Nuclear Adversaries,” 47.

342 prior chapters—interests are neither fixed nor objective facts, and must be understood contextually. H2 performs poorly in this case, to the extent that we can code it.

Chapter VII: Conclusion: Findings and Implications

This dissertation started with a simple question: When do the actions of rival states in a crisis have consequences beyond the crisis in which they occur, if in fact they do? Answering this broad question logically leads to several more specific questions. What, if any, are the consequences of backing down from a rival’s challenge? Does backing down in a rivalry context strengthen future threat credibility, encourage future challenges from rivals, or is it irrelevant?

Does initiating challenges against rivals make future threats more credible? Is a firm response to a rival’s challenge the only way to make future threats more credible, or are there other ways?

The concept of, and debate surrounding, reputation in international relations is central to all these questions—whether, under what conditions, and how past word and deed affects others’ expectations of your word and deed in the future. A wave of scholarship on reputation published over the last 20 years has challenged not only the conventional wisdom that reputations matter, but also the very foundations of deterrence theory, claiming that decision-makers do not consider the past when assessing the credibility of another state’s threats. The real-world consequences of this pessimistic claim about reputation are huge: if statesmen making decisions about issuing threats, resorting to coercive force, or backing down from a fight exist in a world without reputations, then they can decide any fool thing they like in the present without concern for future consequences.

My general argument has been a direct but calibrated rebuttal against Reputation

Pessimists: Not only do decision-makers often consider the past when assessing the credibility of their rival’s threats, but their past word and deed affect the likelihood that a rival will challenge them in the future. This is not to say that statesmen ignore an adversary’s material capabilities or

343 344 the importance of the issue at stake when making calculations about threat credibility; on the contrary, capability and interest have always been considered necessary elements of credibility— but are they sufficient? Not in the recurring crises that punctuated the U.S.-North Korea rivalry.

I chose a rivalry dyad to test competing claims on reputation not only because we can derive multiple observations over time while naturally controlling for other factors like regime type and the balance of capabilities, but also because rivalries fit the circumstances that microeconomics and political science literatures tell us are the circumstances in which reputations matter most: under conditions of uncertainty and repeated play. Examining the U.S.-North Korea rivalry in particular has additional payoffs because it allows us to explore the origins of North Korean provocations—a persistent problem that in recent years have grown in intensity—while helping to advance the reputations debate. It also posits an answer to an outstanding question in the rivalries literature about why weaker states in asymmetric rivalries so often challenge their stronger rivals. But the U.S.-North Korea rivalry is useful for reasons of research design as well.

In terms of research design, focusing on crises within the U.S.-North Korea dyad stack the deck in favor of Reputation Pessimist theories of threat credibility by fitting the “most likely” conditions for their predictions to obtain. A focus on crises favors current calculus theory because it is in the impaired decision-making environment of a crisis that decision-makers are most likely to focus on present factors and pay scant attention to what happened in the past. As an ideal-type asymmetric rivalry, the U.S.-North Korea dyad also favors current calculus theory because the distribution of capabilities takes on such extreme values—that hold essentially constant over time—that calculations of credibility for both sides should be fairly straight

345 forward, especially since in three of the four cases the issues at stake were the same. The U.S.-

North Korea rivalry favors constant calculus theory as well, principally because each side has historically harbored strong out-group feelings of hostility toward the other; it is one of the longest hostile relationships that did not end with the end of the Cold War. If reputational hypotheses explain this set of cases better than current or constant calculus theory even though case selection favors the predictions of the latter theories, then we should be even more confident in the findings here and the role that reputations play in rivalries. If in addition reputational hypotheses can explain even more phenomena than credibility—specifically the likelihood of future challenges—then we have all the more reason to describe reputation optimism as a generally superior explanation, at least under conditions that mirror those examined in the U.S.-

North Korea rivalry.

Summary of Findings

Despite selecting a set of cases that should have proven easy for current and constant calculus theories, we find that reputational theories, in addition to providing a superior overall explanation of each case, more tightly fit the variations in threat credibility that occurred over time. Whereas Reputation Pessimists can only purport to explain how threat credibility was calculated by each side, a reputational approach is able to explain not only calculations of threat credibility in most instances, but also the recurrence of repeated challenges.

Pueblo Crisis

In the years immediately prior to North Korea’s seizure of the USS Pueblo in 1968,

North Korean military forces launched hundreds of small-scale provocations against U.S. and

South Korean forces, most of which involved various kinds of ground incursions across the

346 DMZ, but some of which included attempts to down U.S. reconnaissance aircraft. Prior to the

Pueblo’s seizure, the perception of U.S. decision-makers was that North Korea’s threatening rhetoric far outpaced the small acts of violence to which North Korea had habituated the United

States, leaving U.S. decision-makers incredulous when North Korea began issuing threats that presaged the attack on the Pueblo. North Korea had thus accrued a reputation for bluffing. The

United States, meanwhile, backed down from the hundreds of provocations launched between

1964 and the 1968 Pueblo seizure, leading North Korea to believe that the United States would not retaliate when challenged with military coercion short of all-out war. Once North Korea seized the ship, not only did U.S. officials suddenly begin finding North Korean threats credible, but the United States issued no threats of its own. Instead, the United States spent nearly a year agonizing over how it could accommodate North Korean demands without losing face or giving a false admission of guilt simply to save the Pueblo crew. When enough time had passed, the

United States did precisely that: it satisfied North Korea’s demand for the “three As” in a way that minimized the blow to U.S. prestige.

EC-121 Shoot Down

North Korea’s bold shoot down of a U.S. reconnaissance aircraft in 1969 took place only months after U.S. conciliation to resolve the Pueblo crisis. The EC-121 shoot down was arguably a more aggressive provocation than prior incidents, but it is difficult to claim that it was riskier given that North Korea seemed to believe it could take such actions with impunity— repeated U.S. non-responses or concessionary responses to North Korean challenges gave North

Korea no incentive not to use coercive violence. As with the Pueblo crisis, most U.S. officials

347 did not find the uptick of threatening rhetoric from North Korea in the months prior to the shoot down credible; North Korea’ s reputation for bluffing superseded any reputation it might have accrued for resolve because of willingness to repeatedly confront the United States with violence. Some, however, changed their tune once the shoot down took place. U.S. officials intensely debated whether to retaliate against North Korea’s bold action, and in these debates we see U.S. calculations being made on the basis of capability, interest, and reputation. U.S. decision-makers frequently consulted the balance of capabilities, but were not able to render an assessment of North Korean credibility based on it. U.S. interests were discussed, but they were inextricably tied to questions about North Korean threat credibility and the anticipated likelihood of the consequences of not retaliating for the shoot down, which included explicit considerations of North Korean past behavior. There was no U.S. consensus about what action to take, and the more hawkish officials in the Nixon administration placed disproportionate emphasis on North

Korea’s past behavior following U.S. instances of backing down as an indicator of North Korea’s future behavior if the United States continued to back down when challenged.

This is not to say that reputation is the province of military hawks; hawks and doves alike attributed a reputation for bluffing to North Korea, leaving everyone surprised by North Korea’s provocations. Moreover, it was the hawks in the Nixon administration that most sought to perpetuate hostility against North Korea, feeding the very conditions that would make it necessary for the United States to stand firm in response to a future challenge. Yes, backing down from a rival’s challenge makes future challenges more likely, but only if the state that backs down fails to ameliorate the conditions of dyadic hostility in the meantime. The

348 perpetuation of shared enmity and the U.S. reputation for lacking resolve were both necessary conditions for North Korea’s provocations. Ultimately, the Nixon administration issued no threats in response to the EC-121 shoot down, not only because the preponderance of officials believed North Korea’s post-shoot down threats, but also because neither the U.S. national security machinery nor its leadership had established the goals of any threat-making or punitive retaliation. So while the U.S. decision not to escalate allowed the crisis to fizzle out, North

Korean provocations across the DMZ, including a North Korean attack on a U.S. helicopter, continued in the months after it became clear that the United States would not respond to the EC-

121 shoot down.

Panmunjom Crisis

From 1970 until the bludgeoning of U.S. soldiers at Panmunjom in 1976, the frequency of North Korean provocations against the United States dropped sharply, largely because of on- again, off-again attempts at rapprochement. U.S. and South Korean efforts—however meager— to bridge their differences with North Korea in the early 1970s largely curbed the effects of normal hostile rivalry conditions. By 1975, however, attempts at bridge-building all but collapsed, and North Korean threats and provocations against the United States recurred in the

DMZ area. The United States did not find North Korea’s threats ahead of the murders at

Panmunjom credible; even the direct, explicit threats North Korean soldiers issued to U.S. soldiers as the latter attempted to trim a poplar tree were dismissed as bluster. The consequence was that the United States was once again caught off guard by North Korea’s violent act.

349 The primary difference between the Panmunjom crisis and the prior incidents were in the

U.S. reaction, which involved a large amount of coercive signaling, crossing the “red line” that

North Korea had previously drawn, and then issuing a deterrent threat of its own. In short, the

United States stood firm in response to North Korea’s challenge, and although the United States was prepared for North Korean retaliation, it seemed to believe that North Korea’s threats relating to the poplar tree were not credible, a calculation that was reached by consulting past

North Korean behavior during similar crises, the balance of power on the Peninsula, and U.S. and North Korean relative interests in the issue at stake. The United States also attached political demands to its coercive signaling, requiring that North Korea apologize, take measures to ensure such an incident did not occur in the future, and punish the soldiers responsible for the attack.

North Korea acquiesced to the first two demands, issuing an unprecedented apology from Kim Il

Sung himself, but refused to punish the soldiers involved, instead rewarding them. The crisis came to an end because North Korea was willing to partially meet U.S. demands, and the United

States was willing to declare “victory” by dropping its demand for punishment of the culpable

North Korean soldiers. Following the Panmunjom crisis, North Korean ire manifested violently against South Korea, but almost never again against the United States. Two incidents involving

North Korean attempts at military violence did occur in later years, but one was triggered by the accidental U.S. violation of North Korean airspace, and the other was a failed attempt that resulted in no losses. These military actions pale in comparison to the history of North Korean violence in the decade prior to the Panmunjom crisis.

Nuclear Crisis

350 The nuclear crisis, consisting of a first phase in early 1993 and a second phase in the first half of 1994, was different from the previous three crises. In the nuclear crisis, the principal issue at stake was not the U.S. military presence on the Korean Peninsula as it was in previous crises, but rather North Korea’s nuclear program. The challenger, for once, was also not North

Korea but the United States. From the first time the United States began wielding the threat of conducting the 1993 Team Spirit military exercise with South Korea, North Korea saw a credible threat in U.S. word and deed. North Korea interpreted the threat of the Team Spirit military exercise with perhaps even more malignity than the United States intended because it was viewing U.S. military moves around Korea through the lens of U.S. action in the Gulf War.

Even though the United States backed down to end the first phase of the nuclear crisis, we find that North Korea believed U.S. threats regardless of whether the United States had previously backed down when challenged. While substantial U.S. power was clearly a pre-requisite for

North Korea to believe any threats the United States made, we can convincingly say that at a minimum backing down did not negatively impact U.S. credibility, and that the U.S. decision to stand firm at a different place and time—specifically against Saddam Hussein to defend

Kuwait—did affect how North Korea interpreted U.S. threat credibility. Direct evidence from the case tells us quite clearly that North Korea focused on two factors when assessing North

Korean threats: U.S. capabilities and U.S. conduct of the Gulf War.

As with the previous crises, North Korea started with a reputation for bluffing because of its track record; U.S. decision-makers had trouble separating the “nuggets of reason” from the

“ocean of vitriol” when it came to assessing North Korean credibility. The familiar result was

351 that the United States was caught by surprise when North Korea actually made moves that it had threatened to make, which in this case included the combined threat of war and withdrawal from the NPT. Surprised by the alignment of North Korean word and deed, U.S. decision-makers then updated their perceptions of North Korea, shifting from outright incredulity to a presumption of credibility about North Korea’s threats. The first phase of the nuclear crisis only ended because the United States backed down, agreeing to meet some of North Korea’s demands in exchange for North Korea’s agreement to suspend its NPT withdrawal process. The second phase of the nuclear crisis was on a trajectory to culminate in war before a negotiated solution was reached in which the United States agreed to considerable concessions of both a tangible and intangible character. The U.S. decision to pursue the path of coercion in response to North Korean threats in the second phase of the nuclear crisis did not lead to North Korea backing down but to it engaging in brinkmanship and escalatory threats of its own. Although the exogenous influence of Jimmy Carter offered the United States a way to back down with minimal loss of prestige, the

United States and North Korea were on a path to conflict that was not so much driven by credibility in the end as it was by a nearly irreducible conflict of important interests over North

Korea’s nuclear program, revealing in the process an important lesson: just because you can convince an adversary that your threats are credible does not mean you can convince them to comply with your desires. When hostile rivalry conditions renewed in earnest in 2002, threats and challenges from North Korea renewed as well. The current status quo on the Korean

Peninsula, for better or for worse, is in part the legacy of how the United States chose to respond to North Korean threats and challenges.

352 Overall Summary

The lenses of reputations for resolve, and to some extent reputations for honesty, supply a coherent and accurate narrative of some of the most crucial aspects of the U.S.-North Korea rivalry: repeated crisis occurrence and the frequency of North Korean challenges; why the

United States did not believe North Korean threats before each crisis but did believe them following crisis onset; why North Korea was not deterred from conducting violent provocations by superior U.S. power alone; and why North Korea found U.S. threats credible in the rare occasions when it did make them. Although no alternative to reputational theories can account for this breadth of outcomes—including those of Reputation Pessimists—alternative theories provide insights regarding factors that are important for credibility to obtain.

Current calculus theory tells us that threats are credible “when—and only when—they are backed by sufficient power and serve clear interests.”1 Constant calculus theory tells us that

“Adversaries rarely get reputations for lacking resolve,” and as a consequence a rival’s threats will almost always be credible.2 Both theories of credibility articulate what they claim are sufficient conditions but that may be more accurately considered favoring conditions for threats to be found credible. In each of the cases here we find U.S. decision-makers consulting the balance of power when determining what to make of North Korean threats following a provocation, just as current calculus theory expects. In the nuclear crisis, we even find specifically that U.S. capabilities was one of two major factors on which North Korea based its assessment of U.S. threats. And as I have stated throughout this dissertation, a rivalry context in

1 Press, Calculating Credibility, 143. 2 Mercer, Reputation and International Politics, 213-14.

353 which hostile in-group/out-group conditions obtain is an important enabling condition in which we assess reputations for resolve and honesty affect threat credibility and the likelihood of future challenges; North Korean provocations were predictably most frequent and most intense when dyadic relations were at their most hostile.

Nevertheless, neither current nor constant calculus theory offers sufficient conditions to consistently explain threat credibility given that their predictions cannot explain why both U.S. and North Korean assessments of threat credibility varied across crises. Why would the United

States find North Korean threats incredible prior to a provocation but credible during a crisis?

Why would North Korea be willing to threaten and kill U.S. soldiers over a poplar tree in

Panmunjom on August18, 1976, but completely reverse its position over the same tree only three days later? And why would North Korea find U.S. power unintimidating during the 1960s but not during the nuclear crisis in 1993? These are questions that can best be answered by bringing the concept of reputations into a conversation that Reputation Pessimists would have limited to the balance of power, interests, and in-group/out-group psychology. Even when current and constant calculus theories are at their explanatory best, as during the nuclear crisis, the limited scope of their claims—focusing on threat credibility only—leaves unaddressed other important claims that reputational theories make about the likelihood of future challenges and the effect of bluffing.

Hypothesis Performance

At the beginning of this dissertation, I posited two hypotheses that relied on reputational concepts: the temporal effects of backing down (H1) and the temporal effects of initiating

354 coercive challenges (H2). For H1, I expected that backing down would have three effects: de- escalate the current crisis; strengthen future threat credibility; and increase the likelihood of future challenges. Although I was unable to observe each of these dependent variables in every case, by the end I was able to make at least two observations of each variable. The overall performance of H1 is summarized in Table 3 below.

Table 4 *Boxes shaded in grey match H1’s expectations H1 Results: The Effects of Backing Down in a Rivalry Context* Future Who Frequency or Current Threats Case Backed intensity of future Crisis Ended Found Down challenges increased Credible Pre-Pueblo Period U.S. N/A N/A Yes Pueblo Crisis U.S. Yes N/A Yes EC-121 Shoot Down U.S. Yes Yes Yes Panmunjom Crisis DPRK Yes N/A No Nuclear Crisis Phase U.S. Yes Yes Yes I Nuclear Crisis Phase Unclear Yes N/A No II

As Table 3 shows, backing down helped de-escalate every crisis in which it occurred. In the two instances where the state that backed down (the United States) made threats subsequently, those threats were deemed credible. And in four of the six observations, backing down led to future challenges. The two observations of backing down that were not followed by future challenges were both instances where the period after backing down saw a temporary abatement from rivalry conditions as both sides pursued rapprochement, however fleeting. This high degree of explanatory fit between H1 and the observed cases would be even higher if hundreds of provocations that took place prior to the Pueblo crisis were included as discrete observations.

355 Even though including these pre-Pueblo provocations would skew the overall results even more in favor of my H1 explanation, I excluded them as numerous separate observations because these provocations did not satisfy the case selection criteria I advanced in chapter two.

For H2, I anticipated that initiating coercive challenges would not have a systematic effect on the challenger’s reputation for resolve because reputations for resolve cannot be earned cheaply. Even though standing firm when challenged does impact the defender’s reputation for resolve, I posited that it was harder to accrue a reputation for resolve by initiating challenges.

Table 4, below, reveals mixed outcomes with an interesting potential explanation for why initiating coercive challenges does not systematically affect the challenger’s reputation for resolve.

Table 5 H2 Results: The Temporal Effect of Challenge Initiation Within- Crisis Post-Crisis Did Challenger Case Challenger Threats Threats Found Have History of Found Credible Bluffing? Credible Pre-Pueblo Period DPRK Yes No Yes Pueblo Crisis DPRK Yes No Yes EC-121 Shoot Down DPRK Yes and No No Yes Panmunjom Crisis DPRK No No Yes Nuclear Crisis Phase U.S. Yes Yes No I Nuclear Crisis Phase Unclear Yes N/A N/A II

Whereas the defender found the challenger’s coercive attempt credible in almost every case, that credibility assessment rarely held for any time period beyond the crisis in which it was first made. In fact, we find a noteworthy co-variance: four of five observations depended on whether

356 the challenger had a reputation for bluffing. North Korea’s four observed challenges did not accrue it a reputation for resolve because North Korea had made so many unfulfilled threats; the one instance in which the defender found the challenger’s post-crisis threats credible was when the challenger (the United States) did not have a reputation for bluffing. It is quite plausible that there are other unaccounted for factors that affect whether initiating challenges contributes to future threat credibility, but this study has uncovered one important factor likely to prevent that effect: a reputation for bluffing.

Implications for Theory

In this study, I did not seek to falsify theories advanced by Reputation Pessimists, nor do

I think a standard of falsification is even practical in the social sciences most of the time. As

Imre Lakatos advised, “theories are not only equally unprovable…they are also equally undisprovable.”3 I have instead sought to assess how well theories advanced by Reputation

Pessimists explain real-world outcomes under highly favorable conditions as compared to reputational theories. This kind of assessment is a necessary next step in the reputations debate and in building on the reputations research program. In so doing it also helps fill gaps in a separate research program on rivalries. While I have taken care in my case selection and research design to maximize confidence in my findings—by stacking the deck in favor of

Reputation Pessimist theories— the theoretical implications of this research must be offered with a certain amount of humility. This is only a small set of cases focused on only one rivalry dyad.

More observations of more cases are necessary to have confidence in the generalizability of my

3 Lakatos, “Falsification and the Methodology of Scientific Research Programmes,” 103.

357 research. My findings nevertheless have something meaningful to say about several theoretical topics.

Optimism, Pessimism, and Coercion Theory

The single most significant theoretical implication of this study is that, contra the claims of Reputation Pessimists like Daryl Press and Jonathan Mercer, reputations for resolve and honesty can matter a great deal in a rivalry context. Reputation Pessimists do not have the last word on the sources of threat credibility. Under conditions of uncertainty and repeated play—an ideal description of a rivalry—a state’s past words and deeds are likely to have their strongest effects. During crises in the U.S.-North Korea rivalry, calculations of credibility were most often based on a combination of factors that included the balance of capabilities and the adversary’s past word and deed. The classical literature on coercion theory emphasized the importance of past word and deed in establishing credibility, which is considered crucial for attempts at coercion to succeed.4 This study shows quite clearly that there are at least some times when this classical view is right.

Reputation Pessimists have tried to argue that although they do not deny that states and statesmen reflect on the past when trying to impute the future,5 they do not do so when it comes to assessing threat credibility, especially during a crisis. There are logical and anecdotal reasons to doubt that credibility is somehow exempt from the effects of the past, including admissions from past presidents that they frequently looked to past adversary behavior when formulating

4 Schelling, Arms and Influence; Jervis, The Logic of Images: 46; Kahn, On Thermonuclear War; Snyder, Deterrence and Defense, 36-7; Snyder, “The Security Dilemma and Alliance Politics,” 469-70. 5 Ernest May, “’Lessons’ of of the Past: The Use and Misuse of History in American Foreign Policy (New York: Oxford University Press, 1973); Yuen Foong Khong, Analogies at War: Korea, Munich, Dien Bien Phu, and the Vietnam Decisions of 1965 (Princeton: Princeton University Press, 1992).

358 foreign policy.6 The study of the U.S.-North Korea rivalry provides added empirical reasons to consider that past word and deed will affect not only future threat credibility, but also the likelihood of future challenges from rivals. And while I have focused on the U.S.-North Korea dyad, in the process I discovered some direct evidence from declassified documents that states other than North Korea thought the United States had revealed a reputation for irresoluteness based on how it handled the Pueblo crisis, saying “…After all, it isn’t so risky to defy the United

States—look at North Korea and the Pueblo.”7

This is by no means a rejection of Reputation Pessimism. The positive claims of current calculus theory in particular were mostly valid in this study; decision-makers did consult the balance of capabilities and the importance of the issue at stake when determining whether to belief the adversary’s threats. But the negative claim—that decision-makers will not consult the past when deciding whether to believe an adversary’s threats—is simply untrue when tested against historical episodes from the U.S.-North Korea rivalry. This study provides sufficient evidence to take the connection between reputations and threat credibility seriously, but we should not ignore that power and the value of the issue at stake matter a great deal as well.

Resolve, Honesty, and Cheap Talk

This dissertation also informs a logical tension within the reputation optimism camp, between advocates of a reputation for resolve and those of a reputation for honesty. Each makes

6 Lyndon B. Johnson, The Vantage Point: Perspectives of the Presidency, 1963-1969 (New York: Rinehart and Winston, 1971), 151. 7 Memorandum, “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” April 15, 1969 (6:30pm), Nixon Presidential materials, NSC Files, Korea: EC-121 Shootdown, North Korean Reconnaissance Shootdown, 4/9/69-4/16/69, Vol. 1 Haig, box 434, NARA; U.S. Department of State, April 15, 1969. “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 8: 17-20.

359 theoretically distinct claims that are mostly complementary. The resolve hypothesis tells us that aligning word and deed in a way that signals resolve is necessary to make threats credible. The honesty hypothesis tells us that aligning word and deed makes future threats credible regardless of what the deed is. Empirically, the resolve and honesty hypotheses expect the same outcome when standing firm; it is an honest form of communication and it signals resolve, therefore both hypotheses anticipate the future threat credibility will be strengthened as a result. Where they meaningfully differ is on the consequences of backing down from a challenge. The resolve hypothesis expects that backing down will make future challenges more likely, whereas the honesty hypothesis expects that backing down will make future challenges less likely because it is an honest form of communication. In the U.S.-North Korea rivalry, as Table 3 shows, the logic of reputations for resolve outperformed this aspect of the logic of reputations for honesty in four out of five observations. Every time the United States backed down, it was subsequently faced with more frequent or more aggressive challenges from North Korea.

But simply because reputations for resolve better explains the consequences of backing down than reputations for honesty does not mean that the honesty perspective is altogether wrong or useless. On the contrary, it is the reputations for honesty logic that best explains why initiating coercive challenges may not lead to an accrual of a reputation for resolve. Thanks to the reputations for honesty logic, we can posit that talk is most likely to be “cheap”—that is, threats are most likely to be incredible—when following a history of bluffing. North Korea initiated near countless challenges against the United States, yet the latter only found the former’s threats credible in a crisis induced by one of the challenges, but not after the crisis was

360 over. Moreover, it is the honesty perspective that best accounts for North Korean perceptions of

U.S. threat credibility. By repeatedly backing down when challenged, the United States signaled to North Korea that it could challenge the United States at levels below a threshold of open conflict without suffering direct costs. But because the United States was willing to suffer the imposition of North Korean violence without retaliation, it also signaled that it would not bandy threats about for issues of less than great importance to the United States. In this way, U.S. decisions to back down when its interests were not sufficiently engaged may have strengthened the credibility of its future threats; at a minimum a history of backing down did not impair future threat credibility.

This study is the first to directly test the resolve and honesty perspectives against this dissertation’s case selection. It is not the final word, but rather among the first on the relationship of reputations for resolve and honesty to one another, as well as on which aspects of a reputation for honesty are most causally relevant and under what conditions.

Rivalries

The literature on rivalries—particularly asymmetric rivalries—has suffered from omitted variable bias by its failure to consider reputational approaches to explaining rivalry maintenance, escalation, serial crisis, and war outbreak phenomena under conditions when reputational approaches are quite plausible. This bias is of particular consequence to a major puzzle in studies addressing why the weaker state in an asymmetric rivalry so frequently challenges the stronger state. 8 Various explanations for this phenomenon have been advanced—from

8 Joseph M. Greico, “Repetitive Military Challenges and Recurrent International Conflicts, 1918-1994,” International Studies Quarterly 45, no. 2 (2001): 298.

361 geopolitical alignments9 to adverse shifts in the weaker state’s strategic situation10—but none of these has tested or even controlled for the effects of reputation, and none also simultaneously posits an explanation for why a rivalry would experience repeated crises without escalation to war.

This study has contributed to rectifying this bias by deeply exploring a specific asymmetric rivalry dyad in which not only the rivalry maintained over the entire Cold War, but suffered repeated crises that managed not to bubble over into war. Because the very nature of a rivalry involves two parties who find themselves in the present eyeing their adversary through a filter that was constructed based on past word and deed, the logic of reputation can and does seem to weigh heavily on decision-making; the past may not always be determinative of states’ decisions in the present, but sometimes it clearly does, and when it does not directly, its presence nevertheless permeates the circumstances in which decision-makers find themselves.

For rivalries scholars, this study offers a novel explanation for serial crises between rivals, the outcomes of which support the perpetuation of the rivalry. The United States and

North Korea experienced hundreds of clashes with one another over the course of the last 60 years, four of which constituted serious crises with the potential to lead to war. In three of these crises, as with nearly all of the other several hundred smaller incidents, it was North Korea—the smaller, weaker state—that initiated violence as a coercive challenge. The thousands of pages of evidence combed for insights into motivations and the interaction sequences that led to crisis

9 Steve Chan, “Major-Power Intervention and War Initiation by the Weak,” International Politics 47 (2010): 163- 185. 10 David Lemke, “Investigating the Preventive Motive for War,” International Interactions 29 (2003): 273-92; T.V. Paul, Asymmetric Conflicts: War Initiation by Weaker Powers (Cambridge: Cambridge University Press, 1994).

362 resolution each time strongly supports the claim that while the U.S. presence and military commitment to South Korea was an effective general deterrent against North Korea launching a renewed war campaign, the U.S. decisions to repeatedly back down when faced with North

Korean coercive challenges convinced North Korea that the United States would not retaliate. In

Kim Il Sung’s own words in a conversation about the Pueblo and EC-121 crises, “Americans don’t want to continue this fight. The Americans let us know that it’s not their intention to fight the Koreans again.”11

The United States took no serious or costly steps to ameliorate the hostile rivalry conditions it maintained with North Korea—indeed, some U.S. actions seemed intended to antagonize North Korea—yet it also refused to stand firm when challenged. Shared enmity motivates the weaker-state challenge, which itself is a calculation that the stronger state will not retaliate because it has not in the past. This line of reasoning is supported by the fact that when the United States did stand firm against North Korean provocation in 1976, violent challenges against the United States virtually ceased, though North Korea found other ways to express its virulence, principally against South Korea. Thus, viewed through the reputational lens, serial crises initiated by the weaker power can be seen as a function of persistent rivalry conditions combined with one party to the rivalry (in this case the stronger party) cultivating a reputation for backing down when challenged.

11 “Minutes of Conversation on the Occasion of the Party and Government Delegation on behalf of the Romanian Socialist Republic to the Democratic People’s Republic of Korea,” June 10, 1971, History and Public Policy Program Digital Archive, Archives of the Central Committee of the Romanian Communist Party, Obtained and translated for the NKIDP by Eliza Gheorghe.

363 Implications for Policy

This dissertation is not a study of coercion per se, nor is it intended to be a “how-to” guide for coercion. Neither has it rendered a judgment about the efficacy of U.S. policy toward

North Korea over the decades. Still, the main implication of my research for policymakers is clear and simple, if general: your word and deed in dealing with a rival in the present may affect your rival’s perception of you in the future. States are capable of accruing reputations for resolve/irresoluteness and for honesty/bluffing. Standing firm has its risks, but backing down or doing nothing in response to rival aggression has its own hazards. I cannot say that states should always or never back down from rivals; it depends on how the policymaker weighs the value of present versus future costs and benefits, in addition to calculations of risk propensity associated with any decision to stand firm, though the latter calculation should be highly correlated with the importance of the issue at stake. I have highlighted here only a few of the many factors to which policymakers must attend when rendering judgment calls in a complex decision-making environment. It would therefore be foolish for me to make any unequivocal statements about what must be done or what must never be done. But it would be equally foolish not to highlight the several specific policy relevant insights that logically flow from this study.

Backing Down is a Complex Decision

Backing down from a rival’s challenge is a temporally differentiated cost-benefit proposition: it pays off in the present by de-escalating the crisis in which it occurs; it pays a future benefit in the form of strengthening future threat credibility because you have shown you are willing to back down when your interests are not sufficiently engaged; and it increases future costs in the form of increased challenges from a rival, either in frequency or intensity. There is

364 an intuitive demand for policymakers to value the pressing matters of the present more than the unrealized potentialities of the future; perhaps that is the right call, and perhaps that is myopic.

In the U.S.-North Korea rivalry, this tendency produced the provocation paradox: U.S. decision- makers traded present day stability in each case (by not confronting North Korea when challenged) for more and more intense North Korean challenges in the future. This does not mean that the United States was making “bad” decisions. Is it worth risking war because one or two soldiers were killed? Maybe not. Is it worth risking a war to prevent nuclear proliferation, or worth risking a second war when a first war is already ongoing elsewhere? Again, maybe not.

My goal is not to criticize or validate past judgments. My goal is to ensure that policymakers are fully witting of the consequences of their decisions (or at least as witting as possible), and to be aware that those consequences are not necessarily confined to the time period in which a crisis erupts. Backing down from a rival, especially if repeatedly, makes it more likely that you will accrue a reputation for irresoluteness, and as a consequence more likely that you will experience future coercive challenges from your rival.

This is not necessarily cause for eternal pessimism about conflict between rivals, and it should not lead to the conclusion that you should retaliate every time a rival challenges you. In one of the seminal works on deterrence, Alexander George and Richard Smoke argued that the best thing deterrence could buy you is time; for deterrence to truly “succeed,” time bought must be used to ameliorate the hostile conditions that made deterrence necessary in the first place.12 It might be said that the same principle applies to backing down. The ability to de-escalate a crisis

12 Alexander George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974), 5.

365 is valuable, and if a nation’s interests are not sufficiently engaged then backing down may be the most prudent course. But if a nation takes steps to actively fan the flames of hostility with its former challenger subsequently, it only increases the likelihood that its former challenger becomes its future challenger. Backing down makes the most strategic sense if it occurs as part of a genuine attempt to reform a hostile relationship; one needs to alleviate the rivalry conditions that made a choice of standing firm or backing down necessary in the first place. Rivalry termination is not easy, requiring anything from the strategic use of appeasement to forced regime change or even war. But rivalries make realpolitik tendencies in statesmen more likely than in other contexts. The logic of rivalry places a premium on avoiding the appearance of weakness, which means the best way to avoid the consequences of such a zero-sum context is to break out of the stand-firm/back-down choice structure by finding a way to ameliorate rivalry conditions.

Bullying May Not Pay

Can a reputation for resolve be earned on the cheap? That is, can the frequent use of deterrent and compellent threats around the world contribute to a nation’s security without needing to resort to war? Some think so. Credible voices in U.S. policy circles have even begun heralding the “return of coercive diplomacy” as a way for the United States to exploit its power without incurring the heavy burdens of war.13 This may be dangerously wrongheaded. We saw in the nuclear crisis that U.S. attempts to initiate a coercive challenge did not lead to a desirable outcome; North Korea only escalated the situation, turning it into a crisis. Credibility is only one

13 See, for example, Sam Brannen, “The Return of Coercive Diplomacy,” Defense One (September 12, 2013). Available online, http://www.defenseone.com/ideas/2013/09/return-coercive-diplomacy/70284/.

366 element necessary to make coercion work. When there are irreducible conflicts of vital interest, a credible threat could simply result in conflict. Using coercive threats where interests are not genuinely engaged—a judgment that depends on context—will eventually lead to a costly war or reputational damage; it is therefore often better not to hazard the risk of using coercion if alternatives exist.14 And although the issue needs more study, it seems that a reputation for resolve is not easily earned at any rate. This dissertation has identified one factor that specifically prevents initiating coercive challenges from earning you a reputation for resolve over time: a reputation for bluffing. Despite North Korea’s hundreds of provocations, U.S. officials often viewed North Korea’s threats as nothing more than bluster. North Korea’s demonstration of toughness and a seeming willingness to risk war on occasion was not commensurate with its stream of violent rhetoric.

Bluffing Has Costs, But Not Always the Same Costs

Contemporary public debates about “red lines,” bluffing, and reputation surged with decisions about whether the United States should intervene with military force in Syria in response to Bashar Al-Assad’s use of chemical weapons on rebels in his own country.

Interestingly, the focus of the debate in the United States was not primarily whether the United

States should prevent a civil war, but whether the United States should implement its explicit deterrent threat against Syria the year prior, proscribing the use of chemical weapons. Hawkish pundits arguing for intervention made implicitly reputational arguments; U.S. credibility is on

14 Thies makes this point in the context of the Vietnam conflict. Thies, When Governments Collide, 416. See also Jervis, The Logic of Images.

367 the line, they say.15 Others, often of a libertarian bent, have argued that the United States has no business in Syria, and have even cited some of the reputation pessimism literature to explain that, in effect, bluffing does not affect credibility.16 Academics have even weighed in on the public debate, offering opinions with only the loosest ties to the literature on reputation optimism and pessimism.17 The commentary has gotten so opinionated and analytically unmoored that some scholars even refused to accept direct evidence when it contradicted their opinion, such as the claim of some scholars that U.S. Secretary of Defense Chuck Hagel was exaggerating when he said publicly that South Korean government officials have told him they are concerned about how North Korea will interpret U.S. word and deed on Syria.18

The U.S.-North Korea rivalry makes clear that bluffing has costs, as does accruing a reputation for irresoluteness. Under certain conditions, not only does failing to implement a threat you make weaken the credibility of your future threats, but backing down whether you made a threat or not increases the likelihood of future challenges. Neither of these findings necessarily determines the consequences of President Obama failing to implement threats for

Syrian chemical weapons use because it is not at all clear that the conditions for reputation to

15 Max Fisher, “Syria Crosses Red Line Again: U.S. Will Do Nothing,” Real Clear World (August 24, 2013). Available online, http://www.realclearworld.com/2013/08/21/syria_crosses_red_line_again_us_will_do_nothing_150919.html. 16 Doug Bandow, “Syria: The Only Red Line Should be to Stay Out,” Forbes Online (April 29, 2013). Available online, http://www.forbes.com/sites/dougbandow/2013/04/29/syria-the-only-red-line-should-be-to-stay-out/; Dylan Matthews, “Why Obama Shouldn’t Care About Backing Down on Syria,” Washington Post (September 12, 2013). Available online, http://www.washingtonpost.com/blogs/wonkblog/wp/2013/09/12/why-obama-shouldnt-care- about-backing-down-on-syria/. 17 Daryl G. Press and Jennifer Lind, “Red Lines and Red Herrings,” Foreign Policy (May 6, 2013). Available online, http://walt.foreignpolicy.com/posts/2013/05/06/red_lines_and_red_herrings; Fareed Zakaria, “U.S. Credibility is not on the Line in Syria,” Washington Post (May 8, 2013). Available online, http://fareedzakaria.com/2013/05/08/u-s- credibility-is-not-on-the-line-in-syria/. 18 Daniel Drezner, “What Do Syria and North Korea Have in Common?,” Foreign Policy (September 16, 2013). Available online, http://drezner.foreignpolicy.com/posts/2013/09/16/what_do_syria_and_north_korea_have_in_common.

368 have such effects are present in the U.S.-Syria relationship. Certainty—not uncertainty— characterizes Syria’s intentions to use chemical weapons on the rebels it fights; the weapons have already been used and the United States and United Nations have confirmed it.19 Also, beyond the immediate conflict between Assad and the rebels, notwithstanding the possibility that it could drag on for years, the United States should not necessarily have high expectations of future conflict interactions with Syria.

This is not to say either that it is “okay” to draw red lines that you refuse to implement or that it is always necessary to act simply because you made a prior commitment to do so. Syria does not fit the conditions under which reputational effects are strongest, which implies that perhaps the costs of a U.S. bluff on Syria are low. But as we saw in the conversation between the presidents of Egypt and Jordan reflecting on the lack of U.S. resolve in handling the Pueblo seizure, the consequences of word and deed are not always limited to the dyadic context in which they take place.20 It is therefore risky to be cavalier with threat-making. States do accrue reputations, and the United States risks impairing its own if it develops a reputation for bluffing.

Areas for Future Research

This study hardly represents the final word in the reputations debate. Given the real- world consequences at stake, it cannot afford to be so. This represents only one set of cases

19 Carol J. Williams, “UN Probe Confirms Syria Chemical Weapons Use: Now the Hard Part,” Los Angeles Times (September 17, 2013). Available online: http://articles.latimes.com/2013/sep/17/world/la-fg-wn-un-syria-chemical- weapons-what-next-20130916. 20 Memorandum, “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” April 15, 1969 (6:30pm), Nixon Presidential materials, NSC Files, Korea: EC-121 Shootdown, North Korean Reconnaissance Shootdown, 4/9/69-4/16/69, Vol. 1 Haig, box 434, NARA; U.S. Department of State, April 15, 1969. “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” FRUS, 1969-76, Vol. XIX, Part 1, Korea, 1969- 1972, document 8: 17-20.

369 derived from one rivalry dyad. More must be done. First, the hypotheses generated here need to be tested against a larger number of cases to determine the extent to which my findings are generalizable to a broader universe of cases. Second, the honesty and resolve hypotheses, which previously have not been tested against one another as I have done here, should not only be applied to other rivalry dyads, but to non-rivalry contexts; it is plausible that a reputation for honesty explains non-rivalry interactions far better than a reputation for resolve logic. Third, greater attention must be paid to whether or not initiating challenges actually pays reputational dividends. I have identified reputations for bluffing as one factor that prevents a reputation for resolve accruing to an aggressor, but there may be other factors that create the same effect.

Finally, there are at least three areas of international relations scholarship not addressed in this dissertation where we might find the concepts of reputations for resolve and honesty to be either particularly useful or limited: nonproliferation; cyberspace; and in nuclear dyads.

Chapter five’s discussion of the geopolitical context of the 1976 Panmunjom crisis briefly touched upon South Korea’s clandestine nuclear program, which sprang from fear of U.S. abandonment and various U.S. signals implying that the time would come when South Korea could no longer rely on the United States for its security. The motivations that drove the South

Korean nuclear program were likely multifaceted, but evidence from chapter five suggests that the repeated U.S. unwillingness to retaliate against North Korean violence, as well as its efforts to restrain South Korean retaliation on many occasions, eroded the U.S. reputation for reliability.

It is possible that perceptions of U.S. irresoluteness exacerbated perceptions of U.S. unreliability, which fed the South Korean insecurity that led to its fleeting attempt at nuclear proliferation.

370 Although this episode in South Korea’s history merits further inquiry through a reputational lens, what we already see is an important causal interplay among reputations for resolve with reputations for reliability and preservation of the nuclear taboo. The very logic of nonproliferation as a regime rests on the notion that demonstrating resolve and enforcing proscriptions against the development of nuclear weapons is necessary at all times in order to prevent cascading violations of the nuclear taboo.21 It is reasonable, then, to expect that there is an articulable but as yet unspecified linkage between reputations for resolve and the strength of the nonproliferation norm.

The second area of scholarship where reputations for resolve and honesty could prove useful is in the burgeoning literature on cyberspace. As a domain of international interaction, scholars are just beginning to generate knowledge structures about the logic of actor behavior in cyberspace and how behavioral dynamics in that realm can affect more traditional understandings of war and peace.22 As discussed in chapter one, the concept of reputation has played a crucial role in building the extensive research program on deterrence and coercion that exists today. Yet, our limited and evolving understanding of cyberspace suggests that attribution of cyber attacks is harder than traditional attacks,23 and even more barriers to successful coercion with cyber weapons may exist than in coercion with conventional weapons.24 Reputational

21 Maria R. Roblee, Nonproliferation Norms: Why States Choose Nuclear Restraint (Georgia: University of Georgia Press, 2009). 22 For a review article characterizing the limited state of knowledge on cyber, see Lucas Kello, “The Meaning of Cyber Revolution: Perils to Theory and Statecraft,” International Security 38, no. 2 (2013): 7-40. 23 David D. Clark and Susan Landau, “Untangling Attribution,” in Proceedings of a Workshop on Deterring Cyberattacks: Informing Strategies and Developing Options for U.S. Policy (Washington, DC: National Academies Press, 2010): 25–40. 24 Jon R. Lindsay, “Stuxnet and the Limits of Cyber Warfare,” Security Studies 22, no. 3 (2013): 365-404.

371 logics of resolve and honesty might be a useful way of making sense of a new domain of human interaction, but may also reveal the limits of crucial security concepts in this new domain.

Finally, the role, and perhaps limits of reputation in a nuclear dyad requires examination.

Whereas this dissertation revealed that reputations for resolve and honesty can be a powerful way of understanding conflict-oriented behavior in an asymmetric rivalry, only the United States possessed nuclear weapons during the periods examined. When open hostility between the

United States and North Korea renewed in 2002, the status of North Korea’s nuclear program was uncertain but some suspected that North Korea might have already possessed nuclear weapons.25 By 2006, North Korea had conducted a confirmed nuclear test.26

The timing of North Korea’s nuclear acquisition in relation to renewed hostilities is important for several reasons. First, the fact that U.S.-North Korea disputes since 2002 have not escalated to war since North Korea obtained nuclear weapons could potentially be attributed to speculation that both the United States and North Korea had the ability to “go nuclear,” even though reputation played an important role in explaining why recurring disputes did not escalate to war prior to 2002. Second, “proliferation optimists” have argued that, “…the probability of major war among states having nuclear weapons approaches zero…”27 Particularly during crises, the argument has been made that the overwhelming devastation associated with nuclear weapons incentivizes even the most irrational rivals to restrain themselves when both parties to

25 Yoichi Funabashi, The Peninsula Question: A Chronicle of the Second Korean Nuclear Crisis (Washington, DC: Brookings Institution Press, 2007). 26 Ibid. 27 Kenneth N. Waltz, “Nuclear Myths and Political Realities,” American Political Science Review 84 (1990): 740.

372 the rivalry possess nuclear weapons.28 In a nuclear dyad, therefore, threat credibility becomes a matter of simple risk aversion—threats of war must be taken seriously because the costs of being incredulous but wrong are unacceptably high. Thus, the logic of nuclear crises in which both parties have nuclear weapons may marginalize the importance of reputations in calculations of threat credibility. This suggests at least the possibility of an important scope condition for the applicability of reputation, even in a rivalry context where reputations should have their strongest effects. On the other hand, if future researchers can show that past word and deed did have a direct effect on credibility in a crisis between nuclear powers, this would serve as an even harder test of reputations than the test contained in this dissertation, and would strengthen confidence in the applicability of reputations in a wide range of contexts.

Third, the phenomenon dubbed the “stability-instability paradox,” which describes low- intensity conflict between two nuclear powers that are nevertheless deterred from going to war with one another, could benefit from the analysis of U.S.-North Korea relations since 2002.29

Although North Korea and the United States have, with few exceptions, been on a renewed hostile footing toward one another since 2002, neither has initiated violence against the other; this is incongruent with the predictions of the stability-instability paradox. Further research will be necessary to understand exactly why this is so. It may be possible that although reputations may have almost no effect on credibility during a nuclear crisis, reputations may be particularly

28 See, for example, Oran Young’s treatment of the Berlin and Cuban missile crises. Oran Young, The Politics of Force: Bargaining During International Crises (Princeton, NJ: Princeton University Press, 1968), 311-36. 29 Glenn H. Snyder, “The Balance of Power and the Balance of Terror,” in Balance of Power, ed. Paul Seabury (San Francisco, CA: Chandler, 1965).

373 important for discerning credibility about threats relating to lower levels of violence, below the nuclear threshold, even if the conflict is between nuclear powers.

Bibliography Reputations

Baldwin, David A. “The Power of Positive Sanctions,” World Politics 24 (1971): 19-38.

Bratton, Patrick C. “When is Coercion Successful? And Why Can’t We Agree on It?” Naval War College Review 58, no. 3 (Summer 2005): 99–120.

Brodie, Bernard. Escalation and the Nuclear Option. Princeton: Princeton University Press, 1966.

Byman, Daniel, and Matthew Waxman. The Dynamics of Coercion: American Foreign Policy and the Limits of Military Might. New York: Cambridge University Press, 2002.

Clare, Joe, and Vesna Danilovic. “Reputation for Resolve, Interest, and Conflict,” Conflict Management and Peace Science 29, no. 1 (2012): 3-27.

Copeland, Dale C. “Do Reputations Matter?” Security Studies 7, no. 1 (1997): 33-71.

Crescenzi, Mark J.C. “Reputation and Interstate Conflict,” American Journal of Political Science 51, no. 2 (2007): 382-96.

Danilovic, Vesna. “The Sources of Threat Credibility in Extended Deterrence,” Journal of Conflict Resolution 45, no. 3 (2001): 341-369.

Davis, James W. Threats and Promises: The Pursuit of International Influence. Baltimore, MD: Johns Hopkins University Press, 2000.

Downes, Alexander B., and Todd S. Sechser. “The Illusion of Democratic Credibility,” International Organization 66, no. 3 (Summer 2012): 457-89.

Fearon, James D. “Domestic Political Audiences and the Escalation of International Disputes,” American Political Science Review 88, no. 3 (1994): 577-92.

Fioretos, Orfeo. “Historical Institutionalism in International Relations,” International Organization 65, no. 1 (Spring 2011): 367-99.

George, Alexander, and Richard Smoke. Deterrence in American Foreign Policy: Theory and Practice. New York: Columbia University Press, 1974.

374

Glaser, Charles L. “Political Consequences of Military Strategy: Expanding and Refining the Spiral and Deterrence Models,” World Politics 44, no. 4 (July 1992): 497-538.

Greico, Joseph M. “Repetitive Military Challenges and Recurrent International Conflicts, 1918- 1994,” International Studies Quarterly 45, no. 2 (2001): 298.

Guisinger, Alexandra, and Alastair Smith. “Honest Threats: The Interaction of Reputation and Political Institutions in International Crises,” Journal of Conflict Resolution 46, no. 2 (April 2002): 175-200.

Hopf, Ted. Peripheral Visions: Deterrence Theory and American Foreign Policy in the Third World, 1965-1990. Anne Arbor, MI: University of Michigan Press, 1994.

Huth, Paul K., and Bruce R. Russett. “What Makes Deterrence Work? Cases from 1900-1980,” World Politics 36, no. 4 (July 1984): 496-526.

Huth, Paul K. “Reputation and Deterrence: A Theoretical and Empirical Assessment,” Security Studies 7, no. 1 (1997): 72-99.

Jervis, Robert. The Logic of Images in International Relations. Princeton: Princeton University Press, 1970.

______. Perception and Misperception in International Politics. Princeton, NJ: Princeton University Press, 1976.

______. “Deterrence Theory Revisited,” World Politics 31 (1979): 289-324.

Kahn, Hermann. On Thermonuclear War. Princeton: Princeton University Press, 1960.

______. On Escalation: Metaphors and Scenarios. Baltimore, MD: Penguin, 1965.

Kaufman, William. “The Requirements of Deterrence,” in Military Policy and National Security, ed. William Kaufman. Princeton, NJ: Princeton University Press, 1956.

Knopf, Jeffrey W. “The Fourth Wave in Deterrence Research,” Contemporary Security Policy 31, no. 1 (2010): 1-33.

Kreps, David M., and Robert Wilson. “Reputation and Imperfect Information,” Journal of Economic Theory 27 (1982): 245-79.

375

Kydd, Andrew H. “Sheep in Sheep’s Clothing: Why Security Seekers Do Not Fight Each Other,” Security Studies 7, no. 1 (1997): 114-55.

______. Trust and Mistrust in International Relations. Princeton: Princeton University Press, 2005.

______. “When Can Mediators Build Trust?” American Political Science Review 100, no. 3 (2006): 449-462.

Lemke, David. “Investigating the Preventive Motive for War,” International Interactions 29 (2003): 273-92.

Mercer, Jonathan. Reputation & International Politics. Ithaca, NY: Cornell University Press, 1996.

Milgrom, Paul, and John Roberts. “Predation, Reputation, and Entry Deterrence,” Journal of Economic Theory 27 (1982): 280-312.

Miller, Gregory, D. “Hypotheses on Reputation: Alliance Choices and the Shadow of the Past,” Security Studies 12, no. 3 (2003): 40-78.

______. The Shadow of the Past: Reputation and Military Alliances before the First World War. Ithaca, NY: Cornell University Press, 2012.

Montgomery, Evan Braden. “Breaking Out of the Security Dilemma: Realism, Reassurance, and the Problem of Uncertainty,” International Security 31, no. 2 (2006): 151-85.

Morgan, Forrest E., Karl P. Mueller, Evan S. Medeiros, Kevin L. Pollpeter, and Roger Cliff, Dangerous Thresholds: Managing Escalation in the 21st Century. Santa Monica, CA: Rand Corporation, 2008.

Morgan, Patrick M. “Saving Face for the Sake of Deterrence,” in Psychology and Deterrence, eds. Robert Jervis, Richard Ned Lebow, and Janice Gross Stein. Baltimore: Johns Hopkins University Press, 1986.

Orme, John David. Deterrence, Reputation, and Cold War Cycles. London: Macmillan, 1992.

Patchen, Martin. Resolving Disputes between Nations: Coercion or Conciliation? Durham, NC: Duke University Press, 1988.

376

Petersen, Walter J. “Deterrence and Compellence: A Critical Assessment of Conventional Wisdom,” International Studies Quarterly 30, no. 3 (1986): 269-94.

Pierson, Paul. “Not Just What But When: Timing and Sequence in Political Processes,” Studies in American Political Development 14 (2000): 72-92.

Posen, Barry R. Inadvertent Escalation: Conventional War and Nuclear Risks. Ithaca: Cornell University Press, 1991.

Press, Daryl, G. Calculating Credibility: How Leaders Assess Military Threats. Ithaca, NY: Cornell University Press, 2005.

Rock, Stephen R. Appeasement in International Politics. Lexington, KY: University of Kentucky Press, 2000.

Sartori, Anne E. “The Might of the Pen: A Reputational Theory of Communication in International Disputes,” International Organization 56, no. 1 (2002): 121-149.

______. Deterrence by Diplomacy. Princeton, NJ: Princeton University Press, 2007.

Schelling, Thomas. The Strategy of Conflict. New York: Oxford University Press, 1963.

______. Arms and Influence. New Haven: Yale University Press, 1966.

Schmalensee, Richard. “Advertising and Entry Deterrence: An Exploratory Model,” Journal of Political Economy 91 (1983): 636-53.

Schweller, Randall L. “Bandwagoning for Profit: Bringing the Revisionist State Back In,” International Security 19, no. 1 (1994): 72-107.

Selten, Richard. “The Chain-store Paradox,” Theory and Decision 9 (1978): 127-59.

Sescher, Todd S. “Goliath’s Curse: Coercive Threats and Asymmetric Power,” International Organization 64, no. 3 (2010): 627-60.

Shimshoni, J. Israel and Conventional Deterrence. Ithaca, NY: Cornell University Press, 1988.

Snyder, Glenn H., and Paul Diesing. Conflict among Nations: Bargaining, Decision-Making, and System Structure in International Crises. Princeton: Princeton University Press, 1977.

377

Snyder, Glenn H. Deterrence and Defense: Toward a Theory of National Security. Princeton: Princeton University Press, 1961.

______. “The Balance of Power and the Balance of Terror,” in Balance of Power, ed. Paul Seabury. San Francisco, CA: Chandler, 1965.

______. “The Security Dilemma in Alliance Politics,” World Politics 36 (1984): 469-70.

Snyder, Jack. “Rationality at the Brink,” World Politics 30 (1978): 345-65.

Tang, Shiping. “Reputation, Cult of Reputation, and International Conflict,” Security Studies 14, no. 1 (2005): 34-62.

Thies, Wallace J. When Governments Collide: Coercion and Diplomacy in the Vietnam Conflict, 1964-1968. Berkley: University of California Press, 1980.

______. “Compellence Failure or Coercive Success: The Case of NATO and Yugoslavia,” Comparative Strategy 22, no. 2 (2003): 243-67.

Tilly, Charles. Big Structures, Large Processes, Huge Comparisons. New York: Russell Sage Foundation, 1984.

Tingley, Dustin H., and Barbara F. Walter. “The Effect of Repeated Play on Reputation Building: An Experimental Approach,” International Organization 65, no. 1 (Spring 2011): 343-65.

Tomz, Michael. “Domestic Audience Costs in International Relations: An Experimental Approach,” International Organization 61, no. 4 (2007): 821-40.

______. Reputation and International Cooperation: Sovereign Debt across Three Centuries. Princeton: Princeton University Press, 2007.

Trachtenberg, Marc. A Constructed Peace: The Making of the European Settlement, 1945-1963. Princeton: Princeton University Press, 1999.

Walter, Barbara F. Reputation and Civil War: Why Separatist Conflicts are So Violent. Cambridge: Cambridge University Press, 2009.

Wiegand, Krista E. “Militarized Territorial Disputes: States’ Attempts to Transfer Reputations for Resolve,” Journal of Peace Research 48, no. 1 (2011): 101-13.

378

Wohlstetter, Albert. “The Delicate Balance of Terror,” Foreign Affairs 37, no. 2 (1959): 211-34.

Young, Oran. The Politics of Force: Bargaining During International Crises. Princeton, NJ: Princeton University Press, 1968.

379

Rivalries

Bennett, Scott D. “Security, Bargaining, and the End of Interstate Rivalry,” International Studies Quarterly 40 (1996): 157-83.

______. “Democracy, Regime Change, and Rivalry Termination,” International Interactions 22 (1997): 369-97.

______. “Integrating and Testing Models of Rivalry Duration,” American Journal of Political Science 42 (1998): 1200-32.

Blainey, Geoffrey. Why Nations Go to War. New York: Free Press, 1973.

Chan, Steve. “Major-Power Intervention and War Initiation by the Weak,” International Politics 47 (2010): 163-185.

Colaresi, Michael P., and William R. Thompson. “Hot Spots or Hot Hands? Serial Crisis Behavior, Escalating Risks, and Rivalry,” The Journal of Politics 64, no. 4 (November 2002): 1175-98.

Colaresi, Michael. “Shocks to the System: Great Power Rivalries and the Leadership Long Cycle,” Journal of Conflict Resolution 45 (2001): 569-93.

______. “When Doves Cry: International Rivalry, Unreciprocated Cooperation, and Leadership Turnover,” American Journal of Political Science 48, no. 3 (2004): 555-70.

Crescenzi, Mark J.C., and Andrew J. Enterline. “Time Remembered: A Dynamic Model of Interstate Interaction,” International Studies Quarterly 45, no. 4 (2001): 409-31.

Crescenzi, Mark J.C., Jacob D. Kathman, and Stephen B. Long. “Reputation, History, and War,” Journal of Peace Research 44, no. 6 (2007): 651-67.

Darnton, Christopher. “Rivalry and Rapprochement in Early Cold War Central America,” Security Studies 20, no. 2 (2011): 198-237.

Diehl, Paul F., and Gary Goertz. War and Peace in International Rivalry. Ann Arbor, MI: University of Michigan Press, 2000.

Fearon, James D. “Signaling Foreign Policy Interests: Tying Hands Versus Sinking Costs,” International Organization 41, no. 1 (1997).

380

Ganguly, Sumit, and William R. Thompson, eds. Asian Rivalries: Domestic Politics, Conflict Escalation, and Limits on Two-level Games. Palo Alto: Stanford University Press, 2012.

Geller, Daniel S. “Power Differentials and War in Rival Dyads,” International Studies Quarterly 37, no. 2 (1993): 173-93.

Goertz, Gary, and Paul F. Diehl. “Enduring Rivalries: Theoretical Constructs and Empirical Patterns,” International Studies Quarterly 37 (1993): 147-71.

______. “The Initiation and Termination of Enduring Rivalries: The Impact of Political Shocks,” American Journal of Political Science 39, no. 1 (1995): 30-52.

Goertz, Gary, Bradford Jones, and Paul F. Diehl. “Maintenance Processes in International Rivalries,” The Journal of Conflict Resolution 49, no. 5 (October 2005): 742-69.

Hensel, Paul R. “One Thing Leads to Another: Recurrent Militarized Disputes in Latin America 1816-1986,” Journal of Peace Research 31, no. 3 (1994): 281-98.

______. “An Evolutionary Approach to the Study of Rivalry,” Conflict Management and Peace Science 17 (1999): 175-206.

Jervis, Robert. “Perceiving and Coping with Threat,” in Psychology and Deterrence, eds. Robert Jervis, Richard Ned Lebow and Janice Gross Stein. Baltimore, MD: Johns Hopkins University Press, 1988.

Lebow, Richard Ned. Between War and Peace: The Nature of International Crisis. Baltimore, MD: Johns Hopkins University Press, 1981.

______. “Windows of Opportunity: Do States Jump through Them?” International Security 9, no. 1 (1984): 147-86.

Leng, Russell J. “When Will They Ever Learn? Coercive Bargaining in Recurrent Crises,” Journal of Conflict Resolution 27 (1983): 379-419.

______. Interstate Crisis Behavior, 1816-1980: Realism vs. Reciprocity. Cambridge: Cambridge University Press, 1993.

______. Bargaining and Learning in Recurring Crises: The Soviet-American, Egyptian- Israeli, and Indo-Pakistani Rivalries. Ann Arbor: University of Michigan Press, 2000.

381

______. “Escalation: Competing Perspectives and Empirical Evidence,” International Studies Review 6, no. 4 (2004): 51-64.

Maoz Zeev, and Ben Mor. Bound by Struggle. Ann Arbor: University of Michigan Press, 2002.

Paul, T.V. Asymmetric Conflicts: War Initiation by Weaker Powers. Cambridge: Cambridge University Press, 1994.

Rasler, Karen A., and William R. Thompson. “Contested Territories, Strategic Rivalries, and Conflict Escalation,” International Studies Quarterly 50, no. 1 (2006): 145-68.

______. “Rivalries and the Democratic Peace in the Major Power Subsystem,” Journal of Peace Research 38, no. 6 (2001): 659-83.

______. “Political Shocks and the De-escalation of Protracted Conflicts: The Israeli- Palestine Case,” in Evolutionary Interpretations of World Politics, ed. William R. Thompson. New York: Routledge (2001): 240-600.

Russett, Bruce. Grasping the Democratic Peace: Principles for a Post-Cold War World. Princeton, NJ: Princeton University Press, 1993.

Scobell, Andrew. “China and Taiwan: Balance of Rivalry with Weapons of Mass Democratization,” in Asian Rivalries: Conflict, Escalation, and Limitations on Two-level Games, eds. Sumit Ganguly and William R. Thompson. Palo Alto: Stanford University Press, 2012.

Stein, Janice. “Deterrence and Learning in an Enduring Rivalry: Egypt and Israel, 1948-73,” Security Studies 6 (1996): 104-52.

Stinnett, Douglas, and Paul F. Diehl. “The Path(s) to Rivalry: Behavioral and Structural Explanations of Rivalry Development,” Journal of Politics 63, no. 3 (2001): 717-40.

Thompson, William R. “Identifying Rivals and Rivalries in World Politics,” International Studies Quarterly 45, no. 4 (2001): 557-587.

Vasquez, John A. The War Puzzle. Cambridge: Cambridge University Press, 1993.

Waltz, Kenneth N. Theory of International Politics. Reading, MA: McGraw Hill, 1979.

382

Womack, Brantley. “Asymmetric Rivals: China and Vietnam,” in Asian Rivalries: Conflict, Escalation, and Limitations on Two-Level Games, eds. Sumit Ganguly and William R. Thompson. Palo Alto: Stanford University Press, 2012.

383

North Korea

An, Tai Sung. North Korea in Transition: From Dictatorship to Dynasty. Westport, CT: Greenwood Press, 1983.

Armstrong, Charles K. “America’s Korea, Korea’s Vietnam,” Critical Asian Studies 33, no. 4 (2001): 527-39.

______. “Juche and North Korea’s Global Aspirations,” NKIDP Working Paper #1. Washington, DC: Woodrow Wilson International Center, 2009.

Baek, Dong-ho. Silmido (in Korean). Seoul: Balgeun Saesang, 2004.

Bermudez, Jr., Joseph S. The Armed Forces of North Korea. London: I.B. Tauris, 2001.

Bernstein, Thomas P., and Andrew J. Nathan. “The Soviet Union, China, and Korea,” in The U.S.-South Korean Alliance, eds. Gerald L. Curtis and Sung-joo Han. Lexington, MA: Lexington Books, 1983.

Bolger, Daniel P. “Scenes from an Unfinished War: Low-Intensity Conflict in Korea, 1966-69,” Leavenworth Papers, No. 19. Ft. Leavenworth, KS: Combat Studies Institute, 1991.

Brazinsky, Gregg. Nation Building in South Korea: Koreans, Americans, and the Making of a Democracy. Chapel Hill, NC: University of North Carolina Press, 2007.

Bucher, Lloyd M., with Mark Rascovich. Bucher: My Story. New York: Doubleday, 1970.

Caprio, Mark. Japanese Assimilation Policies in Colonial Korea, 1910-1945. WA: University of Washington Press, 2009.

Cha, Victor D. “Hawk Engagement and Preventive Defense on the Korean Peninsula,” International Security 27, no. 1 (2002): 40-78.

Chae, Ria, ed. East German Documents on Kim Il Sung’s Trip to Beijing in 1975. NKIDP Document Reader. Washington, DC: Woodrow Wilson International Center for Scholars, 2012.

CIA Directorate of Intelligence, November 26, 1968. Kim Il-Sung’s New Military Adventurism. Freedom of Information Act request, released May 2007.

384

Creekmore, Jr., Marion. A Moment of Crisis: Jimmy Carter, The Power of a Peacemaker, and North Korea’s Nuclear Ambitions. Cambridge, MA: PublicAffairs, 2006.

Dingman, Roger. “Atomic Diplomacy during the Korean War,” International Security 13, no. 3 (1988/89): 60-86.

Downs, Chuck.Over the Line: North Korea’s Negotiating Strategy. Washington, DC: AEI Press, 1998.

Drennan, William M. “Nuclear Weapons and North Korea: Who’s Coercing Whom?” in The United States and Coercive Diplomacy, eds. Robert J. Art and Patrick M. Cronin. Washington, DC: U.S. Institute of Peace Press, 2003.

Drezner, Daniel, September 16, 2013. “What Do Syria and North Korea Have in Common?,” Foreign Policy. Available online: http://drezner.foreignpolicy.com/posts/2013/09/16/what_do_syria_and_north_korea_have _in_common.

Federation of American Scientists, June 7, 1994. “Director General’s Statement to the Board of Governors on the Democratic Peoples’ Republic of Korea,” Available online: www.fas.org/news/un/dprk/dgbg1994n01.html.

Ferguson, Bill. Memories of JSA, Sept ’75 – Oct ’76. Date unknown. Available online: http://bill_ferguson.tripod.com/JSA_Memories.htm.

Finley, James P. The U.S. Military Experience in Korea, 1971-1982: In the Vanguard of ROK- U.S. Relations. San Francisco, CA: Command Historian’s Office, Secretary Joint Staff HQ, USFK/EUSA, 1983.

Gerald R. Ford Presidential Library. Information Memorandum, May 28, 1976. “North Koreans Charge ROK Provocations and Threaten Possible Retaliation,” National Security Adviser, Presidential Country Files for East Asia and the Pacific, box 10, Korea (16). Ann Arbor, MI.

______. Intelligence Memorandum, April 22, 1976. “Possible North Korean Initiatives During 1976,” National Security Adviser, Presidential Country Files for East Asia and the Pacific, box 10, Korea (15). Anne Arbor, MI.

______. National Security Adviser, NSC Staff for East Asia and Pacific Affairs, Convenience files, WSAG Meeting, Korean Incident, box 27. Anne Arbor, MI.

385

______. “The Korean Situation and the China Element,” December 3, 1973. NSC East Asian and Pacific Affairs Staff, box 36, Korean Northwest Islands (working file). Ann Arbor, MI.

Foster, Peter. “Richard Nixon Planned Nuclear Strike on North Korea,” The Telegraph (July 8, 2010). Available online: http://www.telegraph.co.uk/news/worldnews/asia/northkorea/7878422/Richard-Nixon- planned-nuclear-strike-on-North-Korea.html.

Funabashi, Yoichi. The Peninsula Question: A Chronicle of the Second Korean Nuclear Crisis. Washington, DC: Brookings Institution Press, 2007.

Gheorghe, Eliza. The Blue House Raid and the USS Pueblo Incident: NKIDP e-Dossier #5. Washington, DC: North Korea International Documentation Project, 2012.

Gills, Barry K. Korea versus Korea: A Case of Contested Legitimacy. London: Routledge, 1996.

Head, Richard G., Short, Frisco W., and McFarlane, Robert. Crisis Resolution: Presidential Decision Making in the Mayaguez and Korean Confrontations. Boulder, CO: Westview Press, 1978.

Hong, Seuk-ryul. “Wigi sog-ui chongjon hyopchong- Pueblo sakeon-gwa P’anmunjôm tokki salhae sakeon,” (“In Crisis: The Pueblo and Panmunjom Axe Murder Incidents”) Yôksa pip’yông 63, no. 3 (2003): 64-66.

Jonsson, Gabriel. Peace-Keeping in the Korean Peninsula: The Role of Commissions. Seoul: Korea Institute for National Unification, 2009.

Jung, Chang-hyun. Kyeot-aeseo Bon Kim Jeong-Il (“Seeing Kim Jong-Il from beside him”). Seoul: Kimyeongsa, 2000.

Kang, Choi and Joon-Sung Park. “South Korea: Fears of Abandonment and Entrapment,” in The Long Shadow: Nuclear Weapons and Security in 21st Century Asia, ed. Muthia Alagappa. Palo Alto: Stanford University Press, 2008.

Kang, David C., and Victor D. Cha. Nuclear North Korea: A Debate on Engagement Strategies. New York: Columbia University Press, 2003.

Kim, Byung-gook and Vogel, Ezra, eds. The Park Chung Hee Era: The Transformation of South Korea. Harvard University Press, 2011.

386

Kim, Chang-hui. “Pukhan, 80 Nyun-dae Jungban Seodok-aeseo Haekmuljil Goo-eep” (North Korea, Purchased Nuclear Material from West Germany in the Mid-1980s), Sindong-a (1995): 134.

Kim, Il Sung. “Report to the Conference of the Workers’ Party of Korea, October 5, 1966,” in Kim Il Sung Works Vol. 20. Pyongyang: Foreign Language Publishing House, 1984.

______. “Congratulations to Officers and Men of Unit 447 of the Korean People’s Army on Shooting Down the Heavy Reconnaissance Aircraft of the US Imperialist Aggressive Forces which Illegally Intruded into the Northern Half of the Republic for Reconnaissance Purposes, April 16, 1969,” Kim Il Sung Works 23. Pyongyang: Foreign Languages Publishing House, 1985.

Kim, Yong-ho. “North Korea’s Use of Terror and Coercive Diplomacy: Looking for Their Circumstantial Variants,” Korean Journal of Defense Analysis 14, no. 1 (2002): 45-67.

Kirkbride, Wayne A. DMZ: A Story of the Panmunjom Axe Murder. Seoul: Hollym, 1984;

Koh, Brian C. “North Korea and the Sino-Soviet Schism,” The Western Political Quarterly 22, no. 4 (1969): 940-962.

______. The Foreign Policy of North Korea. New York: Praeger, 1969.

______. “North Korea in 1987: Launching a New Seven-Year Plan,” Asian Survey 28, no. 1 (1988): 62-70.

Lankov, Andrei. From Stalin to Kim Il Sung: The Formation of North Korea 1945 to 1960. New Brunswick, NJ: Rutgers University Press, 2002.

Lee, Mun-hang. JSA – Panmunjom (1953-1994) (in Korean). Seoul: Doseo Choolpan Seohwa, 2001.

Lee, Suck-ho. “Party-Military Relations in North Korea: A Comparative Analysis,” unpublished doctoral dissertation. Washington, D.C: George Washington University, 1983.

Lerner, Mitchell, and Jong-Dae Shin. “New Romanian Evidence on the Blue House Raid and the USS Pueblo Incident,” NKIDP E-Dossier #5. Washington, DC: Woodrow Wilson International Center, March 2012. Available online: http://www.wilsoncenter.org/sites/default/files/NKIDP_eDossier_5_The_Blue_House_R aid_and_the_Pueblo_Incident.pdf.

387

Lerner, Mitchell B. “A Failure of Perception: Lyndon Johnson, North Korean Ideology, and the Pueblo Incident,” Diplomatic History. 25, no. 4 (2001): 663-5.

Lim, Dong-won. Peacemaker: Twenty Years of Inter-Korean Relations and the North Korean Nuclear Issue. Stanford, CA: Stanford Asia-Pacific Research Center, 2012.

Lyndon Baines Johnson Presidential Library. “Aircraft Carriers in the Sea of Japan,” March 9, 1968. File “Pueblo Crisis 1968, Vol. 15,” NSF, NSC Historical Archives, Pueblo 1968, box 34. Austin, TX.

______. ADMIN CINCPAC, January 23, 1968. “Pueblo incident,” folder “Korea –Pueblo Incident, Vol. 1, Pt. A,” NSF Country Files, Asia and Pacific, box 257. Austin, TX.

______. ADMIN CINCPAC, January 23, 1968. “Pueblo incident,” folder “Korea –Pueblo Incident, Vol. 1, Pt. A,” NSF Country Files, Asia and Pacific, box 257. Austin, TX.

______. CINCPACFLT January 24, 1968. “USS Pueblo Incident,” Folder “Korea-Pueblo Incident-Military Cables, Vol. 1,” National Security Files Country Files, Korea, boxes 263, 264. Austin, TX.

______. National Military Command Center, January 23, 1968. “Current Information Concerning USS Pueblo Incident,” folder “Korea-Pueblo Incident, Vol. 1, Pt. A,” NSF Country Files, Asia and Pacific, box 257. Austin, TX.

______. Responses to Senator William Fulbright, 1968. folder “Pueblo Crisis, 1968, Vol. 10, Fulbright Letter,” NSC Historical Archives, Pueblo Crisis, box 31. Austin, TX.

______. Office of National Estimates of the CIA, “Communist Reaction to Certain US Actions,” January 26, 1968. Pueblo Crisis, 1968, Vol. III, Day-by-Day Documents, Part 3, National Security File NSC History. Austin, TX.

______. Telephone Conversation between President Johnson and the Representative to the United Nations (Goldberg), January 28, 1968. Recordings and Transcripts, Recording of Telephone Conversation between President Johnson and Goldberg, Tape F68.01, PNO 2. Austin, TX.

Mansourov, Alexandre Y. “The Origins, Evolution, and Current Politics of the North Korean Nuclear Program,” Nonproliferation Review 2, no. 3 (1995): 25-6.

388

Mazarr, Michael, J. North Korea and the Bomb: A Case Study in Nonproliferation. New York, NY: St. Martin’s Press, 1995.

Michishita, Narushige. North Korea’s Military-Diplomatic Campaigns, 1966-2008. New York: Routledge, 2011.

Millot, Mark Dean. “Facing the Emerging Reality of Regional Nuclear Adversaries,” The Washington Quarterly 17, no. 3 (1994).

Mobley, Richard A. Flash Point North Korea: The Pueblo and EC-121 Crises. Annapolis, MD: Naval Institute Press, 2003.

Myers, Brian R. The Cleanest Race: How North Koreans See Themselves and Why It Matters. Brooklyn, N.Y.: Melville House Publishing, 2011.

National Archives and Records Administration. “Draft Message OPLAN,” Wheeler records, file “Korea (Pueblo) February 21, 1968 Vol. 3,” box 29, RG 218. College Park, MD.

______. “North Korean Downing of a U.S. Reconnaissance Plane: The EC-121 Incident, April 1969,” October 1969. State Department Historical Studies Division, Office of Executive Secretariat, Folder History Office Research Projects 1969-1974, RG 59, box 6. College Park, MD.

______. “Telegram from the Embassy in Korea to the Department of State,” January 24, 1968. Central Files 1967-69, folder POL 33-6 KOR N-US, RG 59. College Park, MD.

______. Memorandum from Henry Kissinger to Richard Nixon, April 21, 1969. “Pueblo,” National Security Council Files, Pueblo Folder, box 379, Richard M. Nixon Collection. College Park, MD.

______. Memorandum from INR/REA Fred Greene to Claus Ruser, August 1, 1968. “State of Intelligence on North Korea,” Folder Korea 2 of 2, Senior Interdepartmental Group Files 1966-72, box 8, RG 59. College Park, MD.

______. Memorandum from Russell Murray to Secretary of Defense, “Record of Meeting on January 25, 1968,” folder “Pueblo Incident 1968,” McNamara Papers, RG 200. College Park, MD.

______. Memorandum from the Director of the Korean Task Force (Brown) to the Under Secretary of State (Katzenbach), March 15, 1968. “Possible Actions in the Event the

389

North Koreans Announce that They Intend to Try the Pueblo Crew,” Central Files 1967- 69, RG 59, folder POL 33-6 KOR N-US. College Park, MD.

______. Memorandum, “Communist Reactions to Certain U.S. Actions,” April 17, 1969. Nixon Presidential Materials, NSC Files, Country Files Far East, Korea, Vol. 1 to 9-69, box 540. College Park, MD.

______. Memorandum, “Helicopter Crew Members Held by North Koreans,” October 4, 1969. NSC Files, Country Files Far East, Korea, Helicopter Incident, box 540. College Park, MD.

______. Memorandum, “Movement of CVA Task Forces,” April 16, 1969. Wheeler Records, Folder 091 Korea (BP) EC-121 Shootdown,” RG 218, box 31. College Park, MD.

______. Memorandum, “Notes on Conversation between President Johnson and President Park,” December 21, 1967. Folder POL KOR S-US, Central Files 1967-69, RG 59. College Park, MD.

______. Memorandum, “Record of a Telephone Conversation Between President Nixon and the President’s Assistant for National Security Affairs (Kissinger),” April 15, 1969 (6:30pm). Nixon Presidential materials, NSC Files, Korea: EC-121 Shootdown, North Korean Reconnaissance Shootdown, 4/9/69-4/16/69, Vol. 1 Haig, box 434. College Park, MD.

______. Memorandum, “Shoot-Down Incident,” April 16, 1969. Nixon Presidential materials, NSC Files, Korea: EC-121 Shootdown, North Korean Reconnaissance Shootdown, 4/9/69-4/16/69, Vol. 1 Haig, box 434. College Park, MD.

______. Memorandum, Korea Task Force, “Fulbright Letter on USS Pueblo,” March 20, 1968. Folder POL 33-6 KOR N-US 3/15/68, Central Foreign Policy files, box 2269, RG 59. College Park, MD.

______. National Security Council History, Pueblo Crisis 1968, Vol. VII, Day-by-Day Documents, Part 13. LBJ; Memorandum from C. Berger to N. Katzenbach, February 2, 1968. Central Files 1967-69, folder POL 33-6 KOR N-US, RG 59. College Park, MD.

______. NSC Memorandum for General Scowcroft from Thomas J. Barnes, September 29, 1975. “Secretary Schlesinger’s Discussion in Seoul.” FOIA request, released on June 13, 1995. College Park, MD.

390

______. Telegram from the Department of State to the Embassy in the Soviet Union, January 25, 1968. Central Files 1967-69, folder POL 33-6 KOR N-US, RG 59. College Park, MD.

______. Telegram from the Department of State to the Embassy in the Soviet Union, February 6, 1968. Central Files 1967-69, POL 33-6 KOR N-US, RG 59. College Park, MD.

______. Untitled Memorandum from CINCPAC to CJCS, April 17, 1969. Wheeler Records, Folder 091 Korea (BP) EC-121 Shootdown, RG 218, box 31. College Park, MD.

______. U.S. Department of State, Bureau of Intelligence and Research, March 4, 1969. Intelligence Note, “USSR/China: Soviet and Chinese Forces Clash on the Ussuri River,” Folder POL 32-1 Chicom-USSR, Central Files 1967-69, RG 59. College Park, MD.

______. September 18, 1969. Intelligence Note, “Communist China: War Fears and Domestic Politics,” Folder POL Chicom, Files SN 67-69. College Park, MD.

National Security Archive. “Alternative Courses of Action in Response to Korean Attack on U.S. Aircraft,” April 16, 1969. FOIA request, released on April 15, 2002. Washington, DC.

______.“Korea Contingency Options,” September 22, 1969. FOIA request, released on March 29, 2006. Washington, DC.

______. “Review of U.S. Contingency Plans for Washington Special Action Group,” June 25, 1969. FOIA request, released on October 5, 2006. Washington, DC.

______. “Talking Paper for the Chairman, Joint Chiefs of Staff, on an Item to be Discussed at the National Security Council Meeting, 16 April 1969,” April 15, 1969. FOIA request, released on January 9, 2007. Washington, DC.

______. Untitled Memorandum from the Secretary of Defense (Laird) to the President’s Assistant for National Security Affairs (Kissinger), July 15, 1969. FOIA request, released on January 8, 2007. Washington, DC.

Oberdorfer, Don. The Two Koreas: A Contemporary History. Reading, MA: Addison-Wesley, 1997.

Office of the Korea Chair. Record of North Korea’s Major Conventional Provocations since the 1960s. Washington, DC: Center for Strategic and International Studies, 2010. 391

Ostermann, Christian F., and James Person. Crisis and Confrontation on the Korean Peninsula, 1968-69: A Critical Oral History. Washington, DC: Woodrow Wilson International Center, 2011.

Park, Hee-do. Dora Ojiahneun Dari-ae Seoda (standing at the Bridge of No Return). Seoul: Saemteo, 1988.

Perry, William J. “Standing at the Brink in North Korea: The Counterproliferation Imperative,” in Preventive Defense: A New Security for America, eds. Ashton B. Carter and William J. Perry. Washington, DC: Brookings Institution Press, 1999.

Person, James F, ed. Limits of the “Lips and Teeth” Alliance: New Evidence on Sino-DPRK Relations, 1955-1984. NKIDP Document Reader. Washington, DC: Woodrow Wilson International Center for Scholars, 2009.

______. “‘We Need Help from Outside’: The North Korean Opposition Movement of 1956,” CWIHP Working Paper No. 52. Washington, DC: Woodrow Wilson International Center for Scholars, 2006.

______. After Détente: The Korean Peninsula, 1973-1976, Critical Oral History Conference Briefing Book. Washington, DC: CWIHP, 2011.

Pollack, Jonathan, D. No Exit: North Korea, Nuclear Weapons and International Security. London: International Institute for Strategic Studies, 2011.

Pritchard, Charles L. Failed Diplomacy: The Tragic Story of How North Korea Got the Bomb. Washington, DC: Brookings Institution Press, 2007.

Public Papers of the Presidents. “The President’s News Conference: North Korea’s Attack on U.S. Reconnaissance Flights,” April 18, 1969. Washington, DC: The American Presidency Project. Available online: http://www.presidency.ucsb.edu/ws/index.php?pid=2004.

Radchenko, Sergey S. “The Soviet Union and the North Korean Seizure of the USS Pueblo: Evidence from Russian Archives,” CWIHP Working Paper #47. Washington, DC: Woodrow Wilson International Center for Scholars.

Russel, Daniel. “U.S.-North Korean Relations,” in Current Issues in Korean-U.S. Relations: Korean American Dialogue, ed. Lee Man-woo. Seoul: Institute for Far Eastern Studies, 1993.

392

Schaefer, Bernd. North Korean Adventurism and China’s Long Shadow, 1966-1972, Working Paper #44. Washington, DC: CWIHP, 2004.

______. Overconfidence Shattered: North Korean Unification Policy, 1971-1975, NKIDP Working Paper #2. Washington, DC: Woodrow Wilson International Center for Scholars, 2010.

Shen, Zhihua and Danhui Li. After Leaning to One Side: China and Its Allies in the Cold War. Palo Alto: Stanford University Press, 2011.

Shin, Gi-wook. Ethnic Nationalism in Korea: Geneology, Politics, and Legacy. Palo Alto: Stanford University Press, 2006.

Sigal, Leon V. Disarming Strangers: Nuclear Diplomacy with North Korea. Princeton, NJ: Princeton University Press, 1998.

Simmons, Robert R. The Pueblo, EC-121, and Mayaguez Incidents: Some Continuities and Changes. Occasional Papers/Reprints Series in Contemporary Asian Studies 8, no. 20 (1978).

Stefans, John J. “Confessions of a Red Agent,” Army Digest 23 (May 1968): 19.

Suh, Dae Sook. Kim Il Sung: The North Korean Leader. New York: Columbia University Press (1988): 247-8.

U.S. Department of Defense. “EC-121 Fact Sheet,” unclassified report, May 1969. Available online: http://www.dod.mil/pubs/foi/International_security_affairs/other/500.pdf.

U.S. Department of State. Foreign Relations of the United States, 1964-68, Vol. XXIX, Part 1, Korea. Washington, DC: U.S. Government Printing Office, 2000.

______. Foreign Relations of the United States, 1969-1976, Vol. XIX, Part 1, Korea, 1969- 1972. Washington, DC: U.S. Government Printing Office, 2010.

______. Foreign Relations of the United States, 1969-1976, Vol. E-12, Part 1, Documents on East and Southeast Asia, 1973-1976. Washington, DC: U.S. Government Printing Office, 2011.

393

U.S. House Armed Services Committee, Special Subcommittee on the USS Pueblo. Inquiry in to the USS Pueblo and EC-121 Plan Incidents. 91-12. Washington, DC: U.S. Government Printing Office, 1969.

Wit, Joel S. “Clinton and North Korea: Past, Present, and Future,” in Solving the North Korean Nuclear Puzzle, David Albright and Kevin O’Neill, eds. Washington, DC: Institute for Science and International Security, 2000.

______. “North Korea: Leader of the Pack,” Washington Quarterly 24, no. 1 (2001): 1.

Wit, Joel S., Daniel B. Poneman, and Robert L. Gallucci. Going Critical: The First North Korean Nuclear Crisis. Washington, DC: Brookings Institution, 2004.

Woodrow Wilson International Center for Scholars. “The Rise and Fall of Détente on the Korean Peninsula, 1970-1974: A Critical Oral History,” July 2010. Washington, DC.

______. “348th Meeting of the United Nations Command Military Armistice Commission ,” February 28, 1974. History and Public Policy Program Digital Archive, United Nations Command, Military Armistice Commission. Available online: http://digitalarchive.wilsoncenter.org/document/114030.

______. “Information about the Incident with the Ship Pueblo,” January 28, 1968. Archive of the Czech Foreign Ministry, Obtained and translated for the NKIDP by Adolf Kotlik. Washington, DC.

______. “Letter from Government of North Korea” May 13, 1974. History and Public Policy Program Digital Archive, Gerald R. Ford Presidential Library, Ford Vice Presidential Papers, Office of Assistant for Defense and International Affairs, Files 1973- 1974, John O. Marsh, Box 61, North Korea. Obtained for NKIDP by Charles Kraus. Available online: http://digitalarchive.wilsoncenter.org/document/114199.

______. “Minutes of Conversation on the Occasion of the Party and Government Delegation on behalf of the Romanian Socialist Republic to the Democratic People’s Republic of Korea,” June 10, 1971. History and Public Policy Program Digital Archive, Archives of the Central Committee of the Romanian Communist Party. Obtained and translated for the NKIDP by Eliza Gheorghe.

______. “Report on a Stay of a GDR Military Delegation in the DPRK in October 1976,” October 1976. German Federal Archive, Military Archive (BA-MA), AZN 8283. Translated for NKIDP by Bernd Schaefer. Available online, http://ditigalarchive.wilsoncenter.org/document/114292.

394

______. “Report on the ‘Axe Murder Incident’ from the GDR Embassy in North Korea,” August 31, 1976. Political Archive of the Federal Foreign Office, Berlin (PolA AA), MfAA. Translated for NKIDP by Bernd Schaefer. Available online, http://digitalarchive.wilsoncenter.org/document/114291.

______. “TELEGRAM 066.569 from the Romanian Embassy in Pyongyang to the Romanian Ministry of Foreign Affairs,” February 28, 1978. AMAE, Folder 784/1978, Issue 220: Features of political-diplomatic relations between the Democratic People's Republic of Korea and some countries in Europe, Asia, Africa, America (Cyprus, Spain, USA, Bangladesh, Philippines, India, Indonesia, Japan, Pakistan, Sri Lanka, Central African Republic, Egypt, Gabon, Iraq, Iran, Libya, Nigeria, Mozambique, Syria) January 7, 1978 – September 23, 1978. Obtained and translated for NKIDP by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/116417.

______. “Telegram 066678 from the Romanian Embassy in Pyongyang to the Romanian Ministry of Foreign Affairs, US Helicopter Incident on July 14,” July 19, 1977. Archive of the Romanian Ministry of Foreign Affairs, Folder 933/1977, Issue 220/H: Partial US troop withdrawal from South Korea-Discussions regarding the reunification of the two countries, January-October 1977. Obtained and translated for NKIDP by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114874.

______. “Telegram from Beijing to Bucharest, SECRET, Urgent, No. 066.252,” August 25, 1976. Archives of the Romanian Ministry of Foreign Affairs, Matter 200-Relations with the Democratic People’s Republic of Korea, 1976. Obtained by Izador Urian and translated by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114117.

______. “Telegram from New York to Bucharest, SECRET, Urgent, No. 060.387,” August 20, 1976. Archives of the Romanian Ministry of Foreign Affairs, Matter 220-Relations with the Democratic People’s Republic of Korea, 1976. Obtained by Izador Urian and translated by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114115.

______. “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 061.041” February 7, 1973. History and Public Policy Program Digital Archive, Archives of the Romanian Ministry of Foreign Affairs, Matter 220 - Relations with the Democratic People’s Republic of Korea, 1973. Obtained by Izador Urian and translated for NKIDP by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114580.

395

______. “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.043,” February 28, 1976. History and Public Policy Program Digital Archive, Archives of the Romanian Ministry of Foreign Affairs, Matter 220 - Relations with the Democratic People's Republic of Korea, 1976. Obtained by Izador Urian and translated for NKIDP by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114107.

______. “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.212, August 21, 1976. Archives of the Romanian Ministry of Foreign Affairs, Matter 220-Relations with the Democratic People’s Republic of Korea, 1976. Obtained by Izador Urian and translated by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114116.

______. “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.219,” August 26, 1976. Archives of the Romanian Ministry of Foreign Affairs, Matter 220-Relations with the Democratic People’s Republic of Korea, 1976. Obtained by Izador Urian and translated by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114119.

______. “Telegram from Pyongyang to Bucharest, SECRET, Urgent, No. 067.220,” August 27, 1976. Archives of the Romanian Ministry of Foreign Affairs, Matter 220/Year/1976/Country: Democratic People’s Republic of Korea, Telegrams from Pyongyang to the Romanian Ministry of Foreign Affairs, January-December 1976. Obtained and translated by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114120.

______. “Telegram from the First Directorate to Washington, DC, No.01/04493” May 5, 1973. History and Public Policy Program Digital Archive, Romanian Ministry of Foreign Affairs Archives, Matter 210, 1973, Democratic People’s Republic of Korea, Secret, MFA, Folder no. 1495. Obtained and translated for NKIDP by Eliza Gheorghe. Available online, http://digitalarchive.wilsoncenter.org/document/114056.

______. Declassified Memorandum from Ambassador of the GDR in the DPRK to State Secretary and First Deputy Minister of Foreign Affairs, December 8, 1967. Obtained and translated for the NKIDP by Karen Richert. Washington, DC.

______. Minutes of Conversation between Nicolae Ceausescu and the economic delegation from the Democratic People’s Republic of Korea, Bucharest, September 22, 1972. NKIDP Document Reader, “New Evidence on Inter-Korean Relations, 1971-1972,” Document #28. Obtained for NKIDP by Mircea Munteanu and translated by Eliza Gheorghe.

396

______. Speech by Leonid Brezhnev at the CC CPSU Plenum, April 9, 1968. Russian State Archive of Recent History, obtained and translated for the Cold War International History Project by Sergey Radchenko. Washington, DC.

397