ABSTRACT Title of dissertation: WHY DO WE DO TRACK TWO?: TRANSNATIONAL SECURITY POLICY NETWORKS AND U.S. NUCLEAR NONPROLIFERATION POLICY Alexander Thomas Jacobson Lennon, Doctor of Philosophy, 2006 Dissertation directed by: Professor John Steinbruner School of Public Policy As globalization has accelerated, unofficial transnational (a.k.a. track-two) dialogues have proliferated. Do these networks matter? This study examines both their effects in the United States and, by focusing on nuclear nonproliferation, their potential to improve cooperative security as well as conflict resolution. Reviews of relevant theory, secondary literature, and primary materials produced by three case studies?the Council on Security Cooperation in the Asia Pacific (CSCAP), the Northeast Asian Cooperation Dialogue (NEACD), and the Program on New Approaches to Russian Security (PONARS)?supplemented and guided 67 original interviews to help answer the question: Have transnational security policy networks changed U.S. nuclear nonproliferation policies or the perceptions that shape them? These networks have improved intelligence and private as well as public diplomacy, enhancing the analytical capacity and soft power of their participants and interlocutors. They have strengthened otherwise weak ties across countries, areas of expertise, generations, and professions, particularly from inside government to nongovernmental experts, to provide blunter feedback and improve open-source intelligence analysis. These improvements are three-dimensional?delving deeper into overseas foreign policy elite, integrating across wider issues and regions, over longer periods of time?to help understand the implications of political changes, summits, and crises. Diplomatically, they have provided fora for nongovernmental experts and government officials in their private capacity to better understand and convey interests behind official talking points. Although U.S. policymakers will realistically rarely participate, they benefit from one-page or personal briefings by the most effective networks?those that have diverse members, integrate current or former government officials, and focus on ideas and information exchange. Although pressures exist to prove networks changed near-term policy decisions, the diversity that improves intelligence also impedes consensus on policy recommendations, which can be more effectively made by issue-specific cells derived from the network base. Ultimately, these networks empower their members and interlocutors with ideas and information, which enhances their soft power and builds their capacity to diagnose and agree on the root causes of contemporary threats, understand the political pressures shaping national responses to them, evaluate the merits of potential strategies to respond, and explore prospects for cooperative solutions. WHY DO WE DO TRACK TWO?: TRANSNATIONAL SECURITY POLICY NETWORKS AND U.S. NUCLEAR NONPROLIFERATION POLICY by Alexander Thomas Jacobson Lennon Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2006 Advisory Committee: Professor John D. Steinbruner, Chair Professor I. M. Destler Professor Janne E. Nolan Professor George H. Quester Professor Robert H. Sprinkle ?Copyright by Alexander Thomas Jacobson Lennon 2006 ii DEDICATION To three generations of inspiration and support: To my father, a conservative who somehow taught me to change the rules if I didn?t like the game To my wife, Sarah, for never letting me know just how difficult this enterprise was making her life and selflessly encouraging me anyway To my son, Tyler, and his sister-to-be, for constantly reminding me how weekends should really be spent iii ACKNOWLEDGEMENTS My personal thanks are contained on the previous page. Here, I simply briefly wanted to acknowledge my professional admiration and thanks for three sets of people that helped make this project possible and even more enjoyable. First, all the members of my committee have been immensely helpful to organize and clarify my thoughts and presentation of this work. You hear a lot of horror stories about some committees; I have none to add, and only gratitude for each and every one of them and their contributions to make this project better. Second, thanks to the Center for Strategic and International Studies, especially Kurt Campbell, John Hamre, Robin Niblett, and Erik Peterson, for allowing me to pursue this degree part-time, enabling me to keep some perspective on the world outside the dissertation. Third, to the people willing to be interviewed for this study, and especially to Ralph Cossa, Susan Shirk, and Celeste Wallander, who are not only the pioneers that founded these networks but both allowed and even facilitated my research. They may have had nothing to gain by having their life?s professional work subjected to the scrutiny that follows, but I certainly have benefited from their generosity and learned from their experiences. iv TABLE OF CONTENTS List of Tables vii Chapter One: Introduction 1 Key Terms of Inquiry 2 Transnational Security Policy Networks 3 Nuclear Nonproliferation 9 Policies 10 Perceptions 11 Methodology 13 Framework for Analysis 16 Sources and Methods 21 Case Studies 26 Potential Significance 31 Chapter Two: Think Tanks in U.S. Foreign Policy 34 Roles and Functions: What Do Think Tanks Do? 37 The Nature of Governance 38 What Think Tanks Do 41 What Makes the United States Different? 57 How Think Tanks May Wield Influence 65 Models of Influence 66 Channels of Influence 71 When Think Tanks May Have Influence 76 Think Tanks as Models for Transnational Networks 81 Chapter Three: The Council on Security Cooperation in the Asia Pacific 83 Issues CSCAP Has Addressed 89 Proliferation Issues 93 North Korea?s Nuclear Program 97 Nuclear Energy Transparency (PACATOM) 100 Export Controls 105 Future Efforts 107 Results 107 Contacts and Perceptions 107 U.S. Government Access and Interaction 128 Agenda Setting in the United States 135 Policy Options 138 v Observations 141 Drawbacks? 142 U.S. Membership 146 What Happens in a Crisis? 147 Limits of Multilateralism 148 Socializing China and North Korea 150 CSCAP?s Influence in Other Countries 155 Conclusions 158 Chapter Four: The Northeast Asian Cooperation Dialogue 161 Issues NEACD Has Addressed 165 Results 170 Contacts and Perceptions 170 U.S. Government Access and Interaction 182 Agenda Setting 188 Policy Options 189 Observations 191 Drawbacks? 191 Network Membership 193 Socializing China and North Korea 199 NEACD?s Effects in Other Countries 204 Sustainability 204 Conclusions 207 Chapter Five: The Program on New Approaches to Russian Security 209 Issues PONARS Has Addressed 214 Proliferation Issues 217 Loose Nukes 217 Missile Defense 219 Russian Nonproliferation Policy 221 Results 222 Contacts and Perceptions 222 U.S. Government Access and Interaction 230 Agenda Setting 242 Policy Options 244 Observations 249 Drawbacks? 249 Government Participation and Network Membership 250 Maximizing Policy Influence 253 Effects in a Crisis? 259 Socialization, With a Twist 260 Influence in Russia? 263 Conclusions 263 vi Chapter Six: Conclusions 267 Findings 268 Contacts and Perceptions 268 U.S. Government Access and Interaction 273 Agenda Setting 280 Policy Options 288 Recommendations 292 Conclusions 297 Bibliography 300 List of Interview Subjects 310 vii LIST OF TABLES 1.1 Potential Indicators of Influence 20 1.2 Summary of Interviews 25 1.3 Case Studies at a Glance 30 3.1 CSCAP Interview candidates 87 3.2 CSCAP interviews conducted 88 3.3 USCSCAP meetings 114 4.1 NEACD interview candidates 163 4.2 NEACD interviews conducted 164 4.3 Principles of Cooperation in Northeast Asia 166 5.1 PONARS interview candidates 212 5.2 PONARS interviews conducted 213 5.3 PONARS Memo Topics (1997-2005) 215 5.4 Repeated Government RSVPs to PONARS 234 1 CHAPTER ONE: INTRODUCTION Every year, security experts from around the world gather at conferences for days on end to share opinions, ideas, information, stories, and recommendations about threats to global stability and ways to reduce them. Some meet by the hundreds in large, cavernous hotel ballrooms; others gather more intimately in groups of 20-30 around a conference table. Some get together only two or three times and disband; others keep coming back, year after year. Participants are invited from a variety of countries with a range of experiences: life-long government employees, academics, think tank analysts, or consultants might attend. Some have worked at the same place for years; others change jobs or even the countries where they live frequently. Participants might renew acquaintances with colleagues with whom they have worked for most of their professional lives, or they might meet others for the first time. Some participants will not see or talk to each other again until they reconvene; others will be linked by network organizers in cyberspace, or may even make an effort to stay in touch on their own? electronically, over the phone, or face-to-face. Although similar networks have existed at least since scientists from around the world were invited to attend the first Pugwash Conference almost fifty years ago, in 1957, 1 their number has increased dramatically in the last fifteen since the dissolution of the Soviet Union. On Asian political and security matters, for example, one scholar 1 Maureen R. Berman and Joseph E. Johnson, eds. Unofficial Diplomats (New York: Columbia University Press, 1977): esp. 37-41. 2 counted three or four such groups in 1989, and almost 30 just five years later. 2 To try to catalogue all these networks is effectively impossible since they take so many shapes, forming and dissipating without notice or central organization. They are part of the modern, decentralized, agile infrastructure of globalization. A number of reasons explain why they are growing. The post-Soviet acceleration of globalization has facilitated access to a previously shrouded part of the world, while simultaneously making it easier and cheaper to travel everywhere internationally, communicate by e-mail, and distribute working papers or conference reports. The concomitant growth of think tanks in other countries has also increased potential nongovernmental dialogue partners. In today?s more complex, multipolar world, these networks may also be more strategically valuable to help experts sort through myriad challenges. 3 Whatever the reasons for their growth, skeptics counter that these gatherings are simply talk shops, at best: gab fests that produce nothing useful. At worst, they may confuse foreign policy by organizing dissension or sending conflicting diplomatic signals to other countries from individuals not authorized to speak for governments. Yet, advocates point out, participants in and out of government keep coming back to some of these meetings, a few for a decade or more, and they keep getting funding to continue meeting and talking. Why? What do these networks do? Do they matter? Key Terms of Inquiry This study focuses on a piece of this burgeoning universe. It asks: Have transnational 2 Paul M. Evans, ?The Dialogue Process on Asia Pacific Security Issues: Inventory and Analysis,? in Evans, ed. Studying Asia Pacific Security (Toronto: University of Toronto-York University Press, 1994): 299. 3 Louise Diamond and John McDonald, Multi-Track Diplomacy: A Systems Approach to Peace (West Hartford, CT: Kumarian Press, 1996): 23. 3 security policy networks changed U.S. nuclear nonproliferation policies or the perceptions that shape them? What those terms mean and why they have been chosen is the subject of this first section of the introductory chapter, with the methodology and potential significance of the findings following. Transnational Security Policy Networks Starting by having to define what the subject even means can be an inauspicious beginning to any research project, no matter how specialized. Here, at least, it reflects how new the subject area is. No precise definition for transnational policy networks exists. Broadly, transnational relations were defined in 1995 by University of Konstanz Professor Thomas Risse-Kappen as ?regular interactions across national boundaries when at least one actor is a non-state agent or does not operate on behalf of a national government or an intergovernmental organization.? 4 The starting point then is to look beyond traditional national policymaking or diplomatic channels. For those with more spatial minds, Harvard political science professor Robert Putnam equipped analysts with a model of a ?two-level game? in foreign policy, in which each government must negotiate simultaneously with other governments at one (horizontal) level and its own domestic interests at a second (vertical) level. Picture two parallel, horizontal circles: the upper circle simply depicts traditional diplomacy, with each point around its rim representing a country?s government officials (U.S., China, Russia, etc.) who regularly and formally interact. Within each country, government officials must work, or at least contend, with experts out of government (non-state actors) 4 Thomas Risse-Kappen, Bringing Transnational Relations Back In: Non-State Actors, Domestic Structures and International Institutions (Cambridge, UK: Cambridge University Press, 1995): 3. 4 who seek to influence their government?s policy. In this picture, each country?s nongovernmental experts are represented by a point on the second, or lower, level directly underneath its own government. Each country, then, is represented by a vertical pillar around a cylinder connecting the upper circle, representing government policymakers and diplomats, with the lower one, nongovernmental experts. Putnam speculated that on this lower level ?transnational alignments may emerge, tacit or explicit, in which domestic interests [in different countries] pressure their respective governments to adopt mutually supportive policies.? 5 Transnational policy networks are an explicit form of these alignments. To return to the diagram, they connect the dots around the rim of the lower circle, or nongovernmental experts from different countries. The organizers of transnational networks can be represented as points at the center of the lower (nongovernmental) level. The image of the lower tier then resembles a wheel with spokes extending from network organizers at the hub to global nongovernmental experts all around its perimeter. To extend this model for one more paragraph (and mercifully end it for the less geometrically inclined), networks operating within each country are represented by a two-dimensional slice, or right triangle, of this cylinder. The three points of the triangle are network organizers at the hub of the lower wheel, a country?s nongovernmental members on its perimeter (the point at the right angle of the triangle), and its governmental officials directly above it on the upper level. The triangle depicts two paths by which networks may seek to influence governments. The direct path, or shortest distance between networks and governments, is the hypotenuse of the right triangle 5 Robert D. Putnam, ?Diplomacy and Domestic Politics: the Logic of Two-Level Games,? in Peter B. Evans, Harold K. Jacobson, and Robert D. Putnam, editors, Double-Edged Diplomacy: International Bargaining and Domestic Politics (Berkeley, CA: University of California Press, 1993): 445. 5 representing a variety of ways network organizers may reach out to governments (including publications, briefings for officials, or invitations to attend network sessions). The other two sides of the triangle represent an indirect path from network organizers through domestic, nongovernmental experts who participate in transnational networks (represented by the base of the triangle) and seek to influence their country?s officials (represented by the vertical leg of the triangle). A closer look at this vertical leg in the United States, or how U.S. nongovernmental experts seek to influence Washington, is the subject of chapter two. This study is limited to a network?s effects within the United States because of my own limitations and interests. I have neither the language skills nor sufficient knowledge of any other government?s foreign policymaking process to investigate potential effects in other countries. Therefore, the study focuses only on the influence of these networks on U.S. government officials directly, on U.S. nongovernmental members directly, or on the U.S. government indirectly through U.S. nongovernmental participants. In theory, security policy networks face a particularly uphill battle. Traditionally, security policymaking processes are more closed than those for other issues because of the premium on security clearances, inherently giving some government officials the perception that outsiders are not ?in the know,? potentially limiting their influence. 6 The U.S. security policymaking apparatus may be even more difficult to penetrate than other countries because of its relative military power globally. Previous work, drawing on John Ikenberry who is now at Princeton, argues that U.S. networks may be more resistant because of the balance of power: the more powerful a country is, the less they may be 6 See, for example, Alexandra G. Carter, Politics and Physics: Epistemic Communities and the Origins of United States Nuclear Non-proliferation Policy, PhD dissertation, Georgetown University, April 2000, pp. 37, 313-314. 6 influenced by transnational networks. 7 Nevertheless, Cornell Professor Matthew Evangelista looked for empirical evidence of the influence of transnational relationships in security issues in his 1999 book, Unarmed Forces: the Transnational Movement to End the Cold War. In it, he concluded that scientists and physicians in the United States and Soviet Union who primarily populated groups like the Pugwash conference pursued three strategies to seek to reduce the risk of superpower conflict: by providing ideas and information to their like-minded colleagues on the other side of the world; by coordinating policy initiatives to influence their governments; and by appealing to international norms, such as peace and disarmament, that might resonate in their own country. 8 These three strategies? information exchange, policy coordination, and norm construction?are potentially the forerunners of contemporary transnational networks. Today, these networks seek to influence government policymaking processes that have arguably become more receptive to nongovernmental actors over time. Although Evangelista ironically found that Gorbachev empowered transnational movements to change the authoritarian Soviet system more effectively than actors in a decentralized democracy could, 9 the emergence of 24/7 cable news channels and real-time coverage of Congressional hearings on C-SPAN provide prima facie evidence of the eroding barriers between those inside government and out. As early as the 1970s, political scientist Hugh 7 Peter M. Haas, ?Introduction: epistemic communities and international policy coordination,? International Organization 46:1 (Winter 1992): 6-7; and G. John Ikenberry, ?A Word Economy Restored: Expert Consensus and the Anglo-American Postwar Settlement,? International Organization 46:1 (Winter 1992): 289-322. 8 Matthew Evangelista, Unarmed Forces: The Transnational Movement to End the Cold War (Ithaca, NY: Cornell University Press, 1999): especially 381-382. 9 Evangelista, pp. 341-373, 378. Sarah E. Mendelson, Changing Course: Ideas, Politics, and the Soviet Withdrawal from Afghanistan (Princeton, NJ: Princeton University Press, 1998): 127, similarly concludes that certain nongovernmental actors with access to political leaders that controlled resources influenced the Soviet withdrawal from Afghanistan. 7 Heclo was arguing that the traditional U.S. policymaking model of iron triangles dominated by the executive branch, Congress, and vested interest groups overlooked other members of what he called ?issue networks? that were also influencing policy. Subcultures of policy experts outside the iron triangles, some of whom had previously worked in government and were now based in universities and think tanks, were defining public policy issues, shaping the way they were debated, and developing policy proposals, making policymaking more fluid and decentralized. 10 Heclo himself distinguished these larger issue networks, whose members simply exchanged knowledge about a particular subject, from a subset which he called ?shared- action groups? that formed coalitions to do something about it. Other scholars in the 1990s picked up on this distinction, labeling the latter, more directed groups either policy communities or policy networks. 11 This study does not necessarily define policy networks in that way, primarily because the line between issue and policy networks is not as easy to distinguish in practice. Here, the adjective ?policy? is meant to distinguish those networks whose members seek to study and influence policy, even if they do not agree on the definition of the security problem or how to solve it, from transnational advocacy networks, which has simultaneously emerged as an area of study in the 1990s. 12 Advocacy networks are active for normative reasons in debates over issues such as human rights or the environment. Their members are not necessarily recognized professional experts, but can be extremely 10 Hugh Heclo, ?Issue Networks and the Executive Establishment,? ch. 3 in Anthony King, ed. The New American Political System (Washington: AEI for Public Policy Research, 1 st ed., 1978): 87-124. 11 Diane Stone, Capturing the Political Imagination: Think Tanks and the Policy Process (London: Frank Cass, 1996): 90, and Tanja A. Borzel, ?Organizing Babylon?on the Different Conceptions of Policy Networks,? Public Administration 76 (Summer 1998): 253-273. 12 Margaret Keck and Kathryn Sikkink, Activists Beyond Borders (Ithaca, NY: Cornell University Press, 1998), particularly chapter 1, ?Introduction,? pp. 1-38. Also see Diane Stone, ?Introduction: Global Knowledge and Advocacy Networks,? Global Networks 2:1 (2002): 4. 8 effective, primarily by leveraging public attention to those principles through the media or Congressional hearings. One prominent example is the International Campaign to Ban Landmines, for which its founder Jody Williams shared a 1997 Nobel Peace Prize. 13 Members of transnational security policy networks studied here are both more heterogeneous than advocacy networks and seek greater collaboration with actors inside government across countries. They operate at the intersection of an increasing supply of transnational experts and a fluid U.S. policymaking system that demands, or at least accommodates, nongovernmental participation. One set of practitioners at the World Bank and the UN calls this the universe of global public policy networks, of which transnational are a subset of global, and security policy are a subset of public policy, networks. 14 What theoretically enables these networks to be useful to government bureaucracies is their superior flexibility and subsequent capacity to acquire policy-relevant information and even apply it to policy problems. Those networks with the most diverse membership are the ones notionally most likely to be effective because they can take advantage of what economic sociologist Marc Ganovetter in 1973 dubbed ?the strength of weak ties,? 15 more recently popularized 13 See Kenneth Robin Rutherford, NGOs and the International Ban on Anti-Personnel Landmines, PhD dissertation, Georgetown University, 2000. 14 Jan Martin Witte, Wolfgang H. Reinicke, and Thorsten Benner, ?Beyond Multilateralism: Global Public Policy Networks,? International Politics and Society (Politik und Gesellschaft Online) 2/2000 (www.fes.de/IPG/ipg2_2000/artwitte.html (accessed April 16, 2004)); and Wolfgang H. Reinicke, ?The Other World Wide Web: Global Public Policy Networks,? Foreign Policy (Winter 1999-2000): 44-57. Also see Diane Stone, ?Think Tanks Across Nations: The New Networks of Knowledge,? NIRA Review (Winter 2000): 34-39; and Stone, ?Introduction: Global Knowledge and Advocacy Networks,? Global Networks 2:1 (2002): 1-11. 15 Mark S. Granovetter, ?The Strength of Weak Ties,? The American Journal of Sociology 78:6 (May 1973): 1360-80. For a related discussion on the distinction between homogeneous and heterogeneous networks, see Tanja A. Borzel, ?Organizing Babylon?on the Different Conceptions of Policy Networks,? Public Administration 76 (Summer 1998): esp. 257-258 9 in Malcolm Gladwell?s The Tipping Point and Frans Johansson?s The Medici Effect. 16 Diversity arguably both sparks greater learning among members and distributes ideas more widely, even if members do not agree. The key is not necessarily a common agenda like banning landmines, but a common interest and an opportunity to exchange information through transnational security policy networks. Although their structure widely varies, criteria for membership in security policy networks generally limit participation to experts. They are predominantly nongovernmental participants employed in think tanks or universities, although government employees may also be invited to attend network events either in their private capacity, meaning they can express their personal opinions and not be held accountable for views of the governments that employ them, or as guest speakers in their official capacity. The core of each network is a regular, usually annual, conference where various contemporary security issues are discussed. Network members may also interact through smaller working groups, an exclusive website or electronic listserv, memos or briefs distributed through the network, or individual efforts to contact other network members. Nuclear Nonproliferation Rather than look at a colossal array of security threats?traditional causes of war as well as so-called soft security issues such as migration and environmental catastrophe?this study is focused on nuclear nonproliferation for a number of reasons. Proliferation threats take a relatively long time to become acute, allowing official as well as informal dialogue 16 Malcolm Gladwell, The Tipping Point: How Little Things Can Make a Difference (New York: Little, Brown, and Company, 2000) and Frans Johansson, The Medici Effect: Breakthrough Insights at the Intersection of Ideas, Concepts & Cultures (Boston: Harvard University Business School Press, 2004). 10 to develop. 17 Nonproliferation is also a common interest of countries, facilitating multilateral efforts at cooperation, 18 such as the six-party talks on North Korea. A series of international treaties, institutions, and regimes have also been developed to formally buttress these efforts and provide a normative framework for addressing them. 19 Finally, the technology to make nuclear weapons has become so diffuse that cooperative strategies have become at least the first option by default. 20 Even unilateral military strikes to preempt nuclear proliferation threats would be preceded by official and unofficial diplomatic efforts. Nuclear proliferation is not the only threat that fits these criteria, but it is a particularly salient one that does. Especially if the U.S. security policymaking system is resistant to nongovernmental influences, choosing a security issue that networks are most likely to address maximizes the chances of results that show influence, enabling a better understanding of how these networks operate if and when they are useful. Any findings of network influence must be qualified by these enabling factors, however, which may not apply to all security issues. Policies Finding that these networks directly changed policy would be the most direct way to identify influence. But direct causal links from nongovernmental actors to government 17 David Cortright, ?Incentives Strategies for Preventing Conflict,? ch. 11 in The Price of Peace: Incentives and International Conflict Prevention (West Hartford, CT: Kumarian Press, 1997): 277. 18 Cortright, p. 285 19 Thomas Risse-Kappen, Bringing Transnational Relations Back In: Non-State Actors, Domestic Structures and International Institutions (Cambridge, UK: Cambridge University Press, 1995): 6-7 20 Ashton B. Carter, William J. Perry, and John D. Steinbruner, A New Concept of Cooperative Security (Washington D.C.: Brookings Institution Press), 1992: 31; and Robert O. Keohane and Joseph S. Nye, Power and Interdependence: World Politics in Transition (Boston: Little, Brown, and Company, 1977): 233. 11 policy decisions can be elusive. Ideas or recommendations from outside government are usually intangible, while inside government, decisions take time, diffusing influence through an opaque process particularly in national security?all problem s that recur in chapter two. 21 Thankfully, theory provides other potential benchmarks to evaluate. In his classic policymaking study, University of Michigan professor John Kingdon widens the policy lens into four procedural aspects: setting the agenda, developing options to address the problem, the policy decision itself, and implementing it. He focuses on agenda setting and option development as preliminary points in the policy process where what he calls policy entrepreneurs, which include a variety of nongovernmental actors, might more realistically have influence, although they still risk being intangible, diffuse, and opaque. 22 Kingdon concluded that ?policy windows? may open at certain times, such as either during political change or a crisis, enabling some policy entrepreneurs a greater opportunity to get an issue or a policy option onto the agenda and possibly have it chosen at those times. 23 Perceptions Theoretically, network effects are unfortunately even more likely to be found in the amorphous world of perceptions. Evangelista, Heclo, and Reinicke et al. all discuss information sharing and issue definition among the benefits of transnational security 21 David M. Ricci, The Transformation of American Politics: The New Washington and the Rise of Think Tanks (New Haven: Yale University Press, 1993): 24; Diane Stone, Capturing the Political Imagination: Think Tanks and the Policy Process (London: Frank Cass, 1996): 3-4, 112, 219; and Donald E. Abelson, Do Think Tanks Matter? Assessing the Impact of Public Policy Institutes (Ithaca, NY: McGill-Queens University Press, 2002): 5, 87, 164. 22 John W. Kingdon, Agendas, Alternatives, and Public Policies (New York: Harper Collins Publishers, 1995), esp. 16-18, 196-201. 23 Kingdon, pp. 165-195, 203-4 12 policy networks. Three other authors reinforce the need to better understand the ethereal benefits that the dialogue process, or even just contact facilitated by networks, may provide and give ideas where to look. Similar to Putnam?s two-tier model, Joseph Montville coined the term ?track-two diplomacy? to describe efforts to help address the core psychology of conflicts between Arabs and Israelis, Indians and Pakistanis, Croats and Serbs, and others where stereotypes exacerbate strife. 24 Montville explains that unofficial, problem-solving workshops among nongovernmental leaders can help break down negative stereotypes. These workshops are examples of track-two diplomacy. Others have called dialogues that include government employees, even if in their private capacity, ?track one and a half,? denoting their close collaboration with ?track one,? or traditional diplomacy. 25 Whoever attends, Montville postulates that workshops can help participants develop workable personal relationships, understand the perspective of other members, and ?at some point? develop joint strategies for cooperation. While transnational security networks include but are not limited to bringing adversaries like North Koreans and Americans together, they could provide value to its members if a) new and useful contacts are made, b) understanding is improved, or c) joint strategies are developed. Roger Fisher and William Ury of the Harvard Negotiation Project are two other analysts focused on addressing the psychology behind conflicts, at least diplomatic ones. In Getting to Yes, Fisher and Ury advise negotiators to focus on the interests behind stated government positions, not the statements themselves, to reach negotiated 24 Joseph V. Montville, ?The Arrow and the Olive Branch: A Case for Track Two Diplomacy,? in Vamike D. Volkan, Joseph V. Montville, and Demetrios A. Julius, eds. The Psychodynamics of International Relationships: Unofficial Diplomacy at Work (Lexington, MA: Lexington Books, 1991): 161-175. 25 David Capie and Paul Evans, The Asia-Pacific Security Lexicon (Singapore: Institute of Southeast Asian Studies, 2002): 209-220. 13 agreements. 26 Because transnational dialogues theoretically enable participants to interact as individuals, not as officials, they may help participants discover the interests behind the talking points used in official negotiations, helping improve understanding and facilitate agreements. The informality of dialogues, then, can theoretically aid diplomacy and change perceptions. These networks are also examples of what Peter Haas calls epistemic communities or ?a network of professionals with recognized expertise and competence in a particular domain and an authoritative claim to policy-relevant knowledge within that domain or issue-area.? 27 Haas highlights four reasons why governments may seek help from epistemic communities: nongovernmental experts may provide perspective to help explain why crises emerged (what might be called their root causes); they may predict whether certain policies will help alleviate the crises; they may help the state define its interests in a crisis; or they may provide intellectual support, or political cover, to justify a policy decision. 28 Methodology These models, descriptions, or observations of how transnational security policy networks operate (Putnam, Evangelista, Reinicke et al., Ganovetter) in a more fluid U.S. policymaking system (Heclo and a literature survey in chapter two of how think tanks operate in U.S. foreign policy) with an expanded view of the policy process (Kingdon) and directions to look for perceptions potentially changed (Montville, Fisher and Ury, 26 Roger Fisher and William Ury, Getting to Yes: Negotiating Agreement without Giving In (New York: Penguin Books, 1981): particularly pp. 41-57. 27 Haas, p. 3 28 Haas, pp. 14-16 14 Haas) can be combined into a framework for analysis, or a set of benchmarks to guide the study. Such a framework is crucial for a subject that is fundamentally subjective. My own interest in these networks stems from personal experience. Any dissertation or book- length project is a labor of love, or at least passion to find the answers to the questions asked. My own passion derives from the unsurpassed benefits that I believe I have professionally gained from traveling and learning from colleagues on week-long professional programs to Germany, Japan, Spain, and China as well as my 18-month service at the U.S. State Department as a foreign affairs officer. These experiences are not examples of transnational security policy networks for a variety of reasons, but they have inspired me to look for as many opportunities as possible to meet with colleagues from the countries that I study, particularly because I have never lived abroad, while employed at the Center for Strategic and International Studies (CSIS), where I have worked for most of the last fifteen years. The chance to share opinions, ideas, and better understand how counterparts from China, Europe, Japan, India, Russia, the Middle East, Australia, or anywhere else see the world is indispensable to being a foreign affairs analyst or scholar-practitioner. That passion led me to join CSCAP, one of the case studies which will be discussed later, and informally begin this study. The question that remained in my mind was whether my experience was unique. What did others get from these networks, if anything? Did government officials benefit from them, or are benefits limited to experts outside government straining to get in? To be fully transparent, CSCAP and one of the other case studies (PONARS) have loosely or partially been affiliated with CSIS, 15 although not the division at which I work. The director of PONARS was also a professor of mine as an undergraduate fifteen years ago. There have been benefits from my experience at CSCAP and this affiliation, including a more intimate understanding of those networks, access to its members, and potentially a certain candor to interviews. It also meant that I had a constant challenge throughout this study to balance the advantages that familiarity brought with the need to remain objective, analytical, and even skeptical of various claims. One way I did that was to constantly remind myself that this study would not evaluate the fundamental goals these networks set for themselves. It focused on nonproliferation (while networks also addressed other issues) and tested their impact in the United States alone. Therefore, finding no relevance for nonproliferation or in the United States should not lead me or anyone else to conclude that these networks have no value, mitigating any potential bias in their favor in this study. If these networks have no impact in the U.S. government, it is also in my selfish interest to know that now, before wasting any subsequent time seeking that effect as a scholar-practitioner based in a U.S. think tank or university. Finally, lessons learned about any limits of these networks or ways to improve them are not intended to be criticisms of the network organizers, but could be passed on to them, offering to help them improve network influence rather than blithely overlooking any deficiencies. Nevertheless, ignoring the potential bias in favor of these networks would be na?ve, particularly because any study of perceptions is inherently subjective. The rest of this section is divided into three parts to describe the methodology?or the set of benchmarks, sources and methods, and case studies?employed to help limit the potential bias, guide the inquiry, and seek to answer whether transnational security policy networks 16 have changed U.S. nuclear nonproliferation policies or the perceptions that shape them. Put another way, the rest of this chapter answers: what is this study looking for to evaluate the impact of these networks, how and to whom it will turn for answers, and where (or to which case studies) it will look. Framework for Analysis Because the contemporary use of these networks is unexplored territory, no single theory is available to guide the inquiry. Instead, the models, observations, and descriptions previously discussed in this chapter can be combined and condensed to form a hybrid analytical framework to focus on four potential areas of influence, or benchmarks. First are the contacts and perceptions facilitated by networks, the most intangible but potentially promising aspect of study. Ganovetter?s emphasis on weak ties makes the search for contacts important, while those contacts can be a source for the new perceptions, ideas, information, and opinions which Evangelista, Heclo, Reinecke et al., Montville, Fisher and Ury, and Haas all emphasize. 29 What, if any, contacts and perceptions did networks facilitate? Did networks help U.S. participants think of new ideas, reexamine assumptions, see different perspectives, or change their understanding of the interests behind certain foreign government positions in ways that only looking at policies might not make immediately obvious? Second are the ways that networks sought to interact with government participants. How did networks seek to influence U.S. policymakers? Putnam?s two-level game, Evangelista?s conclusion that Gorbachev?s Soviet system was more permeable 29 Also see Diane Stone, ?Think Tanks Across Nations: The New Networks of Knowledge,? NIRA Review (Winter 2000): 36. 17 than the United States to transnational networks, Haas? hypotheses that epistemic communities might help policymakers, and Kingdon?s concept of policy windows opening at certain times all raise questions about the timing and pathways of a network?s access to government. Did networks publish recommendations, or summaries or transcripts of annual meetings? Did U.S. government officials receive them? Did they read them? Did they participate in network activities? At what level? Did they receive briefings about network activities? Did they request them? Did they ask networks to consider certain issues? Did government officials previously know network participants? Did network members become government officials through what is known as the revolving door, in and out of U.S. government? Because some network members are based in U.S. think tanks and because the ways that U.S. think tanks seek to influence U.S. policy could provide a potential model for transnational networks, the literature on the timing and ways that think tanks seek to influence U.S. foreign policymaking will be surveyed in chapter 2. At this point, it is simply worth highlighting that an incentive may exist for nongovernmental actors to hide public attention from any contribution they may have made to policies recommended by government officials. That contribution is not necessarily unseemly, but nongovernmental actors may prefer to allow policymakers to take credit for any ideas, recommendations, or policies that they generate, seeking to preserve their long-term influence as a consultative staff. There is no incentive for government officials to give that recognition. 30 If such influence is not publicized, consistent access to policymakers may provide an alternate indicator of their value to those policymakers and potential influence. 30 See, for example, Kenneth Lieberthal, ?Initiatives to Bridge the Gap,? Asia Policy 1 (January 2006): 14. 18 The third potential benchmark is any network effect on the U.S. policy agenda, what John Kingdon calls agenda setting or Hugh Heclo calls issue shaping. How a problem is defined or how public debate is framed can theoretically have an effect on whether any action is taken, or even what policies are subsequently favored. 31 In Harvard University Professor Joseph Nye?s terms, agenda setting is a form of soft power, or the ability to shape the preferences of others through noncoercive means, and persuade them to seek the same goals. 32 What proliferation issues did the networks discuss, such as North Korea, Iraq, Iran, Pakistan, the 1998 Indian nuclear tests, inadequately secured nuclear materials, the A.Q. Khan network, or terrorist threats? Do government officials acknowledge networks in private or in public, such as briefings or press releases, 33 for drawing their attention to aspects of those issues? Do issues uniquely or principally discussed by networks subsequently appear on the political agenda in a way that networks might credibly claim some credit? Fourth are potential policy options that networks may develop. Evangelista?s policy coordination, Montville?s joint strategies, Kingdon?s policy options, and Haas?s policies to alleviate crises all provide theoretical support for this line of inquiry. What nonproliferation policy options did networks discuss, such as export controls, economic 31 For more on agenda setting, see Andrew Rich, Think Tanks, Public Policy, and the Politics of Expertise (Cambridge, England: Cambridge University Press, 2004): esp. 111-112; David A. Rochefort and Roger W. Cobb, ?Problem Definition: An Emerging Perspective,? ch. 1 in Rochefort and Cobb, eds. The Politics of Problem Definition: Shaping the Policy Agenda (Lawrence, KS: University Press of Kansas, 1994): esp. 9, 15-26; and Joseph G. Peschek, Policy Planning Organizations: Elite Agendas and America?s Rightward Turn (Philadelphia: Temple University Press, 1987): 7. 32 Joseph S. Nye, Jr. Soft Power: The Means to Success in World Politics (New York: Public Affairs, 2004): esp. x, 5, 31, 91. 33 Margaret E. Keck and Kathryn Sikkink, in Activists Beyond Borders: Advocacy Networks in International Politics (Ithaca, NY: Cornell University Press, 1998): 24, argue that such statements can often be used to force a government to address a particular issue, once it has acknowledged that a problem exists. 19 sanctions, military operations, missile defense, arms control, security assurances, trade incentives, conflict resolution, or crisis management? Did networks reach a consensus supporting any options? Do government officials acknowledge networks in public or private for drawing their attention to particular options? Does evidence exist that those policies were considered by government? Were they chosen? If so, can networks credibly claim some credit for it? Alternatively, did network efforts generate tension, antagonize, complicate, or otherwise interfere with government activities? Table 1.1 summarizes these four benchmarks, some of the potential indicators of network influence for each, and some hypothetical conditions under which results might be more likely. 20 Table 1.1 Potential Indicators of Influence Benchmark Indicators of Influence Could be more likely when? Contacts and Perceptions - new contacts formed - contacts maintained - anecdotal claims from interviews - ties otherwise weak - former foreign govt officials participate - more useful for nongovernmental members? Access to Government - officials claim to read network publications - number of govt attendees at meetings - seniority of govt attendees at meetings - briefings to government officials - network members become govt officials - briefings convened at government initiative - higher level attendance - govt officials know and trust nongovt network member Agenda Setting - claims from interviews - media coverage - discursive statements - Congressional hearings - after elections - policymakers change Policy Options and Selection - claims from interviews - media coverage - discursive statements - Congressional hearings - changes in institutional procedures 34 - recommended policy considered, adopted (or dropped) - after elections - policymakers change - crises - network consensus recommendations 34 Keck and Sikkink, p. 25, argue that network efforts may lead governments to change institutional procedures or bureaucratic structure to address network concerns. 21 Sources and Methods These four benchmarks point to what network influence to look for, but not how to answer those questions. What is the data to be examined, and how can it be acquired? Primary network materials?such as publications, conference summaries or transcripts? of selected case studies are instrumental to understand how networks operate and what they seek to do. But to try to address how governments perceive and interact with them or ascertain anybody?s perceptions, original interviews are necessary. Interviews raise some understandable methodological concerns. The data acquired are inherently unverifiable unless a reader asks the author for interview transcripts. Even then, interviews do not have the objective power of proper statistical models, when they are available, because they are the product of conversations conducted by an individual. Different interviewers will ask different questions and get different responses. They are inherently subjective, imprecise, and vulnerable to a range of exogenous influences and biases. Network directors are potentially inclined to exaggerate their influence, with government officials potentially inclined to dismiss them. The interviewer?s potential bias, a subject?s memory, and even their mood during the interview may play a part. Nevertheless, there is no better way to get at perceptions, which are by definition subjective. Ultimately, topics lend themselves to be examined in certain ways, and determine the methodology, not the other way around. Although conclusions drawn must be mindful of the vulnerabilities of interview-based studies, not asking these questions will simply perpetuate ignorance about the operations and potential value of part of the emerging foreign policymaking architecture in a globalizing world. At the cutting edge of research, exploratory methods are often necessary. In this case, interviews are an 22 essential, if imperfect, way to get at perceptions and other intangible effects networks may have. Some of a reader?s uncertainties about interviews can be tempered by doing more than just citing ?interview with author? while also protecting the confidentiality and identity of some interview subjects. When only one source makes a particular claim, that is noted in this study in the text, not just a footnote. Interviews are also coded to preserve the identity of the subject and cited (i.e. ?interview A1?), unless their personal background or view is relevant and they did not request anonymity in which case they are cited by name. Coding enables readers to see how many and which (coded) subjects make a particular claim. The text also includes relevant professional descriptions of sources, such as whether they are government employees. An open-ended set of questions was used to loosely guide interviews. This allowed conversations to flow and focus on the most relevant topics for each subject, particularly since many had limited time available. Rather than using a survey, a more flexible format both increased the chances of participation, since policymakers are extremely unlikely to fill out a form, and allowed for greater exploration of a topic that has been practically untouched. The goal is to seek to better understand how networks operate, when they might have influence, and on what aspects of U.S. security policy, so that hypotheses can be reached for subsequent, potentially more rigorous examination. As of now, analysts can?t even be sure where to look. The interview questions were loosely based around some preliminary background queries (how long subjects had been members or participants in networks, whether they 23 had worked in government, etc.) and examined the four benchmarks of potential influence, by asking: Without necessarily mentioning anyone by name, have you made or maintained any professionally valuable contacts through the network? If so, why were they valuable? Was the network trying to influence U.S. policy? How? Was the network useful to you? To public policy? Are network activities harmful for any reason? How could this network improve its utility? Because none of the networks were limited to nonproliferation, in practice the interviews often discussed networks broadly. In cases where individual subjects focused on nonproliferation, they were also asked whether perceptions about or attention to any nonproliferation issues or policy options were derived from network activities. In practice, however, nonproliferation mainly served to limit the review of primary written materials more than interviews, which ranged more widely. Five different groups of people were sought to be interviewed, each of whom bring different biases, perspectives, and experiences to the networks: 24 Directors of the networks, who in all cases also founded them, were consulted. Although they have an obvious bias in favor of the relevance of the networks, they have the largest reservoir of experience and knowledge from which to draw. Funders were sought for their perspective as well as ways they evaluate each network?s success. U.S. nongovernmental members were asked about their experiences and contacts in each network. Permanent nongovernmental members may have a bias in favor of networks, but those who are part of the revolving door rotating in and out of government may have less of a bias. U.S. government members may have a bias against significant policy impact of networks, as it could diminish their own influence, and have insight into how networks may have complicated their work or the work of their colleagues. U.S. senior government policymakers, meaning Deputy Assistant and Assistant Secretaries in the Departments of State and Defense as well as directors in the National Security Council (and higher, if relevant and possible) for relevant regions and issues, were sought whether they attended network events or not. Interviews were conducted with 67 subjects, mostly ranging from 30-60 minutes with a few as long as 90 minutes or as short as 15. Although the cases will be introduced 25 momentarily and discussed in greater detail in subsequent chapters, table 1.2 summarizes the interviews conducted with nongovernmental members, those members of the revolving door that worked both in and out of government, and government officials. The level of government officials interviewed ranged from current or former assistant secretaries through deputy assistant secretaries, policy planning staff members, office directors, desk officers, staff from Capitol Hill, and members of the intelligence community. Table 1.2 Summary of Interviews Subjects interviewed Case 1 (CSCAP) 2 (NEACD) 3 (PONARS) Total Only NGO members 6 4 10 20 Revolving door 11 4 6 21 Only govt employees 5 9 12 26 Total 22 17 28 67 Interviews were transcribed and direct quotes are liberally preserved in the results presented in chapters 3-5 to maintain the voice of the subjects. In all cases, except for some more senior government policymakers, interviews were sought with people who had experience with the network. This may give a bias, even among government participants, in favor of network influence as it seeks to better understand the experience of those that keep coming back. Findings, then, arguably indicate how networks may have maximized their influence in ideal circumstances, when government officials are interested. 26 Finally, to examine more tangible policy impacts, such as agenda setting and policy options, a method called ?process-tracing,? 35 can be used, which pretty much sounds like what it is. Explorations of how networks interact with governments can help to trace any claimed agenda setting or policy option results back to their network foundations. However, even if experiences, perceptions, or recommendations can be traced from policies back to networks, that does not necessarily prove that transnational dialogues caused those effects. Tracing can only reveal a relationship, or a correlation, between networks and tangible results. Successful correlations also need to reexamine the context, for example to see if exogenous crises may explain both network and government activity independently. Even epistemic community theory acknowledges that they may be cited for their influence by politicians or decisionmakers as political cover for choices they had already reached anyway. Only correlations can be determined; causal conclusions are unlikely and unrealistic. In historical research, this is known as the ?critical method,? 36 where scholars attach probabilities to inferences about events. Any correlations, however, can at least help conclude this research with hypotheses that would give some insight and merit further research to evaluate the strength of those correlations. Case Studies The final step of the methodology is to determine where to look, or to what case studies. Although networks are decentralized and not catalogued anywhere, a vast number of 35 Alexander L. George and Andrew Bennett, ?Process-Tracing and Historical Explanation,? ch. 10 in Case Studies and Theory Development in the Social Sciencies (Cambridge, MA: MIT Press, 2005): 205-232; and Mendelson, p. 11. 36 Judith Goldstein and Robert O. Keohane, ?Ideas and Foreign Policy: An Analytical Framework,? in Goldstein and Keohane, eds. Ideas and Foreign Policy: Beliefs, Institutions, and Political Change (Ithaca, NY: Cornell University Press): 26-29. 27 options are available. To limit the study to relevant transnational security policy networks and a reasonable number of differences among them, a number of guidelines were used to select cases. These guidelines are not the only ones that could be chosen but they were selected for various analytical reasons. As mentioned earlier, policy networks among experts with similar interests but not necessarily similar goals that often worked with governments were selected, not advocacy networks that sought to externally and publicly pressure governments to make a particular change. Networks that have operated since the end of the Cold War were selected, to analyze their contemporary role in a globalizing world. Only networks that have existed for at least five years and continued to meet during this research were chosen, providing sufficient data to examine the enduring effect of broad networks, as opposed to numerous discrete subjects that may be discussed in a two-year project. Although this study only tests the effects of these networks in the United States, whom U.S. members interact with is crucial to what members might learn and a network?s potential value. For this study, networks that included Chinese and/or Russian participants were selected for four reasons. First, U.S. relations with these two countries have been dynamic since the end of the Cold War. The political space for nongovernmental actors to help navigate these changes has potentially been greater than if relations with those countries were static. Second, the countries are not historical U.S. allies, also potentially generating more interest in nongovernmental channels than with allies who have well-developed formal government channels with Washington. Third, as global powers, the agenda with China and Russia may include security challenges where those countries have similar interests with the United States, such as preventing North 28 Korean nuclear proliferation, as well as potentially competing interests, such as how to cease Russian nuclear cooperation with Iran or Chinese missile sales to Pakistan. Choosing these countries enables this study to test the value of these dialogues as both cooperative security management and conflict resolution tools. Finally, testing participation by more than one other country reduces the risk that the single partner country studied might hold security policy networks hostage. Case studies primarily comprised of scientists or others with particular technical knowledge not likely to be prevalent among government policymakers were avoided. Choosing cases other than the Pugwash conference or the National Academy of Science?s Committee on International Security and Arms Control helped test the inherent value of transnational efforts, rather than the technical consultative role that scientific networks may provide. Finally, although each of the network directors is well qualified and may have government experience in their own right, networks that were not dominated by a high- profile chair were selected. Other cases?such as former Secretary of Defense Bill Perry and former Assistant Secretary of Defense Ash Carter?s efforts through the Preventive Defense Project at Stanford and Harvard, or the Aspen Strategy Group, currently led by former National Security Advisor Brent Scowcroft and former Assistant Secretary of Defense Joseph Nye?are more likely to test an extension of the directors? influence rather than the network?s inherent value and effect. Nothing would necessarily be wrong with changing any of the above guidelines for case selection. If anything, they emphasize the potential scope of the universe of 29 transnational security policy networks. Although other cases may also fit these guidelines, three were selected for study here. Arguably the largest and most well-known transnational security policy network, the Council on Security Cooperation in the Asia-Pacific (CSCAP), established in June 1993, focuses on security issues in the Asia-Pacific and is comprised of 21 national member committees, including China, Russia, and the United States. An international Steering Committee of its members governs the organization?s activities. Each country has its own member committee, hosted by a national think tank. The U.S. committee of nearly 200 members is administered by the Pacific Forum/CSIS, based in Hawaii. The founder, Ralph Cossa, is currently the USCSCAP executive director. CSCAP attempts to provide direct support to the ASEAN Regional Forum (ARF), a governmental multilateral security organization. CSCAP also hosted several issue-oriented international working groups focusing, among other issues, on confidence and security building measures particularly to reduce weapons of mass destruction proliferation, and a North Pacific Working Group concentrating on Northeast Asian security cooperation especially on the Korean peninsula. The second case study, the Northeast Asian Cooperation and Dialogue (NEACD), was founded and is currently directed by Susan Shirk of the Institute of Global Conflict and Cooperation at the University of California at San Diego. It consists primarily of an annual meeting of five members from each of six countries: China, Japan, South Korea, North Korea, Russia, and the United States. Generally annual meetings are convened with each country bringing one policymaker from each of the foreign and defense ministries, as well as a uniformed military officer and two members from think tanks or 30 universities for a total of up to 30 attendees. NEACD has held seventeen meetings since it was founded in 1993. Occasional special studies are also commissioned. The Program on New Approaches to Russian Security (PONARS) was founded in 1997 by Celeste Wallander at Harvard University. PONARS has since moved with Dr. Wallander to CSIS, where it is currently organized. It consists of approximately 80 members who study Russian and Eurasian security issues and the region?s role in international affairs. Most of those members teach in the United States, Russia, or Ukraine and/or work at think tanks in those countries but are not invited or identified by their country of origin. In addition to its annual meeting, which government officials and other interested observers are invited to attend, scholarly working papers and shorter policy memos are published throughout the course of the year. PONARS members also can communicate through an exclusive e-mail network, or listserv. These networks differ from each other in at least five ways. If results differ among networks, these variables may provide initial hypotheses about how network structure affects their ability to make contacts and change perceptions, effectively interact with government decisionmakers, change the policy issue agenda, or promote policy options. The variables are summarized in table 1.3: Table 1.3 Case Studies at a Glance Network National Participation # Intl(US) Members Primary Members Primary Govt. Outreach Other Activities CSCAP Regional (21 countries) About 1,000 (200) Think Tank ARF (multilateral)/ participation Working groups NEACD Subregional (6) 30 (5) /year 60% govt, 40% NGO Core participation Occasional studies PONARS 85-90% U.S., Russia (2+) 80 (44) Academic Conference observers/memos Electronic listserv 31 A greater number of nations participating may either increase the global influence of the network, or may encumber it with an inflexible bureaucracy which plagues many multilateral exercises. The number of network members or their principal occupation might affect network influence. Frequent and direct interaction with governments may improve network influence, but it may reduce their creative brainstorming, or agenda setting, potential. Operations other than the annual conferences?such as small, focused working groups or electronic networks to facilitate member interaction?may have effects. Potential Significance Why does any of this matter? The proposed study seeks to improve the understanding of the potential influence of these three case studies on U.S. nonproliferation policies and the perceptions that shape them. In preliminary research, a number of experts were very interested in the results, even if they themselves frequently participated in transnational networks. As one expert informally explained: ?we all go to these things and they seem useful, but sometimes I am not really sure why.? I expect that such findings can be useful and original, even if they might be initial and limited. Completion of this study seeks to improve the ability to see through the fog of foreign policymaking by shedding some light on how these networks might change the policies and perceptions of either government officials directly, nongovernmental experts directly, or U.S. government officials indirectly through their interaction with U.S. nongovernmental network members. This study seeks to conclude with informed 32 hypotheses about the value of these transnational security policy networks from these three case studies that can be further tested. Within these potential limitations, the research can conceivably make three other contributions to new knowledge: First, this study seeks to test the potential for these networks to serve not just as conflict resolution tools, which currently dominate the literature, but also to evaluate their potential as cooperative security tools to assist contemporary great powers in managing security challenges. Second, this study may derive a set of best practices for current and future networks. Findings may include insight into the organizational structure that would best serve to change policy or perceptions: the types of members who should be included; the outreach to most effectively include and inform government policymakers; and even possibly the types of issues most likely to be successfully addressed by transnational security policy networks. It is not necessarily the case that what may, or may not, work for nuclear nonproliferation will work for other security challenges or with other countries. Nevertheless, this study could develop recommendations for best practices that may be tested by other networks. Finally, the study may serve as a model for other researchers seeking to improve the understanding of at least two other types of nongovernmental activities. First, how might U.S. think tank employees improve U.S. foreign policy? Existing literature has been dominated by process tracing, often without conducting interviews, and has been frustrated by an inability to draw a line from a think tank?s idea to an adopted policy. Findings in this study of U.S. think tanks? transnational cousins may help develop hypotheses about the influence of think tanks to be tested by others. Second, researchers 33 with language and regional expertise may be able to apply this study?s methods to improve the understanding of transnational network influence in other countries. Although the questions that shape the methodology have already been elaborated, at its core, this study asks three deceptively simple things: How do these networks tackle nonproliferation issues? What value, if any, do they serve for public policy? How should they be structured to maximize their value? Each of the case studies will be discussed separately in chapters 3-5, with chapter 6 drawing conclusions across the three case studies. What remains to complete this framework for analysis is the literature review of the role of think tanks in U.S. foreign policy. Chapter two contains that literature review. 34 CHAPTER TWO: THINK TANKS IN U.S. FOREIGN POLICY Theoretically, how might transnational security policy networks seek to influence U.S. nuclear nonproliferation policies or the perceptions that shape them? What might networks seek to do? How and when are they most likely to have influence? Does existing literature draw conclusions or propose hypotheses that can be tested by further study? Partly because transnational networks are so new, existing literature attempting to answer these questions is limited and was discussed in the previous chapter. Are there other models that might shed light on why, how, and when networks can maximize their influence? As depicted in the introductory chapter, transnational policy networks might seek to influence U.S. policy or perceptions in two ways: either directly through the U.S. government or indirectly by affecting individual U.S. nongovernmental members who subsequently seek to change U.S. policy or perceptions. Although the limited existing literature on transnational networks examines the direct channel, other literature may help explain indirect influence. The key may lie in understanding the organizations that permanently employ many of these nongovernmental network members: think tanks. What these think tank-based members seek to do in their full-time capacity as well as how and when they may be most influential could help enlighten any indirect impact. Beyond individual member activity, what purposes think tanks organizationally serve, as well as how and when they influence U.S. foreign policy, might also provide models for how transnational networks might maximize their own influence. Clearly, there are differences between the two types of organizations. Think tanks are more 35 established institutions, occupying buildings with full- time employees, which have existed for decades and are usually based in the country they are trying to influence. Networks are more malleable, often exist only virtually except for one annual conference, have fewer resources, and potentially face greater opposition from skeptical government officials worried about foreign penetration of national security policy. Yet both organizations are comprised of nongovernmental experts seeking to shape security policies and perceptions; may seek similar goals, channels, and timing of influence; and even have many of the same members. At a minimum, think tanks could provide some useful lessons for networks to consider emulating. The literature on think tanks in U.S. policy, much less U.S. foreign policy, is only slightly larger, still emerging in its own right. Much of it has been plagued with difficulties trying to define what think tanks are. In their broadest sense, think tanks are defined as non-profit, nongovernmental organizations whose primary purpose is public policy research and influence, 37 a definition which sufficiently helps as a guide for transnational networks. Despite the fact that think tanks have proliferated in recent years (an exhaustive study by University of Pennsylvania Ph.D. James G. McGann estimated that 1200-1400 think tanks existed in the United States in the mid-1990s, half of which were established since 1970), 38 no more than eight books have been written on think tanks focusing exclusively or primarily on U.S. policy, with only one explicitly on U.S. foreign policy. Methodological problems about defining them and attempting to prove 37 This is a hybrid of, and consistent with, definitions used by James G. McGann, The Competition for Dollars, Scholars and Influence in the Public Policy Research Industry (Lanham, MD: University Press of America, 1995): 31-32; and Howard J. Wiarda, ?New Actors: Think Tanks and Foreign Policy,? ch. 8 in American Foreign Policy: Actors and Processes (New York: HarperCollins, 1996): 138. 38 McGann, The Competition for Dollars, Scholars and Influence, p. 4 36 they caused a certain policy provide few empirical examples of the impact think tanks may have had and have left them otherwise neglected. 39 Although limited, the existing literature does outline the purposes that think tanks seek to serve in U.S. foreign policy as well as how and when they might theoretically maximize their influence. The goal of this chapter is not to prove whether think tanks have actually influenced policy as the literature describes. Although it is tempting to investigate each speculated purpose, conduct original interviews, and examine empirical evidence, that would be a distinct research project on think tanks. The goal in this chapter is merely to provide a map of the theoretical roles, channels, and timing of think tank influence that can be integrated into the transnational security policy network research framework outlined in chapter one, and help test whether the three transnational policy network case studies have changed U.S. nuclear nonproliferation policies or perceptions. To construct that map from existing literature on think tanks in U.S. policy, this chapter examines: the purposes think tanks serve in U.S. policy; what uniquely about the U.S. system may enable think tanks to have influence; how, or through what channels, think tanks seek to influence U.S. policy; and when think tanks are most likely to be influential. 39 Donald E. Abelson, ?Think Tanks in the United States,? ch. 5 in Diane Stone, Andrew Denham, and Mark Garnett, eds. Think Tanks Across Nations: A Comparative Approach (Manchester: Manchester University Press, 1998): 123. 37 Roles and Functions: What Do Think Tanks Do? Think tanks are a heterogeneous lot. They vary in size, political orientation, policy scope, influence, and purpose. Yet as a group, a variety of potential roles exist, even if no single think tank pursues all of them. They may seek to be strategic planners, looking beyond political cycles, or instant analysts, explaining current events to the public. At times they are government?s obstreperous adversaries, criticizing priorities or policies, and at times they are valuable allies, serving as sounding boards for policymakers. Think tanks can also serve as debating societies, providing places and people to examine events or recommendations in depth; as houses of convalescence, for ex-officials whose political party has lost power; or as interdisciplinary bridges, linking various experts and interest groups with each other and with policymakers. The starting point is to dispense with the notion that think tanks only seek to influence final policy decisions. 40 While in their most vainglorious moments they certainly claim that ability, there are too many factors in policymaking beyond their control. 41 A more accurate view widens the policy lens to include, and focus on, 40 Alexander L. George, Bridging the Gap: Theory and Practice in Foreign Policy (Washington, D.C.: U.S. Institute of Peace Press, 1993): xix, 34, 74; David M. Ricci, The Transformation of American Politics: The New Washington and the Rise of Think Tanks (New Haven: Yale University Press, 1993): 198; James Allen Smith, The Power Brokers: Think Tanks and the Rise of the New Policy Elite (New York: Free Press, 1993): 207, 229-30; Donald E. Abelson, American Think Tanks and their Role in U.S. Foreign Policy (New York: St. Martin?s Press, 1996): 74; Diane Stone, ?Private Authority, Scholarly Legitimacy, and Political Credibility: Think Tanks and Informal Diplomacy,? in Richard A. Higgott, Geoffrey R.D. Underhill and Andreas Bieler, eds. Non-State Actors and Authority in the Global System, (London: Routledge, 2000): 223; and Andrew Rich, Think Tanks, Public Policy, and the Politics of Expertise (Cambridge, England: Cambridge University Press, 2004): 111. 41 George, p. 133; Abelson, American Think Tanks and their Role in U.S. Foreign Policy, p. 1; Diane Stone, Capturing the Political Imagination: Think Tanks and the Policy Process (London: Frank Cass, 1996): 15; Rich, p. 153; and Alexander L. George and Andrew Bennett, Case Studies and Theory Development in the Social Sciences (Cambridge, MA: MIT Press, 2005): 276, 284-5. 38 informing and shaping the earlier stages of the policy process in a wide variety of ways, all before governments have time or the political pressure to address those issues. 42 The Nature of Governance For inherent structural reasons, the reality is that it is difficult for governments to perform many of the roles that think tanks seek to provide. For starters, politics make it difficult to plan over the horizon. Elected officials are governed by 2-6-year terms, limiting their ability to think strategically beyond their next election or pay much attention to long-term effects of policy recommendations. 43 To support and advise these politicians and political appointees that come and go, a cadre of permanent, government employees in the executive branch provides experience and institutional memory, stabilizing even drastic electoral upheavals. 44 Although there are times when this role is beneficial, there are others when it can be a liability. Employees that succeed in government positions are often risk averse, 45 naturally support existing programs into which they may have invested years of their own career, are rewarded for their ability to build consensus before changing policy, 46 and risk developing a ?group think? mentality among colleagues who have shared similar professional experiences or even have worked together directly for much of their 42 Richard Higgott and Diane Stone, ?The Limits of Influence: Foreign Policy Think Tanks in Britain and the USA,? Review of International Studies 20 (1994): 15. 43 Stone, Capturing the Political Imagination, p. 15 44 James G. McGann, Responding to 9/11: Are Think Tanks Thinking Outside the Box? (Philadelphia, PA: Foreign Policy Research Institute, July 31, 2003): 9. 45 Ronald D. Asmus, ?Having an Impact: Think Tanks and the NATO Enlargement Debate,? U.S. Foreign Policy Agenda 7:3 (November 2002): 29-30. 46 Hugh Heclo, ?Issue Networks and the Executive Establishment,? ch. 3 in Anthony King, ed. The New American Political System (Washington: AEI for Public Policy Research, 1 st ed., 1978): 120-121. 39 careers. 47 The advantage of such a system is that it prevents overreacting to what turn out to be transient problems. In times of truly radical change, such as the last fifteen years however, such stability can ossify governments and impede their ability to change with the times. The end of the Cold War, globalization, regional integration, and democratization are just some of the recent geopolitical transformations that have led to calls for necessary and often radical changes, straining the structural stability that governments traditionally provide. 48 This is not the first time that think tanks have called for change during geopolitical transformations. Previously, at the end of World War II, think tanks apparently served a similar role by providing foreign policy advice to prod the United States to adjust its policies, at the time helping Washington navigate its way out of its previously more isolationist shell and into its new global responsibilities. 49 During times of change, when an old consensus has decayed and a new one has yet to form, think tanks can help counter government?s inherent stability. 50 The dawn of the information age has also exacerbated government?s inability to be agile or to look ahead. Over twenty-five years ago, Hugh Heclo was already writing about the end of the knowledge oligopoly?previously held in ?iron triangles? by the executive branch, the legislature, and interest groups?and the increasing flow of information facilitated by emerging issue networks, discussed in chapter one. More information requires more government staff to process it, bloating the bureaucracy even 47 McGann, Responding to 9/11, p. 57 48 R. Kent Weaver and James G. McGann, ?Think Tanks and Civil Societies in a Time of Change,? ch. 1 in McGann and Weaver, eds. Think Tanks & Civil Societies: Catalysts for Ideas and Action (London: Transaction Publishers, 2002): 13; and Asmus, p. 31. 49 Donald E. Abelson and Christine M. Carberry, ?Policy Experts in Presidential Campaigns: A Model of Think Tank Recruitment,? Presidential Studies Quarterly 27:4 (Fall 1997): 681. 50 Stone, Capturing the Political Imagination, p. 92 40 further. 51 The arrival of e-mail and the internet has aggravated information overload as interest group lobbies have more data to wield, investigative media have found more sources, constituents have electronic channels to register complaints more easily, and other actors?including think tanks themselves?produce more reports to create a constant onslaught of bureaucratic fires that have to be put out. 52 Even without these external responsibilities, the increasing volume of information challenges the fundamental ability of governments internally to process it, to decipher trends, and to evaluate policies. 53 These massive information flows have subsequently forced government portfolios at all but the most senior levels to become more specialized, limiting the number of people and the ability of government to integrate information and understand broad trends or predict the effects of policy change. 54 Put it all together, and a limited number of officials with increasing demands on their time in an era of global transformation are left to process a dizzying volume of information in institutions that are structured to resist change and by their nature are confined to the time horizons of political cycles. 51 Ricci, pp. 41, 139-140 52 James G. McGann, Think Tanks: Catalysts for Ideas in Action: An International Survey (Philadelphia: Foreign Policy Research Institute, October 1999): 56-57, and Weaver and McGann. p. 1 53 Jan Martin Witte, Wolfgang H. Reinicke, and Thorsten Benner, ?Beyond Multilateralism: Global Public Policy Networks,? International Politics and Society (Politik und Gesellschaft Online) 2/2000 (www.fes.de/IPG/ipg2_2000/artwitte.html (accessed April 16, 2004)). 54 Maureen R. Berman and Joseph E. Johnson, ?The Growing Role of Unofficial Diplomacy,? ch. 1 in Berman and Johnson, eds. Unofficial Diplomats (New York: Columbia University Press, 1977): 30; Craufurd D. Goodwin, ?The Fifth Estate: Research for Informed Debate in Democratic Society,? ch. 3 in Jeffrey Telgarsky and Makiko Ueno, eds. Think Tanks in a Democratic Society: An Alternative Voice (Washington, D.C.: The Urban Institute, 1996): 24; Heclo, pp. 99, 115; and Ricci, p. 45. 41 What Think Tanks Do To help governments cope with the demands of the information age, think tanks in their simplest role can filter, synthesize, and systematically respond to information in ways that might be useful for policymakers. 55 Unencumbered by the daily demands of governance, think tanks have the time and expertise to provide such analysis. The most effective think tanks at this role are ones perceived to be impartial evaluators of information, unbiased with a particular interest or political axe to grind. 56 Beyond simple information analysis, however, think tanks inherently seek to improve policy by helping governments as well as the public better understand, react to, and shape events. Sometimes this assistance is cooperative, working with governments; at others, it is combative, challenging or criticizing official priorities or policy. Either way, think tanks can perform at least four functions: to employ experts that inject new ideas into the policy process; to debate policy proposals, eventually even helping build public support for those that meet expert approval; to serve as houses of political convalescence, providing time and space for former (and future) government officials to improve policy to be implemented in a future administration; and to convene an interdisciplinary variety of experts to improve or build public support for policies. Generate Ideas In the earliest stages of the policy process, think tanks serve to inject new ideas into it. They can do this through a variety of roles: sometimes proposing grand strategic plans; at times, providing original and unconventional thoughts; and at times 55 Heclo, p. 99, and James G. McGann, Think Tanks: Catalysts for Ideas in Action: An International Survey, pp. 56-57 56 Ricci, pp. 46-47 42 repackaging the ideas or proposals of others in ways that policymakers can better process and understand. One way think tanks put new ideas into the policy process is by thinking ahead, or over-the-horizon. Political officials, as discussed earlier, are often constrained by short- term political pressures for re-election. Think tanks can compensate for those pressures, by serving as early warning systems, 57 anticipating problems beyond political cycles, or as strategic planners, analyzing long-term policy implications. 58 These roles can help transcend the political constraints of re-election by providing the time and perspective to look beyond immediate problems or instant effects of policy proposals. A second way think tanks put new ideas into the policy process is by being provocative, providing unconventional ideas to help balance the inherently consensus- driven nature of governance. 59 Sometimes these novel ideas are specific, detailed policies, designed to respond to problems differently, and are presented to lower-level officials for the machinery of government to consider and incorporate. 60 In their grandest forms, provocative individuals outside government may perceive the world in a different way, attempting to get the most senior officials to alter the policy agenda or even grand strategy itself. A commonly cited example comes from the 1950s when Albert Wohlstetter and a team of RAND researchers were credited with using game theory and systems analysis to highlight American strategic bomber vulnerability, which in turn is attributed to be a foundation for the subsequent evolution of deterrence theory and 57 Strobe Talbott, ?The Brookings Institution: How a Think Tank Works,? U.S. Foreign Policy Agenda 7:3 (November 2002): 21. 58 William Wallace, ?Between Two Worlds: Think-tanks and Foreign Policy,? ch. 8 in Christopher Hill and Pamela Beshoff, eds., Two Worlds of International Relations: Academics, Practitioners and the Trade in Ideas (London: Routledge, 1994): 158. 59 Wolfgang H. Reinicke, ?The Other World Wide Web: Global Public Policy Networks,? Foreign Policy (Winter 1999-2000): 44. 60 Ricci, p. 194 43 nuclear strategy. 61 As former State Department policy planning director Richard Haass wrote, ?[o]riginal insights can alter conceptions of U.S. national interests, influence the ranking of priorities, provide roadmaps for action, mobilize political and bureaucratic coalitions, and shape the design of lasting institutions.? He cited a Council on Foreign Relations project, called the War and Peace Studies project, immediately after World War II as helping explore the foundations for post-war peace including the creation of the United Nations, the occupation of Germany, and the strategy of containment itself. 62 Clearly, there are political limits to how radical new ideas might be. 63 Media attention to unconventional ideas and even financial compensation provide independent incentives for this creative, brainstorming role, 64 but to actually influence policy, there is a certain undefinable political threshold beyond which some ideas are simply too radical for governments to consider. What may appear ingenuous to one person is an insane or outrageous proposal to another. This sometimes means that the most provocative ideas will be stranded outside political circles, left out in an intellectual wilderness. Over time, ideas may gain popularity in think tanks, serving as a sort of intellectual incubator until events change or policymakers become desperate enough to reconsider an idea that may have been prematurely aired before its political time. 65 The most novel ideas will often have to wander in this political purgatory, limited to think tanks and possibly universities, before passing through the gates of political feasibility at some point in the future. 61 Smith, pp. 117-121, and Stone, Capturing the Political Imagination, pp. 191-194 62 Richard N. Haass, ?Think Tanks and U.S. Foreign Policy: A Policy-Maker?s Perspective,? U.S. Foreign Policy Agenda 7:3 (November 2002): 6. 63 Wallace, ?Between Two Worlds,? pp. 159-160; Gregg Easterbrook, ?Ideas Move Nations: How Conservative Think Tanks Have Helped to Transform the Terms of Political Debate,? The Atlantic (January 1986); and Ricci, p. 218. 64 Heclo, pp. 120-121 65 R. Kent Weaver, ?The Changing World of Think Tanks,? PS: Political Science and Politics 22:3 (September 1989): 568. 44 Not all ideas conveyed by think tanks are novel, however. Think tanks may merely market or repackage ideas originally hatched elsewhere, particularly in academia. While certainly a generalization, professors David Ricci and James McGann separately characterized academics as researchers who often seek new knowledge about issues that may or may not be making headlines, and who convey their findings using models or a language intended for other academic scholars and experts that can sometimes be difficult for the general public or politicians to understand. 66 Although the divisions vary, specialization and the requirements of in-depth research lead most original researchers to form narrow groups of experts, sometimes developing a language so technical and complex that communicating with the public becomes nearly impossible. 67 Universities themselves exacerbate the problem by dividing faculty and research into departments, complicating their ability to be truly interdisciplinary. 68 Stanford professor Alexander George argues that the incentive structure of academia also rewards those that differentiate their own ideas from others, rather than rewarding those that communicate to the general public, build consensus, or work well in groups. 69 The more novel and potentially groundbreaking the research, the greater the potential to be confined in elite, academic circles. Think tank officials, in contrast, focus on more contemporary problems and how to communicate them as well as their potential solutions to a broader political and public audience. 70 In such a role, think tanks may repackage original ideas into a form that government can digest. This role is sometimes described as serving as ?research brokers? 66 Ricci, pp. 223-224 and McGann, The Competition for Dollars, Scholars and Influence, p. 27 67 Ricci, p. 223 68 Stone, Capturing the Political Imagination, p. 211 69 George, pp. 15-16 70 Ricci, pp. 223-224 and McGann, The Competition for Dollars, Scholars and Influence, p. 27 45 to translate academic findings, serving as a bridge between the world of ideas that universities generate and the world of action that policymakers and the media inhabit. 71 In practice, this means that think tanks can serve as idea middlemen, taking complex theories and developing strategies to make them comprehensible and enhance their appeal in political and popular circles. In the field of economics, for example, public choice theory in academics has reportedly been translated into policy through privatization studies. 72 Acting as research brokers to translate ideas to policymakers raises a fundamental question: what type of research is useful for policymakers? Professor George argues that policy-relevant knowledge can do two things: diagnose a problem and shape responses to that problem. Of these two, George argues that diagnosing problems is a more important potential contribution since diagnosis is a prerequisite to prescribing policy solutions. Although he does not cite it, one current example might be the North Korean nuclear issue. Why does North Korea seek nuclear weapons? Does Pyongyang seek security it believes a nuclear arsenal can provide, or are they North Korea?s bargaining chips to cash in for better relations with the United States and an economic soft landing out of its anachronistic economic and political system? George draws an analogy to doctors, who must diagnose illnesses correctly before being able to treat them. He also contends that this role is often overlooked by those who study think tanks, concentrating instead on the fame generated by inventing policy ideas that governments eventually adopt. 73 71 McGann, The Competition for Dollars, Scholars and Influence, pp. 13-16; Higgott and Stone, p. 28; Stone, Capturing the Political Imagination, p. 122; Haass, p. 5; and Paul Dickson, Think Tanks (New York: Atheneum Press, 1972): 28. 72 Stone, Capturing the Political Imagination, p. 166 73 George, pp. xx, 17, 132, 139 46 Evaluate and Communicate Policy Generating or conveying new ideas, however, is merely the beginning of the policy process and only one role that think tanks can play. They also debate policy proposals after they are generated. The exact way that think tanks perform this role varies considerably among such a diverse set of organizations. Sometimes think tank experts deliberate in elite circles; at others, they set the parameters of policy debate for the general public and participate in that debate. Some think tanks approach issues with no conscious preconceived political agenda; others seek to promote a particular political philosophy or policy strategy. At times, think tanks criticize government policies, and at others, think tanks can serve as government allies, helping to build public support for the policy ideas that emerge with expert approval from this intellectual gauntlet. As policy deliberators, in other words, think tanks can act as agents of intellectual Darwinism, debating and refining both the analytic content and political feasibility of ideas in a variety of elite or public circles, so that only the fittest ideas survive to become policy. Think tank experts vet policy proposals and provide opinions on policies or crises in a variety of forums. In public, for example, think tanks provide news stations with analysts, such as on the Middle East with Rachel Bronson from the Council on Foreign Relations on MSNBC, Tony Cordesman from CSIS on ABC, and Ken Pollack from Brookings on CNN. In private, if governments wish, think tanks can also provide intellectual sounding boards or second opinions on potential proposals before they are launched for public consumption. In one of the few instances in which the literature provides an empirical example of the role of think tanks, transatlantic expert Ron Asmus 47 relayed that think tanks became part of the government debates over NATO enlargement in the early-mid 1990s: As a result, a number of think tanks became, for a period of time, an informal but nonetheless real part of an extended interagency process and debate within the U.S. government on NATO?s future. Their briefings and memos became an integral part of the intellectual and policy debate. Think tank analysts worked closely with, and were often invited in to brief, senior officials. 74 In addition to their employees, think tanks can help provide a broader set of elite nongovernmental opinion to policymakers by convening a wider array of experts from related fields?such as academia, the media, legislative and executive branches? throughout what Harvard professor Ernest May calls the ?foreign policy public.? 75 Such an extensive base of opinions can help government fine-tune proposals before officially offering them as policy. Participants in these nongovernmental groups and even the think tanks themselves are certainly not all centrist or politically unbiased. Some think tanks can also be advocates, making sure that a particular political perspective, policy strategy, or even a particular region of the world is considered. Conservative think tanks, for example, emerged in the 1970s to relentlessly drive their point of view through the perceived liberal media after the Vietnam War broke the previous foreign policy consensus. 76 These advocacy tanks, such as the Heritage Foundation, ?transformed the terms of public policy debate? in the 1970s, according to at least one observer. 77 More recently, Democrats have countered with their own think tanks?such as the Progressive Policy Institute that Bill 74 Asmus, pp. 30-31 75 Haass, pp. 7-8 76 Wiarda, p. 155 77 Gregg Easterbrook, ?Ideas Move Nations: How Conservative Think Tanks Have Helped to Transform the Terms of Political Debate,? The Atlantic (January 1986). 48 Clinton used during his run to the 1992 presidency or the recently founded Center for American Progress?to seek a similar role: to insure that their political viewpoint is represented in all public policy debate. Differences between these advocacy tanks and other types of think tanks will be discussed in greater detail later in the chapter. For now, it is simply important to recognize that think tanks participate in elite and public debates with diverse objectives. As events or policies are considered in these debates, they may either generate fierce disagreements or produce a general, although rarely completely universal, consensus. The ideas that find consensus are more likely to be influential, according to Yale Ph.D. Andrew Rich?s study of think tank influence in health care, telecommunications, and tax reform from the 1990s to 2001. Specifically, Rich found that although experts agreed that problems with health care existed, they could not agree on how to fix them. As a result, their collective influence declined in the policy debates. 78 It is reasonable to assume that such a conclusion applies to foreign policy cases as well. On some foreign policy ideas, think tanks will collectively generate a din of controversy, signaling to policymakers that an idea may not be ready for policy. In other cases, near- expert consensus should give officials more confidence that policy consensus may be viable. As ideas emerge from these elite debates, think tank experts help define the boundaries of ever broadening circles of public debates, providing intellectual margins for them in a democracy. These analysts may explain global events or government proposals not just to other foreign policy elite or the ?foreign policy public,? but also to the broader informed public, who watch cable news or read newspapers like The New 78 Rich, pp. 107, 121-122, 139-141 49 York Times or The Wall Street Journal, and eventually to the more general public who watch network news or read their major regional or local paper. If objective and unbiased, think tanks can perform a sort of educational civic duty, explaining events and policies to the public. 79 Whether biased or unbiased, think tank experts may not act simply as educators in policy debates, but they may also serve as policy advocates. Although some advocates support policies for preconceived political reasons, others advocate their own conclusions based on the results of objective study. In either case, wittingly or unwittingly, think tank advocates may even test run ideas for government officials, allowing them both to examine elite, public, or even foreign reaction and, if the idea fails, to maintain plausible deniability, avoiding any subsequent political fallout. In the NATO enlargement debate during the mid 1990s, for example, then-Council on Foreign Relations senior fellow Ron Asmus states that think tank officials were asked to ?cross the Atlantic and test-market ideas and policy options with West European allies or Central European partners in order to provide feedback before final decisions in Washington were made.? 80 In this role, think tanks may be envisioned as a sort of analytical secret service, taking any potential diplomatic bullets from abroad or the public for policymakers if proposals are not well received. To this point, roles discussed for think tanks as policy evaluators have described participation in debates prior to governments making policy decisions. Once these policy decisions are made, however, think tank experts can also help explain and communicate those decisions, potentially building public support for or against them. Here again, 79 Berman and Johnson, p. 21, Shulock, pp. 226-7, 240-1, and Haass, p. 8 80 Asmus, pp. 30-31 50 Asmus explained that think tank experts helped defend to the broader public the decision to extend NATO membership to former Warsaw Pact countries after the government debate had been resolved. 81 Building public support can be particularly important as confidence in governments throughout the world wanes. In recent years, evidence has emerged that public confidence has declined, potentially exacerbated by the constant criticism conveyed through 24-hour news media and even, in some corners of the world, by the shocking collapse of once seemingly invincible governments in eastern Europe and the former Soviet Union. Accompanying radical geopolitical change has led the public to be skeptical about the ability of their representatives to cope with such a transformation and adequately protect their citizens or promote their interests during it. In this environment, think tanks can provide a perceived unbiased second opinion, sometimes legitimating government decisions and increasing public confidence and support. 82 Provide Political Convalescence To this point, discussion has not addressed what types of people think tanks employ, other than assuming that they are experts at generating policy ideas, evaluating them, or marketing them for policymaking audiences. In practice, some of these employees are ex-officials who develop foreign policy strategies while their political party is out of power. In other words, think tanks often serve as a stop in a revolving door in and out of government for former (and future) officials. 83 81 Asmus, p. 31 82 James N. Rosenau, ?Relationships Revised: Proliferating Subgroups, Weakened Governments, and Eroded Authority,? ch. 14 in Turbulence in World Politics: A Theory of Change and Continuity (Princeton, NJ: Princeton University Press, 1990): 397; and James G. McGann, ?How Think Tanks are Coping with the Future,? The Futurist 34:6 (November 2000). 83 Donald E. Abelson, ?Think Tanks and U.S. Foreign Policy: An Historical Perspective,? U.S. Foreign Policy Agenda 7:3 (November 2002): 11-12; and Higgott and Stone, p. 33. 51 The revolving door was facilitated, according to historian James Allen Smith, by the National Security Act of 1947 and other reforms in the late 1940s that created the National Security Council, the policy planning staff at the State Department, and the Central Intelligence Agency, all of which created senior advisory posts for nongovernmental experts. 84 As a result, think tank employees were no longer dependent on indirect channels of influence such as the media, published reports, or briefings to influence government policy. They could enter government themselves and directly implement preferred policies. 85 These future political appointees would become empowered to balance inherent bureaucratic interests in preserving particular programs, and shake up the status quo. 86 While employed at think tanks, these ex-officials can participate in studies to improve their knowledge of particular issues, their strategic perspective freed from the political pressure of re-election, and their reputation as issue experts. 87 Recent examples include current Undersecretary of State Paula Dobriansky, the former director of the Washington office of the Council on Foreign Relations; Richard Haass, who went from director of foreign policy studies at Brookings to be the State Department?s director of policy planning in the first term of the current administration to now president of the Council on Foreign Relations; and Clinton Administration Deputy Secretary of State Strobe Talbot, who is now the president of the Brookings Institution. Although out of political power, some ex-officials find they benefit from and even enjoy the time to think about vexing policy issues without the demands of governance. 84 Smith, p. 113 85 McGann, The Competition for Dollars, Scholars and Influence, p. 65 86 Weaver, p. 569 87 Heclo, p. 112, and Ricci, p. 166 52 Far beyond a small band of political aspirants, participating in this revolving door permeates American think tanks. Having former employees in government positions helps the think tank itself, anointing it with prestige and opening potential channels of influence when its former employees move to government. 88 The Heritage Foundation perceived that having its alumni employed in government was so prestigious and important that it created a talent bank for the transition to the Reagan administration in 1980, and expanded its efforts in 1989 to send 2500 resumes of its employees, affiliates, and ideological allies to the Bush transition team. 89 The obvious drawback of this role as houses of convalescence for expert political exiles is that it takes time to see results. Convene Interdisciplinary Networks Nearly all of the previous discussion has highlighted the purposes that experts employed by think tanks seek to serve. At times, though, think tanks can also provide places to improve policy by serving as convention centers for interdisciplinary networks, or simply as a forum for officials to present their ideas. In a corollary to the role that individual employees may play as research brokers, think tanks institutionally serve to bring together experts with common interests or objectives from government, politics, law, the military, diplomacy, international agencies, interest groups, business, the elite media, academia, and other think tanks to facilitate information, idea, and proposal sharing. 90 These interdisciplinary gatherings may make recommendations to refine original research, improving its political viability, or they may simply introduce government officials around the table to new ideas, potentially catalyzing that official to 88 Gregg Easterbrook, ?Ideas Move Nations: How Conservative Think Tanks Have Helped to Transform the Terms of Political Debate,? The Atlantic (January 1986). 89 Smith, p. 207 90 Diane Stone, ?Think Tanks Across Nations: The New Networks of Knowledge,? NIRA Review (Winter 2000): 38; and Wallace, ?Between Two Worlds,? p. 146. 53 carry the idea back to the office with them. 91 One way to think about it is a form of intellectual entropy, giving experts from one sector of the conversation entirely different insights on a problem or policy that enables them to add their own expertise and devise an approach to diagnose or solve the problem differently. At a minimum, such interdisciplinary meetings and networks can take place off-the-record, seeking to minimize the political tension that a topic may generate and facilitate its ability to be discussed among experts. 92 Certain think tanks can also serve a political function, providing a venue for government to announce policy and present ideas. 93 One very high-profile recent example is the National Endowment for Democracy, which hosted President Bush on its twentieth anniversary when he explained that democratization of the Middle East had been a major goal of the war in Iraq. 94 At its most basic level, providing a podium and an audience for a new policy may allow officials from any political party to reach an audience of experts. It may simply provide the space to provoke initial reactions from the interested public, similar to the role that individual think tank employees play. 95 More than that, in some cases think tanks may even provide a legitimating role for governments, helping policymakers reach out beyond partisan divisions if they present their ideas at a think tank that is perceived to be bipartisan or particularly concerned about the issue being addressed. 96 91 Stone, ?Private Authority, Scholarly Legitimacy, and Political Credibility,? p. 211 92 Smith, p. 212 93 Stone, ?Private Authority, Scholarly Legitimacy, and Political Credibility,? p. 219 94 For text, see http://www.whitehouse.gov/news/releases/2003/11/20031106-2.html (accessed March 15, 2006). 95 Haass, p. 8 96 Higgott and Stone, p. 30 54 Altogether, these convening roles enable think tanks to transcend the talent of their individual employees. The organizations themselves can stimulate interdisciplinary connections and research, expose ideas and policies to a diverse but expert audience to be refined for public consumption, and introduce ideas to policymakers off-the-record within the walls of its meeting rooms. Its podiums and conference halls also allow government officials to test ideas or improve their political feasibility by convening these diverse but expert audiences, or even occasionally lend their reputation to government officials when they announce proposals. These potential roles of think tanks have been characterized optimistically, or even idealistically, reflecting the majority of the literature written by scholars or members of the revolving door themselves who move in and out of think tanks. That does not mean that everyone values the roles think tanks can play. In particular, career government employees often believe that think tanks merely produce a cacophony of criticism that increases the already overwhelming demands on governance. Any fair assessment of think tanks can not overlook their potential negative effects, even if the literature does not discuss them in great detail. Potential Drawbacks Think tanks or their employees can sometimes seek to conduct government policy themselves. As early as 1798, the U.S. government began its struggle with nongovernmental experts. At the time, diplomatic relations with France had been severed under President John Adams. Dr. George Logan, a Jeffersonian, went to France with a letter of introduction from Adams?s political rival, and met with French officials who had previously refused President Adams?s mission. Logan returned to Washington to 55 report that France wanted better relations with the United States. Livid with the perceived interference in their right to govern, the Adams administration responded by passing the Logan Act, prohibiting any U.S. citizen without authorization to ?directly or indirectly commence or carry on any correspondence or intercourse with any foreign government or any officer or agent thereof, with intent to influence the measures or conduct of any foreign government or any officer or agent thereof, in relation to any disputes or controversies with the United States.? 97 Over 150 years later, President Truman faced a similar problem when a handful of American experts went as ?diplomatic tourists? to Moscow, and others had entertained the visiting Soviet Deputy Premier in Washington. Truman subsequently wrote an editorial in The New York Times on January 19, 1959, publicly reminding readers that only the President made foreign policy. 98 Although these are extreme examples of the extent to which nongovernmental experts may go if they disagree with government policy, they exist. In fact, they potentially may become more common now, as think tanks seek to serve as conflict mediators in a globalized world. 99 Domestically, the criticisms that think tanks may wield in public can compound the inherent challenges of governance by adding more distractions that governments must deal with in the information age. Government officials often resent the swarms of think tanks that pile controversy on, complicating policymaking. 100 Most media producers can find so-called experts from think tanks, willing to enhance their individual stature by 97 Berman and Johnson, pp. 2-3 98 Cited in Berman and Johnson, p. 1 99 Haass, p. 8 100 Stone, Capturing the Political Imagination, p. 105 56 criticizing government proposals. 101 Often, the media tends to present two extremist views, some of which other experts would not regard seriously, elevating those that are willing to say or write outrageous, provocative conclusions that generate controversy, get their own name in lights, and raise the entertainment value of the news. 102 The general public, however, may not always be able to discriminate between the highly regarded expert and the provocative extremist. 103 Nongovernmental Checks and Balances In the end, one can not generalize about whether think tanks improve or confuse policy. It quite simply depends at least on what role the think tank plays, and how much is already known about the issue. Ideally, think tanks may improve democracy itself by generating public debate and educating the public in a free society. 104 At other times, such debate risks confusing, undermining, or even misrepresenting government proposals, complicating governance and exacerbating the inherent conflict between governments that are structured to provide stability and think tanks that seek change. On the bright side, such a tense relationship can serve to prod traditionally staid bureaucracies to diagnose a problem in a different way, particularly during periods of transformation, or to consider different policies. They can help governments cope with information overload by filtering information and presenting it in a useful manner, assisting officials who must cope with the daily demands of governance. Think tanks can look beyond political cycles to help plan strategically, serve as research brokers to 101 Ricci, p. 95 102 Rochelle L. Stanfield, ?The Golden Rolodex,? The National Journal 22:10 (March 10, 1990). 103 Heclo, pp. 119-121 104 Stone, Capturing the Political Imagination, p. 119 57 repackage theoretically obscure academic research, or convene a wide array of issue experts to discuss issues out of the politically charged limelight. In some instances, even if they may be rare, think tanks can provide experts to float trial balloons that the government is tentatively considering or provide a forum for officials to present their ideas. At other times, think tanks may not seek to help governments but may serve the public interest by injecting unconventional ideas into the policy process, complicating governance but potentially improving policy even if the benefits emerge long after the debate begins. Debates themselves may educate the public on public policy. Even research may be improved through interdisciplinary connections that think tanks convene. Clearly, the benefits depicted in the literature on think tanks are idealized but that is precisely the purpose of this chapter: to explore the roles, channels, and timing that think tanks might play. What think tanks seek to do, however, is only half the battle. The other half is to consider what is unique about the U.S. system to better understand how think tanks, and subsequently transnational networks, may seek to influence Washington and how they may fit into that system. What Makes the United States Different? Make no mistake about it. The literature universally concludes that think tanks are more influential in the United States than in any other country in the world. This could be either because think tanks themselves are more established institutions in the United States or because the U.S. system of governance is more responsive to them. Although 58 analytically distinct, the reality is that the two are intertwined and existing literature contends both are true. For starters, the funding for U.S. think tanks is more secure than anywhere else. From the creation of the first U.S. think tanks?such as the Russell Sage Foundation, the Brookings Institution, and the Carnegie Endowment, all named after their founders? philanthropists have been willing to support policy-related research. In no other country have the resources for independent think tanks been so richly available. 105 The fact that the United States has a well developed economy certainly helps; it means that the country can afford to have think tanks, unlike some developing countries that have had a difficult time. 106 But the size of the U.S. economy itself or of the hearts (and wallets) of philanthropists are not the only financial incentives to support think tanks. The U.S. tax code encourages funding think tanks by making corporate and individual contributions tax exempt under section 501(c)(3) of the Internal Revenue Code. Such a structural financial incentive helps create a massive funding source for non-profit organizations including think tanks in the United States. 107 Partially because of these tax incentives, smaller gifts from corporations 108 and from philanthropic foundations 109 over time supplemented the larger financial support of individual benefactors, diversifying the financial pipeline for think tanks and enhancing their independence even further. 105 Carol H. Weiss, ?Helping Government Think: Functions and Consequences of Policy Analysis Organizations,? ch. 1 in Weiss, ed. Organizations for Policy Analysis: Helping Government Think (London: Sage Publications, 1992): 8. 106 McGann, Think Tanks: Catalysts for Ideas in Action, p. 2 107 Weaver and McGann, p. 16; Donald E. Abelson and Evert A. Lindquist, ?Think Tanks in North America,? ch. 2 in McGann and Weaver, eds. Think Tanks & Civil Societies: Catalysts for Ideas and Action (London: Transaction Publishers, 2002): 54; and Stone, Capturing the Political Imagination, p. 45. 108 Goodwin, p. 25 109 Smith, p. 37 59 Some of the motivations of these financial sponsors reflect the nature and history of U.S. democracy and its climate of fierce independence to check the potential tyranny of information held by governments. 110 Creating think tanks and other similar institutions are part of a vibrant civil society to help preserve democracy. 111 A free press is a particularly vital asset for think tanks, enabling them to criticize government without fear of being censored and giving think tanks access to the public, through which they can pressure government to change policy. 112 Arguably more than other democracies, the culture of American democracy, including a limited role for the state and a strong culture of private entrepreneurship, helps diversify financial and moral support for think tanks and may help explain their relative influence in the United States. 113 The literature points to three other aspects of American democracy that uniquely facilitates avenues of influence for think tanks into the U.S. government: the width and depth of the revolving door, the weakness of U.S. political parties, and the independence as well as atomization of the U.S. Congress. Although the revolving door was discussed earlier, its relevance here is its uniqueness in the United States. Even in democracies such as France and Japan, the ability for outside analysts to get jobs in government is limited. These other countries rely much more heavily on a permanent, professional civil service. Such separation reduces the chances that nongovernmental analysts will have personal connections to government officials, narrowing the channels of access to those officials from the outside. This 110 Weaver, p. 571; James G. McGann, ?Think Tanks and the Transnationalization of Foreign Policy,? U.S. Foreign Policy Agenda 7:3 (November 2002): 16; and Donald E. Abelson, Do Think Tanks Matter? Assessing the Impact of Public Policy Institutes (Ithaca, NY: McGill-Queens University Press, 2002): 70. 111 Weaver and McGann, p. 17 112 McGann, The Competition for Dollars, Scholars and Influence, p. 42 113 Abelson, Do Think Tanks Matter?, p. 69 60 division also potentially reduces the substantive ability of think tank employees to advise governments because those employees would not have government experience themselves. In contrast, each U.S. political transition can mean hundreds of mid-level and senior executive branch personnel as well as congressional staff moving in and out of government, opening up positions for think tank alumni and potential avenues of influence to new government employees. 114 Part of the reason that new U.S. government employees might actually listen to advice from independent think tanks can be attributed to the weakness of American political parties. In this context, calling American political parties weak only means that they do not play a strong role in policymaking itself, particularly foreign policy, and that party platforms are merely symbolic documents. Parties may very well raise and distribute campaign contributions effectively, but in between elections, their role is limited. 115 In many other countries, such as Germany and Japan, the research and policy divisions of political parties provide ideas to political candidates and guidance to legislative and executive branch officials. In the United States, the parties do not play such a role, creating a void that think tanks and other nongovernmental organizations can fill. 116 Party weakness also widens the revolving door, since parties are not powerful enough to require elected officials to fill senior or even mid-level positions with loyal party officials. Although individual candidates will certainly hire campaign staff and volunteers once elected, the party itself can not force officials to do so. In fact, Congress 114 Haass, p. 7 115 ?The Charge of the Think Tanks,? The Economist, February 13, 2003; and Wiarda, p. 137 116 Weaver, p. 570, Weaver and McGann, p. 15, McGann, The Competition for Dollars, Scholars and Influence, p. 41, Abelson, American Think Tanks and their Role in U.S. Foreign Policy, p. 51, Gellner, p. 504, and Heclo, pp. 116-117 61 reportedly provided a budget of more than $2 million for the 1992 Clinton administration transition to help locate and assess potential appointees, including cabinet members, between the election and inauguration. Such a budget demonstrates the ability that officials have to find the best qualified candidates, not just the most loyal ones to the party. 117 The relative independence and atomization of the Congress are a third reason why even among democracies think tanks are relatively influential in the United States. 118 Several dimensions of government structure open the U.S. legislature to outsiders. The separation of powers between the President and Congress creates a structural competition that leads Congressional actors, particularly relevant committee members, to search for independent sources of information and advice beyond the executive branch. When the government is divided (with the majority in Congress from a political party other than the president?s), as it was for most of the time from 1970-2000, both branches of government are even more skeptical and more anxious to find independent sources of information and ways to criticize the other. 119 Within Congress itself, the U.S. Senate and House operate separately and are each powerful in their own right, unlike some democracies which effectively have one legislative body. Particularly after the 1970s, the atomization of Congress further enabled nongovernmental influence. After Vietnam and the Watergate scandal, Congress became more skeptical of the executive branch in general and instituted a number of reforms to become more decentralized, enabling more independent sources of oversight on future 117 Ricci, p. 135 118 Abelson, Do Think Tanks Matter?, p. 162 119 Stone, Capturing the Political Imagination, p. 40 62 administrations. 120 One result was to create a variety of committees and subcommittees, empowering a large number of individual Congressional actors, each of whom may have their own relationship with nongovernmental institutions. 121 The behavior of Congress also changed. Prior to the 1970s, Congressional order was preserved by relying heavily on seniority. New members of Congress deferred to senior members and followed party advice on how to vote to help secure their party?s support for re-election. In the 1970s, the orderly process of seniority was replaced by a system that dispersed Congressional authority to many individual members of Congress and empowered more individuals who chaired subcommittees and simply didn?t follow old rules to defer to senior members. 122 Combined with already weak political parties, these structural and behavioral changes increased the number of influential Congressional members who think tanks could either influence directly, or who might generate their own initiatives that think tanks could debate. 123 Another result of post-Watergate Congressional reforms increased the number of personal and committee staff able to provide research and guidance to Congressional members. Each of those staffers, in turn, became a potential person to whom nongovernmental organizations could reach out and advise. 124 In his research on how Congress uses information, then-associate professor David Whiteman explained that long before the mid-1990s, when his study was published, members of Congress had expanded their own advisory councils, ceasing to make isolated decisions as individuals. Instead, he 120 Brad Roberts, Stanton H. Burnett, and Murray Weidenbaum, ?Think Tanks in a New World,? The Washington Quarterly 16:1 (Winter 1993). 121 Higgott and Stone, p. 33, Gellner, pp. 505-506, Stone, p. 40, and Weaver and McGann, p. 15 122 Ricci, pp. 6, 23 123 Weaver, p. 570, and Abelson, Do Think Tanks Matter?, pp. 60-61 124 Whiteman, Communication in Congress: Members, Staff, and the Search for Information (Lawrence, KS: University Press of Kansas, 1995): 2. 63 called these advisors ?enterprises? consisting of a member and their associated personal and committee staff. Each of those staff members serving on each of the 535 Congressional enterprises might seek to diagnose problems differently or consider alternative policies to respond to those problems. 125 Think tanks, then, could essentially act as intellectual mercenaries to a number of Congressional members and employees. 126 Although other nongovernmental institutions, such as interest groups, seek to reach out to Congress, those groups carry obvious biases, leading their information and advice to be viewed skeptically by many. Think tanks, in contrast, may uniquely be perceived to have no ulterior motives and be able to provide the objective information Congressional actors seek. 127 Although some of this independence and atomization may have declined in more recent years under a united government (with Republicans controlling the executive branch and legislative majorities) and as the current administration has sought to recentralize executive power, the U.S. system remains more independent and atomized than most other countries. Although the independence, behavior, and atomization of Congress describe how nongovernmental analysts might gain access to one branch of government, they do not explain how Congress might use think tanks. In the late 1980s, Harvard education professor Carol Weiss published a study that looked at precisely that question, detailing the ways that the legislative branch uses external analysis. 128 Based on an array of interviews with Congressional staff, Weiss concluded that Congress does not use external analysis to set broad policy, relying instead on political guidance for that purpose. 125 Ibid. 126 Berman and Johnson, p. 21 127 Wallace, p. 151, and Weiss, ?Helping Government Think,? pp. 9-10 128 Carol H. Weiss, ?Congressional Committees as Users of Analysis,? Journal of Policy Analysis and Management 8:3 (1989): 424-428. 64 Congressional members and their staff did, however, use external analysis to help justify policy decisions after they were already made, as a source of warning to make sure that they or their bosses were not caught off guard by pending policy crises, and even occasionally to evaluate the details of specific policies. 129 The primary use of external analysis, based on the interviews, was to find seemingly independent sources that supported conclusions already reached by the Congressional member. In other words, think tank studies could be used as ex post facto justifications for policy conclusions already reached. This role clearly casts think tanks as pawns in a grand political chess game, rather than as the knights or major advisors that some may see for themselves. 130 Although Weiss?s study was limited to Congress, and speculated that the broad policy guidance think tanks seek to provide was more likely to influence the executive branch, Congress may not be the only institution that seeks to use think tanks to justify decisions already made. Weiss found other ways that Congress sought to use nongovernmental analysis. Seeking warning about veiled policy problems was the next most common use. Congressional staff, according to the study, was aware of many of the publications or proposals by think tanks to help insure that their member was not caught off guard by any looming policy problem. Finally, think tanks might also provide specific, expert guidance about how to write legislative details like what mathematical formula to use to guide funding distribution. Other more recent studies have also concluded that Congress is influenced by think tanks, or at least that Congressional staffers perceive that think tanks are somewhat, although not necessarily very, useful and influential. None of the 129 Weiss, ?Congressional Committees as Users of Analysis,? pp. 426-429 130 Also see Ricci, p. 210 65 subsequent studies have looked into foreign policy specifically and frequently sought a more general evaluation of whether, not how, think tanks were valuable to Congress. 131 Altogether, these are the major reasons explaining why think tanks are credited with being more influential in the United States than in other countries. They are both stronger institutions and the U.S. government is more permeable to nongovernmental organizations. The institutions themselves have been supported by a philanthropic tradition, motivating individual donors and further encouraging them with a tax code that makes donations to think tanks tax-exempt. That provides a stable supply of funding for a variety of competitive think tanks that have been able to operate within the U.S. system. Meanwhile, as a democracy, the U.S. system of government has provided a steady demand for think tanks to perform the functions outlined in the first section of this chapter while they help strengthen U.S. civil society and democracy itself by providing informed, free, and independent voices to evaluate government policy. Furthermore, the active revolving door into and out of U.S. government service, the minimal role in policymaking that American political parties play, and the independence as well as atomization of the U.S. Congress all provide opportunities for nongovernmental organizations to contribute policy advice and influence. How Think Tanks May Wield Influence Equipped with a better understanding of the purposes that think tanks may seek and the U.S. system of governance that may uniquely enable them to influence policy, the next step is to try to understand how they step into these opportunities provided by the U.S. system to achieve various goals or, in other words, how think tanks operate. The 131 See Whiteman, p. 186, Shulock, pp. 226-244, and Rich, p. 77 66 literature discusses it both in broad terms?depicting three different models or general types of think tanks?and in specific terms?describing the products, such as reports, and channels, such as the Congress, through which these organizations seek to influence government. These models and channels may each be useful to serve as precedents or to provide maps that might explain how transnational networks can seek to influence U.S. policy. Models of Influence To this point in the chapter, think tanks have been generalized as diverse actors, describing a variety of functions that they could perform while trying to make clear that no one think tank necessarily pursues all these goals or uses all the available tactics to influence government policy. With this diversity in mind, Brookings scholar R. Kent Weaver nevertheless usefully divided think tanks according to how they seek to influence policy into three different patterns of behavior or models: so-called ?universities without students? that seek to provide objective and usually public policy advice, consultant think tanks that directly and proprietarily respond to government requests for policy analysis, and advocacy tanks that promote a particular perspective on policy issues in public debate. 132 Although it is possible to generalize that certain think tanks behave more frequently like one of the three models, the same think tank could act like an objective public analyst on one project (particularly if funded outside the government) and act like a consulting tank on another (for a government contract). 132 Weaver, pp. 564-567; and Andrew Rich and R. Kent Weaver, ?Advocates and Analysts: Think Tanks and the Politicization of Expertise,? ch. 11 in Allan J. Cigler and Burdett A. Loomis, eds. Interest Group Politics (Washington, D.C.: Congressional Quarterly, 5 th edition, 1998): 240-242. 67 The first model, ?universities without students,? probably represents the most common stereotype for how think tanks behave. It certainly is how the existing literature assumes that they do. 133 These first think tanks, such as the Brookings Institution, emerged after World War I as the global demand for U.S. involvement in foreign affairs increased but the public remained more isolationist. A small foreign policy elite set out to educate the public, and to a lesser extent policymakers, about foreign policy. 134 In this model, think tanks perform independent, non-partisan research and help to define the middle ground for what foreign policies are politically feasible. In some ways, its employees act like professors without the teaching responsibilities required at universities. 135 Another way to envision their role is to think of these think tank employees as private investigators, seeking to find the truth. Philosophically speaking, this model assumes that some form of truth does exist, in the scientific tradition that Francis Bacon maintained, and could be found if experts were simply empowered with time to research and access to inform policymakers. In an extremely idealized form, think tank employees in this role might be envisioned as part of a Platonic Republic, acting as the advisers that could inform public officials, helping them become philosopher-kings and rationalize the often inefficient democratic societies. 136 In less philosophical terms, think tanks in this model depend on being perceived as independent analysts, relatively free of ulterior motives or inherent biases. 137 Their 133 See, for example, Smith, p. 213. 134 McGann, The Competition for Dollars, Scholars and Influence, pp. 46-47; and Donald E. Abelson, ?Think Tanks and U.S. Foreign Policy: An Historical Perspective,? U.S. Foreign Policy Agenda 7:3 (November 2002): 10. 135 Weaver and McGann, p. 9 136 Ricci, pp. 12-14, 22, 212, and Smith, p. 51 137 Ricci, p. 221, and Stone, Capturing the Political Imagination, p. 117 68 independence notionally enables them to find and deliver the truth, not obscure it for other objectives. Behaving in this manner, think tanks are able to determine their own agenda, constrained only by the ability to draw on an endowment or find specific funding, and identify topics that they believe experts should be thinking about. In contrast, think tanks behaving as consulting tanks set their agenda in response to requests from a particular branch of the government. This role for U.S. think tanks emerged after World War II, when the international system once again placed new demands on the U.S. government. It sought help and direct advice from particular think tanks on questions that the government asked without the intervening media pressure that the ?universities without students? generated. 138 The most frequently cited example of a think tank that often behaves in this fashion is RAND. 139 In this model, think tanks literally perform like consulting firms. 140 They respond to requests directly from a client (such as the government or, in other countries such as Germany, a political party) to research a particular problem. 141 As a result, think tanks that predominantly perform a consulting role do not use the media to pressure the government; they have a direct channel of access. 142 Although the bulk of the literature implicitly or explicitly assumes that think tanks behave as independent actors similar to the first model, at least one recent study has concluded that government officials actually pay more attention to research conducted by consulting tanks than either of the other models discussed here. 143 Although this result may initially seem surprising because the 138 Abelson and Lindquist, p. 41 139 Higgott and Stone, p. 25 140 Weaver and McGann, p. 9 141 McGann, The Competition for Dollars, Scholars and Influence, p. 74 142 Rich, p. 103 143 Rich, p. 101 69 media generally pays less attention to consulting tanks or research conducted for a proprietary contract, it also makes sense that the government would heed the result of a study that it funded without the accompanying public furor. Because of the financial resources available to the U.S. government, consulting contracts are unusually large. On the bright side, this potentially means a large source of revenue for tanks behaving in this fashion, 144 in addition to the direct channel of influence for its findings. On the other hand, it also can limit the intellectual freedom of research performed under consulting contracts. The questions asked, or the agenda, is determined by the client (the government) not the think tank. 145 The answers themselves may also be restricted. A consulting tank may believe that the client seeks a particular answer which, if not provided, could jeopardize future work and the large revenue streams that pay for it. 146 Whether future contracts would actually be jeopardized is another matter, but the potential conflict of interest exists. The third and final model, advocacy tanks, dispense entirely with the idea that some form of objective truth exists. Instead, when they behave as advocacy tanks, think tanks act like interest groups. 147 What these tanks advocate varies, ranging from attention to a particular issue such as human rights, to a region of the world such as Africa, or a political philosophy such as libertarianism. The think tank most frequently cited in the literature as behaving in this fashion is the Heritage Foundation although others such as Amnesty International or the Cato Institute also fit. 144 Rich, p. 12 145 McGann, The Competition for Dollars, Scholars and Influence, p. 141 146 Ricci, p. 20 147 Weaver and McGann, p. 9 70 This model emerged in the United States in the 1960s in response to ideological divisions and a societal call for greater values, not just objective science, in public policy debates. 148 The sources of these greater ideological divisions in the 1960s are not relevant for this study. What is relevant is that this model emerged in response to these societal pressures, in part because new funding for nonprofit organizations began to go to specific projects that supported one side or another of an emerging ideological debate, rather than the institutions more generally. 149 If other, investigative tanks are based on a Platonic-Baconian premise that truth exists and can be found, advocacy tanks are philosophically based on the ideas of John Stuart Mill, who rejected the idea that some absolute truth exists. Instead, Mill conceived of a marketplace of ideas. 150 Guided by their philosophical principles, think tank employees chose their product in this market and engaged in debates to defend it. Rather than having the institution itself act as a private investigator to find the truth, think tanks in this model behave like lawyers in a public hearing, arguing for their perspective in front of a judge (in the analogy, either the public or a government official). The practical application of this philosophical foundation is that advocacy tanks devote relatively more of their resources to working through the media, particularly compared to consulting tanks that would rely much more on private channels of influence. 151 Because they use the media and have greater (although still financially limited) freedom to choose their own agenda, advocacy tanks are more similar to 148 McGann, The Competition for Dollars, Scholars and Influence, p. 23, and Rich, pp. 30-31 149 McGann, The Competition for Dollars, Scholars and Influence, p. 51 150 Ricci, p. 215 151 Abelson, Do Think Tanks Matter?, p. 74 71 ?universities without students,? but they seek to promote a particular viewpoint, not to find an objective truth. At this point, it is only important to know that these three models for think tanks exist and that their characteristics, particularly who determines the agenda and how direct the channels of influence might be, could be useful for later discussion of transnational networks. In the meantime, examining in more detail the specific ways and times that think tanks may most successfully seek to influence the policy process in the existing U.S. system may also be useful. Channels of Influence Many more sources in the available literature examine the specific channels, or medium, that think tanks may use to seek to influence policy. Although it is nearly impossible to find direct evidence proving that ideas originating outside government migrated their way into public policy, the opportunities through various channels, such as briefings or reports, do exist. Through these channels, think tanks might approach the executive branch directly, the legislative branch, or use the media to seek indirectly to inform, influence, and ultimately shape policy. One set of channels think tanks use targets government officials directly. A particular channel, the revolving door, has already been discussed. What is noteworthy here is its depth. While some may serve as cabinet officials, others also rotate into mid- level positions, expanding the range of potential jobs into which think tanks employees might move. Although the revolving door started with the 1947 National Security Act, it developed through the late 1940s and 1950s during the Truman and Eisenhower 72 administrations as more policy advisers were appointed to the NSC, the CIA, the Council of Economic Advisers, or the State Department, particularly in the policy planning office. 152 In the early 1960s, President Kennedy then took the influence of outsiders to new heights by flooding middle levels of the bureaucracy with outside appointees in an attempt to have more influence over policy, circumventing the perceived stasis of career civil or foreign service officers. 153 The revolving door allows nongovernmental experts at least two different ways to influence policy. The first, and more obvious, is that think tank employees could be hired into government themselves. The second, more subtle, channel of influence is that employees that remained outside the government had new contacts rotating through public service. This increased the chances that a colleague may be in a position of influence within the government or, at least, that a friendly ear that might heed advice from outside government would hold influence somewhere. 154 The U.S. government has also set up a number of advisory boards, such as the President?s Foreign Intelligence Advisory Board (PFIAB) or the Defense Policy Board, that enable nongovernmental employees to have a direct channel to the White House. 155 How influential such boards are undoubtedly varies from administration to administration, but the point here is that they exist. The revolving door and advisory boards are transparent means of potential influence, opened by government itself; other channels are more subtle and initiated by think tanks. Government officials from either the executive or the legislative branches are 152 Stone, Capturing the Political Imagination, p. 42 153 Smith, p. 125 154 Smith, p. 113. Also see Mitchell B. Reiss, Remarks to the German Marshall Fund, Washington, D.C., December 6, 2004. Available at http://www.state.gov/s/p/rem/39495.htm (accessed February 17, 2005). 155 Abelson, American Think Tanks and their Role in U.S. Foreign Policy, pp. 72-75 73 invited to attend various conferences or lectures to speak or listen to the proceedings, taking in whatever they may, or even to participate in more intimate seminars and working groups. 156 Warwick University lecturer Diane Stone argues that exclusive groups in particular, requiring invitations to attend, facilitate the potential for influence by promoting contacts among a variety of people working on similar topics from inside government and out, building trust among them over time. 157 Stone even goes so far to argue that such groups are the ?primary means by which think tanks make ideas matter.? 158 Participation in such groups might not only build trust, but also might give government officials a sense of ownership over the ideas. Congressional staffers or members, for example, may be more likely to translate those ideas into legislation that they sponsor or support. 159 Such a process of directly including government staffers or senior officials in studies can increase the likelihood that such ideas will become policy even as it reduces the likelihood that think tanks will get credit for them, creating some of the research dilemmas about how to assign credit for ideas becoming policy identified in the first chapter. At the end of the day, think tanks actually serve their own long-term interests by shedding credit for original policy ideas, instead giving it to political officials. In this way, think tanks may gain a powerful and potentially valuable political ally for the long run even if deprived of public credit in the short run. 160 156 Abelson, American Think Tanks and their Role in U.S. Foreign Policy, p. 67 157 Stone, Capturing the Political Imagination, pp. 122-134 158 Stone, p. 126 159 Carol Matlack, ?Marketing Ideas,? The National Journal 23:25 (June 22, 1991). 160 Nelson W. Polsby, Policy Innovation in America: The Politics of Policy Initiation (New Haven, CT: Yale University Press, 1984): 171-172. 74 The more traditional, and transparent, way that think tanks are believed to seek influence is through the books and papers that they publish. It is a fair question, though, to ask how often these publications are read by busy policymakers. The busier the official is, the less time available to read, particularly books. Some written products produced by think tanks may garner additional influence by generating subsequent debate among experts or being condensed into op-eds more likely to be read. 161 But the think tank community has evolved to adjust to the limited time available for policymakers to read. Since the mid- 1970s, and the arrival of advocacy tanks more geared toward public debate, longer reports have been supplemented or even replaced by shorter, often one page, reports or by oral briefings. 162 The compressed amount of time available to read may be more pronounced in Congress, where officials have much wider portfolios than most executive branch aides with narrower, deeper sets of responsibilities. 163 Even staffers, who have relatively more narrow sets of responsibilities than Congressional members, reportedly prefer briefings, which can be more concise and quickly followed up with direct questions to authors. 164 The most efficient briefings for Congress, testifying in hearings, offer an intellectual economy of scale, potentially reaching all the subcommittee or committee members at once. 165 In addition to all these direct channels, think tanks seek to influence policy indirectly through the media. Appearing on television or writing op-eds puts indirect 161 Smith, p. 193 162 Smith, p. 121; Louise Diamond and John McDonald, Multi-Track Diplomacy: A Systems Approach to Peace (West Hartford, CT: Kumarian Press, 1996): 157; and Abelson and Carberry, p. 682. 163 Smith, p. 194 164 Weiss, ?Congressional Committees as Users of Analysis,? p. 414 165 Abelson, American Think Tanks and their Role in U.S. Foreign Policy, p. 71, and Haass, p. 6 75 pressure on policymakers to change policy. 166 Particularly if an array of experts expresses concern about a developing crisis or supports a policy response, it becomes more difficult for government to ignore the problem or oppose the policy which that apparent consensus of experts publicly supports. In other words, think tank experts help the media define the parameters, and even the terms, of public policy debate. 167 Particularly with the development of the internet and the emergence of 24-hour cable news channels, 168 a constant stream of news programs and outlets request experts from think tanks to comment on current events and debate policy responses. 169 There is a danger that frequent media appearances can be overestimated, however, leading some to mistake prevalence for policy influence. 170 Nevertheless, it may be possible for think tanks to generate support for broad approaches or attention to overlooked issues through the media. In other words, the media can act as a force multiplier, not just defining the limits of public policy debate but improving the ability of think tanks to help set the policy agenda itself. 171 The clich? is that the media may not be successful in telling people (or the President or Congress) what to think, but they are successful in telling them what to think about. This often means the earliest, agenda-setting stages of the policy process before policymakers have made decisions and commitments, allowing think tanks and the 166 Rochelle L. Stanfield, ?The Golden Rolodex,? The National Journal 22:10 (March 10, 1990). 167 Goodwin, pp. 25-29 168 Ricci, p. 101 169 Abelson, ?Think Tanks in the United States,? p. 119 170 Donald E. Abelson, ?Policy Experts and Political Pundits: American Think Tanks and the News Media,? NIRA Review (Summer 1998). 171 Abelson, Do Think Tanks Matter?, p. 82 76 media to draw attention to issues, diagnose problems, propose policies, and frame the boundaries of public policy debate. 172 When Think Tanks May Have Influence The heightened impact in the agenda-setting stage begs the broader question about when think tanks are influential in the policy process. Two studies may shed some light on the answer. In his previously described study of health care, telecommunications, and tax policy, Yale Ph.D. Andrew Rich concluded that ?the role of experts tends to be greater in debates that take on a high public profile, that move at a relatively slow pace, and that do not elicit the mobilization of organized interests with much to lose in the decisions under consideration.? 173 In the telecommunications case specifically, he found evidence that even when special interests are heavily invested in debates, a consensus among experts can still be influential in the early stages of agenda setting, confirming the relative opportunity available at that time. 174 It is, Rich argues, ?during agenda setting that think tanks and experts generally can often have their greatest substantive influence.? 175 Although it may take ?years, if not decades, to become evident? and ?be diffuse and difficult to trace,? Rich concludes that experts can make a difference when they warn about problems and recommend solutions to them, playing ?a critical role in how issue debates take shape and are initially defined.? 176 Once nongovernmental experts have drawn sufficient attention to a problem, they subsequently become invaluable sources of information on that problem for 172 Weiss, ?Helping Government Think,? pp. 10-11 173 Rich, p. 107 174 Rich, p. 148 175 Rich, p. 28 176 Rich, pp. 108, 138-139 77 lawmakers or other policymakers who face pressure to learn about this unfamiliar subject and then do something about it. In particular, Rich found evidence that experts on health care were sought out by policymakers once the issue had developed a high public profile. 177 The longer that an issue is debated among policymakers however, Rich found, the more difficult it becomes to actually influence them. As deliberations proceed, officials become committed to positions and external expertise increasingly shifts from being truly influential to become ammunition justifying decisions that have already been taken. 178 This conclusion is identical to the one that Carol Weiss reached in the late 1980s, although Rich takes the additional step of specifying the timing of the shift from experts being truly influential in the agenda-setting phase to simply providing ex post facto justifications in the policy deliberation phase. Rich certainly acknowledges that it is still possible for think tanks to be influential in later phases, but it is more difficult. 179 By the time a policy decision is being made, Rich concludes that ?the opportunities for contributions of policy research were minimal. Frequently, the minds of policymakers had been made up. Even more often, though, the specific issues that remained to be resolved tended to be narrow provisions that had the greatest consequences for interests heavily vested in the outcomes.? 180 In other words, the power of interest groups becomes greater as policy deliberation proceeds, likely overwhelming the ability of think tanks with more limited resources to influence the final decision itself. 177 Rich, p. 143 178 Rich, pp. 108, 139 179 Rich, pp. 197, 202 180 Rich, p. 146 78 In the final stages of the policy process, as decisions are being implemented, Rich found that experts become useful to evaluate those decisions, providing ?assessment and further guidance.? 181 In essence, experts once again become agenda setters, raising the profile of issues that may have seemed to have been addressed but, in fact, are not solved. The policy cycle starts all over again, and the influence of experts is heightened in setting the agenda, particularly if interest groups are not opposed. A second study highlights a particular time in the policy process when think tank influence is maximized: during presidential election campaigns and the transition period that follows an election. In a sense, every presidential election is an exercise in agenda setting. Political opponents raise issues that have not successfully been addressed by an incumbent president or their party. In his study of think tank influence on U.S. foreign policy particularly, University of Western Ontario associate political science professor Donald Abelson even concluded that ?the best opportunity for think tanks to influence U.S. foreign policy is during a presidential election and the transition period that follows.? 182 (emphasis added) Abelson contends that American presidential elections uniquely open up channels for think tank influence not just because of the depth of the revolving door and the weak policy influence of U.S. political parties, but also because the United States has a predictable electoral cycle, unlike other countries where elections may be called by votes of no confidence or, conversely, by the incumbent during the 181 Rich, pp. 108-109 182 Donald E. Abelson, ?Trends in Search of Policy Influence: The Strategies of American Think Tanks,? NIRA Review (Spring 1998). See also Abelson, American Think Tanks and their Role in U.S. Foreign Policy, pp. 66, 122. 79 height of their popularity. 183 This allows more time for think tanks to organize and engage in the extended pre-election debates. Specifically, Abelson argues that two characteristics of presidential candidates can help predict whether they will use think tanks extensively during the campaign: whether they are Washington outsiders, increasing their use of think tanks, and whether they are strongly ideological, increasing the likelihood that candidates will rely on employees of think tanks if they are, particularly ideologically similar advocacy tanks. 184 He found that ?Jimmy Carter, Ronald Reagan, and other presidential candidates who spent little time in Washington before assuming the presidency recognized the enormous benefits of relying on think tanks that had many former high-profile policymakers and leading policy experts.? 185 Anecdotally, former U.S. State Department policy planning director Richard Haass also cited the perceived impact of the 1980 Heritage Foundation publication, Mandate for Change, which provided a blueprint for the incoming Reagan administration, and 1992 publications by a number of think tanks proposing a National Economic Council, which was adopted by the Clinton administration, as evidence of the influence of think tanks during and immediately after presidential elections. 186 In their most basic role during campaigns, experts may identify potential problems or suggest solutions to aspiring presidential candidates to consider highlighting. Alternatively, the relationship may work in reverse, with candidates proposing the ideas, particularly in the early stages of a campaign, and running them by subject experts in informal, private settings. Depending on the candidate, experts might be consulted widely 183 Abelson, Do Think Tanks Matter?, pp. 162-163 and Ricci, p. 123 184 Abelson and Carberry, pp. 679, 684-685 185 Abelson, Do Think Tanks Matter?, p. 125 186 Haass, p. 7 80 or might be limited to campaign staff loyal to a particular candidate. Even doors to the incumbent administration itself may open if a think tank?s ideas are co-opted to prevent an opposing candidate from using them. In addition to these behind-the-scenes roles, perceived experts can play a more public role, endorsing a candidate?s ideas to provide expert credibility and validation for them. 187 Once the campaign has completed, the floodgates may open up for think tank staff, particularly those that volunteered for a victorious candidate, to become part of the revolving door either on a transition team or in the administration itself. 188 Scholars can also recommend other like-minded colleagues from outside government to fill out the ranks of hundreds of positions throughout the U.S. government in a new administration. 189 Other than the agenda-setting stage of the policy process and presidential campaigns, a third time when think tanks might be influential is during crises. 190 There is no question that think tank experts get greater media exposure as they comment on developing events. These public appearances, however, may not necessarily mean they are influencing policy. Unlike other times when media appearances help raise issues for governments to consider, the crisis itself has already set that agenda, particularly in national security crises. It seems instead that think tanks turn to the media in a crisis to help explain rapidly unfolding events to the public. Although that may boost their long- term notoriety and credibility in public and policy circles, none of the think tank literature speculates on any short-term impact on crisis management. In security policy in 187 Abelson and Carberry, pp. 679-684 188 Abelson, ?Think Tanks in the United States,? p. 120 189 Abelson, Do Think Tanks Matter?, p. 63 190 John W. Kingdon, Agendas, Alternatives, and Public Policies (New York: Harper Collins, 2 nd ed., 1995): 85-88. 81 particular, this seems to make sense as governments retreat into classified circles during security crises. Think Tanks as Models for Transnational Networks Collectively, think tanks act as public investigators, private consultants, and issue advocates to convene interdisciplinary networks, inject new ideas into the policy process, debate issues in-depth, and provide over-the-political-horizon perspective through the media or directly to legislative or executive branch officials to shake up the inherently stabilizing nature of governance. Their potential influence is maximized, although potentially overlooked, during the early, agenda-setting phase of the policy process and particularly during presidential election campaigns, especially through the revolving door if the incumbent loses. The facts that such a revolving door is uniquely deep, that U.S. political parties are weak, and that Congress is relatively independent and atomized may uniquely empower U.S. think tanks?supported by tax exempt, independent, philanthropic funding sources?more than in other countries, even democracies. These roles, channels, and timing of think tank influence may serve either to describe the indirect impact that their transnational cousins, policy networks, have through individual nongovernmental members, or serve as precedents to predict how U.S. policy or perceptions might be influenced by networks, as another policy-oriented nongovernmental organization. Plugged into the four-part research framework devised in chapter one, the literature on think tanks both reaffirms a potential agenda-setting role (the third part of the research framework) and presents three particular questions about the channels of access to government policy (the second part) that may be useful: 82 How do networks incorporate and engage the U.S. Congress? How do networks utilize the media to inform the U.S. public and put policy pressure on the U.S. government? and How do networks use the revolving door to influence U.S. policy? Beyond the research framework, the literature on think tanks also points to four broad positive roles that networks might serve?to convene interdisciplinary (and, in their case, global) networks; to inject new ideas into the policy process; to debate the diagnoses of problems and their potential solutions; and to provide strategic, over-the- horizon perspective?as well as to ask if networks adversely affect or complicate U.S. policy. What remains is to test the theoretical framework constructed in these first two chapters with three case studies?CSCAP (chapter 3), NEACD (chapter 4), and PONARS (chapter 5)?focusing on nonproliferation policy activity within each network, and to review the results in a concluding chapter. 83 CHAPTER THREE: THE COUNCIL ON SECURITY COOPERATION IN THE ASIA PACIFIC The Council for Security Cooperation in the Asia Pacific (CSCAP) was formally established in June 1993 ?to contribute to the efforts toward regional confidence building and enhancing regional security through dialogues, consultations, and cooperation.? 191 Originally founded by nongovernmental representatives from ten countries throughout the Asia Pacific region, it has since expanded to over 1,000 individual members 192 from 21 national committees, represented by a think tank from each country including Australia, China, India, Japan, North Korea, South Korea, Russia, each of the ASEAN countries, Canada, Europe, and the United States, among others, to become ?the broadest Track II security organization in the world.? 193 CSCAP (commonly pronounced see' skap) has a relatively formal and well- developed infrastructure that holds four different types of meetings and produces or distributes a small number of written products. Business is coordinated and guided by an overarching international steering committee, with one or two representatives from each national member committee meeting semiannually. A second, larger or general international conference has been convened biannually since 1997 to broadcast individual views as well as provoke thought and discussion on regional security among all CSCAP members in attendance and, in recent meetings, the general public. A third type of 191 ?The Kuala Lumpur Statement establishing the Council for Security Cooperation in the Asia Pacific (CSCAP),? June 8 1993, available at http://www.cscap.org/kl.htm (accessed October 29, 2004). 192 Electronic message from Brad Glosserman to author, October 31, 2005 193 Sheldon Simon, ?Evaluating Track II Approaches to Security Diplomacy in the Asia-Pacific: the CSCAP Experience,? The Pacific Review 15:2 (2002): 173. For further details on its history, see also Desmond Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP): Its Record and Prospects, Canberra Papers on Strategy and Defense 139 (Canberra: Australian National University, 2000): 5-32. 84 meeting, working groups ?to undertake policy-oriented studies on specific regional and sub-regional political-security problems? 194 conduct the bulk of CSCAP?s international substantive work. International working groups on maritime cooperation, concepts of cooperative and comprehensive security, the North Pacific, and confidence and security building measures (CSBMs) were established in 1994 with a fifth working group, on transnational crime, added in 1997. Of particular relevance to this study, the bulk of work on nonproliferation was conducted in the North Pacific (almost exclusively on North Korea) and CSBM groups. In 2004, all five working groups were dissolved with six new working groups taking their place: capacity building for maritime cooperation, prospects for a multilateral security framework in Northeast Asia, countering WMD proliferation, human trafficking, regional peacekeeping and peacebuilding, and enhancing the campaign against international terrorism. These and subsequent groups now have two years to complete their work unless the international steering committee extends them. Unlike the first three types of international meetings, the fourth is held within a single country. USCSCAP meetings are convened once or twice a year in Washington to inform members about CSCAP activities and address any national administrative matters, while simultaneously gathering about 40-50 USCSCAP members to listen to one or two government briefings and discuss recent events. As of 2004, U.S. membership had been granted to about 200 people who expressed interest in joining and were asked to pay a $75 annual membership fee. 195 U.S. members include nongovernmental analysts from universities or think tanks, some of whom are former government officials, and current 194 CSCAP Charter, Article VIII, section 1, available at http://www.cscap.org/charter.htm (accessed on October 29, 2004) 195 Author?s interview with Brad Glosserman, Washington, D.C., December 14, 2004 85 government officials in their private capacity, formally labeled ?ex officio.? A typical USCSCAP meeting itself will start with a briefing on the year?s CSCAP activities, followed by a briefing by a State Department official on ASEAN Regional Forum (ARF) activities, continue with one or two roughly one-hour panel discussions on recent events (often China or North Korea), and conclude with an off-the-record lunch presentation and discussion led by a senior government policymaker, usually an assistant secretary or deputy assistant secretary (DAS). Funds to cover meeting costs have come from a variety of sources, with each working group, principally its cochairs, and national committee responsible for raising their own funds. Initial money from the United States around 1993-94 was provided by the Pew Foundation, and $75,000 a year by the Ford Foundation from about 1994-2000 to cover both the costs of the U.S.-cochaired CSBM working group and, in addition to members? dues, USCSCAP meetings. More recently, the U.S. Department of Energy?s Office of Nonproliferation and National Security has become the primary funder of the CSBM working group, ?accounting for 90% of support? with an additional $15,000 provided by the Asia Foundation. A new grant from the Carnegie Foundation was provided in 2005 to supplement the DoE funding and cover roughly one-third of the countering WMD proliferation working group?s expenses. 196 CSCAP also produces or distributes a handful of written products. Eight CSCAP memos, the network?s most formal product, have been issued, most recently in April 2004. Typically reporting a working group consensus, memos are approved by the international steering committee and are usually intended for the ARF, although they are publicly distributed. Two other products, PacNet and Comparative Connections, are also 196 Electronic message from Ralph Cossa to the author, August 12, 2005 86 electronically distributed to over 3,000 recipients worldwide. 197 Both compiled by the Pacific Forum, PacNet is a series of weekly editorials and Comparative Connections is a quarterly publication systematically reviewing many of the region?s key bilateral relationships. PacNet, in particular, was frequently raised in interviews conducted for this study and will be discussed later in this chapter. Twenty-two original interviews were conducted for this study. As outlined in the introductory chapter, interviews were initially sought with staff, funders, participants (nongovernmental and governmental), and senior government policymakers. Participants (either nongovernmental or governmental) were targeted for interviews if they attended two or more North Pacific or CSBM working groups or international steering committee meetings, or had presented remarks at two or more USCSCAP meetings. 198 Although some candidates participated in multiple CSCAP events, such as working groups and the international steering committee, table 3.1 outlines the potential interview candidate list according to my assessment of their primary CSCAP participation, with no candidate counted in more than one category (or box). Some interviews were not conducted either because candidates had retired or otherwise could not be found, declined to be interviewed (usually because they felt they did not have enough experience with CSCAP), or did not respond to requests to be interviewed. 197 Author?s interview with Brad Glosserman, Washington, D.C., December 14, 2004 198 Participants lists at USCSCAP meetings are not collected, so presenters were identified through agenda for all USCSCAP meetings from 1997-2005 (including semiannual meetings in 2001 and 2003) which were provided by USCSCAP staff. 87 Table 3.1 CSCAP Interview Candidates NGO USG Total CSCAP staff 4 - 4 CSCAP funders 2 0 199 2 CSCAP participants 17 10 27 Attended 2 or more working groups 9 5 14 Attended 2 or more steering committees 4 1 5 Presented at 2 or more USCSCAP mtgs 4 0 4 DASs who participated once 200 - 4 4 Senior government policymakers 201 - 3 3 Referrals from other interviews 3 5 8 Total 26 18 44 Table 3.2 presents a matrix of descriptions of the subjects actually interviewed. Ten of the 22 conducted were with U.S. government officials, five of whom were members of the revolving door but were in government either during the bulk of their affiliation with CSCAP, or were in policymaking positions (deputy assistant secretary or higher) during at least part of their affiliation. The other twelve interviews were with members of nongovernmental organizations, although six of those twelve were also members of the revolving door who had government experience before their affiliation with CSCAP. In other words, of the 22 interviews conducted, five were with permanent government employees (civil or foreign service), eleven with members of the revolving door, and six with nongovernmental employees who had not been employed in government. Throughout this chapter, opinions expressed in only a single interview will 199 Although the Cooperative Monitoring Center of Sandia National Labs, and subsequently DoE, funded some CSCAP activities, those interview subjects are already accounted for as either government participants in CSCAP or as NEACD funders. Although the box in which a candidate is placed is somewhat arbitrary, the exact placement is less important since the purpose is simply to systematically determine whom to interview. 200 Because of their position in the policymaking chain, deputy assistant secretaries were sought for interviews even if they only attended one meeting. 201 Senior government policymakers were defined as senior directors of the NSC staff or Assistant Secretaries of State for East Asia and the Pacific. 88 be specifically noted as one person?s opinion (whether cited anonymously or specifically). Table 3.2 CSCAP Interviews Conducted NGO USG Total CSCAP staff 2 - 2 CSCAP funders 0 0 202 0 CSCAP participants 8 5 13 Attended 2 or more working groups 5 3 8 Attended 2 or more steering committees 1 1 2 Presented at 2 or more USCSCAP mtgs 2 0 2 DASs who participated once - 1 1 Senior government policymakers - 2 2 Referrals from other interviews 2 3 5 Total 12 10 22 In addition to the interviews conducted and a handful of secondary sources (bibliography is included in appendix A) that have evaluated CSCAP for their own purposes, primary written materials analyzed for this chapter include relevant PacNet publications; CSCAP memos; a summary of the fourth general conference (the only one of the four available); publications from, and the summaries of, meetings from the North Pacific and CSBM working groups; and summaries of USCSCAP meetings. 203 The North Pacific working group held eleven meetings between its first meeting in Tokyo in April 1995 and its last in Seoul in 2003 with detailed summaries for four, participants lists for six, and a detailed agenda for eight meetings either available from the CSCAP website or obtained from the Canadian or Japanese cochairs. The CSBM working group held 22 meetings between its first in Washington in October 1994 and its last in Hanoi in May 202 Government funders for CSCAP, providing part but not all of its support, who also funded the second case, NEACD, were interviewed and are counted in that chapter because of their principal funding role for that network but not here to avoid redundantly counting their participation in the study. 203 Notes from the international Steering Committee meetings are not publicly available and generally only include administrative matters and condensed summaries of the working groups. 89 2004, with detailed summaries for fourteen, participants lists for ten, and a detailed agenda for fifteen meetings either available from the CSCAP website or obtained from the U.S., Singaporean, or South Korean cochairs. Brief, one sentence summaries identifying the primary agenda of all working group meetings are available on the CSCAP website. In addition to these interviews and written materials, the author was able to attend three of the international steering committee meetings between 2001 and 2004, and most of the ten USCSCAP meetings between 1998 and 2005. Having introduced CSCAP?s infrastructure and the primary materials utilized, the subsequent three parts of this chapter will present the issues the network has covered from 1993-2004, when the original working groups were dissolved; an evaluation of the network based on the four-part framework established in the introductory chapter; and finally a series of observations and conclusions about CSCAP. The concluding chapter (chapter six) will include further discussion comparing the network to the other two case studies. This chapter will be limited to CSCAP itself, starting with the issues the network has addressed. Issues CSCAP Has Addressed A mere eighteen months after the dissolution of the Soviet Union, CSCAP began to tackle a multilateral security agenda for the Asia-Pacific. A quick glance at the issues that consumed the network?s initial deliberations and the fact that China did not even officially join CSCAP until 1996 are reminders of how fluid the Asian security agenda has been over the last 10-15 years. Although this study is focused on the influence on US nuclear nonproliferation policies and the perceptions that shape them, overlooking 90 CSCAP?s other work would distort the view of its efforts. The first part of this section briefly reviews that other work, with a more detailed discussion of the network?s nonproliferation agenda following. In the early-mid 1990s, the regional security landscape was dominated by concerns about how to accommodate the rise of Japan and China, tensions between China and Taiwan, and the future of the Korean peninsula. Those issues generally remain today but others, such as the potential for conflict in the South China Sea and whether the United States would remain engaged in Asia without the Soviet enemy, have significantly faded, at least for now. 204 In this environment, CSCAP first and foremost simply sought to help build the infrastructure for multilateralism in Asia, which didn?t yet exist. Now part of the institutional landscape, organizations like APEC (1989) and particularly security fora like the ARF (1994), were only just being established. In part, CSCAP sought to provide analytical support to the ARF, subsequently struggling throughout the network?s history with the question of how closely it should align itself with the official forum, 205 a topic which will be discussed in greater detail in the observations section of this chapter. As it mentioned in its charter, CSCAP initially sought simply to establish its own multilateral forum for dialogue and to increase regional transparency which could ?provide reassurance and help build trust and confidence,? as a 1994 CSCAP memo 204 For early CSCAP-sponsored surveys of the regional security scene and concomitant potential confidence and security building measures, see Ralph A. Cossa, ?Asia Pacific Confidence and Security Building Measures,? (ch. 1, pp. 1-18) and Robert A. Manning, ?Building Community or Building Conflict?,? (ch. 2, pp. 19-40) in Cossa, ed. Asia Pacific Confidence and Security Building Measures, Significant Issues Series 17:3 (Washington D.C.: Center for Strategic and International Studies, 1995); as well as Desmond Ball, ?CSCAP: Its Future Place in the Regional Security Architecture,? (pp. 289-325) and Jusuf Wanandi, ?The Future of the ARF and CSCAP in the Regional Security Architecture,? (pp. 281-288) in Bunn Nagara and Cheah Siew Ean, eds. Managing Security and Peace in the Asia-Pacific (Kuala Lumpur, Malaysia: ISIS Malaysia, 1996). 205 Wanandi, p. 284 91 argued. 206 Dialogue was uncommon at the time, but a necessary foundation for this broader agenda of allaying regional suspicions and fears derived from misperceptions. 207 Just talking at least initially ?may be as (or more) important than the product,? as the 1994 memo put it. Even in its founding stages, however, CSCAP itself recognized that soon ?dialogue without a focused, defined purpose can be difficult to sustain and a waste of precious resources and effort.? 208 To extend transparency further, CSCAP initially tried to promote two particular confidence-building efforts: a regional arms registry and the publication of national Defense White Papers. The former efforts sought an independent regional initiative, but eventually settled for promoting regional accession to the UN Conventional Arms Register. 209 For White Papers, CSCAP drafted a template, adopted by the ARF?s Intersessional Group (ISG) on CBMs, that countries could use as a model for their own efforts. Mongolia, Vietnam, and Taiwan subsequently used the draft, and China even partially used it for its second and third white papers. 210 As the 1990s progressed, CSCAP continued to seek to promote regional transparency in two other specific areas, working with the ARF in both cases: maritime 206 ?Asia Pacific Confidence and Security Building Measures,? CSCAP Memorandum No. 2 (June 1995): 4. See also ?The Security of the Asia-Pacific Region,? CSCAP Memorandum No. 1 (April 1994): 3; as well as Brad Roberts and Robert Ross, ?Confidence and Security Building: A USCSCAP Task Force Report,? ch. 8 in Cossa, ed. Asia Pacific Confidence and Security Building Measures, Significant Issues Series 17:3 (Washington D.C.: Center for Strategic and International Studies, 1995): 145 for early CSCAP arguments in favor of transparency. 207 Ball, ?CSCAP: Its Future Place in the Regional Security Architecture,? p. 295; Roberts and Ross, ?Confidence and Security Building: A USCSCAP Task Force Report,? p. 140; and Paul M. Evans, ?Assessing the ARF and CSCAP,? ch. 10 in Hung-mao Tien and Tun-jen Cheng, eds. The Security Environment in the Asia-Pacific (London: M.E. Sharpe, 2000): 165, 170. 208 CSCAP Memorandum 2, p. 3 209 See Cossa, ?Asia Pacific Confidence and Security Building Measures,? pp. 10-11; as well as CSCAP Memorandum 2, pp. 4-5 for initial CSCAP efforts. Reviews of those efforts include Desmond Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP): Its Record and Prospects, Canberra Papers on Strategy and Defense 139 (Canberra, Australian National University, 2000): 15; and Simon, ?Evaluating Track II Approaches,? p. 188. 210 Simon, ?Evaluating Track II Approaches,? p. 188. Also see CSCAP Memorandum 2, p. 8. 92 cooperation and preventive diplomacy. The network has ventured into a wide variety of regional, particular Southeast Asian, maritime concerns such as territorial disputes, law and order at sea, smuggling, and piracy, 211 working with the ARF, and in one case with APEC, to help develop regional maritime cooperation. 212 But CSCAP?s greatest success has arguably been in the field of preventive diplomacy, where the network?s tangible effect is most evident, at least on the ARF. Itself a new concept in the 1990s, little agreement existed on what preventive diplomacy actually meant. Some in Southeast Asia even feared that this new-fangled idea was simply a way to justify interfering in another country?s domestic affairs. Over three days in the early spring of 1999, CSCAP convened a workshop in Bangkok with many of its participants immediately attending a subsequent ARF CBM ISG on preventive diplomacy. The CSCAP workshop produced a statement on ?Preventive Diplomacy: Definitions and Principles? to define the concept in a regional context, and forwarded it to the ARF meeting. 213 Subsequently, the ARF formally agreed to a three-stage road map toward comprehensive security for Asia, starting with confidence-building, moving through a stage of preventive diplomacy, and concluding with conflict resolution. 214 Despite the initial success in helping the ARF reach consensus 211 Simon, ?Evaluating Track II Approaches,? pp. 179-180 212 For a description of CSCAP?s efforts working with the ARF on maritime cooperation, see Summary of Discussions, USCSCAP general membership/board of directors meeting (Washington, DC, June 14, 1999): 2; and Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), p. 50. For APEC, see Summary of discussions, USCSCAP meeting, USCSCAP Special Report 10 (Washington, DC, October 24, 2003). 213 Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), pp. 49-50. For primary summaries of the concerns about preventive diplomacy and the original CSCAP work, see Summary of Discussions, Ninth meeting, CSCAP CSBM International Working Group (Manila, Philippines, 12-13 December 1998): 2; and Summary of Discussions, CSCAP Workshop on Preventive Diplomacy (Bangkok, Thailand, 28 February-2 March, 1999): 4. 214 Sheldon W. Simon, ?The ASEAN Regional Forum Views the Councils for Security Cooperation in the Asia Pacific: How Track II Assists Track I,? Managing Security Challenges in Southeast Asia, National Bureau of Asian Research Analysis 13:4 (July 2002): 9. 93 on the terms of preventive diplomacy and adopt it as a regionally desirable goal, subsequent implementation by the ARF has disappointed at least some observers. 215 Finally, it is worth mentioning recent efforts on terrorism. Within months of the September 11 attacks, the CSCAP international steering committee convened a special study group on terrorism with the intent of preparing a memo for an ARF Senior Officials Meeting (SOM). 216 Subsequent efforts, including a March 2003 workshop in Laos, 217 and a panel as part of the 2003 general international CSCAP conference in Jakarta, 218 concluded that CSCAP might best focus its efforts on long-term issues, not short-term operational recommendations. Based on this evidence, it seemed that CSCAP?s working group on enhancing the campaign against international terrorism may focus on investigating, and ideally defining and building consensus around, the root causes of terrorism and how to address them. Proliferation Issues Although CSCAP has tackled a range of Asian security issues, the proliferation of nuclear and other weapons of mass destruction has been on its agenda since the network?s inception, particularly because U.S. participants have put it and kept it there. CSCAP?s very first memo, ?The Security of the Asia Pacific Region,? identified the North Korean nuclear issue, a nuclear weapons-free zone in Southeast Asia, and ?cooperative measures 215 Simon, ?Evaluating Track II Approaches,? p. 190. See also Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), pp. 81-82 for earlier skepticism. 216 Joyce Kallgren, letter to USCSCAP members, ?Summary of 2001 CSCAP and USCSCAP Activities,? (January 2002): 2. 217 Co-chairs? Summary Report of the Workshop on ?Countering the New Terrorism: Options and Strategies for Policy-Makers,? Vientiane, Lao, 25 March 2003, particularly p. 4. 218 See Summary of Discussions, CSCAP Steering Committee (Jakarta, Indonesia, December 6-7, 2003), sent as electronic message to all USCSCAP members, USCSCAP Special Report #1, January 12, 2004, p. 4. 94 to control new technologies and weapons of mass destruction, especially ballistic missiles, nuclear and chemical weapons? as potential areas to develop confidence and security-building measures. 219 Those initial efforts were led by U.S. members, however, and faced resistance from regional participants. CSCAP initially scoped out the potential confidence and security-building agenda through a U.S. task force, coauthored by Brad Roberts and Robert Ross, whose findings were then presented to the first CSCAP international CSBM working group in 1994. That USCSCAP task force even explicitly noted that, ?We recognize that many in the Asia Pacific do not share the sense of urgency about [unconventional weapons proliferation, arms races, and their attendant instabilities] that prevails in the United States.? 220 Initial efforts focused on getting all regional states to vote to indefinitely extend the NPT in 1995 as well as join and support other global arms control regimes such as the Chemical Weapons Convention (CWC), Biological Weapons Convention (BWC), and Missile Technology Control Regime (MTCR). 221 Proponents of CSCAP?s nonproliferation efforts, however, discovered that there was little regional support for going beyond existing global efforts, just as they had encountered with efforts to go beyond conventional transparency measures such as the UN Conventional Arms Registry. 222 Roberts recalled that despite the initial regional reaction that nonproliferation was ?America?s job,? the ARF had by 1997 ?adopted a communiqu? that gave strong praise to nonproliferation well beyond what the U.S. delegation had gone there expecting or 219 CSCAP Memorandum 1, pp. 1, 5. Discussion of nuclear weapon free zones in both the Korean peninsula and Southeast Asia continued in CSCAP Memorandum 2, p. 6. 220 Roberts and Ross, ?Confidence and Security Building: A USCSCAP Task Force Report,? p. 152 221 Roberts and Ross, ?Confidence and Security Building: A USCSCAP Task Force Report,? pp. 152-155; and Brad Roberts, ?Evaluating Global Non-Proliferation Instruments,? ch. 3 in Ramesh Thakur, ed. Keeping Proliferation at Bay (Jakarta: Center for Strategic and International Studies, 1998): 60-74; and CSCAP Memorandum 2, p. 5. 222 Ball, ?CSCAP: Its Future Place in the Regional Security Architecture,? p. 294 95 even hoping to promote. Out of this came an even stronger ARF interest in nuclear nonproliferation and a decision by the CSCAP working group to focus in on the nuclear question: one, because we saw the interest; two, because we saw less interest in the rest of our [CSBM] agenda; and, third, because we saw money? available to fund those efforts. 223 To this day, regional resistance still exists, at least partially because of perceptions that nuclear powers need to go beyond nonproliferation and make more progress toward their own disarmament, or so-called Article VI issues of the NPT, specifically including objections to: U.S. opposition to ratifying the comprehensive test ban treaty, its withdrawal from the ABM Treaty as well as subsequent national missile defense deployment, and its nuclear modernization. 224 The final CSBM working group in May 2004 even speculated that ?indeed, it can be asked if nonnuclear states really see proliferation as a problem, at least as far as their national security is concerned. They seem more focused on the inequalities in the NPT, rather than the nuclear threat per se.? Nevertheless, the report concluded that ?states may question the ?inequities? in the [NPT], but they still appreciate the benefits?both direct and indirect?they get from it.? 225 Although CSCAP subsequently has focused its nuclear nonproliferation efforts on North Korea, nuclear energy transparency, and export controls, the network initially 223 Interview with author, Arlington, VA, December 20, 2004. For subsequent U.S. efforts to raise the profile of nonproliferation in the region, also see Ralph A. Cossa, ?Proliferation and Regional Security,? ch. 7 in Ramesh Thakur, ed. Keeping Proliferation at Bay (Jakarta: Center for Strategic and International Studies, 1998): 107-120. 224 Report on First CSCAP CSBM Working Group (Washington, DC, October 11-13, 1994): 6; Summary of Sixth Meeting of the CSCAP North Pacific Working Group (Ulaanbaator, Mongolia, 14-16 June 2000): 4-5; and Report of Twenty-second meeting, CSCAP CSBM International Working Group (Hanoi, Vietnam, May 25-28, 2004): 1-2. 225 Report of Twenty-second meeting, CSCAP CSBM International Working Group, p. 4. For earlier opposition to the inequity embedded in the NPT, see Summary of Fourth Meeting of the North Pacific Working Group (Beijing, November 8-10, 1998): 2. 96 considered tackling missile defense, particularly after the August 1998 3-stage North Korean missile test but before the September 11, 2001 attacks and subsequent U.S. withdrawal from the ABM treaty. In at least June 2000 and June 2001 sessions of the North Pacific working group, Chinese participants expressed concerns that U.S. plans for national missile defense were actually targeted at China, not at rogue states such as Iran and Iraq as the United States had declared. 226 Citing the absence of government dialogue on the subject, CSCAP considered holding a workshop exclusively on missile defense, 227 although it was not held before the 9/11 attacks. At least one participant believed that CSCAP was reluctant to focus significant effort on missile defense because they believed that major governments, presumably the United States, would not be receptive to the network?s advice on the subject. 228 Cossa replied that the network did not avoid missile defense but put the issue in a broader context about regional missile proliferation until it just stopped being an issue. 229 CSCAP also had to deal, at least tangentially, with the 1998 Indian and Pakistani nuclear tests. Although CSCAP had always been reluctant to bring India or Pakistan into the network for fear of expanding the scope of its activities into South Asia, India and Europe had both joined as associate members in 1994, and seemed to be on track for full membership in 1998. After India?s May 1998 nuclear tests, the CSCAP Steering Committee condemned the nuclear tests at their June 1998 meeting, subsequently leading 226 Summary of Sixth Meeting of the CSCAP North Pacific Working Group, p. 5; and Summary of Eighth Meeting of the CSCAP North Pacific Working Group (Paris, France, June 27-29, 2001): 3. Also see Summary of Discussions, USCSCAP general membership meeting (Washington, DC, May 8, 2001): 4. 227 Summary of Sixth Meeting of the CSCAP North Pacific Working Group, pp. 6, 8; and Summary of Discussions, Sixteenth meeting, CSCAP CSBM International Working Group (Washington, DC, October 29-31, 2001): 4. 228 Simon, ?Evaluating Track II Approaches,? p. 189 229 Electronic message to author, February 28, 2006 97 India not to come to CSCAP meetings for a year. Although Europe was granted full membership in December 1998, India?s was delayed until June 2000. 230 More recently, significant parts of the CSBM working group?s final two meetings, including a May 2004 joint meeting with the CSCAP maritime working group, debated the origins, effectiveness, international legality, and implications for regional security of the proliferation security initiative (PSI), which President Bush had announced in May 2003. 231 CSCAP brought in Lew Dunn, a U.S. nonproliferation expert from SAIC, to give a presentation about the origins and intentions of the PSI. Following the presentation, a North Korean participant strenuously objected to being targeted by U.S. hostility, a South Korean speaker discussed Seoul?s concern that the PSI targeted Pyongyang as well as its own technical ability to implement PSI, and other participants expressed their skepticism about the initiative?s international legality, including whether it was authorized by April 2004?s UN Security Council Resolution 1540. North Korea?s Nuclear Program Beyond their work on nonproliferation generally?including promoting accession to global regimes, missile defense, India?s nuclear tests, and PSI?CSCAP?s work on nuclear nonproliferation until 2004 focused on three areas: North Korea, nuclear energy transparency, and export controls. 230 Summary of Discussions, USCSCAP annual meeting, Washington, DC (May 21, 1998): 2, and author?s interview with Ralph Cossa, Washington, D.C., December 16, 2004. 231 For CSCAP?s work, see Ralph A. Cossa, ed. Countering the Spread of Weapons of Mass Destruction: the Role of the Proliferation Security Initiative, Issues and Insights 4: 5 (Honolulu: Pacific Forum CSIS, July 2004). Also see Report of Twenty-first meeting, CSCAP CSBM International Working Group (Singpaore, December 10-11, 2003): 3; and Report of Twenty-second meeting, CSCAP CSBM International Working Group, pp. 6-7. 98 A nascent CSCAP was not able to meaningfully engage in the initial 1994 North Korean nuclear crisis, when Pyongyang announced its intention to withdraw from the NPT and the Agreed Framework was subsequently reached. North Korea (which joined CSCAP in 1994) and China (1996) were not CSCAP members yet, and the network was dispersing its focus across the spectrum of Asia-Pacific security issues to scope out and determine where it might best add value. Subsequently, the initial work on North Korea was conducted in the North Pacific working group. Cochaired by Canada and Japan, this working group did not hold its second meeting until 1997, after North Korea (which did not attend the inaugural 1995 working group meeting) and China had joined CSCAP. The group was always ?less confident about its capacity to produce policy recommendations,? as its own background statement declared, and ?tasked with promoting the institutionalised dialogue necessary for the development of cooperative policies.? 232 It was the only one of the initial CSCAP working groups never to put forward findings as a memorandum. 233 In 2002, North Korea?s nuclear program returned to the headlines with President Bush calling North Korea part of the ?axis of evil? in his January State of the Union address. Assistant Secretary of State James Kelly also confronted North Korea in early October with evidence of a secret DRPK highly-enriched uranium (HEU) program, violating at least the spirit if not the letter of the earlier Agreed Framework, which the United States subsequently terminated later that month. Although a U.S. administration official expressed frustration that ?everybody and their mother has something to say 232 ?Background,? CSCAP North Pacific Working Group, available at http://www.cscap.org/pacific.htm (accessed October 31, 2004) 233 Evans, ?Assessing the ARF and CSCAP,? p. 164, as well as interview B37 99 about North Korea,? 234 CSCAP?s work shifted from the North Pacific to the CSBM working group in 2003 which began, without North Korean participation, to look into the network?s possible role including: establishing a working group (which was subsequently held) to evaluate progress in each round of the six-party talks; conducting case studies of denuclearization in South Africa or the former Soviet republics to evaluate their potential application to North Korea; recommending ways to sequence the potential incentives to North Korea and their concomitant obligations; or studying and proposing potential verification mechanisms. 235 At a 2004 CSBM working group at Hanoi, North Korean representatives attended. One delivered a presentation, explaining their perception that U.S. hostility had caused the current crisis, with subsequent group discussion including questions from Americans about what Washington could do to convince Pyongyang that it was not ?implacably hostile? to the DPRK. Other discussion focused on how to structure a potential grand bargain, or package deal, to solve the nuclear crisis including a Chinese participant?s suggestion to explore complete, verifiable, irreversible assurances (CVIA) as incentives to parallel U.S. demands for complete, verifiable, irreversible disarmament (CVID) as the formula to solve the nuclear problem. 236 After 2004, when the CSCAP working groups were restructured, work on North Korea continued in a new working group investigating prospects for a multilateral security framework in Northeast Asia. 234 Interview A14 235 Report of Twenty-first meeting, CSCAP CSBM International Working Group, pp. 3, 5 236 Report of Twenty-second meeting, CSCAP CSBM International Working Group, pp. 9-11 100 Nuclear Energy Transparency (PACATOM) Throughout the 1990s, along with their work on preventive diplomacy, nuclear energy transparency dominated the CSCAP CSBM agenda. Out of CSCAP?s CSBM scoping work, nuclear energy had emerged as a regional concern that potentially lent itself to multilateral solutions. Unlike most of the rest of the world at the time, Asian countries including China, India, Japan, Russia, South Korea, and of course North Korea as well as Taiwan were and are heavily investing in nuclear power. Among each country?s neighbors, these investments raised a variety of concerns such as operational safety, the environment, vulnerability to terrorism, and long-term options for the spent fuel that nuclear power produces, particularly if it was plutonium. A common concern was that countries might use spent fuel to build their own nuclear weapons. 237 Funding was available from the U.S. government, first through Sandia National Laboratories and eventually the U.S. Department of Energy, for CSCAP to address these concerns ?without [the U.S. government] trying to direct the answers,? according to Cossa. 238 At the time, the U.S. government was not seized with the problem. Bob Manning, one of the codirectors of U.S. efforts in PACATOM who subsequently went to work in the U.S. State Department, speculated that it was such ?a difficult issue bureaucratically?. It intersects Asian security, proliferation, and a whole range of 237 Roberts and Ross, ?Confidence and Security Building: A USCSCAP Task Force Report,? pp. 149, 155, Kazuko Hamada, CSCAP?s NEEG: Exploring Nuclear Energy Transparency as a Regional Confidence and Security Building Measure, Issues and Insights 3:7 (Honolulu: Pacific Forum/CSIS, September 2003): v, 7. Ralph A. Cossa, PACATOM: Building Confidence and Enhancing Nuclear Transparency (Honolulu, HI: Pacific Forum Occasional Papers, October 1998): 7. For specific concerns about Japan?s plutonium program, see Robert A. Manning, ?PACATOM: A Nuclear Cooperation Regime as Asian CSBM,? ch. 3 in Cossa, ed. Asia Pacific Multilateral Nuclear Safety and Non-Proliferation: Exploring the Possibilities (Honolulu, HI: Pacific Forum Occasional Papers, December 1996): 30-31. 238 Interview with author, Washington, D.C., December 16, 2004 101 technical nuclear issues that sort of complicates it bureaucratically.? 239 Although CSCAP was not the only organization to begin investigating the problem at the time, it did believe it could provide a unique forum with ?a policy-oriented perspective? where ?all current and prospective nuclear energy producers can meet together to discuss security-related concerns in an unofficial but highly-informed setting.? 240 Although managing spent fuel was the ultimate ideal, CSCAP limited its sights at the outset. Simply raising policymakers? and the nuclear energy community?s awareness about regional security concerns was the first step. 241 More ambitious objectives would be sought over time. CSCAP formally defined the PACATOM?s project goals as: to identify and articulate, and then help to address or alleviate, nuclear energy-related regional concerns; to identify and help institute both information collection and dissemination and a series of confidence building measures aimed at reducing current nuclear energy-related concerns while setting the stage for more formalized multilateral cooperation; and to assess the feasibility and define the likely parameters of an institutionalized regional regime aimed at promoting greater safety, security, and transparency in nuclear energy production and research operations. 242 To begin its work, the CSBM working group identified six potential areas of cooperation?safety, energy, research, regional safeguards, the front end of the nuclear fuel cycle, and the back end of the nuclear fuel cycle?and concluded that CSCAP should concentrate its efforts on safety cooperation and the back end of the fuel cycle, or the 239 Interview with author, Washington, D.C., December 20, 2004 240 Cossa, PACATOM, pp. 1-2, 13-14 241 Summary of Discussions, Sixth meeting, CSCAP CSBM Working Group (Washington, DC, May 21-23, 1997): 2; Summary of Discussions, Seventh meeting, CSCAP CSBM Working Group (Fukushima, Japan, October 30-31, 1997): 1; and Cossa, PACATOM, pp. v-vii. 242 Cossa, PACATOM, p. vi 102 nuclear fuels and other waste by-products that nuclear power produces. Of those two, safety was a much less controversial area to begin. 243 The goal, however, was to build confidence among CSCAP?s participants and, more generally, in CSCAP?s efforts to enable it to return to the more controversial and strategically significant back end issues. 244 At the same time, Sandia National Laboratory?s Cooperative Monitoring Center (CMC) had been investigating various technologies that could be used to improve regional transparency and security. According to CMC?s John Olsen, who was working on Northeast Asian issues at the time, he became aware of CSCAP?s work and contacted Ralph Cossa about introducing network members to Sandia?s findings. Although the CMC funded these outreach efforts, Olsen said that Cossa and other CSCAP members advised the Sandia team how to give the presentation in a regionally appropriate way that would not be too pushy. The partnership essentially capitalized on mutual interests in improving nuclear energy transparency by marrying Sandia?s technical work with CSCAP?s political connections. 245 The increasingly technical discussions became a subset of the CSCAP CSBM working group known as the Nuclear Energy Experts Group (NEEG). By March 2000, the group had designed and posted a website (www.cscap.nuctrans.org) to provide a single virtual forum for as much relevant data as they could post, enhancing regional nuclear energy transparency. According to Olsen, the site started by displaying airborne radiation data, following the CSCAP formula to begin incrementally with nuclear safety 243 Cossa, PACATOM, pp. vi-vii. For initial scoping efforts, see Manning, ?PACATOM,? pp. 29-40; and Roberts and Ross, ?Confidence and Security Building: A USCSCAP Task Force Report,? p. 150. 244 Hamada, pp. vi, 2, 27 245 Author?s interview with John Olsen, Washington, D.C., April 1, 2005, as well as B3 103 data that infringed on no country?s national security concerns. In this case, Japan was already displaying some form of this data in local communities, while South Korea and Taiwan were also planning on displaying similar data on national internet systems. CSCAP stepped in to provide a single site combining the data, seeking eventually to add other relevant information, such as on research reactors, to enable interested parties to learn what was really going on in neighboring countries and try to establish a regional habit, or norm, of transparency that others would follow. 246 Within three years, the site contained at least some data on nuclear energy producers and programs from Australia, Canada, China, India, Japan, South Korea, Russia, and the United States as well as Taiwan, including virtual tours of several Japanese and U.S. nuclear facilities. 247 Among the website?s contents, the NEEG also developed and placed a template for an ?Asia Pacific Nuclear Energy Data Book,? which would promote transparency among nuclear energy producers. 248 Resembling CSCAP?s previous initiative to encourage countries to produce defense white papers, these nuclear energy white papers went one step further, as they were designed so that they could also be produced by interested nongovernmental parties including businesses themselves. 249 To this point, however, that is as far as the nuclear energy transparency efforts have gone. Despite continuing efforts to return to spent fuel concerns, such as an internationally monitored retrievable storage system (IMRSS) or a regional spent fuel 246 Author?s interview with John Olsen, Washington, D.C., April 1, 2005. For expansion plans, see also Summary of Discussions, Sixteenth meeting, CSCAP CSBM International Working Group, p. 4. 247 Hamada, p. v; and Summary of Discussions, Seventeenth meeting, CSCAP CSBM International Working Group (Winnipeg, Canada, May 22-25, 2002): 1. 248 Simon, ?Evaluating Track II Approaches,? p. 179 249 Hamada, p. 20; and Summary of Discussions, Thirteenth meeting, CSCAP CSBM International Working Group (Washington, DC, May 22-24, 2000): 2. 104 disposal site, CSCAP has not been able to reach agreement on how to proceed, 250 at least not yet. One obvious problem is that no country wants to volunteer to collect potentially hazardous spent fuel in their country. In early discussions, China had apparently considered offering a potential regional spent fuel disposal site in the Gobi desert, but by 1998 that option was off the table. 251 John Olsen contends that the group was beginning to make progress when Taiwan apparently granted permission to have a low-level waste site monitored. Although low-level waste is not a proliferation concern, Olsen argued that the project would have been ?a toehold into the back end of the fuel cycle,? but its funding was terminated in 2003. 252 One fairly compelling reason why the funding was cut off is that the September 11 attacks heightened concerns about putting any sort of nuclear reactor information on the web, 253 although participants protested that none of the information on the CSCAP site is uniquely sensitive at all. One CSCAP member argued that 9/11 simply gave organizations that were used to being secretive an excuse to withdraw their participation. The same member argued that the post-9/11 environment did lead Chinese, North Korean, Russian, and Vietnamese participants to be denied visas for a 2003 NEEG workshop at Yucca Mountain, Nevada, despite having apparently received approval from even the State Department Korea desk. 254 Other governmental officials familiar with CSCAP, including one previously involved in funding decisions, argued that the NEEG 250 See, for example, Brad Roberts and Zachary Davis, ?Nuclear Cooperation in the Asia-Pacific: A Survey of Proposals,? ch. 1 in Cossa, ed. Asia Pacific Multilateral Nuclear Safety and Non-Proliferation: Exploring the Possibilities (Honolulu, HI: Pacific Forum Occasional Papers, December 1996): 4; and Cossa, PACATOM, p. 15. 251 Summary of Discussions, Eighth meeting, CSCAP CSBM International Working Group (Washington, DC, May 7-9, 1998): 5. 252 Author?s interview with John Olsen, Washington, D.C., April 1, 2005 253 Hamada, p. 20 254 Interview A27 105 was not making enough progress, questions were continuing to be raised about whether regional governments would ever allow CSCAP to step into the back end of the fuel cycle, and nuclear energy transparency simply seemed to have run its course. 255 Interest had shifted to a range of other issues like export controls, halting enrichment and reprocessing overseas, and counterterrorism. 256 Whatever the reason, or combination of reasons, may have been for terminating the funding, two of the original leaders of CSCAP?s PACATOM work independently volunteered that the broader spent fuel and back end of the nuclear fuel cycle concerns are not going away, predicting that the PACATOM concept would be resurrected, although presumably not with as much focus on nuclear safety and reactor transparency. 257 In the meantime, some of the technical experts involved in the NEEG?s safety transparency efforts began to put their expertise to work to try to design a verification scheme for North Korea as a contingency, if political agreement is reached. 258 Export Controls Export controls appeared on the CSCAP agenda as early as the initial 1995 USCSCAP task force?s scoping effort on CSBMs, 259 but the issue did not really gain traction until relatively recently, after the September 11 attacks raised fears of WMD terrorism, the August 2003 public revelations of the extent of the A.Q. Khan network and its links 255 Interviews A20 and B33 256 Interviews B3, B33, and B38 257 Author?s interviews with Bob Manning, Washington, D.C., December 20, 2004; and Brad Roberts, Arlington, VA, December 20, 2004 258 Brad Glosserman, ?A Verification Regime for the Korean Peninsula,? PacNet Newsletter 19 (May 15, 2003), available at http://www.csis.org/pacfor/pac0319.htm (accessed March 9, 2005); and Hamada, pp. vi, 33-35. 259 Roberts and Ross, ?Confidence and Security Building: A USCSCAP Task Force Report,? p. 151 106 through Malaysia, and April 2004?s UN Security Council Resolution 1540 required all states to put export controls in place. 260 In addition, Ed Fei, the former director of DoE?s Nonproliferation and National Security office and original funder of CSCAP?s nonproliferation and nuclear energy transparency efforts, had moved within DoE to work on export controls beginning in 2002. 261 The Department of Energy subsequently has funded CSCAP to look into promoting Asian export controls even though, as one CSCAP staff member put it, ?frankly we didn?t know diddly about it at the time.? 262 Much of the initial work has been consciousness-raising exercises about the value of export controls generally and the specific systems put in place, including presentations by participants from the United States, Europe, India, China, Japan, Singapore, and Thailand about their country?s export control systems. 263 At the May 2004 meeting, CSCAP also brought in experts from the University of Georgia?s Center for International Trade and Security to outline a potential CSCAP action plan, including potentially producing an export control template outlining the elements of an effective regime, similar to the model CSCAP had applied to defense white papers and nuclear energy transparency. Beyond enabling participants to learn from one another?s efforts, the multilateral initiative also sought to ensure that no country would economically benefit by implementing weaker export controls or not implementing them at all. 264 260 Report of Twenty-second meeting, CSCAP CSBM International Working Group, p. 1 261 Author?s interview with Ed Fei, Washington, D.C., July 18, 2005, as well as B33 262 Interview A11 as well as A20 263 Brad Glosserman, ?Export Controls: Fighting Proliferation and Building Confidence,? PacNet Newsletter 2003-35A (August 22, 2003), available at http://www.csis.org/pacfor/pac0335A.htm, accessed on March 9, 2005; Report of Twenty-first meeting, CSCAP CSBM International Working Group, p. 5; and Report of Twenty-second meeting, CSCAP CSBM International Working Group, p. 4 264 Executive Summary and Report of Twenty-second meeting, CSCAP CSBM International Working Group, pp. 1, 4-5 107 Future Efforts After the June 2004 working group restructuring, CSCAP planned to continue and expand ongoing work in the back end of the fuel cycle, export controls, PSI, and the six- party talks on North Korea through its new working groups. 265 CSCAP also planned to consider initiating work on increasing regional accession to the NPT Additional Protocols and on developing an East Asian Action Plan for Nonproliferation, loosely modeled after a similar EU Action Plan, both of which will be discussed further in this chapter. 266 Results Contacts and Perceptions Whether CSCAP participation established valuable new networking connections, or useful new information or perceptions, often depended on the previous experience of the individual participants. The effect it may have had on a think tank researcher is very different from a State Department employee who interacts with foreign officials regularly. For that reason, this section separately reports the experiences of nongovernmental members from government participants, also distinguishing between regionalists who focused on East Asia from experts on functional issues such as nonproliferation. Nongovernmental Contacts Regional experts on East Asia comprise the bulk of CSCAP members, with functional or issue experts brought in on particular topics such as 265 Summary of Discussions, CSCAP Steering Committee, Jakarta, Indonesia, December 6-7, 2003, sent as electronic message to all USCSCAP members, USCSCAP Special Report 1, January 12, 2004; and Report of Twenty-first meeting, CSCAP CSBM International Working Group, p. 5. 266 Executive Summary and Report of Twenty-second meeting, CSCAP CSBM International Working Group, pp. 2-4, 11 108 nonproliferation or very specific issues like export controls. A few experts, either regional or functional, participate in working groups or even international steering committee meetings, but most U.S. members do not attend any meetings or only attend the USCSCAP meetings. Nevertheless, even infrequent participants cited a number of benefits from simply being members or attending those U.S. meetings. For some, membership or USCSCAP participation is a way to get credibility within the community of American experts on East Asia. In separate interviews, a former military officer and a young, rising regional expert said that CSCAP membership and participation helped them be accepted into the community and, in one case, even affirmed in their own mind that they belonged. 267 For the majority of regional experts, however, CSCAP has not established new contacts among Americans but helped maintain old ones. 268 One member said that when they were an academic outside Washington, the USCSCAP meetings were particularly important to get up to speed on U.S. politics and policy toward Asia. 269 But this benefit seemed to be rare because travel funds are not available, and most attendees end up being those already based in Washington. Occasionally, those members said they would get to know someone from government better, or meet a young, rising scholar. Mostly, CSCAP has provided a network to meet periodically with old friends. CSCAP distributes the Pacific Forum?s PacNet weekly newsletter, which was cited as a useful tool to create a virtual network. 270 One cited its utility to enable members from either the United States or Asia to share their findings and impressions of recent 267 Interviews A8 and A9 268 Interviews A2, A3, A5, A7, A12, and A18 269 Interview A5 270 Interviews A6, A12, and B37 109 trips in the region, or their perceptions of recent developments, with the whole PacNet network on a weekly basis without having to be physically in the same room or even city together. The key to a piece?s value, in their mind, was who wrote it, with a small set of well-respected authors being the most worthwhile. 271 Another cited the ability to keep up on other parts of the region, particularly Southeast Asia, which they did not follow as closely and could get from very few other places, such as the Asia Times or the Far Eastern Economic Review. 272 Both these members also cited PacNet?s role in facilitating regional security debates, such as whether President Clinton should have gone to North Korea at the end of his administration, by sparking responses and subsequently printing dissenting pieces for the entire, 3,000-person audience. For functional specialists who do not focus on East Asia, CSCAP?s overall value is different: they do meet new people, mostly the Asian regional experts. One member described it as an opportunity for ?one-stop shopping for getting up to speed on what the currents in the region are, [and] what the concerns in the region are,? 273 while another cited CSCAP uniquely as the only venue they have seen ?that compels the regionalists to come together with the security generalists.? 274 Some generalists simply benefited from meeting American experts on Asia, but others also had the opportunity to meet people from the region itself. For example, after CSCAP?s May 2004 CSBM working group meeting on export controls, one nonproliferation expert said they not only learned about Southeast Asia?s efforts, which they did not know existed, but they began to think of 271 Interview A12 272 Interview A6 273 Interview A16 274 Interview A15 110 Singapore ?taking the lead in helping their neighbors strengthen their export controls?as an interesting model that you could try in other regions?? 275 While some CSCAP benefits are different for regional and functional experts, other values are cited by nongovernmental participants generally, including the network?s role facilitating connections across the Pacific. Both regional and functional members cited CSCAP as the vehicle through which they met many government and pseudo- government, or think tank, officials from Asia. Although it varies from country to country, Asian think tanks typically behave more like the consulting tanks described in chapter two, more similar to an organization like RAND in the United States, which receive significant amounts or even all its money and projects from a part of that country?s government. Many of the host think tanks that make up the 21-member CSCAP network are the national foreign ministry?s think tank, giving CSCAP members some access into Asian governments and their perspectives, at least in their foreign ministries. 276 In particular, members valued CSCAP?s ability to introduce them to Chinese experts. 277 In the early 1990s such contact was rare to nonexistent. As China has engaged the region, CSCAP has provided a base through which China?s security policy experts can reach out to both Asians and Americans, with CSCAP activities actually serving as a catalyst for additional follow-on opportunities. John Olsen from Sandia?s CMC, for example, cited the years of interaction through CSCAP as the reason why a Chinese visiting fellow was subsequently allowed to do a six-month study on counterterrorism 275 Interview A22 276 Interviews A14, A16, and A22 277 Interviews A15, A16, and A27 111 cooperation at CMC. 278 An annual bilateral Sino-US nongovernmental security affairs network, co-hosted by CSCAP director Ralph Cossa and a Chinese scholar, has also been established. In addition, Brad Roberts said that he was able to use his CSCAP connections to invite Chinese scholars to participate in a global security cooperation project among analysts from the five permanent members of the UN Security Council, held at Wilton Park from 1999-2001. 279 Nongovernmental Perceptions and Information Even when CSCAP helped establish some connections, the question remains: why are they valuable? One benefit cited by CSCAP members was that it gave them a forum to explain U.S. policy and its intent to analysts from Asian think tanks. In some cases, such as PSI, this was the explicit purpose of CSCAP?s work: to further explain the U.S. initiative to the region, CSCAP brought Lew Dunn to brief a CSBM working group. 280 In other cases, individual U.S. members took the initiative themselves. One former military officer, for example, felt that after the 2000 Quadrennial Defense Review (QDR) was released ?in Asia and the United States, people had missed some of the key elements of what the QDR was saying in so many words about Asia.?I was essentially in the mode of instructing, or giving my interpretation, of what the administration was thinking about Asia.? 281 A second value cited was that CSCAP enabled members to integrate views from across the region, building a broader picture of Asian security threats and challenges. It enabled members to transcend countries?broadening those who generally attended 278 Interview with author, Washington, D.C., April 1, 2005 279 Interview with author, Arlington, VA, December 20, 2004 280 Interview A16 281 Interview A8 112 bilateral meetings on just Korea, Japan, China, Taiwan, or Southeast Asia?and issues? such as maritime security, nonproliferation, or even more specific issues like the nuclear industry or export controls. Country or issue specialists could therefore use CSCAP to help look outside their issue and put it in a larger context, while regional or global security generalists could use the network to help them see how certain security challenges interrelate and construct those bigger pictures. 282 The third benefit was simply the bits of information derived from attending CSCAP events in Asia for those who were able to travel. To some, the connections enabled members to elicit regional perceptions. At a CSBM working group, for example, a Chinese member gave a presentation about the Shanghai Cooperation Organization (SCO)?a multilateral organization founded by China, Russia, and four Central Asian states?that had been perceived to seek to balance U.S. influence in the region. The presentation sought to explain the history and purpose of the SCO, presumably to assuage some of the U.S. concerns. 283 One member also recalled a discussion with a Chinese nongovernmental participant about North Korea, leading them to conclude that Beijing ultimately might live with a nuclear DPRK. 284 Beyond those few who were able to travel for CSCAP events, members also cited the USCSCAP meetings as useful places to gather information by discussing and comparing notes on recent events, 285 particularly three types: crises, regional domestic political changes such as elections or generational transfers of power, and high-level government visits including summits. Ten USCSCAP meetings were held from 1998, 282 Interviews A1, A9, and A16 283 Report of Twenty-first meeting, CSCAP CSBM International Working Group, p. 3 284 Interview A11 285 Interview A9 113 when discussion panels were first convened in this forum, through the summer of 2005, with each holding one or two discussion panels usually on China or Korea. Table 3.3 shows that all 10 USCSCAP meetings included panels on crises (4 of 10), the foreign policy implications of domestic political changes in the region or the United States (6 of 10), or recent or upcoming summits, high-level meetings, or negotiations (5 of 10). 114 Table 3.3 USCSCAP Meetings Panel Title(s) Crises Political Changes High-Level Visits 1998 China roundtable Korea roundtable Deng died (PRC) 286 Impact of Asian financial crisis 287 ROK elects Kim DJ who announces Sunshine Policy US-PRC Summits: After Jiang to US but before Clinton to PRC 288 1999 US-China Relations: What?s Next? Belgrade embassy bombing 2000 Cross-Strait Relations ROC elects Chen Shui- Bian May 2001 Implications of ?Spy Plane Incident? EP-3 incident Nov 2001 Implications of War on Terrorism on US-Asia Relations Bush APEC Visit Sept. 11 attacks PRC 2002 transition 289 Bush keeps trip to APEC after 9/11 2002 US Asia Policy after the Bush Visit Bush?s Asia Trip: How Successful Was It? (Non-US Views) Bush visit to Asia May 2003 Developments on the Korean Peninsula US-China Relations: Impact of Domestic Developments ROK elects Roh Hu Jintao (PRC) acceded Political impact of SARS in PRC 290 Roh (ROK) to US Oct 2003 Recent Developments on the Korean Peninsula First six-party talks 2004 US-China Relations: Impact of Taiwan and Korea Changes in Japanese Defense and Security Policy ROC re-elects CSB May 2005 Nationalism in Northeast Asia Anti-Japanese riots in PRC 286 Summary of Discussions, USCSCAP annual meeting (Washington, DC, May 21, 1998): 3 287 Summary of Discussions, USCSCAP annual meeting (Washington, DC, May 21, 1998): 3 288 Summary of Discussions, USCSCAP annual meeting (Washington, DC, May 21, 1998): 2 289 Summary of Discussions, USCSCAP general membership meeting (Washington, DC, November 2, 2001): 2 290 Summary of Discussions, USCSCAP general membership meeting (Washington, DC, May 15, 2003): 2 115 Why these discussions may be valuable to members is another story. One reason cited is because the USCSCAP meetings serve as a clearinghouse for members who had recently traveled to the region to share their perceptions and experiences, relaying perspectives from the region to other U.S. members. 291 One expert on China particularly remembered a 1999 USCSCAP discussion where attendants compared their impressions from visiting China after the Belgrade embassy bombing with others who had traveled at about the same time, with experts reaching different conclusions about how people in China perceived the bombing, namely whether it was accidental or deliberate. 292 Others cited the value of the opportunity simply to debate recent events with smart people and learn from them. 293 Although many interviewed seemed to believe that CSCAP provided a uniquely valuable forum for these debates, either because of the intellectual quality of the participants in USCSCAP meetings or the breadth of the network, 294 a few questioned whether these types of exchanges didn?t happen around town all the time, even if CSCAP was valuable. 295 What experts may do with the information varies. In fact, when pressed, most of those interviewed could not spontaneously come up with specific examples of information they gathered or how that information may have changed projects they worked on or their perspectives on issues. Although some did?such as the recollection of the 1998 Belgrade embassy bombing debate or the nonproliferation expert who thought about exporting Singapore?s model of leading regional export controls to other areas of the world?members seemed to believe that the threads of information from 291 Interviews A6 and A7 292 Interview A7 293 Interviews A6, A7, A11, and A14 294 Interviews A1, A11, and A15 295 Interviews A6, A12, and A14 116 regional contacts, peers? recent trips to the region, and thoughtful debates about recent events such as crises, political changes, and summits weaved their way into a professionally useful tapestry for those nongovernmental members. A final, significant value mentioned by U.S. nongovernmental members is the opportunity to be briefed at USCSCAP meetings by U.S. government officials about regional policy and recent events, as well as the opportunity to ask them questions in an off-the-record setting among about 40 experts. 296 Some cited the unique opportunity to be briefed by a State Department official about ARF activities or policy toward Southeast Asia, which gets very little attention elsewhere. 297 One member recalled gaining greater insight into U.S. government perceptions, remembering a State Department official presenting a fairly benign impression of a Chinese-sponsored vice defense minister?s meeting among Asian countries, which had previously raised concerns in Washington because it excluded the United States. 298 Another member simply valued meeting some U.S. officials in person, in at least one case gaining new respect for an official with whom they nevertheless disagreed. 299 Overall, however, subjects valued the opportunity to get beyond talking points, which government officials used with the media, to understand the intent and process behind U.S. policy. USCSCAP essentially provided a halfway house between the bland policy statements for public consumption and the gold mine of information potentially generated from private one-on-one conversations with government officials if they personally knew and trusted individual analysts. 300 296 Interviews A1, A7, A8, A11, and A12 297 Interviews A6, A7, and A9 298 Interview A7 299 Interview A12 300 Interviews A7 as well as A1 117 U.S. Government Participants From its conception, CSCAP viewed including government participants in their private capacities as particularly important, both to increase the influence of network activities and to secure government funding for them. 301 As it turns out, these government participants also valued CSCAP, but for different reasons than their nongovernmental counterparts. For some permanent government employees (in the civil or foreign service), CSCAP introduced them to experts outside the U.S. government, either from the United States or other countries, and even in some cases to foreign diplomats, particularly for U.S. government officials who were not in the foreign service. 302 Within the United States, while one U.S. official cited the value of subsequently being able to call on individual analysts for ?ideas and input,? 303 another conversely cited that USCSCAP was the only opportunity they had to interact with U.S. nongovernmental regional experts. 304 Either way, the contact was perceived to be valuable. Although one might think that criticism from nongovernmental experts might be perceived as annoying or frivolous, some officials valued the sanity check on U.S. policy, soliciting smart opinions and feedback on how it is perceived outside of government. 305 One former deputy assistant secretary said that even hearing the types of questions that U.S. experts were asking at a USCSCAP meeting gave them a sense of what concerns existed outside government. 306 Government officials interviewed also believed that the experience of many USCSCAP members who were former government officials, or part 301 Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), p. 6 302 Interviews A20, A28, B1, and B36 303 Interview A28 304 Interview A20 305 Interviews A17, A18, A20, A22, and A28 306 Interview A18 as well as B1 118 of the revolving door, made them more useful either because they still had security clearances or because they understood what type of information and ideas were relevant to government officials as well as the political constraints they operated within. 307 A second value cited by government officials is that these nongovernmental experts have become informal extensions of the U.S. intelligence and diplomatic services, conveying opinions and feedback from the region to U.S. government officials that those officials cited as useful, in some cases even better than what they received through official government channels. 308 One former deputy assistant secretary explained ?you?re overestimating intelligence?.Actually getting out and having a lot of conversations with people is very valuable, and it?s remarkable how little of it is done in government?.They just don?t do as much of these things as they should do.? 309 In other words, as another former deputy assistant secretary elaborated, CSCAP members provide human intelligence: [A] lot of these people have very good contacts and channels of communication, and are certainly experienced and skilled enough to put analysis into their communication in a way that even people in the intelligence community can?t.?[E]veryone is capable of human intelligence if they have a good conversation with somebody who is actually in the know and is willing to come back and report it in some fashion. Certainly a lot of the participants in CSCAP are able to do that. 310 One nongovernmental member reasoned that CSCAP members can get beyond foreign ministry contacts that the diplomatic corps seem to rely on, going deeper into society to the foreign policy elite at think tanks and universities, the military, influential 307 Interviews A4 and A28 308 Interviews A3, A4, A5, A18, A20, B36, and B301 309 Interview B36 310 Interview A18 119 interest groups, the party in countries like China, or the younger generations that are increasingly shaping foreign policy throughout Asia. 311 Why the foreign service does not have contact with these people is unclear, but one former government official suggested that the ?transaction cost??such as filling out clearance forms and receiving approval before official meetings, and then filing reporting cables after?of formal contacts is so burdensome that officials become ?myopic.? 312 USCSCAP members have stepped into that void, relaying their findings back to U.S. government officials without all the clearance and reporting requirements. One career government official, however, argued that e-mail has helped government officials break down those narrow corridors to other countries, enabling U.S. officials to ?have offline conversations with people in other countries, in other agencies, without going through the embassy.? 313 A second potential value to the intelligence provided is that nongovernmental CSCAP members can also provide a wider regional array of information, integrating the perceptions of countries throughout Asia not just from one U.S. bilateral relationship. One former government official explained that the Department of ?State and embassies are organized to bilaterally approach problems which are increasingly global and regional.? 314 Those in CSCAP that focused on the whole region were cited as being able to see the big picture and integrate, for example, Chinese and Korean perceptions on U.S.-Japanese bilateral issues to U.S. officials focusing on Japan. 315 Beyond providing deeper and wider intelligence, a third value cited is that nongovernmental members can provide blunter, or more honest, perceptions than that 311 Interview A7 312 Interview A5 313 Interview A4 314 Interview A5 315 Interview A4 120 available through official channels. This information might be relayed either directly from the region at CSCAP events or, more often, indirectly through U.S. nongovernmental members. One member of the revolving door said that they would ?get a certain presentation of the facts? when they were in government. ?[P]eople will tell you things that you want to hear or, a better way to put it, avoid telling you things you don?t want to hear?.You?re more likely to hear it when you?re out of government.? 316 In at least one other case, a government official felt that their own participation in CSCAP helped them really understand regional concerns about U.S. policy: ?I mean I had read things and heard things but it was really?you could feel it. It came across very clearly. I had heard it previously through press reports. I had not heard it through official channels. Officially, countries downplay that kind of stuff. You don?t get in people?s faces officially.? 317 U.S. Government Regional Engagement At its most basic level, CSCAP can even provide a forum to demonstrate U.S. engagement in the region. 318 In other words, government officials do not just passively listen at CSCAP meetings, but some have also actively used CSCAP as a forum: to conduct public diplomacy, clarifying and explaining U.S. policy to help shape the regional security agenda; to test run ideas or unofficially develop contingency plans for regional relations; and to help manage relations with North Korea and to a lesser extent Taiwan, neither of which have had official diplomatic relations with the United States. 316 Interview B36 317 Interview A20 as well as B36 318 Interviews A13 and A14 121 Although at least one senior government policymaker felt that official diplomatic interactions made track-two public diplomacy superfluous, 319 others at lower levels felt that government messages could usefully both be expanded to overseas elite foreign policy scholars and even reinforced directly with officials through unofficial contacts. CSCAP allowed U.S. participants to explain existing U.S. policy, such as Lew Dunn?s presentation on PSI. Clearly, any U.S. nongovernmental experts would have to generally agree with U.S. policy to be useful to the U.S. government but, in cases where they do, CSCAP provides an opportunity to clear up misconceptions, such as whether the PSI violates international law, and in some ideal cases even hopefully to get others to agree with it. 320 Although CSCAP has not eliminated initial regional opposition to the PSI, it can be, as Dunn said, ?one more drop of water on the rock? of those perceptions. 321 In South Korea, there is even a circumstantial paper trail that CSCAP might have affected Seoul?s deliberations. Several weeks after the May 2004 CSCAP CSBM working group in Hanoi discussed PSI, a South Korean government-affiliated think tank, IFANS (the Institute for Foreign Affairs and National Security) published a paper encouraging Seoul to join. 322 The author that wrote the paper had attended the CSCAP meeting and, according to one CSCAP staffer, ?there are many parallels? between the CSCAP discussion and the South Korean document. 323 Although this is far from definitive evidence that CSCAP changed the mind of the South Korean author, much less Seoul which has yet to officially fully support PSI, it seems that the CSCAP meeting 319 Interview A17 320 Interviews A18 and A20 321 Interview with author, Arlington, VA, February 2, 2005 322 "Recent Developments in PSI and Key Issues," IFANS Policy Brief No. 3 (Seoul, South Korea: October 2004). 323 Interview A13 as well as A11 122 might have provided at least additional ammunition or an opportunity for the author to preliminarily test his own thinking on the subject. Occasionally, U.S. government officials themselves may even use the forum to brief about U.S. policy. In May 1997, for example, Secretary of Defense Cohen used a CSCAP CSBM working group meeting in Washington with about 100 participants to brief the recently-concluded Quadrennial Defense Review to a foreign audience for the first time. 324 Cossa emphasized that although CSCAP does not necessarily defend U.S. policy all the time, the key is that it does provide a forum to explain it and to elaborate on U.S. policy, dispelling rumors and misconceptions: ?if they don?t like it, it?s at least for the right reasons.? 325 In some cases, government participants even mentioned that CSCAP provided an opportunity to brief its own domestic nongovernmental experts at USCSCAP meetings, particularly on issues like ASEAN or the ARF which do not get as much public attention, 326 although another former government official argued that those opportunities were not as unique with the advent of electronic communications. 327 Government support, however, is primarily geared toward the region itself. As a particularly ambitious form of public diplomacy, one CSCAP staffer explained that the network provides a mechanism to help set the regional security agenda and frame issues for regional consumption: ?What we see as our role at Pacific Forum and what CSCAP is as far as we?re concerned is a vehicle to get these agenda issues on?.How do you get that [issue] to the right people?s attention, and how do you keep their attention once 324 Summary of Discussions, USCSCAP Board of Directors Meeting (Washington, DC, October 22, 1997): 1. 325 Interview with author, Washington, D.C., December 16, 2004 326 Interviews A4 and A17 327 Interview B34 123 you?ve got them?? 328 While at first this may sound sinister to some, that is neither its intent nor its apparent result. The key for CSCAP staff is to see if the network can help frame the issue in a way that truly addresses regional concerns. One example is the basics of nonproliferation itself: while many regional governments and analysts want to talk about nuclear disarmament among the declared nuclear powers, the United States may want to talk about counterproliferation options to deter new proliferating states. The solution: CSCAP provided a regional forum to take a step back and discuss WMD threat perceptions, helping to build a consensus among U.S. and regional foreign policy elite on the threat so that further action can be taken later. While that example is obviously broad, other more specific empirical examples emerged, such as CSCAP?s work on developing a contingency plan if the six-party talks were successful for North Korean verification measures, raising awareness of their difficulties or investigating what it might mean to apply the ?Libyan model? to North Korea. While ideally the role CSCAP plays is to help clarify misperceptions or to build a regional consensus, in some of these other cases the network might simply more precisely define areas of disagreement so that alternate solutions could be found. 329 Although this study did not include interviews with foreign subjects, what is remarkable is that there is almost no evidence that Asian participants dismissed CSCAP as a U.S. tool despite the transparency of U.S. government funding, giving it plenty of potential to be perceived as controlling the network?s agenda. Although one U.S. nongovernmental participant expressed some concerns about whether the region was 328 Interview A11 as well as A20 329 Interview A11 124 actually interested in the U.S. agenda, 330 most did not. In fact, at least one U.S. governmental participant believed that because Cossa was so transparent about the funding source, he helped build trust in the network and in Cossa himself, assuaging any concerns about government cooption while pre-empting the nightmare that the funding source might somehow be revealed later, destroying trust in the network. 331 One CSCAP staffer pointed out that the funding actually also gave participants confidence that network findings were more likely to be influential in the U.S. government, since they had a captive (paying) audience. 332 Another CSCAP staffer said that the fact that the number of Americans attending meetings was not disproportionately larger than other countries, and that those Americans rarely agreed with each other, probably helped assuage concerns that a U.S. agenda was being forced on the region. 333 Another former governmental participant explained that CSCAP was careful to have two cochairs, one from Southeast Asia and from another country, who rotated, helping to reinforce the perception that CSCAP was led by its members, not by Washington. 334 Perhaps most importantly, every indication is that Cossa does actually control the agenda within broad Department of Energy-defined parameters such as nonproliferation or export controls, allowing him to more specifically set the agenda with input from regional participants while avoiding some subjects that are simply too controversial to yield agreement. 335 Finally, at the end of the day, the reality is that CSCAP does not have the capability to brainwash members. In fact, as the previously-mentioned Chinese presentation on the 330 Interview A9 331 Interview A20 332 Interview A11 333 Interview A13 334 Interview A1 335 Interviews A11 and A20 125 Shanghai Cooperation Organization demonstrates, public diplomacy is a two-way street, allowing the regional foreign policy elite from countries like China and South Korea to explain as well as clarify their perceptions and policies to its neighbors and the United States, 336 while providing participants on both sides of the Pacific the opportunity to engage in discussion and debate no matter who funds the network?s operations. Beyond public diplomacy and regional issue framing, CSCAP also provided a forum for U.S. government participants to unofficially discuss and test ideas, as the theory in chapter two predicted. Although no government officials interviewed for this study offered specific examples that they injected into the network, many mentioned that CSCAP allowed them to ?float ideas in a much more informal atmosphere,? 337 ?to test the art of the possible? that higher-level officials ?don?t have the time or interest? to consider, 338 ?to brainstorm,? or as a forum for ?exploring ideas that are not ready for prime time.? 339 At face value, the inclination might be to believe that these officials are referring to previews of dramatic changes in U.S. policy, such as whether to bilaterally engage North Korea or to consider changing the U.S. one-China policy. 340 The reality is much more incremental and might better be described as contingency planning for future political changes, such as the work that CSCAP did on what verification regimes or security assurances for North Korea could specifically look like if general political agreements were reached. One government participant explained that: It relieves the doom and gloom a little bit and allows people to imagine that there could be a solution?.That?s where you have people daring to 336 Interview A20 337 Interview A28 338 Interview A2 339 Interview A20 as well as A28 and B33 340 Interview A18 126 imagine what if we could solve this mess: how would we implement it, what would it look like, who would pay what, how would you verify it, what would be the nature of the security assurance, would it be a treaty, would it be multilateral, would it be bilateral, would it be a UN blessed thing or not, would it involve the IAEA or not. That?s where the concrete visioning is going on: to imagine a brighter future. 341 Finally, some government participants also valued their ability to use CSCAP to manage relations with entities with whom the United States has no official diplomatic relations, particularly North Korea. Although Taiwan is another possibility and one government policymaker even mentioned that they thought they met the man who would eventually become Taiwan?s Foreign Minister at a CSCAP meeting, 342 Taiwanese participation at CSCAP has been limited. By the terms of China?s 1996 agreement to join CSCAP, only a few individuals from Taiwan have been allowed to participate in working groups with subsequent network participation to be decided by all members of the CSCAP Steering Committee, including China. 343 In contrast, North Korean participation is encouraged and sought by CSCAP. Although some government officials had misgivings about CSCAP providing North Korea a public diplomacy forum (a potential drawback which will be discussed later in this chapter), one government official who worked on nonproliferation explained that the tenor of the discussions and the absence of the invective rhetoric that Pyongyang uses in public gave them hope, at least at the time, that North Korea really did simply hold on to nuclear weapons because they perceived a hostile world around them, particularly from 341 Interview A21 342 Interview A18 343 Exchange of letters from SC-CSCAP co-chairs Jusuf Wanandi and Nobuo Matsunaga to Ambassador Li Luye, June 6, 1996; from Ambassador Li Luye to Wanandi and Matsunaga, November 15, 1996; and from Wanandi to all members of the CSCAP Steering Committee, November 21, 1996. Also see Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), pp. 7, 14; and Evans, ?Assessing the ARF and CSCAP,? p. 163, as well as interview B37 127 the United States. 344 Assistant Secretary of State Jim Kelly was more pessimistic. Although he felt it was useful to have these unofficial sessions to help ?bring familiarity, at least at some level, with the issues to the general public,? he also cautioned against expecting too much candor from the North Koreans: [Y]ou really have to realize that there is a serious fatuous nature to track- two diplomacy as pursued by academics who, with incredible naivete, continue to believe that, in the informal sessions, they are likely to get North Korean track-two participants to say ?what they really think.? This simply does not resonate with the reality of what the DPRK is, and the interchangeable nature of Foreign Ministry officials and people from the Institute of Disarmament and Peace [the North Korean think tank] who essentially have two sets of business cards?any participants that got off the track would be subject to terrible things, even worse for example than loss of tenure, such as loss of life and limb and, given that punishments in North Korea tend to be visited on three generations, the punishment would be rendered on children and the parents too. 345 Rather than focusing on whether North Koreans would say something different at these meetings, Cossa relayed that ?State has been appreciative of the opportunity to be able to sit on the sidelines and talk with the North Koreans? at occasional CSCAP meetings without much public furor when there were no formal channels or diplomatic contacts. One particular example was a March 2003 North Pacific working group meeting at Berkeley, held just days before the United States invaded Iraq to oust Saddam Hussein, weeks after South Korea had elected Roh Moo-Hyun as its new president, two months after Pyongyang had withdrawn from the NPT, and five months after Assistant Secretary Kelly had confronted North Korea with U.S. evidence of its highly-enriched uranium program. 346 A member of the North Korean delegation, which included their 344 Interview A20 345 Interview with author, Washington, D.C., January 21, 2005 346 Interviews A13 and B37 128 ambassador to the UN as well as employees of the DPRK?s Institute for Disarmament and Peace, gave a presentation blaming the hostile U.S. policy toward North Korea as the source of the proliferation problem, and called for bilateral talks and a nonaggression treaty with the United States as the way to solve the nuclear crisis. 347 U.S. nongovernmental members who attended the meeting reported that the new government in Seoul sent a higher-level delegation than their predecessors had previously sent, including the new director of policy planning from the South Korean Ministry of Foreign Affairs. He gave a presentation, in his private capacity, declaring for the first time that the North Korean nuclear crisis needed to be addressed in a multilateral forum, not bilaterally between North Korea and the United States as the previous Seoul government had supported. The next morning, the Chinese attendants changed their position to also support the six-party format. According to U.S. members present, the North Koreans were visibly upset by the pressure from the South Koreans and the Chinese. 348 Five months later, the first round of the six-party talks was held. U.S. Government Access and Interaction While CSCAP has been appreciated for its utility as a forum to conduct public diplomacy, to shape the regional agenda, to develop contingency plans for different political circumstances, and to manage relations with North Korea, as well as an 347 See ?North Koreans to Join Berkeley Nuclear Meetings: Unofficial Talks Aim to Jump-Start Communications,? San Francisco Chronicle, March 14, 2003, p. A12 (http://www.sfgate.com/cgi-bin- article.cgi?file=/chronicle/archive/2003/03/14/MN259349.DTL, accessed July 18, 2005); ?North Korean Official Explains Rationale for Nuclear Arms,? San Francisco Chronicle, March 15, 2003, p. A6 (http://www.sfgate.com/cgi-bin/article.cgi?file=/chronicle/archive/2003/03/15/MN155609.DTL, accessed July 18, 2005); and ?North Korea Fears U.S. Attack, Say Delegates at Talks at UC,? San Francisco Chronicle, March 16, 2003, p. A18 (http://www.sfgate.com/cgi- bin/article.cgi?file=/chronicle/archive/2003/03/16/MN252963.DTL, accessed July 18, 2005). 348 Interviews A13 and B37 129 intelligence channel to provide information and feedback to U.S. government participants, government participants including former Assistant Secretary Kelly himself, previously the director of USCSCAP, were much more pessimistic about its ability to influence U.S. policy. 349 With the exception of a mid-level government official (usually NSC Director, Senior Director, Deputy Assistant Secretary or Assistant Secretary of State or Defense) who speaks at USCSCAP lunches and the 1997 QDR address by Secretary of Defense Cohen, actual U.S. government participation in CSCAP activities since 1993 has been limited to a handful of desk officers and office directors. At higher levels, government officials simply seem too busy to participate at all. Bob Manning, a member of the revolving door who is on the policy planning staff in the current Bush administration and the former co-director of CSCAP?s efforts on nuclear energy transparency, said that since he went back into government, he has hasn?t been active in the network because ?it?s hard to get that much time: too many meetings in this town.? 350 Presumably, policymakers at the deputy assistant secretary and assistant secretary levels are at least as busy as the policy planning staff, if not more so. The question then is whether government officials who do not participate in CSCAP meetings receive the intelligence benefits the network might provide. Three channels?the PACNET weekly newsletter, briefings from Ralph Cossa, and revolving door members who participate in CSCAP before entering government?seem to be the answer. Despite the very real sense of information overload among government officials, who ignored the longer (usually 20-page) CSCAP memos, 351 a surprising number of current and former government officials interviewed said they received and read, or at 349 Interview with author, Washington, D.C., January 21, 2005 as well as interviews A2 and A4 350 Interview with author, Washington, D.C., December 20, 2004 351 Interviews A4 and A21 130 least scanned, the weekly one-page PACNET opinion piece. 352 Some said that they looked to see who the author was to decide if they wanted to read the piece, reading those written by American analysts they knew and generally viewing the pieces written by foreign analysts as inferior. 353 Others sought generally overlooked issues not discussed elsewhere 354 or to monitor how policy debates were moving in the nongovernmental community. 355 One Hill staffer specifically said they looked for information on bilateral or multilateral contacts among other countries in the region that did not include the United States, such as between China and Japan or Korea and Japan, that therefore did not make it into U.S. papers. 356 One former government official quite simply said, ?Frankly, I?m not mesmerized by the quality of government intelligence?.if a piece is well-written, sometimes it?s superior to government analysis.? 357 But no government official suggested that anything in PACNET changed U.S. policy, and even Cossa concluded that although recommendations were made through PACNET ?that have ended up coinciding with policies that were taken, whether or not they were driven by it [PACNET recommendations] would be too hard [to prove].? 358 Cossa himself provides a second channel into government. Many subjects interviewed expressed personal affinity and respect for Cossa, his writing, and the advice he gave because of the number of miles he travels, the diversity of issues he monitors, the quality of his analysis, his engaging style, and the regional and U.S. government connections he has developed. To many, he is the reason, or the catalyst, that has enabled 352 Interviews A2, A4, A5, A12, A17, A18, A21, A28, B1, B34, and B36 353 Interviews A3, A5, A12, A18 , B34, and B36 354 Interviews A17 and A18 355 Interview A2 356 Interview A21 357 Interview A5 358 Interview with author, Washington, D.C., December 16, 2004 131 CSCAP to exist, arguably to succeed, and to provide any influence it may have. 359 Yet while Cossa grew CSCAP, the network has also synergistically helped empower him by facilitating those connections and the flow of ideas through its meetings. 360 As Cossa himself said, ?It would be hard to come up with something that I?ve written that didn?t at least in some part draw from what I?ve gotten in these various meetings.? 361 Because they trusted and respected him, being briefed saved those government officials the time and travel of sitting through CSCAP meetings. 362 Although at least one nongovernmental member had concerns about how much power that channeled to the director or the small set of Americans that could participate internationally, 363 Brad Glosserman, a CSCAP staffer, maintained that it was essential for at least one person, as dedicated as Ralph is, to understand the history and full scope of CSCAP activities to avoid duplication of work in an underfunded, nonprofit organization. 364 As a result of its activities, CSCAP in general and Cossa in particular have become a resource to brief U.S. government officials cycling into new positions on East Asian security affairs. 365 Cossa was asked in 2004 to brief the new director of regional affairs in the State Department?s East Asia bureau, who had previously been working on economic issues, and apprised newly-appointed ambassador to South Korea Chris Hill, who subsequently has become the Assistant Secretary of State for East Asia, of some 359 Interviews A1, A4, A5, A7, A11, A12, A18, A20, A27, B3, and B36 360 Interview A7 361 Interview with author, Washington, D.C., December 16, 2004 362 Interview A9 363 Interview A12 364 Interview with author, Washington, D.C., December 14, 2004 365 Interview A16 132 regional issues for which the ambassador later thanked Cossa because he had not heard them in his transitional briefings through government channels. 366 CSCAP?s potential for policy influence was arguably never better than during the first term of the Bush administration when the network?s previous director and Cossa?s former boss, Jim Kelly, served as the Assistant Secretary of State. Kelly?s move through the revolving door from CSCAP into government provides a tailor-made test of the advantages that a personal connection might provide. Although numerous subjects interviewed cited the continuing good relations between the two men and the frequency with which they either assumed or knew that Cossa e-mailed or briefed Kelly, 367 the CSCAP staff downplayed the uniqueness of the connection to Kelly. Both Glosserman and Cossa emphasized that Cossa reached out to many other policymakers?such as deputy assistant secretary of defense Richard Lawless, then deputy assistant secretary of state Randy Schriver, and then policy planning director Mitchell Reiss?in the current administration (which was in its first term at the time of these interviews), and to Kelly?s predecessors, assistant secretaries of state Stanley Roth and Winston Lord, as well as other officials such as deputy secretary of defense Kurt Campbell in the Clinton administration. 368 Cossa did, however, concede that ?Jim [Kelly] and I are very close and we e-mail back and forth a couple times a week. If I?m somewhere and hear something I think he?d be interested in, I pass it along to him and we share information?? 369 On policy influence, however, Kelly began the interview for this study quite directly saying he was 366 Author?s interview with Ralph Cossa, Washington, D.C., December 16, 2004 367 Interviews A6, A7, A8, A20 368 Author?s interviews with Brad Glosserman, Washington, D.C., December 14, 2004, and with Ralph Cossa, Washington, D.C., December 16, 2004 369 Interview with author, Washington, D.C., December 16, 2004 133 ?pretty negative about the effect [on U.S. policy].? He continued that ?the fact is that we?ve seen each other remarkably little. I really don?t have a lot of time and he?s 5,000 miles away and I?m not that good of an e-mail correspondent so I read these things, but?[Ralph?s] speaking much more for himself [to CSCAP audiences], I suspect, than some people think.? 370 Still, that does not necessarily mean that the revolving door connections are useless. For one thing, Kelly?s professional experience presiding over CSCAP and his personal connection to Cossa both may have made him more likely to read PACNET, which he claims he did, and to listen to briefings from CSCAP members, particularly Cossa, for information and feedback more generally, even if policies were not adopted because they were recommended by CSCAP. The effect, therefore, may exist but be much more marginal than a direct policy recommendation, or the policy equivalent of the magic bullet. Second, that previous connection between Cossa and Kelly may have led unclassified information to travel more readily from government, or Kelly, through Cossa to CSCAP members. As one nongovernmental CSCAP member speculated: What I tend to see is looking the other way. If I read something that Ralph has written on PACNET or something, you can always see the understanding of the situation, like what?s going on in the six-party talks is coming from Jim or somebody who works for Jim. He?s getting knowledge from them and putting that into what he?s writing, but it?s a big black box to me as to how much impact what he?s writing or sending off to the government is being accepted or used in any way. 371 Kelly himself also pointed to a third advantage of the revolving door and his role in CSCAP prior to government service: 370 Interview with author, Washington, D.C., January 21, 2005 371 Interview A7 134 It enabled me to be a lot more familiar with the issues. My presence at those sessions was of considerable benefit to me and I?m struck now, as I finish up four years, that?I don?t think I could have possibly been more familiar with the issues. And I think I was more familiar with the range of issues than any of my foreign service colleagues who were here. Just as in a lot of other things, they really know a great deal about the particular countries they?re involved in, but sometimes the interactions with other countries are way down there. The CSCAP process, even aside from working groups, bring a broad familiarity with what people are interested in, not interested in, and worried about throughout the region. At least for me, it was great preparation. 372 Although an unhealthy obsession to separate interrelated, synergistic forces into dependent and independent variables can sometimes exist, it does seem that CSCAP?s ability to get information and feedback to government officials and policymakers may be more dependent on how much those officials want to turn to the network, based on their personal experiences and preconceived notions, than not. In other words, CSCAP seems dependent on government to open its doors to the network?s input. While Cossa and CSCAP appear to have been able to prove their value to some previous administrations and officials, some members and analysts have suggested that the network?s influence in government may vary on the individual official?including their willingness to listen to outsiders in general or their perceptions of Cossa or CSCAP in particular, 373 whether cooperative processes are perceived to be able to address the challenges and threats that dominate an administration?s attention, 374 and an administration?s general attitude toward multilateralism in Asia. 375 372 Interview with author, Washington, D.C., January 21, 2005 373 Interviews A7, A12, and A16 374 Ball, ?CSCAP: Its Future Place in the Regional Security Architecture,? p. 311 375 Interviews A8, A16, and A27 as well as Simon, ?The ASEAN Regional Forum Views the Council for Security Cooperation in the Asia Pacific,? p. 8 135 Agenda Setting in the United States Glosserman described the CSCAP method as a continual process of agenda setting, option development, and evaluation, similar to the framework that John Kingdon proposed for government policymaking in general. For example, Glosserman said, ?once you get the discussion on a) the need for transparency, and b) how you do it, and then c) now that you?ve done it, let?s evaluate that process. For each of those [issues], you can continually move this forward.? 376 As discussed before, that process seems to be about setting issues like nonproliferation, nuclear energy transparency, and export controls on the regional agenda by the United States government, or by CSCAP?s nongovernmental members, but not in the United States. There are a couple of potential exceptions. One government official claimed that, through their participation in CSCAP, a Chinese foreign ministry employee expressed interest in developing a bilateral export control agenda: ?We had an inkling therefore they wanted to do something on export controls and I was able to follow up on that.? 377 PACATOM is another potential exception. Because the issue did not neatly fit into any one U.S. government portfolio, as the project?s co-director Bob Manning pointed out, CSCAP?s work not only raised the issue within the region but, as it developed, Cossa and project co-director Brad Roberts both felt that the United States also advanced its own nuclear energy transparency: ?Actually, when we started doing this transparency stuff, we started discovering that the U.S. wasn?t all that transparent either. We were able to get some of the labs to start providing more radiation data onto the CSCAP website just based on the argument that we?ve got the Japanese, and the Koreans, and the Taiwanese, 376 Interview with author, Washington, D.C., December 14, 2004 377 Interview A20 136 and even the Chinese considering it, and we need to have the same example.? 378 Similarly, Roberts contended that, ?I think the nuclear-related work didn?t change a policy but it created an opportunity to formulate a policy.? 379 There is little other evidence in this study that CSCAP successfully engaged in, or even facilitated, agenda-setting in the U.S. government in any of its nonproliferation, missile defense, North Korea, nuclear energy transparency, PSI, or export control efforts. Although it is possible to contend that CSCAP feedback from the region served an agenda-setting role in the United States by drawing policymakers? attention to regional reactions that might need to be addressed, no specific example was raised in this research. One nonproliferation expert argued that even within the region, CSCAP played more of a regional agenda setting role ten years ago, when the network first sought to raise the regional commitment to nonproliferation in Asia during the campaign to indefinitely extend the NPT in 1995. Since then, they argued that CSCAP activities, such as nuclear energy transparency and export controls, had become increasingly technical. 380 Ultimately, a case can be made that CSCAP?s connections to, and influence in, government have led the network to behave more like a consulting tank than a traditional public think tank. This is not entirely surprising as analysts, including Cossa himself, have previously written about the potential trade-off between gaining influence with government and publicly criticizing them. 381 One possible explanation lies in the experience of CSCAP members. Those with government experience may be more hesitant than career academics or other nongovernmental analysts to publicly criticize 378 Author?s interview with Ralph Cossa, Washington, D.C., December 16, 2004 379 Interview with author, Washington, D.C., December 20, 2004 380 Interview A15 381 Ball, ?CSCAP: Its Future Place in the Regional Security Architecture,? p. 309; and Ralph Cossa, ?Track Two Diplomacy: Promoting Regional Peace, Stability,? U.S. Foreign Policy Agenda (January 1998). 137 government policy because they either understand the constraints on government policymaking or because they are more pessimistic about the media?s potential effect. Another possibility is that the two roles trade off with one another, leaving think tanks and transnational networks with a potentially difficult balancing act to manage: between advising the government and publicly criticizing it. In CSCAP?s case, one nongovernmental member speculated that Cossa?s close relationship with Kelly may have come at the expense of the network publicly pushing the U.S. government?and either directly or by implication criticizing its chief regional policymaker, Kelly. 382 But more fundamentally, CSCAP as an institution has never used the media to draw attention to overlooked issues, the primary tool of agenda-setting public think tanks. Cossa said that is a conscious decision: Our view has generally been that everything we do in CSCAP is off-the- record. We don?t have public meetings; we don?t have press conferences at the meeting. We prefer to do things quietly behind the scenes. Many of us in CSCAP write op-eds and things like that, but I?m not representing CSCAP when I write an op-ed, although I may, in the course of the op-ed, talk about conclusions that I?ve drawn based on conversations there. 383 Some have criticized this habit, arguing that the lack of publicity hurts CSCAP?s influence. 384 Ultimately, the reality is that CSCAP has not really tried to set an agenda in the U.S. government. Instead, Cossa and the CSCAP staff have served more as brokers between network members, both in the region and the United States, and the U.S. government to explore areas of mutual interest. 382 Interview A6 383 Interview with author, Washington, D.C., December 16, 2004 384 Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), p. 87 as well as interviews A6 and A8 138 Policy Options Even though CSCAP?s members include many former government officials, the network has produced very few policy recommendations (although it may have shaped the recommendations that some individuals made on their own), with even fewer directed at the U.S. government. Some general exceptions exist, such as the preventive diplomacy recommendations to the ARF. In other cases, as one government member recalled, CSCAP participants may have conveyed specific suggestions from the region, such as trying to get the United States to be more involved in developing a code of conduct on the South China Sea. 385 Even in this case, however, it was an individual or small group in their capacity as CSCAP participants approaching the government attendants, rather than the entire network as an institution seeking to get government to change its policies. Another possible example was the May 2004 effort by a Chinese analyst to link the U.S. CVID formula with similar security assurances to North Korea. At least one U.S. nongovernmental nonproliferation expert interviewed for this study remembered that statement as an interesting and innovative proposal. Over the subsequent months, Washington even appeared to publicly drop its emphasis on the CVID formula. The September 2005 joint statement released after the fourth round of the six-party talks also juxtaposed ?the goal of the Six-Party Talks is the verifiable denuclarization of the Korean peninsula in a peaceful manner? with a security assurance that ?the United States affirmed it has no nuclear weapons on the Korean Peninsula and has no intention to 385 Interview A28 139 attack or invade the DPRK with nuclear or conventional weapons.? 386 Although the CSCAP discussion may have had some effect, and little incentive exists for government officials to acknowledge any, it does not appear that this was the only forum through which such a linkage was being discussed. Kelly stated that similar messages were also being relayed through official channels, 387 and even Cossa admitted that although these points were discussed in CSCAP meetings ?it is hard to tie any specific conversation or meeting to a point in the statement.? 388 At a minimum, though, it seems that CSCAP was a forum through which the idea was first broadcast to some nongovernmental members for their consideration and in turn might shape their thinking as well as elite and public U.S. debates. The 2003 North Pacific working group meeting at Berkeley is another example of potential policy impact, although in this case the South Korean government apparently used CSCAP as a forum to encourage the six-party format. Their actions may have provoked China also to shift its position to support six-party talks, even if CSCAP as a network recommended neither. In either case, the policy options developed or changes made were outside of the United States. Finally, other working groups have sought contingency plans, such as the types of verification provisions for North Korea, but those are plans for the future if the political conditions are right, rather than recommendations for the present. Meetings such as USCSCAP have not even tried to come up with U.S. policy prescriptions, focusing instead on ?analysis or diagnosing problems rather than coming 386 Joint Statement of the Fourth Round of the Six-Party Talks,? Beijing, September 19, 2005, http://usinfo.state.gov/eap/ARchive/2005/Sep/19-210095.html (accessed February 16, 2006) 387 Telephone interview with author, March 1, 2006 388 Electronic message to author, February 28, 2006 140 up with solutions.? 389 Even some of the international working groups, including the North Pacific working group, 390 have not produced policy recommendations, relying instead on developing a better understanding of the perceptions of other countries that attended the meetings, and enabling individual members to subsequently develop their own ideas based on that information, as Cossa himself cited. It is beyond the scope of this study, however, to attempt to track the op-eds published by individual members and attempt to trace what individual ideas developed from what CSCAP meetings, if any. There are simply too many members with too many ideas and op-eds. When pressed during interviews, most could not cite specific examples on the spot to help limit the search. Rather, they emphasized the value CSCAP has contributed to them individually in broad terms, as drops of information within a broader sea of their knowledge. Clearly, CSCAP tries to develop recommendations as an institution, or as individuals if that fails, that are useful for the government. It is a source of pride that Cossa cites and part of the network?s utility to the government, helping the network to secure funding, influence, and access to government officials. As one government participant emphasized, the 2004 restructuring of CSCAP networks has this goal in mind: to enhance its relevance to governments, at least in the region if not Washington. 391 But ultimately, the policy options developed are much more likely to be promoted by individual members in their full-time jobs, as employees of whatever think tank or institution pays their salary, or to evolve from CSCAP activities, such as the general support for six-party talks, than to get consensus recommendations from CSCAP as an organization. Efforts, such as attempts to get members from so many different countries 389 Interviews A6 as well as A7 390 Interviews A9 and A20 391 Interview A20 141 to agree on how to structure security assurances to North Korea and how to sequence those assurances with DPRK commitments to dismantle their nuclear program, have not reached agreement and are unlikely to do so. In fact, as other analysts have previously noted, part of the strength of CSCAP as an information-sharing network is that it has such a diverse membership of ?insiders and outsiders, generalists and specialists, those close to government and those more independent in their orientation? 392 that ?there are more differences than similarities among the participants,? 393 maximizing the potential for individuals to learn something new but usually reducing recommendations such as CSCAP memoranda to become more bland. Clearly the network strives to overcome these impediments, but it is an uphill battle. In the end, as Arizona State professor Sheldon Simon previously concluded, ?CSCAP dialogues may be less important for the recommendations they generate?than for the suspicions they allay and the norms they reinforce.? 394 Observations Beyond the structured part of this study, a number of other issues related to the network?s operations emerged during interviews. Rather than ignore them entirely or leave them unreported, they will be briefly mentioned here for others who may examine different aspects of the network?s utility. Although many of these issues have already been touched upon earlier, six will be discussed further here: the network?s potential drawbacks, its U.S. membership, effects during a crisis, the focus on regional multilateralism, potential 392 Evans, ?Assessing the ARF and CSCAP,? p. 164 393 Simon, ?The ASEAN Regional Forum Views the Council for Security Cooperation in the Asia Pacific,? p. 14 394 Simon, ?Evaluating Track II Approaches,? p. 175 142 socialization of Chinese and North Korean participants over time, and the possible impact on other countries? policies. Drawbacks? Although every subject interviewed for this case study was directly asked whether CSCAP was harmful in any way, not one came up with any examples or even suspicions of past problems. That being said, a few subjects raised some theoretical concerns to watch. One that no one expressed, but one nongovernmental member pre-emptively addressed, is that U.S. analysts might be duped by propaganda, presumably from North Korean or Chinese participants, such as the justification and explanation for the Shanghai Cooperation Organization (SCO). That argument, however, assumes that U.S. participants are pretty malleable, not the critical experts that this member had encountered. 395 A second potential concern is whether U.S. nongovernmental members might be perceived as representing official U.S. views when they actually were not. In some cases, CSCAP invited an expert to give a presentation on the PSI or export controls; in other cases, members took the initiative themselves such as the former military officer who wanted to explain the regional implications of the 2001 QDR. One U.S. government official raised this hypothetically, 396 but since foreign subjects were not interviewed, this study did not test whether signals were misperceived. Although even government officials attend these meetings in their private capacity, making it clear they are speaking as individuals not for governments, misperceptions are always possible. Presumably, the 395 Interview A15 396 Interview A2 143 burden is on all U.S. members, governmental or nongovernmental, to be clear that they express their own views not those of the U.S. government (except when they directly cite government statements). In one case, a former U.S. military officer went even further to guard against potential misperceptions claiming that: ?I go there mindful that most of the other people around the table are official sort of representatives of their government. I may not be particularly happy of some of our government?s views but I?m not going to highlight that. Generally I try to be in sync with where our government is.? 397 A third, related but more nuanced, potential concern is whether USCSCAP members might be perceived to be floating a trial balloon for the U.S. government when they are actually not. As previously mentioned, former Assistant Secretary Kelly himself suspected that others might have guessed that Cossa spoke for Kelly much more often than he did. But when asked during the interview for this study if that concerned him, Kelly responded that: ?Frankly, I hadn?t thought about it that much.? 398 A former deputy assistant secretary went one step further, recalling that ?I can?t produce a specific instance but it?s not just hypothetical. I know that we have faced that a couple of times? with China. Still, that official concluded that the networks were ?not harmful,? and although you had to ?be mindful? of the potential for foreign participants to think they see a trial balloon where one does not exist, ?the value of having these types of contacts outweighs the potential downside.? 399 One nongovernmental member wondered how likely such a misperception really was. For one thing, they argued that the fear was more likely with interlocutors who had limited contacts with the United States, which was true of very few CSCAP members. They also argued that the actors the United States would 397 Interview A16 398 Interview with author, Washington, D.C., January 21, 2005 399 Interview A18 144 be most worried about were very tightly controlled autocracies, who because of their own system, ?would never see a policymaker deigning to float some precious idea out through some venue where they?re not in control.? 400 A fourth and the most commonly voiced concern is that CSCAP and similar venues might provide North Korea with an outlet when the U.S. government was seeking to isolate and pressure Pyongyang. In other words, the very role that CSCAP could serve?to continue dialogue with North Koreans?might be counterproductive. As mentioned earlier, the State Department had previously asked that North Korea not be invited to a May 2003 CSCAP nuclear energy experts group meeting held at Yucca Mountain in Nevada. After the October 2002 diplomatic confrontation over North Korea?s HEU program and subsequent U.S. withdrawal from the 1994 Agreed Framework, State did not want North Korea to get the impression that the United States supported North Korea receiving nuclear energy assistance. 401 Whereas the theory says that the value of track-two diplomacy is to provide an opportunity to reach out to a country like North Korea with whom Washington has no official diplomatic contact, four separate government officials interviewed for this study said that either inviting the North Koreans or sending U.S. government officials to sit on the margins of unofficial meetings when North Koreans were present was, at some point, counterproductive. 402 Given that the bulk of CSCAP interviews for this study were conducted 4-6 months after North Korea had suspended its participation after three rounds of the six- party talks, U.S. frustration was clearly high. Compounding that frustration, Selig Harrison, a senior scholar at the Woodrow Wilson Center who had independent contacts 400 Interview A15 401 Author?s interview with Ralph Cossa, Washington, D.C., December 16, 2004 402 Interviews A2, A3, A14, and A17 145 with the North Korean government, had published a controversial article right around the time of the interviews in the January-February 2005 edition of Foreign Affairs, accusing the United States of having distorted information on North Korea?s uranium enrichment program to exaggerate the threat. One government analyst complained that ?they go to Sig Harrison and all these people and say one thing, and then these guys [like Harrison] go out [and help the North Koreans].? 403 Although Harrison had not met the North Koreans through CSCAP, the U.S. concern was that North Korea would use track-two as a propaganda tool. Cossa felt that North Koreans sought to do just that, since they ?have always seen [CSCAP] as an opportunity to get their views out and to get feedback about whether anyone is buying a particular story or a particular angle of attack. I think that?s been one of the real values. It?s continued to keep the North Koreans engaged. Even when they don?t go to six-party talks, they still go to CSCAP meetings.? 404 At least two of the government officials 405 specifically did not have a problem with Cossa?s efforts, which they speculated had been used to pressure North Korea and make it clear that their approach was not supported in the region such as the March 2003 Berkeley meeting which increased support for the six-party talks. Nevertheless, even at those meetings, the concern was that other members attending the meeting might confuse the message and 403 Interview A14. For the article, see Selig S. Harrison, ?Did North Korea Cheat?? Foreign Affairs (January/February 2005), available at http://www.foreignaffairs.org/20050101faessay84109/selig-s- harrison/did-north-korea-cheat.html (accessed July 27, 2005) and the bipartisan response it provoked from former Bush administration policy planning director and former Clinton administration lead negotiator of the 1994 Agreed Framework with North Korea, respectively, Mitchell B. Reiss and Robert L. Gallucci, ?Red-Handed,? Foreign Affairs (March/April 2005), available at http://www.foreignaffairs.org/20050301faresponse84214/mitchell-b-reiss-robert-gallucci/red-handed.html (accessed July 27, 2005). 404 Interview with author, Washington, D.C., December 16, 2004 405 Interviews A2 and A14 146 loosen the pressure on North Korea at a critical time. 406 As it turns out, participants at the Berkeley meetings reportedly sought to assure North Korean participants that Iraq, which was on the verge of having its regime changed by U.S. military force, was different from North Korea. Some delegates apparently even reminded the North Koreans that, as the San Francisco Chronicle reported, their ?ability to strike South Korea and Japan, two U.S. allies in the region, poses a great deterrent to pre-emptive American military action.? 407 Presumably, those are exactly the types of messages that the U.S. government, particularly at the Pentagon, did not want sent. In essence, what some view as a valuable outlet to help bring North Korea back to negotiations is, at least at certain times, viewed as a way for North Korea to undermine U.S. isolation and pressure. Put another way: one analyst?s safety valve is another?s leak. U.S. Membership Although about 3,000 people receive the PACNET newsletter, CSCAP meetings other than the biannual conference are limited to about 35-40 people to enhance interaction among members. Getting the right participants at those meetings is therefore important to the network?s value. Chapter two highlighted the significance in the U.S. system for think tanks to use the revolving door, which CSCAP clearly did; the media, which CSCAP intentionally does not; and the Hill. When asked, CSCAP staff acknowledged that Hill participation was rare but they were not particularly concerned. 408 Although theory says 406 Interview A2 407 ?North Korea Fears U.S. Attack, Say Delegates at Talks at UC,? San Francisco Chronicle, March 16, 2003, p. A18 (http://www.sfgate.com/cgi- bin/article.cgi?file=/chronicle/archive/2003/03/16/MN252963.DTL, accessed July 18, 2005) 408 Author?s interviews with Brad Glosserman, Washington, D.C., December 14, 2004, and with Ralph Cossa, Washington, D.C., December 16, 2004 147 they should be, one Congressional staff member interviewed for this study agreed with the CSCAP staff, emphasizing that ?Congress is not operational in the management of foreign affairs. We set broad policy guidelines?? 409 Because CSCAP has focused more on regional agenda setting, not in the United States, and because it has gotten increasingly more technical into issues like nuclear energy transparency and export controls, the value of Hill participation actually seems more limited for CSCAP, at least to this point. If the network were to become more active in setting the U.S. foreign policy agenda, Hill participation would theoretically be more important at that time. A separate membership issue that did concern at least two members is whether CSCAP is getting enough young members and whether younger, rising experts even knew about CSCAP, an issue which the network will have to address over time. 410 What Happens in a Crisis? Although the theories about epistemic communities and policy windows discussed in chapter one predicted a heightened role in a crisis for these networks, the reality is that CSCAP does not try to predict crises, and no subjects interviewed for this study expressed any confidence that CSCAP could or should play such a role. Those with government experience discouraged it, arguing that crises were the wrong time to attempt to get involved. 411 Although one former government policymaker stated that questions they were asked at a USCSCAP meeting focused their attention on a potential approaching crisis over a proposed referendum associated with Taiwan?s spring 2004 409 Interview A21 410 Interview A11 and B36 411 Interviews A4 and A5. Also see Simon, ?Evaluating Track II Approaches,? pp. 191-2. 148 elections, 412 other CSCAP experiences are less promising. For example, over half the experts present at an October 2003 USCSCAP meeting predicted that North Korea would test a nuclear weapon before the 2004 U.S. elections, which obviously turned out to be wrong. 413 For now, it will just be noted that crises focused discussion within USCSCAP meetings and sometimes stimulated network activities, such as the working group on terrorism after the 9/11 attacks, but they were not predicted by the network. The concluding chapter will contain further discussion on the role of networks during crises, after examining NEACD and PONARS. Limits of Multilateralism CSCAP?s founding purpose has been to foster multilateral approaches to Asian security issues?an approach which may ironically impede its ability to influence policy. Fundamentally, trying to coordinate the perceptions, attitudes, and recommendations of 21 different people, much less from different countries, is complicated. Particularly when dealing with Northeast Asian issues such as North Korea, a format that automatically includes so many countries from Southeast Asia provides little apparent value. Added to that challenge is the culture of ASEAN, embodied in the so-called ASEAN way which emphasizes an opaque, unstructured decisonmaking process that emphasizes noninterference in the internal affairs of other countries and requires unanimity to take action. 414 Regardless of other merits of that approach, it makes decisionmaking difficult. In addition to these macro-structural challenges, many countries are represented at a 412 Interview A18 413 Summary of Discussions, USCSCAP meeting (Washington, DC, October 24, 2003), sent as electronic message to all USCSCAP members, USCSCAP Special Report #10, November 7, 2003 414 Wanandi, p. 281, and Evans, ?Assessing the ARF and CSCAP,? p. 158 149 micro-level by nascent think tanks that at least one U.S. government official characterized as not experienced or helpful. 415 Taken together, CSCAP?s charge to seek multilateral solutions to Asian security problems has been daunting. Set against this background, CSCAP has struggled with its relationship with its official counterpart, the ASEAN Regional Forum (ARF). 416 At times, especially during its early existence, CSCAP tailored all its work to seek to support the ARF, starting with CSCAP structuring three of its four initial working groups on subjects of interest to the official forum, and submitting its first memo to the inaugural ARF meeting in July 1994. 417 There is some evidence that CSCAP has been sporadically successful in its efforts, such as the definition of preventive diplomacy and work on transnational crime. 418 Among U.S. members, some support even exists for CSCAP?s efforts to influence the ARF, 419 with one government official going so far as saying that CSCAP helped sustain existential support for the ARF. 420 Yet at various times, especially more recently, CSCAP has debated whether it should follow ARF?s agenda, try to lead it, or even how important it is for CSCAP to influence the ARF at all. 421 For many reasons, including the ARF?s inability to implement the preventive diplomacy work that CSCAP provided and a general regional decline in confidence in multilateral institutions after 415 Interview A4 416 For a review of the connections between CSCAP and the ARF, see Simon, ?The ASEAN Regional Forum Views the Council for Security Cooperation in the Asia Pacific,? pp. 11-20. 417 Ball, ?CSCAP: Its Future Place in the Regional Security Architecture,? pp. 289, 310. 418 Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), p. 49, and Simon, ?Evaluating Track II Approaches,? pp. 193, 198 (fn43), as well as interviews A6 and A14. Also see Summary of Discussions, USCSCAP general membership/board of directors meeting (Washington, DC, June 22, 2000): 1. 419 Interviews A4, A15, and A16 420 Interview B1 421 Simon, ?The ASEAN Regional Forum Views the Council for Security Cooperation in the Asia Pacific,? p. 13 as well as interview A13. 150 their perceived helplessness in response to the Asian financial crisis in 1997-99, 422 the ARF is frequently viewed as an ineffective institution. 423 If that perception continues to exist, there may be a danger for CSCAP associating too closely with it. Then-University of Northern British Columbia Associate Professor Lawrence Woods, in his research on other nongovernmental efforts to build Asian regional institutions, warned that NGOs might be dragged down or even scapegoated by the declining reputation of an official body it sought to influence. 424 That being the case, maintaining an informal linkage 425 that has persisted between CSCAP and the ARF may turn out to be to CSCAP?s benefit, enabling it to continue to seek to influence the ARF, insulate itself from ARF?s potential ineffectiveness, and free it to pursue the network?s initiatives through other channels. Socializing China and North Korea One of the other benefits ascribed to organizations like CSCAP is that they may have been able to socialize China during the mid-late 1990s and now, hopefully, be able to lead North Korea to follow in their neighbor?s footsteps. To socialize simply means that by involving participants in these networks from initially reluctant countries, they may become familiar both with the interests of other countries and with the practice of working with them in a multilateral framework, changing their own interests and 422 Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), p. 45 423 Interviews A1, A8, A9, A14, and B1 as well as Report of Twenty-first meeting, CSCAP CSBM International Working Group, p. 1 424 Lawrence T. Woods, ?Learning from NGO Proponents of Asia-Pacific Regionalism: Success and its Lessons,? Asian Survey 35:9 (September 1995): 816. 425 See letter from Saroj Chavanaviraj, Permanent Secretary, Ministry of Foreign Affairs, Bangkok, to CSCAP Co-Chairs, 28 January 2000, cited in Desmond Ball, The Council for Security Cooperation in the Asia Pacific (CSCAP), p. 52. 151 practices in the process. 426 In some particularly difficult bilateral relationships?such as the United States and North Korea, or South and North Korea, or China and Taiwan?the unofficial status of networks like CSCAP allows a unique opportunity for both nongovernmental and governmental participants in their private capacities to interact and exchange views when a country or political entity is not recognized. 427 A problem with socialization is that it is extremely difficult to prove what increment one network or organization may provide, if any. This study also did not interview foreign participants or attempt to observe their individual participation over time, although some U.S. participants had their own observations about the potential socialization of Chinese and/or North Korean participants. Although this study focuses on their effects in U.S. policy, others have asked the question: Do networks really socialize? 428 Socialization might arguably have more potential with governments like China and North Korea because the linkages between their think tanks and governments are so tight. 429 These think tanks are essentially tied to a particular ministry, such as the Foreign Ministry or, in China?s case, the State Council, and provide advice to the government directly. They effectively serve as a consulting tank, as defined in chapter two, rather than a traditional public or advocacy think tank. As a result, CSCAP experiences may feed more directly into their government deliberations. One U.S. member who participated in CSCAP international working groups over some period of time noted that nongovernmental participants from those countries tended to stay on, providing a base of 426 Wanandi, p. 284 427 Interview A11 428 For an excellent example of one effort to study socialization in general, see Alastair Iain Johnston, ?Socialization in International Institutions: the ASEAN Way and International Relations Theory,? ch. 3 in G. John Ikenberry and Michael Mastanduno, eds. International Relations Theory and the Asia-Pacific (New York: Columbia University Press, 2003): 107-162. 429 Simon, ?Evaluating Track II Approaches,? p. 186 152 familiarity and linkages among nongovernmental members even if the government officials in both countries changed jobs. 430 In CSCAP?s case, although interaction between China and Taiwan has existed, it is limited by the conditions of China?s accession. China?s broader interaction with the region has been another matter. U.S. nongovernmental and governmental participants mentioned during interviews that, in their experience, the nature of Chinese participation was both much more open at these unofficial sessions than government negotiations, and had rapidly increased and become even more open in less than a decade, 431 to the point where Chinese participants in CSCAP publicly disagreed with one another over at least CSCAP matters, such as the workability of a multilateral spent fuel regime. 432 Although the ability to prove that CSCAP participation made any difference in Chinese regional security policies or even its attitudes toward multilateralism remains beyond this study, the apparent Chinese use of CSCAP over the network?s first twelve years yields some potential insight into Beijing?s perception of the network?s value. One U.S. participant felt that China?s interaction with CSCAP went through three phases. First, China wanted to coopt the organization to make sure that it did not become a forum to criticize Beijing, and sent a set of experienced diplomats as its representatives. Second, in the late 1990s, that set of members began to use CSCAP to help construct a cooperative agenda, including trying to flesh out exactly what a strategic partnership with the United States would look like. Finally, and most recently, as those initial members have gotten older, China may have sought to reinvigorate its network participation with a 430 Interview A9 431 Interviews A8, A11, A27, and A28. See also Evans, ?Assessing the ARF and CSCAP,? pp. 167-170. 432 Summary of Discussions, USCSCAP annual meeting (Washington, DC, May 21, 1998): 1 153 younger generation of analysts. 433 One other frequent U.S. participant was not as generous about China?s view of CSCAP, arguing that Beijing may have been using CSCAP as a retirement pasture to place its aging diplomats, while integrating its younger generation more swiftly into government itself. 434 While this study does not shed additional light on China?s use of CSCAP, there is no doubt that China?s engagement in CSCAP and multilateralism in Asia more generally has rapidly changed over the last decade. Some seek to lead North Korea through a similar process. Some nongovernmental CSCAP participants believe that North Korea is already being socialized, although they freely admitted there was no way to prove it, 435 or at least that North Korean participants in CSCAP were beginning to elaborate on talking points more than they previously had. 436 Scott Snyder, an expert on North Korea and a frequent nongovernmental CSCAP participant, argued that North Korea initially used informal dialogues like CSCAP when it did not have direct contact with the United States to prepare and train its diplomats for official contacts in the future. Once direct contacts opened up with Washington after the 1993-94 nuclear crisis and the Agreed Framework, Pyongyang de-emphasized the informal networks. Similarly, Snyder argued that North Korea has subsequently used CSCAP to train its younger diplomats, allowing them to take the lead at network activities while older, more experienced North Korean officials observed their performance. 437 In the end, it is too early to tell whether North Korea will integrate more closely into East Asia as China did before it, much less whether CSCAP would facilitate 433 Interview A15 434 Interview A11 435 Interviews A9 and A11 436 Interview A22 437 Interview with author, Washington, D.C., December 13, 2004 154 North Korean socialization. Although Pyongyang joined CSCAP earlier, it has only joined the ARF in 2000 and the six-party talks are still proceeding. Experiences with China and now North Korea do raise two questions about CSCAP?s role and future. First, does a trade-off exist between CSCAP?s potential role socializing North Korea and its other purposes, such as policy coordination and discussion among other participants? Two CSCAP participants separately mentioned that the agenda of any meeting that North Korea attended was substantially diluted to facilitate DPRK attendance, even if the conversation became somewhat more pointed once they were actually in the room. Some topics were not raised when North Koreans did attend, for fear of offending them or sparking an unproductive response from DPRK participants. 438 Second, if North Korea is integrated into East Asia, as China has been, will CSCAP still have a purpose or will its function be served, having helped establish a regional multilateral forum and practice? As far back as the mid-1990s, Southeast Asian governments were already reportedly beginning to question the utility of CSCAP once the ARF was set up. 439 Many of CSCAP?s nongovernmental participants, including Cossa himself, acknowledged that the initial premise of CSCAP had been served: to create a security community of nongovernmental analysts in the Asia-Pacific, and that simply meeting for the sake of building contacts was probably no longer a sufficient goal. Ultimately, CSCAP decided it could evolve and adapt to those new goals. 440 The 438 Interviews A9 and A17 439 Ball, ?CSCAP: Its Future Place in the Regional Security Architecture,? p. 310 440 Author?s interview with Ralph Cossa, Washington, D.C., December 16, 2004 as well as interviews A9, A11, and A15 155 working groups were reorganized in 2004 precisely to refocus the network on having a regional policy impact, not just meeting. 441 CSCAP?s Influence in Other Countries There is no question that CSCAP tries to influence the agenda in other countries. One former U.S. ambassador even said that, while they served as the ambassador overseas, they separately sought to engage and diplomatically influence other countries through Asian CSCAP members because think tanks were so close to their governments. 442 Although no foreign participants were interviewed for this study, a number of American participants in and out of government believe that CSCAP has made more of a difference overseas than it does in the United States. 443 Some specific examples in nonproliferation such as the six-party talks format were previously cited. South Korean attitudes toward joining the PSI may also have been affected by CSCAP. In January 2006, not long after the May 2004 CSCAP working group on the subject, Seoul ?partially? joined PSI. 444 More generally, CSCAP?s initial efforts tried to get regional governments to care about nonproliferation at all, including China which only joined the NPT in 1992. Twelve years later, the network was seeking ways for the region to strengthen the NPT and get any emerging ASEAN security community to take up nonproliferation as well. 445 John Olsen believes that CSCAP?s 441 Summary of Discussions, CSCAP Steering Committee (Jakarta, Indonesia, December 6-7, 2003) sent as electronic message to all USCSCAP members, USCSCAP Special Report 1, January 12, 2004, as well as author?s interview with Ralph Cossa, Washington, D.C., December 16, 2004. 442 Interview A28 443 Interviews A3, A12, A15, and A16 444 ?New Obstacle to Nuclear Talks: Inter-Korean Relations Likely to Face Serious Setback,? Korea Times, January 25, 2006 445 Report of Twenty-second meeting, CSCAP CSBM International Working Group, p. 4 as well as interview A17 156 PACATOM project that he developed may have helped develop a regional norm toward nuclear transparency in China, Japan, South Korea, and Taiwan, 446 while a nongovernmental member with government experience asserted that CSCAP?s export controls project ?is not designed to influence [U.S.] government policy but to figure out how to use CSCAP to push ideas abroad.? 447 CSCAP over time has also sought to transfer lessons to Asia from Europe, mostly through U.S. and European CSCAP members. In its initial meetings, Asian participants deeply resisted any lessons that might be applied from the continent which had previously colonized many Asian states. 448 Within the network?s first decade, Europe had joined as a CSCAP member and hosted activities analyzing the relevance of European cooperative security experiences?including CSBMs, the OSCE, and even German- Polish reconciliation after World War II?for Asia. 449 Within nonproliferation, EURATOM?s lessons for PACATOM followed a similar course, facing initial resistance in Asia. 450 Nevertheless, at least one Japanese participant teased out lessons from EURATOM that might be useful for Asia?s efforts. 451 Experts brought in to discuss export controls included lessons from EU efforts to overcome many obstacles that Asian 446 Interview with author, Washington, D.C., April 1, 2005 447 Interview A6 448 CSCAP Memorandum 2, pp. 2-3, 7; Report on First CSCAP CSBM Working Group (Washington, DC, October 11-13, 1994): 3, as well as interview A16 449 Summary of Discussions, USCSCAP general membership/board of directors meeting (Washington, DC, June 22, 2000): 4, and Summary of Eighth Meeting of the CSCAP North Pacific Working Group, pp. 1-2 450 Roberts and Davis, p. 5; Summary of Discussions, Sixth meeting, CSCAP CSBM Working Group, p. 2; Cossa, PACATOM, p. 14; and Simon, ?Evaluating Track II Approaches,? p. 178. 451 Tatsujiro Suzuki, ?Lessons from EURATOM for Possible Regional Nuclear Cooperation in the Asia- Pacific Region (ASIATOM),? ch. 2 in Cossa, ed. Asia Pacific Multilateral Nuclear Safety and Non- Proliferation: Exploring the Possibilities (Honolulu, HI: Pacific Forum Occasional Papers, December 1996): 21-22. Also see Summary of Discussions, Sixth meeting, CSCAP CSBM Working Group, p. 2. 157 governments encountered while constructing their own systems. 452 Most recently, the CSBM working group began to analyze whether Asia could adopt a nonproliferation and export controls action plan similar to the EU?s. 453 A former foreign service officer explained that papers from these international CSCAP meetings served in Asia as the functional equivalent of ?non-papers,? or a diplomatic name for non-binding, idea-exploring documents shared among governments. The key is that they explore ideas, not agree on them, among think tanks in other countries which serve as extensions of the government. 454 CSCAP might also influence foreign governments through PACNET. Anecdotally, Cossa and Glosserman have both heard from the region about articles they wrote on PACNET, or seen the principles resurface in other media. Cossa relayed that he has been told that op-eds he has written have been sent to the South Korean leadership, while Glosserman got an e-mail from the ASEAN head secretary of the ASEAN+3, assuring him that U.S. concerns that Glosserman had expressed in PACNET were being addressed in the forum. The PACNET also seemed to help shape regional debates. In one, Cossa coined a term ?Japan surpassing,? playing on the terms ?Japan bashing? and ?Japan passing,? to refer to a new phase in U.S.-Japanese relations where Japan was surpassing expectations. According to Glosserman, that phrase soon reappeared in a column by Yoichi Funabashi, a widely respected strategic thinker who might be described as Japan?s version of Tom Friedman. One of Glosserman?s own PACNET?s, referring to Japan as Asia?s Switzerland because 452 Brad Glosserman, ?Export Controls: Fighting Proliferation and Building Confidence,? PacNet Newsletter 2003-35A (August 22, 2003), available at http://www.csis.org/pacfor/pac0335A.htm, accessed on March 9, 2005. 453 Report of Twenty-first meeting, CSCAP CSBM International Working Group, p. 4, as well as interviews A11 and A20 454 Interview A1 158 of their economic potential, was cited and used as the basis for a cover story in Newsweek International. 455 Whether CSCAP?s efforts actually made a difference in these countries is beyond the scope of this paper, but the interviews conducted for it revealed some evidence that they may have been successful. Interviewing foreign participants might reveal more. Conclusions The 2004 reorganization of CSCAP?s working groups could mark either the beginning of the end of the network?s value as multilateralism becomes a regional habit, or it could be the beginning of its transition to even greater, more focused regional influence. Whichever emerges, CSCAP in just its first dozen years has had value for U.S. participants even though not a single American exclusively works on or for CSCAP. Although at least one member wondered whether CSCAP was useful as an organization and how many people actually benefited from it, 456 the key value is primarily to individuals and what they get out of it. Or to be more precise, what they take from it. If utilized, the network benefits both government and nongovernmental participants. For the U.S. government, CSCAP provides: intelligence: honest feedback on U.S. policy both from the region and from U.S. nongovernmental experts; 455 Author?s interviews with Ralph Cossa, Washington, D.C., December 16, 2004; and with Brad Glosserman, Washington, D.C., December 14, 2004. See Yoichi Funabashi, "From Japan Bashing to Japan Surpassing," International Herald Tribune/Asahi Shimbun, March 30, 2004; Brad Glosserman, ?Japan?s ?Swiss? Option,? PacNet Newsletter 11 (March 15, 2002) at http://www.csis.org/pacfor/pac0211.htm; and George Wehrfritz with Hideko Takayama, "Japan: What to Wear," Newsweek International, May 13, 2002, p. 28. 456 Interview A12 159 integration: a tool to help government officials, usually with narrow portfolios, to provide an integrated regional perspective on security issues; public diplomacy: a forum to reach out both to U.S. experts off-the-record to explain U.S. policy in more depth and to the Asian foreign policy elite; regional agenda setting: a way to help shape the Asian security agenda through their own foreign policy elite; an avenue to contact North Korea; a safety net for issues such as PACATOM that slip between bureaucratic cracks; and contingency planning to begin to think through the implications of different political circumstances, such as if an agreement between Washington and Pyongyang were to be reached. Outside the U.S. government, CSCAP provides information and ideas? from thoughtful debates, from the region directly via Asian experts as well as indirectly through peers? recent trips, and from U.S. government off-the-record briefings?on U.S. policy, events in Southeast Asia, and recent events such as crises, political change, and 160 summits. Quite simply, CSCAP?s primary influence is not as an organization but to empower individuals who are members of CSCAP with information and sharper opinions. 457 The influence individuals may wield is rarely policy recommendations, but the ability to give feedback to the U.S. government and to the public about how U.S. policies are perceived, the repercussions of a crisis or summit, or what the foreign policy implications might be of domestic changes overseas, as well as, for the handful that travel overseas, to explain the United States to the region. As Cossa himself simply stated, it is about ?perceptions not policies.? 458 457 Interviews A1, A2, and A4 458 Interview with author, Washington, D.C., December 16, 2004 161 CHAPTER FOUR: THE NORTHEAST ASIAN COOPERATION DIALOGUE The Northeast Asian Cooperation Dialogue (NEACD) was founded in 1993 to ?continue to reduce mistrust within the North Pacific region and to avert conflicts through confidence and security building measures? by periodically convening a handful of government officials in their private capacity with nongovernmental experts from China, Japan, North Korea, South Korea, Russia, and the United States in an informal, multilateral forum. 459 From 1993-2005, NEACD (commonly pronounced nee? sid) has convened sixteen times in plenary sessions which have been held at least annually (two were held in 1996, 1997, and 2002). The host rotates among the six countries except for North Korea, which was a founding member but has only attended three dialogues from 2002-04. NEACD has been organized by the University of California?s Institute on Global Conflict and Cooperation (IGCC) under the direction of Susan Shirk, except from 1997-2000 when she left to be a U.S. Deputy Assistant Secretary of State for East Asia and the Pacific and Stephan Haggard temporarily stepped in. Since its founding, the network has been funded almost exclusively by the U.S. Department of Energy?s Office of Nonproliferation and National Security, although substantial assistance is sometimes provided by the host country. 460 Each meeting notionally consists of 30 participants, with one representative from the Foreign Ministry, one policymaker from the Defense Ministry, one uniformed military officer?all participating in their private or unofficial capacity?and two 459 Summary of Northeast Asian Cooperation Dialogue, available at http://www- igcc.ucsd.edu/regions/northeast_asia/neacd/ (accessed June 8, 2002), hereafter referred to as NEACD summary. 460 Electronic message from Susan Shirk to author, July 24, 2005 162 nongovernmental experts from universities or think tanks from each of the six countries attending. Although no session has exactly met this formula, they have been quite close except for North Korea?s participation. The agenda for each meeting contains a tour d?table of national perspectives on regional security, at which one participant from each country (almost always from the Foreign Ministry) gives a 5-10 minute presentation from their country?s perspective, ideally emphasizing issues since the last NEACD meeting. After each presentation, a question and answer as well as discussion period follows for up to 50 minutes on each country. 461 Through the years, the rest of the agenda has been filled by a variety of subjects including presentations of occasional NEACD study groups, national military perspectives on regional security, the future of NEACD itself, and special discussion sections on a particular topic. Because it seeks to create an informal environment where government officials can speak freely, NEACD publishes very few documents. Participants are assured that sessions are off-the-record and there are no public transcripts of any meeting. Although a handful of papers prepared for the first six meetings were published by IGCC, no paper has been published since 1997. A general, 1-2 page newsletter summary has been prepared and posted on a website after all but one meeting, along with the agenda and participants list for all sixteen NEACD plenary meetings. Conclusions of any study groups, ranging from 1-3 pages long, have also been prepared. In 1998, at the eighth meeting, the IGCC also established an online infrastructure, entitled Wired for Peace (www.wiredforpeace.org), to support NEACD by posting the newsletter summaries, agenda, participant lists, and establishing a password-protected bulletin board enabling alumni to conduct ?virtual diplomacy.? Since at least 2003, Shirk and two or three other 461 NEACD summary 163 U.S. participants have also held a public briefing for the D.C. community at the University of California?s Washington Center. For this case study, seventeen original interviews with staff, funders, participants, and senior U.S. government policymakers were conducted. Participants, either nongovernmental or governmental, were targeted for interviews if they attended two or more NEACD meetings. Table 4.1 contains descriptions of the candidates for interviews, divided by their role in NEACD and whether they were primarily NGO or government participants. Some interviews were not conducted either because candidates had retired or otherwise could not be found, declined to be interviewed, or did not respond to requests to be interviewed. Table 4.1 NEACD Interview Candidates NGO USG Total NEACD Staff 7 462 - 7 NEACD Funders 0 3 3 NEACD Participants 4 15 19 Attended 2 or more NEACD meetings 3 8 11 Policymakers who attended once 463 1 464 7 8 Senior government policymakers 465 - 5 5 Total 11 23 34 462 NEACD staff sought for interviews included Susan Shirk, Stephan Haggard, and five assistants to Professors Shirk or Haggard, most of whom were graduate students during their affiliation with NEACD. Only one of those assistants has actually been interviewed for this study. 463 Because of their positions in the policymaking chain, Deputy Assistant Secretaries, Principal Deputy Assistant Secretaries, and policy planning staff members were sought for interviews even if they only attended one meeting. 464 Former Assistant Secretary of State for Nonproliferation Robert Einhorn, who only participated once and had already left government at that time, was interviewed for this study and is oddly categorized simultaneously as an NGO participant and a policymaker who attended once. The other former policymakers attended NEACD when they were in government and are included as USG officials. 465 Senior government policymakers are defined as senior directors of the NSC staff, Assistant Secretaries of State for East Asia and the Pacific, or the chief U.S. coordinator for the six-party talks on North Korea?s nuclear weapons program. Two subjects were already recorded in the CSCAP chapter because they participated in that study. Although they also discussed NEACD, they are not tallied here to avoid redundantly counting those interviews. 164 Eleven of the seventeen interviews conducted were with U.S. government officials, two of whom are members of the revolving door but were in government when they participated in NEACD. The other six interviews were with members of nongovernmental organizations, although two of those six, including Shirk herself, also had government experience before and/or after their NEACD participation. In other words, of the seventeen interviews conducted, nine were with permanent or retired government employees (civil service, foreign service, or military), four with members of the revolving door, and four with academics who had not been employed in government. Descriptions of the interviews actually conducted are contained in table 4.2. Throughout this chapter, opinions expressed in only a single interview, whether cited anonymously or specifically, will be noted as one person?s opinion. In addition to these interviews, NEACD was kind enough to allow the author to attend the 2004 La Jolla meeting (NEACD-15), which the North Koreans attended, and the 2005 Seoul meeting (NEACD- 16), which the North Koreans did not. Table 4.2 NEACD Interviews Conducted NGO USG Total NEACD Staff 3 - 3 NEACD Funders 0 3 3 NEACD Participants 3 7 10 Attended 2 or more NEACD meetings 2 5 7 Policymakers who attended once 1 2 3 Senior government policymakers - 1 1 Total 6 11 17 Having introduced NEACD?s infrastructure as well as the primary materials and interviews utilized, the subsequent three parts of this chapter will include descriptions of the issues NEACD has covered from 1993-2005; an evaluation of the network based on 165 the four-part framework presented in the introductory chapter; and finally a series of observations and conclusions about NEACD. The concluding chapter (chapter 6) will include comparisons with the other two case studies. This chapter will be limited to NEACD itself, starting with the issues the network has addressed. Issues NEACD Has Addressed In addition to the central tour d?table of national perspectives on regional security, early NEACD sessions typically included a focused session on one non-security issue and/or updates on the progress of one or more of four special NEACD study groups. The non- security sessions explored potential economic cooperation, the environment, agricultural cooperation and trade, energy security, or maritime trade. 466 Two special study groups were established at the 3 rd NEACD meeting (NEACD- 3), held in 1995. At each study group, one member from each participating country would contribute, aiming to issue a series of recommendations for the main plenary meeting to discuss further. 467 One project focused on ?Principles of Cooperation in Northeast Asia,? ultimately agreeing upon eight, which were presented to and endorsed by NEACD-7 participants, held in Tokyo in December 1997 (see box 4.3). Although it seems fair to say that the principles were quite bland, it was arguably both a significant step given regional politics at that time and equally important, NEACD statements emphasized, not to overlook what participants learned through the process of discussing these issues, even if agreement was not reached on some issues such as human rights. 468 466 NEACD summary 467 Ibid. 468 Summary of Northeast Asian Cooperation Dialogue IV, Beijing, China, January 8-10, 1996, available at http://www-igcc.ucsd.edu/regions/northeast_asia/neacd/neacd1.hmtl (accessed June 12, 2002) and 166 The second study group focused on mutual reassurance measures (MRMs), or ?broader measures to promote a basis for mutual confidence and reassurance that include but are not limited to military-related measures.? 469 Although NEACD members did Summary of Northeast Asian Cooperation Dialogue V, Seoul, Korea, September 9-12, 1996, available at http://www-igcc.ucsd.edu/regions/northeast_asia/neacd/neacd5.hmtl (accessed June 12, 2002) 469 NEACD summary Table 4.3: Principles of Cooperation in Northeast Asia The states of Northeast Asia share the common objectives of peace, prosperity, and security in the region. To achieve these ends, they advance the following principles of cooperation in Northeast Asia. 1. The states of Northeast Asia respect each other?s sovereignty, territorial integrity, and equality; accept that other countries have different political, economic, social and cultural systems and the right to determine their own laws and regulations as well as other domestic affairs. They also recognize that they are obliged to abide by and implement international agreements to which they are a party. 2. The states of Northeast Asia will refrain from the threat or use of force against each other; will settle disputes through peaceful means; and pledge to use consultation, negotiation, and other peaceful means to prevent conflict between and among each other. 3. The states of Northeast Asia express their commitment to the protection and promotion of human rights in accordance with the purposes and principles of the UN Charter. 4. To prevent misunderstanding and develop trust, the states of Northeast Asia will promote dialogue, information exchange, and transparency on security issues of common concern. 5. The states of Northeast Asia respect the principle of freedom of navigation based on international law. 6. The states of Northeast Asia will promote economic cooperation and the development of trade and investment in the region. 7. The states of Northeast Asia will cooperate on transnational issues of common concern, such as organized crime, drug trafficking, terrorism, and illegal immigration. 8. The states of Northeast Asia will cooperate in the provision of humanitarian assistance, such as food aid and disaster relief. 167 agree on the statement that ?MRMs are vital to the maintenance of peace and stability,? 470 the group bogged down, deciding within 18 months ?to proceed in an incremental way, focusing first on dialogue and exchanges to cultivate a sense of cooperation and lay a sound foundation for future efforts.? 471 Participants at NEACD-5, held in September 1996 in Seoul, decided to focus the MRM efforts on Defense Information Sharing (DIS), creating a third NEACD project which continues to this day. The purpose of the DIS effort, which includes only civilian defense officials and uniformed military officers, ?is to provide a forum for discussion of current efforts at information sharing and systematic discussions of perspectives on military doctrine, to clarify information currently found in information sharing documents, and to encourage voluntary and unilateral information sharing.? 472 In essence, it is a multilateral military exchange among officials in their private capacity, ideally enabling study participants to discuss military doctrine informally and tour facilities such as the PLA Navy North China Sea Fleet headquarters, stationed in Qingdao. 473 Although they were initially held at a different time and place than the NEACD plenary session, the DIS talks have become a permanent addendum to the NEACD format, being held on the margins of every NEACD plenary session since 2001, and have effectively replaced the national military perspectives on regional security that were presented in the plenary session itself from 1995-99. 470 Summary of Northeast Asian Cooperation Dialogue IV 471 Summary of Northeast Asian Cooperation Dialogue V 472 Summary of Northeast Asian Cooperation Dialogue VI, Harriman, New York, April 2-4, 1997, available at http://www-igcc.ucsd.edu/regions/northeast_asia/neacd/neacd6.hmtl (accessed June 12, 2002) 473 Meeting Report of Northeast Asian Cooperation Dialogue 14, Qingdao, China, 1-2 September 2003, available at http://www.wiredforpeace.org/neacd14.php (accessed November 21, 2003) 168 The fourth and final NEACD study group convened in 2001 to discuss the potential for a formal, track-one Northeast Asian multilateral dialogue. The group produced a concept paper slightly longer than two pages, outlining four possible ways to establish such an organization: ?establish a new forum; expand an existing [governmental] forum (e.g. The Four Party Talks); convert a track-two forum into a track-one forum; or use an occasion when officials of all six countries are present.? To its credit, the group avoided the temptation to promote NEACD overtly and concluded, ?The last approach seems preferable?.At the next ASEAN Regional Forum in July 2001 the foreign ministers of the six countries of Northeast Asia could meet informally for breakfast or tea.? 474 In addition to any study group reports, NEACD-8 began to include a special session on recent events, effectively supplanting the earlier non-security oriented session. At that November 1998 meeting, participants discussed the security implications of the 1996-97 Asian financial crisis, as well as the August 1998 North Korean missile test and its potential implications for regional or national missile defense programs. 475 The November 2000 meeting (NEACD-10) focused on more positive developments: rapprochement on the Korean peninsula after the June 2000 Korean summit between South Korea?s Kim Dae Jung and North Korea?s Kim Jong Il. 476 After the September 11 attacks, the 2001 and April 2002 NEACD sessions discussed regional counterterrorism cooperation, including the expanding role of the Japanese Self-Defense Forces in 474 ?Northeast Asian Track-One Multilateral Dialogue,? A Concept Paper of the NEACD Study Project, June 21, 2001. 475 Summary of Northeast Asian Cooperation Dialogue VIII, Moscow, Russia, November 11-12, 1998, available at http://www-igcc.ucsd.edu/regions/northeast_asia/neacd/neacd8.hmtl (accessed June 12, 2002) 476 Summary of Northeast Asian Cooperation Dialogue X, Seoul, Korea, November 9-10, 2000, available at http://www-igcc.ucsd.edu/regions/northeast_asia/neacd/neacd10.hmtl (accessed June 12, 2002) 169 assisting the United States. 477 Finally, in 2002, attention turned back to Korea after North Korea began to attend NEACD for the first of three consecutive years. Coincidentally, the North Korean nuclear crisis reignited almost as the October 2002 NEACD was being held, when Assistant Secretary of State Kelly confronted North Korea with U.S. evidence that the DPRK was pursuing a highly-enriched uranium program in violation of the 1994 Agreed Framework. While these were the formal NEACD topics during its first thirteen years, subjects interviewed portrayed the network as informally going through three stages. 478 In the mid-1990s, NEACD initially appeared to focus on seeking to integrate, or socialize, China into a regional, multilateral framework particularly after incidents like the 1995 Taiwan Strait missile crisis?when China launched missiles off the coast of Taiwan to seek to deter the Taiwanese from electing Lee Teng-hui, who Beijing believed was pursuing a secessionist agenda. In the late 1990s, a second informal stage, keeping multilateralism alive, seemed to be a primary goal as the immediate post-Cold War optimism for its potential began to fade. Finally, the North Korean nuclear issue has become the primary focus since 2002. Although missile defense was briefly discussed, particularly potential Japanese participation in theater missile defense after the 1998 North Korean nuclear test and before the September 11 terrorist attacks, the proliferation issues at NEACD have almost exclusively revolved around the North Korean nuclear crisis, even when North Korea did not attend. Two former deputy assistant secretaries and one NEACD staff member interviewed for this study separately relayed that, in its earlier years, NEACD seemed to 477 Meeting Report of Northeast Asian Cooperation Dialogue 11, Honolulu, Hawaii, 8-9 November 2001, available at http://www.wiredforpeace.org/neacd11.php (accessed November 21, 2003) 478 Interviews B2, B34, and B36 170 have almost a desperate quality to it at times, searching for topics to tackle in lieu of North Korean attendants, who sometimes canceled at the last minute, while attending members discussed how to entice North Korea to attend the next meeting. 479 Results Contacts and Perceptions Similar to CSCAP, whether NEACD established valuable networking connections and whether participants gained useful new information or perceptions depended on the previous experience of each participant, with nongovernmental contributors potentially gaining different connections and insights than government officials. Even among nongovernmental members, academics and think tank employees, and within government, State Department and military officers seemed to garner different benefits. Nongovernmental Participants For some U.S. nongovernmental participants, NEACD introduced them to new, professionally valuable contacts. In those cases, NGO participants cited the opportunity to follow up individually after the conference as valuable for their own research and writing, particularly with participants from China or government officials from all participating countries. 480 Bob Einhorn, formerly the Assistant Secretary of State for Nonproliferation who participated in NEACD after he left government, directly explained: ?When I was in government, I had all the government contacts I could ever want, even with the North Korean government. I could have as 479 Interviews B4, B36, and B39 480 Interviews B30, B32, and B37 171 many nongovernmental contacts as I wanted?.In terms of networking, frankly, it?s more valuable for nongovernmental people than for governmental people.? 481 From the plenary sessions themselves, U.S. nongovernmental participants mentioned at least two insights they gained. First, North Korean participation seemed to make an impression on academics who had not served in government, and had no previous contact with Pyongyang. In the formal session, for example, the character of the North Korean rhetoric struck one academic participant: [T]he North Koreans were talking at great length about?how all of this was predicated on the perception of changing, and increasingly hostile, U.S. policy toward them. It wasn?t that this was particularly new or shocking to me, but the clarity with which the message was conveyed by them, and the continued resonance of the same approach, made it very clear to me that much of what the American government had been presenting as its policies was just running into a terribly solid brick wall with the North Koreans that made me much more pessimistic about the possibility for resolving the differences than I was, perhaps, going into it. 482 On the margins of the formal plenary sessions, another academic engaged the North Koreans and found that, although they certainly were not critical of their own government, they seemed interested about what kinds of questions were being posed to them, and were willing to open up about their own perspectives on North Korean relations with other countries around the table. 483 Second, U.S. NGO participants said they learned some things about the contemporary policies and perspectives of other countries around the table besides North 481 Interview with author, Washington, D.C., June 2, 2005 482 Interview B37 483 Interview B30 172 Korea and the United States. 484 Although participants did not remember specifics when pressed, they recalled that they felt they were getting unique insights about current policy from government officials in other countries. As one participant put it: ?In general, academics express their views about what should be policy. Governments express what is policy.? 485 Participants in and out of government stressed that the ability of government officials to speak in their personal capacity did not reveal any hidden opposition to current policy from within governments. The insights were more nuanced, allowing for a deeper explanation of current policy, rather than disagreeing with it. 486 One government official went so far as to say that, even in the United States, the idea that a government official would express an opinion distinct from official policy ?is a fallacy. No government official ever speaks on their own unless their own personal opinion is the government opinion. It?s ludicrous!? 487 [emphasis in original interview] Government Participants One of NEACD?s unique aspects is that government contributors outnumber nongovernmental ones, sometimes leading it to be labeled a ?track one-and-a-half? dialogue. Those officials generally already knew each other, limiting NEACD?s potential networking value for government participants. One observer mentioned that many conference participants seemed to hang out in between formal sessions with their own profession 488 ?diplomats with diplomats, military officers with military officers, and professors with professors?further reducing the opportunity for 484 Interviews B30, B32, and B37 485 Interview B30 486 Interviews B29, B30, and B34 487 Interview A4 488 Electronic message from B4 173 government attendees to make new contacts. Although one did mention that they met valuable NGO contacts, 489 and two others working in the U.S. State Department valued meeting Chinese military officials for the first time, 490 just as many government participants interviewed for this study believed they did not make professionally valuable contacts through NEACD, either because they did not value them or did not follow up after the conference. 491 One exception seemed to be that officials?either at low levels or even in policymaking positions?who were either new to the region entirely, to a particular element of the region (such as a Korea expert who was beginning to work on China), or even to diplomacy in general were able to gain introductory insights into the region through NEACD. 492 A second potential exception is if any contacts made or reinforced through NEACD might eventually rise to an even more influential position within their own government. 493 Fu Ying, a Chinese participant in NEACD who subsequently became Beijing?s principal participant in the six-party talks, was cited as an example. Notionally, these specific personal ties built at NEACD could plant the seeds for a better future working relationship. If political relations among countries were to improve in the future, one government participant cited the potential for these early investments in personal relationships to become even more valuable. 494 They could be, in other words, down payments in case relations improved, even if NEACD had no immediate effect on relations at all. 489 Interview B3 490 Interviews B39 and B40 491 Interviews B29, B34, and B36 492 Interviews B1, B39, and B41 493 Interviews B29 and B39 494 Interview B38 174 Even if most government officials did not believe that they established valuable new contacts through NEACD, those officials still might have learned something new from their participation, or used NEACD for diplomatically useful purposes. Specifically, three primary characteristics were cited that made NEACD different from official government contacts: the presence of NGOs; the informality, or ability for officials to speak unofficially in their private capacity; and the multilateral format. The question is: do these differences matter? Do government officials believe they learned something new because of the presence of scholars, the informality of the dialogue, or its multilateral dimensions? Nongovernmental analysts seemed to provoke the discussion in otherwise overlooked directions, according to government participants. 495 Robert Scalapino, a highly respected Asian expert from Berkeley who has attended thirteen of NEACD?s sixteen meetings, was specifically cited as someone who could frame current issues in an historical context, provoking at least one U.S. diplomat to look at current problems in a new, deeper historical way. 496 One participant, however, felt that nongovernmental participants used to be more valuable. That participant felt that NGO participants used to speak more when NEACD first started, 497 increasingly deferring to government officials as those officials have become more comfortable in an unofficial setting. Nevertheless, at one of the more recent meetings, NEACD-15 in La Jolla in 2004, NGO participants did seek to stimulate the plenary discussion, which North Koreans attended, by asking questions like: why Pyongyang objected to the U.S.-proposed CVID formula (comprehensive, verifiable, irreversible disarmament) for solving the nuclear crisis; 495 Interviews B29, B36, and B39 496 Interview B39 as well as B4 497 Interview B31 175 which countries around the table supported CVID; what North Korea felt the United States should do differently if it wanted to prove it did not have a hostile policy toward the DPRK; and what various countries? positions were on the proliferation security initiative (PSI). 498 The informality, NEACD?s second difference from official talks, seemed to allow discussion to flow in a distinct direction that got at the roots, or core motivations, behind various governments? policies. As mentioned earlier, being able to speak in their private capacity did not embolden government officials to raise objections to their government policies. It did, however, change the character of the dialogue to make NEACD more relaxed and informative, according to current and former government officials, stimulating an exchange of views rather than talking points. 499 Although one diplomat disagreed, particularly arguing that ?with North Koreans, there really is no such thing as informal dialogue,? 500 others cited a difference, especially with countries other than North Korea. One senior government policymaker listed a range of differences?including lesser time constraints on talks which might include field trips to allow for broad discussions; an open, unconfined agenda in the formal sessions themselves; the absence of media scrutiny; the lack of extremely detailed reporting requirements; and the fact that agreed decision points are not required at the end of each meeting?all of which allowed participants to explore questions they might have about other governments? positions and 498 Author?s notes from NEACD-15, La Jolla, California, April 5-6, 2004 499 Interview B32 and B301 as well NEACD summary 500 Interview B34 176 to be more expansive about their own answers. 501 Another former senior policymaker even cited what resembles peer pressure as a potential factor: In your track-1 context, everyone is constrained. There?s not terribly much premium in being too interesting. It?s better to be safe than too interesting?. [In track one-and-a half], why should I look like the stilted, rigid schnook in the group, who is reading talking points? Let me be a little bit more spontaneous. There?s an incentive in that setting, I think, to be somewhat more candid. 502 That same former official also mentioned a relaxed setting, such as the La Jolla hotel right on the beach, away from normal diplomatic conference rooms, as a factor that could accelerate informality and the learning process even further. Another government official cited the personal relationships developed over time, particularly with Asian countries who they argued culturally rely more on personal and less on contractual or professional relationships. 503 NEACD?s third distinct factor, its multilateral format, allows participants to see how other countries relate to one another, enabling them to become a proverbial fly on the wall, and had not otherwise been available for just these five or six countries. Participants with U.S. government experience cited the window it gave even regional experts into how frank and tense relations could be between China and Japan, 504 or South Korea and Japan. The United States could also use the forum to try to get those parties to talk to one another, particularly U.S. allies Korea and Japan. 505 Jim Steinberg, the State Department?s policy planning director at the time who would go on to be the deputy national security advisor, said the 1996 dialogue he attended even led him to dispense 501 Interview B301 502 Interview B32 503 Interview B3 504 Interviews B39 and B40 505 Interviews B3 and B36 177 with the notion that Sino-Japanese relations tended to balance Sino-U.S. relations. Prior to the meeting, he sensed that bad Sino-U.S. relations drove China toward Japan, while better Sino-U.S. relations allowed Sino-Japanese relations to deteriorate. After the NEACD meeting he attended, he came away with the impression that Sino-Japanese relations were driven more by endogenous factors, heavily constrained by ?strategic and historic animosity and mistrust.? 506 Such a multilateral forum for the five or six Northeast Asian countries was not available anywhere else, at least for government officials, until the six-party talks were first held in August 2003, ten years after NEACD started. Although the ASEAN Regional Forum (ARF) did exist, it included a number of other Southeast Asian countries that complicated dialogue. Government officials argued that NEACD foreshadowed the more focused official talks, acting as a trial to see if such a multilateral security organization might work. 507 To further enhance opportunities for government officials to interact, NEACD hosts a luncheon just for foreign ministry officials during its plenary session. Two of the U.S. State Department participants, however, each said that the wrong people attended those luncheons to make any sort of meaningful progress and, in at least one case, the agenda focused more on ?the timing, venue, participation, and objectives of the next [NEACD] meeting? rather than a substantive agenda. 508 Nevertheless, for the 2005 NEACD meeting in Seoul, South Korea proposed and hosted an expanded version into a full afternoon, giving some indication that they felt the official forum was useful, or at least that it could be if more time was available. 506 Interview with author, Washington, D.C., August 22, 2005 507 Interviews B3, B33, B36, B38, B39, and B40 508 Interviews B34 as well as A4 178 While some government officials believed they learned new things through NEACD, U.S. officials also emphasized the potential to use the forum to send a variety of diplomatic signals, namely: to conduct multilateral diplomacy, to engage in public diplomacy both narrowly with NGO participants and more broadly as a sign of willingness to engage North Korea, and to help reinforce the track-one dialogue at the six-party talks. Different government officials used NEACD for different multilateral diplomatic purposes, depending on the state of relations at the time. In the early stages of NEACD, one government policymaker said he used a 1996 session to help inform the other countries about ongoing bilateral talks between North Korea and the United States and where they might be going, at a time when South Korea was uncomfortable with those talks, Beijing was much closer with Pyongyang, and no other multilateral format existed. 509 Similarly, in 1996-97, another policymaker was primarily seeking to get the United States, Japan, and Korea to work more closely together on a variety of issues, while simultaneously seeking to get China to bring North Korea to the negotiating table. 510 More recently, others sought to explain elements of U.S. policy to the region as a whole, such as its reasons for concern about the North Korean nuclear program, why it might seek to refer North Korea to the UN Security Council, and the likely priorities of a new diplomatic team at the State Department when Condoleezza Rice had just begun as Secretary of State and Ambassador Christopher Hill was replacing Jim Kelly as Assistant Secretary of State. 511 It also provided an opportunity, as one senior policymaker mentioned, for various pairs of countries, such as the United States and South Korea, to 509 Interview B39 510 Interview B36 511 Interviews B38 and B301 and author?s notes from NEACD-16, Seoul, Korea, April 12-13, 2005 179 clarify their policies toward North Korea and explain to others that they were not diverging as media stories might portray. 512 NEACD also provided an opportunity for more public diplomacy. At the talks themselves, one participating government policymaker said it gave them an opportunity to explain U.S. policy, and to answer North Korean accusations in front of nongovernmental participants who might, in turn, shape their own country?s opinions. 513 Others with government experience said that the dialogue served as a signal to reassure everyone, even beyond the participants in the meeting, such as U.S. allies in South Korea that the United States was interested or at least willing to engage North Korea. 514 A third benefit that has emerged more recently is an opportunity to follow up directly with North Korea, when they participate in NEACD, and reinforce the formal six-party diplomatic channel. While the theoretical argument is that channels like NEACD can be useful when formal government contacts do not exist, government officials argued that the network was more useful when it served to supplement nascent contacts. 515 One senior government policymaker argued that building personal contacts was somewhat useful, but realistically those were limited with so much mistrust built up between Pyongyang and Washington. Because there were almost no supplementary bilateral channels (with the possible exception of contacts that could be made on the margins of the UN) like those that exist between the United States and the other four parties, NEACD allowed the U.S. government to potentially foreshadow an upcoming 512 Interview B301 513 Interview B34 514 Interviews A21 and B36 515 Interviews B301, B34, and B38 180 round of the six-party talks, giving North Korea a ?heads up? about what issues would be particularly important to the United States. 516 NEACD could also be used to follow up after a round of the six-party talks. By sheer coincidence, NEACD-13 was held in China, just a week or two after the first official round. At least two NEACD participants, Fu Ying from China and a Russian, had actually attended the six-party talks and were able to bring that direct knowledge to NEACD. Don Keyser, who was the U.S. Principal Deputy Secretary of State for East Asian and Pacific Affairs at that time, explained that he was able to pursue a wide range of objectives at the NEACD meeting, including reiterating that the United States did not have a hostile policy toward North Korea, 517 and elaborating on the benefits of an executive agreement for any potential solution, as opposed to a treaty which would require Congressional consent. 518 Keyser was also able to correct some of the media coverage immediately after the six-party talks, addressing any potentially inaccurate reporting such as whether the United States was demanding that North Korea make all the concessions first. 519 Keyser also said that the network meeting allowed him to see what messages had been absorbed by North Korea and the other four parties. 520 In turn, it allowed the United States to demonstrate the solidarity among the parties other than North Korea in the negotiations. 521 Finally, NEACD uniquely allowed the United States to clearly convey bipartisan support for certain aspects of the administration?s position. Keyser said that at the 2004 516 Interview B301 517 Electronic message to author, August 19, 2005 518 Author?s notes from NEACD-14 briefing, Washington, DC, October 9, 2003 519 Interview with author, Fairfax, VA, June 15, 2005, and electronic message to author, August 19, 2005 520 Author?s notes from NEACD-14 briefing 521 Interview with author, Fairfax, VA, June 15, 2005, and electronic message to author, August 19, 2005 181 NEACD meeting in the United States, Bob Einhorn, the former Assistant Secretary of State for Nonproliferation from the Clinton administration, was able to convey solidarity for many aspects of the administration?s policies, such as PSI, and dissuade North Korean participants from the notion that Pyongyang might get a better deal from a future Kerry administration. Keyser emphasized that Einhorn could credibly go into greater detail about the reasoning behind U.S. policy because he was out of government but had previously served in an authoritative position. 522 Although another former official highlighted the potential difficulty that nongovernmental participants might theoretically publicly disagree with U.S. policy toward North Korea, 523 in this case Einhorn actually did the opposite: using the network to emphasize that media coverage highlighting controversy should not give foreign governments the impression that disagreements within the U.S. government were deeper than they actually were. Forums like NEACD allowed officials and even nongovernmental officials to clarify that these differences were merely nuances, played up to sell copies, but agreement on the policy fundamentals was solid. 524 Whether these messages were getting through to North Korea in NEACD or in direct talks, for that matter, is anybody?s guess, Keyser emphasized. He said that the state of our knowledge about North Korea is so minimal that it is not even clear whether DPRK diplomats communicate with Pyongyang by cables, the traditional diplomatic method, or exclusively by direct, oral briefings. Keyser even believed that the North Korean delegation at the 2003 NEACD meeting in China did not know what had actually happened at the formal talks that had just concluded. The key, he said, is therefore to get 522 Interview with author, Fairfax, VA, June 15, 2005, and electronic message to author, August 19, 2005 523 Interview B36 524 Interview with author, Washington, D.C., June 2, 2005 182 as many different streams of information as possible to Pyongyang, reinforcing the same message to maximize the chances that it gets through, and provide opportunities to answer any questions that may have arisen through whatever information-sharing process Pyongyang uses. 525 When asked whether any sort of trial balloon may have been floated by the United States to the North Koreans, however, two former officials serving at separate times, one from the State Department and one from the Pentagon, said there were none in their experience. 526 In the broadest sense, as mentioned earlier, NEACD was an organizational trial for the six-party talks, but no specific policy trial balloons were mentioned in interviews conducted for this study. In sum, although experienced U.S. government officials did not seem to make professionally valuable contacts through NEACD, some were able to learn about relations among other parties in the network because of the catalytic presence of nongovernmental participants, the informality of the dialogue, and the multilateral format. They were also able to send diplomatic messages to all the parties simultaneously, to conduct public diplomacy in order to shape elite as well as public opinion through nongovernmental participants, and to reinforce nascent diplomatic contacts with North Korea. U.S. Government Access and Interaction Because NEACD generally meets only once a year, includes only five participants from each country, and integrally relies on confidentiality, the level of participation within the 525 Interview with author, Fairfax, VA, June 15, 2005 526 Interviews B34 and B36 183 U.S. government has been crucial to its potential effect in the United States. 527 From at least the U.S. State Department, the consensus seems to be that a deputy assistant secretary (DAS), and ideally the principal deputy assistant secretary (PDAS), is the appropriate level of participation. 528 As one former policymaker who worked their way up to become an assistant secretary said: ?The level of deputy assistant secretary is the most important level in the government. DASs are the most junior senior people and the most senior junior people. In other words, they have to have the detail of an action officer, but they also have to think in policy terms and be able to translate that policy to the action officer. There is kind of a critical bridge.? 529 The primary problem seems to be the time demands on a policymaker of that level, particularly because in four out of every five years, NEACD meetings are held in Asia, requiring extensive travel. Even if a government official commits to going to a meeting, it is always possible that something will come up at the last minute, requiring them to stay in Washington and leave NEACD with either a lower level or no U.S. State Department representative. 530 In the first two NEACD meetings, the U.S. State Department initially committed to send a PDAS, but it has varied over time. After those two initial meetings, the U.S. State Department sent the DAS responsible for China because of its initial focus on NEACD as a forum to help integrate China into the region. 531 As that happened and the State Department?s perception of the network shifted to use NEACD as a building block for regional multilateralism in the late 1990s, while simultaneously becoming more skeptical about regional multilateralism?s prospects, State downgraded its NEACD 527 Interviews B34 and B301 528 Interviews B2, B32, B38, and B301 529 Interview B32 530 Interviews A4, B34, and B41 531 Interview B34 184 participant to the office director for regional security policy. 532 Finally, after North Korea reemerged as the network?s focus after 2002, State once again sent the PDAS, at least in part because the office director for regional security policy did not have Korea as part of their portfolio, and the PDAS could cover the entire range of regional issues, including North Korea. 533 Altogether, over the sixteen meetings, State has sent a PDAS almost half the time (7 of 16), a DAS a quarter of the time (4 of 16), an office director or below a quarter of the time (4 of 16), and no participant from State went to the most recent meeting, principally because the confirmation hearings for Assistant Secretary of State Chris Hill were being held at the same time. A member of the policy planning staff has also attended twice. The civilian representative from the Defense Department has generally been at a lower level, and less frequent, than State. At the inaugural meeting, the Pentagon sent the deputy assistant secretary of defense (DASD), the highest ranking official responsible for Asia. In the subsequent fifteen meetings, only one DASD attended: Kurt Campbell, who attended twice. In six of the other NEACD meetings, an office director or lower attended with the deputy director for regional affairs and Congressional relations attending three of those times, and no civilian from the Pentagon attended the other seven meetings, including all six of the sessions since 2000. Altogether, a DASD attended less than a quarter of the time (3 of 16), an office director or below about one-third of the time (6 of 16), and no civilian from Defense went almost half the time (7 of 16). Although a uniformed military officer was not part of the initial participation formula, a member of the U.S. military has attended twelve of the subsequent fifteen 532 Interviews B1, B31, and B34 533 Interviews A4 and B34 185 meetings, with the Joint Staff (J-5) sending a representative to ten of the last twelve sessions. Rear Admiral William Sullivan has attended three NEACD sessions including the most recent one in Seoul as the Vice Director for Strategic Plans and Policy of the Joint Staff (J-5). He said that, despite an institutional reluctance in the military to attend a meeting that produced no decisions, he personally found the sessions very valuable. Nevertheless, he said he probably would never have attended in the first place but he accompanied his boss, Admiral Blair who was the Commander in Chief of the U.S. Pacific Command (CINCPAC), who encouraged him to attend. 534 Sullivan has subsequently either attended or sent a member of his staff to four of five meetings. Although interrelated effects can sometimes be artificially separated into dependent and independent variables too readily, it does seem that NEACD?s ability to get government participation depends more on government?s interest and exogenous events, not NEACD?s efforts. 535 One factor is certainly the attitude of other countries, principally whether North Korea has even been willing to participate. 536 But even within the United States, some cited that the preeminence of U.S. power, or the Bush administration specifically, has recently made the United States less interested in listening to others, perceiving this and other fora simply as a way to lecture to other countries. 537 Others detected more of a difference within departments or among personalities. At the State Department, for example, the initial optimism and confidence of the Clinton administration?s first term for multilateralism, when Winston Lord was the Assistant 534 Interview with author, Seoul, Korea, April 12, 2005 535 Interviews B33 and B38 536 Interview B38 537 Interviews B3 and B37 186 Secretary, faded in the second term under Stanley Roth. 538 In the Bush administration, despite its professed skepticism of multilateralism, the State Department?s participation has generally actually been more senior than the second term of the Clinton administration. It seems that Don Keyser and/or Jim Kelly, who had previously run CSCAP, were the sources of the State Department?s recent support. 539 Shirk acknowledged that her connections from her service with the State Department may have also helped with career employees at lower levels of the bureaucracy. 540 Meanwhile, the Bush administration?s Pentagon has never even sent a participant, and NEACD staffers have perceived little to no interest from their attempts to get Bush administration Pentagon participation. 541 Beyond those members who attend the meetings, how else might other government officials benefit from NEACD? Although whether and how individual attendees brief their colleagues varies, one government participant cited the potential for things learned from the meeting to shape their talking points that are vetted up through their own bureaucracy and into the interagency process. They cited an example in which a brief discussion they had on the margins of the most recent NEACD meeting?where a participant questioned whether it was helpful to publicly call North Korea ?an outpost of tyranny,? a term Secretary of State Condeleezza Rice had used in her January 2005 Senate confirmation hearings?might affect how future talking points are shaped. 542 Whether it did, and whether it would have made it through the interagency process, is not known. 538 Interviews B1 and B31 539 Interview B38 as well as author?s interview with Keyser, Fairfax, VA, June 15, 2005 540 Author?s interview with Susan Shirk, Washington, D.C., May 17, 2005 541 Interviews B4 and B31 542 Interview B29 187 A second possible channel is if government officials are briefed by NGO participants. One Senate staffer interviewed for this study mentioned that they had been briefed by Shirk as well as some of the other U.S., South Korean, and Chinese participants. 543 Another potential channel is for alumni of the process to be informed about more recent meetings, although two former participants?Tom Hubbard, a former PDAS at the State Department who went on to become ambassador to South Korea, and Jim Steinberg, the former policy planning director and deputy national security adviser? each said that they had lost track of NEACD, despite their continued engagement in Northeast Asian affairs. 544 A final potential channel is the revolving door. Here, however, the effect practically has been limited to Shirk herself. In no other case has a U.S. nongovernmental NEACD participant subsequently gone on to a government policymaking position. (Steinberg and Campbell participated in the other direction: as government policymakers who then left, and Einhorn participated after he left government.) Scalapino did not remember specifics but said that, although he has not served in government, in cases where he has been asked to brief government officials ?there is no doubt that contacts with individual government and NGO representatives in non-governmental agencies like NEACD help shape the views which I pass on.? 545 Shirk herself explained that what she was able to take into government from the first five or six NEACD meetings was constrained because her State Department portfolio was limited to China, not to developing multilateralism or a formal government 543 Interview A21 544 Author?s phone interview with Thomas Hubbard, July 28, 2005, and author?s interview with Jim Steinberg, Washington, D.C., August 22, 2005 545 Interview with author, Seoul, Korea, April 12, 2005 188 dialogue on Northeast Asia. Nevertheless, her experience with NEACD?s informality and the fact that it did not require decisions at the meeting?s conclusion, she said, helped convince her that such a process might help shape, or at least better understand, the evolution in China?s strategic thinking. Shirk subsequently was one of a number of government officials that helped to embed ?more informal discussion time with the Chinese, that we called a strategic dialogue, which was not about bilateral issues per se but what?s going on in the world, and what do we think about it, and what are our interests? at every level of government interaction. 546 Separately, two former government participants in NEACD each mentioned another effect of the revolving door on the network: they felt that Shirk?s return from government helped her better understand what government officials were interested in learning, making NEACD even more valuable for government participants in recent years. 547 Despite the potential for talking points to be shaped, current policymakers to be briefed, or future policymakers to have their views influence by NEACD, it seems that most of NEACD?s access to the U.S. government is dependent on official participation, which has varied over time depending on the administration, the government department, and even individual interests. Agenda Setting Although NEACD meetings are very loosely structured to allow participants, particularly Foreign Ministry speakers who kick off the tour d?table of perspectives on regional security, to raise whatever issues they want, there is little evidence that the network put 546 Interview with author, Washington, D.C., May 17, 2005 547 Interviews B34 and B36 189 any nonproliferation issues on the policy agenda. One participant mentioned that missile defense had been a big issue at one time, 548 but North Korea has clearly become the focus of the organization. Instead, NEACD has followed recent events, not anticipated unforeseen ones or focused on overlooked items, as an agenda setting organization would do. 549 The initial presentations by government speakers has tended to focus discussion on current policy, rather than setting a new agenda. 550 One former government policymaker raised his doubts about the agenda setting potential for meetings like NEACD: Every day?12 hours a day, well 24 hours a day, really?but everyday we?re getting inputs from wherever it comes from. While it?s not 100 percent impossible that a significant new input would arise from this kind of a meeting, it would be very unusual because you have to hypothesize that somebody has just finished a seminal study in something who decided to make it known for the first time there, at that forum, to produce an interesting conversation about it. Is it possible? Sure, it?s possible but, again, there are so many inputs every single day from so many different sources?from foreign governments, intelligence, U.S. and foreign academics and people, from business community people, the parliamentarians of the world, you name it?you?re basically deluged with information so, at least speaking for the U.S. government, I would say no [new issues are raised]. 551 Policy Options Similarly, there is little evidence that policy options on any nonproliferation issue, including North Korea, have been raised. 552 According to one former DoE employee involved in funding decisions, in its early stages the network may have helped shape a 548 Interview B31 549 Interviews B1, B4, B29, and B30 550 Interviews B30, B32, and B37 551 Interview B34 552 Interviews B3, B33, and B34 190 program to can and store spent fuel under the Agreed Framework, 553 but that is the only evidence that the network has had any specific policy impact. 554 From the funder?s perspective, according to Ed Fei who headed DoE?s nonproliferation policy shop and initially funded the network, NEACD was not founded to perform traditional think tank roles like setting agendas or developing policy options. The key was to develop a learning process about, and dialogue with, China and North Korea, at a time when there had been little contact with either, and subsequently to help build regional stability. 555 Shirk herself said NEACD has not really tried to develop policy options, change, or even discuss specific policies. It has been more of a quasi-official opportunity to allow diplomats to talk to each other informally. 556 Although one DoE official involved in funding emphasized that networks should always strive to develop creative policy options or otherwise change policy to be relevant, it was extremely difficult to fault NEACD for not having had that effect, at least yet, because of the complexity of the North Korean nuclear issue. 557 Other governmental and nongovernmental participants also emphasized that the key was not developing policy options or changing policy, but the process itself, or socialization: to develop people-to-people relationships, share information as well as opinions, and gain a better understanding of the perspectives of the other Northeast Asian countries among mid-level officials who might later rise to even more prominent positions within their own government. 558 Part of the appeal and uniqueness of the dialogue, in fact, was that decisions were not even sought. As one nongovernmental 553 Interview B33 554 Interviews B2 and B34 555 Interview with author, Washington, D.C., July 18, 2005 556 Interview with author, Washington, D.C., May 17, 2005 557 Interview B33 558 Interviews B2, B37, B38, and B41 191 participant said, ?it?s not as if a conclusion or joint position is expected to emerge out of that. That?s not the purpose of the exercise.? 559 Observations Although this study focuses on whether transnational security policy networks have changed U.S. nuclear nonproliferation policies or the perceptions that shape them, a number of other issues related to the network?s operations emerged during interviews. Rather than ignore them entirely, they will be briefly mentioned here for others who may examine different aspects of the network?s utility. Although many of these issues have been touched on earlier in this chapter, five issues related to NEACD will be discussed further here: the network?s potential drawbacks, its membership, potential socialization of Chinese and North Korean participants, the potential impact on other countries? policies, and NEACD?s sustainability. Drawbacks? Every subject interviewed for this case study was directly asked whether NEACD was harmful in any way. Not one came up with any empirical examples or even suspicions that led them to believe that NEACD was harmful. Although many did not believe there were any negative implications at all, 560 a few raised hypothetical concerns worth mentioning. The most common concern was the potential opportunity cost of either money invested in NEACD or the time officials spent traveling for the meetings, although none suggested specific alternatives where money or time would be better 559 Interview B41 560 Interviews B1, B3, B36, B38, B39, and B301 192 spent. 561 Other than that, one participant mentioned that discussions leaking to the media was always a concern, although none had to this point. 562 Another nongovernmental participant wondered whether it might theoretically be possible if NEACD, or any other similar organization, disappointed its participants, it might undermine multilateralism or track-two in general. 563 A final set of concerns raised is the potential for misperceived signals. Although one participant mentioned the possibility that U.S. participants might misperceive a foreign individual?s opinion as a shift in policy, 564 the more frequently discussed contingency in this study was whether foreigners, particularly North Koreans, misperceived signals from Americans. At least one participant thought it was theoretically possible, 565 but other current or former government officials believed that even North Korea was unlikely to perceive any NEACD conversation as a shift in U.S. policy. In fact, they said, NEACD was more likely to give Americans the opportunity to clarify apparent dissent in the U.S. media than it was to reveal some sort of dissension. 566 They also argued that even North Koreans were beginning to understand U.S. politics enough to understand the perspectives that certain individuals might bring to network discussions. 567 The one common bias that all participants might convey, one government staff member said, was that all participants assumed that: it is at least possible that negotiations could result in a denuclearized Korean peninsula. There?s another whole subset of the policy community that has lost faith utterly in a negotiated solution to this mess, and does not participate either in 561 Interviews B31, B33, B40, and B41 562 Interview B30 563 Interview B37 564 Interview B2 565 Interview B31 566 Interviews A21, B29, B32, B34, and B301 567 Interviews A21 and B32 193 track-two or formal dialogue, and is preparing for other contingencies. So, the voice that is missing is the voice of the true skeptic who sees most of this as a waste of time.? 568 [emphasis in interview] Network Membership Because the benefits of NEACD are so confined to its participants, membership is a crucial issue, both from the United States and other countries. The level of U.S. participation has already been extensively discussed, but five sets of questions about participation were raised during interviews: the continuing value of the participation of the defense community, potential Congressional participation, the experience of nongovernmental members, foreign participation, and efforts to broadcast the value of meetings beyond participants. Initially, including military officers and even defense ministry officials may have had substantial benefits, but lower-level or nonexistent Pentagon participation, limited engagement in dialogue by attending officers and officials from all countries, reportedly limited informal interaction with non-military officers, and the emergence of the parallel Defense Information Sharing (DIS) process since 2001 raised questions among at least among some participants about the value of the defense community?s continued participation in the main NEACD meeting. 569 On the other hand, one academic and both U.S. military participants interviewed for this study reported that they learned from the entire spectrum of perspectives within the government?foreign, defense, and military? as well as nongovernmental members. 570 One other nongovernmental participant also emphasized the potential value of ?socialization to diplomacy? for military officers, even 568 Interview A21 569 Interviews B4 and B31 570 Interviews B2, B29, and B30 194 if they did not speak, who otherwise would presumably have few diplomatic opportunities. 571 One U.S. military officer?s claim that they personally were much less comfortable speaking at the first NEACD meeting, compared to more recent ones they attended, provides an example of this potential military socialization. 572 Whether that value justifies continuing to include the defense community at the plenary sessions as well as the DIS process, and whether the DIS process sufficiently socialized military participants, seem reasonable questions given the benefits that others could gain from attending NEACD and the countervailing pressure to keep discussions intimate and manageable. Theoretical evidence presented in chapter two led to a question about the potential participation of an entirely different branch of government: should a Congressional member or staff member participate in NEACD? The answer was no. Even a Congressional staff member argued against it. They explained that Congress does not play enough of an operational role in diplomacy to get involved in fora like NEACD. That staffer did argue, however, that it might be in NEACD?s interest to invite, or offer to brief, a member like Senator Biden, Bingaman, or Lugar in order to raise awareness of a forum like NEACD, even if it is unlikely that the Senator or a member of their staff would come. If NEACD wanted to approach a Senator in the future with an idea derived from the meetings, the Senator may have heard of the network, making it easier to gain access. 573 In the meantime, because network discussions have been so detailed and because Congress? role in promoting think tank ideas rests in agenda setting, which NEACD does not seem to do, Congressional involvement makes little sense. Briefing or 571 Interview B41 572 Interview B29 573 Interview A21 195 even just offering to brief Hill members, however, may be a useful down payment on future access and influence. A third membership issue raised during interviews was the experience of U.S. nongovernmental participants, particularly whether it might be beneficial if more had government experience. Previously in this chapter, the unique role that Bob Einhorn was able to play with the North Koreans was cited, because of his governmental experience and the credibility he brought to the discussion of U.S. nonproliferation policy. In contrast, one former career governmental policymaker went so far as to say about those without government experience: ?I don?t want to be harsh, but I could not point to any value from [the academics].? 574 Other participants disagreed, at least in principle, arguing that academic participants such as Robert Scalapino might stretch a policymaker?s thinking in new directions, potentially providing a unique perspective rather than just having another policymaker around the table, particularly one who was not currently serving in government. 575 Beyond Scalapino, though, a former government policymaker who even supported academics as theoretically more useful to the discussions said that, in practice when he attended in the mid-1990s: ?I don?t even remember who [the other academics] were, so they obviously didn?t leave any impression on me.? 576 What makes career academic participants different seems to be their more distant, and arguably strategic, view of contemporary challenges and policies. Yet that difference may also have drawbacks, limiting or altering the benefits that some of them might get from NEACD. One academic participant who had not worked in Washington described 574 Interview B34 575 Interviews B31 and B39 576 Interview B39 196 their individual concerns as much more strategic than tactical, and explained they found the specifics of current policy less relevant to their own interests: [Y]ou need to understand that even though I participate in these track-two diplomatic activities, and a lot of the focal points of their discussion are contemporary events and contemporary security policy and so forth, deep down inside I?m basically looking for broad structural causality, and big changes and big processes?Intuitively, I tend to be more interested in 5- year, 10-year shifts. 577 In contrast, those participants from the revolving door did focus more on the specifics of policy. One argued that government experience led them to approach issues differently, in a way that was more useful to government policymakers: ?No matter how sophisticated somebody who has never been in government is about these things, they just don?t quite see it the same way or understand it with the same kind of intuitiveness that former officials have.? The simplest explanation that person gave is that those with government experience understood the political pressures that might make an ?intellectually elegant? solution practically infeasible. Those with government experience might also go back into government again, making them both potentially useful assets who might take NEACD lessons learned with them, and even raise the incentive for current government participants to talk to them, 578 increasing the likelihood that NEACD would facilitate frank discussions. Another revolving door member pointed out that they might have individually worked with certain officials, either from the U.S. or other governments. Those personal bonds formed might make it more likely that officials would open up either in formal NEACD sessions or at least informally at meals in between sessions. Even if they hadn?t 577 Interview B37 578 Interview B40 197 personally worked with them, that same participant explained that: ?I know better how to extract [information] from a government official, having been a government official. I know how government officials approach kinds of things like that. I know what defense mechanisms they put up. I know how to break down those defense mechanisms. In that respect, I?ve benefited from being on both sides.? Finally, that participant explained that revolving door members would know more about what motivated U.S. policy from experience and would be at liberty to explain it, particularly to foreign participants, since they were out of government. In turn, sharing that information about the United States might even build trust with foreign officials, and lead them to reciprocate with insight into their own country?s political dynamics. 579 In the end, those that had served in government and those that had not each brought different perspectives to NEACD meetings. Part of the issue revolves around the perceived role of nongovernmental participants: one former government policymaker asserted that their role was exclusively to ?listen and learn? what was really going on in formal negotiations, 580 while others pointed to a catalytic role that they could contribute to NEACD discussions. There seems to be a legitimate debate about what background would be more valuable for NEACD: those who are not as involved in current events and might theoretically look at an issue a different way, or those who follow current policy more closely but are not burdened with the day-to-day demands or constraints of governance while being able to lend their own experience. Less debate, though, seems to exist about who would take away more from the meetings: those who more closely follow current events do. That also makes them a 579 Interview B32 580 Interview B34 198 better audience for public diplomacy, as they may be more likely to spread that information around their own networks. The issue is not whether participants without government experience found the network valuable or contributed anything; they did. The issue is whether at least one nongovernmental U.S. participant with government experience may contribute, and certainly derive, different and potentially even greater benefits from NEACD meetings. A fourth participation issue raised was the potential to distribute network benefits beyond each country?s five participants and their own personal networks. 581 The trick is: how can you distribute NEACD?s findings more broadly without jeopardizing its confidentiality? One nongovernmental participant suggested that meetings be held more frequently to help build personal ties even further. 582 It seems, however, that proposal would exacerbate the strenuous demands already placed on senior government participants. Alternatively, NEACD has experimented with different ways to try to disseminate its experiences. Since at least 2003, the University of California Center in Washington has hosted a briefing by attendees of the most recent NEACD meeting, including Shirk herself. It is unclear, however, how many government officials or even members of the D.C. think tank and university communities have attended. Part of the problem may be that the UC Center does not otherwise frequently host members of the Asian expert community in Washington. Another outreach effort is the virtual network, Wired for Peace, that UCSD established in 1998 to sustain dialogue among conference participants and NEACD alumni in a password-protected forum. According to network staff members, however, participants simply did not engage in the online forum either 581 Interviews A20, B3, and B41 as well as Lawrence T. Woods, ?Learning from NGO Proponents of Asia- Pacific Regionalism: Success and its Lessons,? Asian Survey 35:9 (September 1995): 826 582 Interview B37 199 because they were too busy or were not comfortable posting anything interesting, and Wired for Peace withered. As one NEACD staff member said: ?it was an interesting experiment that didn?t work. We have pretty much given up on it now.? 583 Distributing NEACD?s benefits beyond its participants remains a challenge for the network. The fifth, and final, membership issue raised dealt with foreign members. The biggest problem expressed about non-U.S. participation has been Russian governmental attendance, which generally has not been high-level or in some cases even knowledgeable at all, 584 while those that have attended have been restrained in open sessions. 585 One participant mentioned that at least the Russian academics were useful, providing potential insight from bilateral Soviet Cold War and current Russian ties with North Korea. 586 Beyond Russia, South Korean Foreign Ministry participation was also brought up as a problem by one participant because that Ministry does not control North Korean policy in Seoul; the Unification Ministry does. Although a representative from the Unification Ministry has attended NEACD in recent years, the South Korean Foreign Ministry participants have had limited utility. 587 Socializing China and North Korea Although these questions were raised about some Russian and South Korean participation, NEACD participants praised Shirk for her ability to get useful network participation from China over the years. 588 In the beginning, Beijing?s participation was 583 Electronic messages from B31 as well as B4 584 Interviews B2 and B4 585 Interviews B30 and B34 586 Interview B34 587 Interview B38 588 Interview B3 and B34 200 reportedly extremely rigid, 589 but has evolved to the point where U.S. participants believe the Chinese are the ?most serious,? well prepared, independent (at least non-American), and informative NEACD members. 590 One participant noted that some limits to Chinese candor certainly still existed. For example, they would not openly criticize North Korea in the formal sessions, but would privately criticize Pyongyang on the margins of NEACD meetings, particularly for not undertaking political and economic reform quickly enough. 591 Nevertheless, the question remains: to what extent has, or can, Chinese and North Korean NEACD participation helped integrate these countries into Northeast Asian dialogue and multilateral approaches to subregional problems? For some, this process of socialization has been the major purpose of NEACD. 592 The analytical question often asked about socialization is whether fora like NEACD helped change China, or whether China?s participation in NEACD demonstrated its engagement in regional affairs after the collapse of the Soviet Union. In other words, it?s the analytical equivalent of the chicken or the egg: did Chinese participation change their engagement, or did a Chinese decision to engage cause their participation? In NEACD?s specific case, the limited participation to five members makes the case for broad socialization more difficult to make. Most Americans interviewed for this study felt that any socialization was not obvious from one meeting to the next, and that the limited NEACD participation would have made widespread Chinese socialization difficult. 593 Nevertheless, change over time has been clear. One participant argued that 589 Interview B36, and electronic message from former Deputy Assistant Secretary of State for East Asian and Pacific Affairs Charles Kartman, July 11, 2005 590 Interviews B34, B37, and B40 591 Interview B30 592 Interviews B31 and B41 593 Interviews B3, B34, and B36 201 the biggest effect fora like NEACD has had on China is to develop its nongovernmental participants ?to the point now that their think tankers are as good as anybody in the world.? That participant elaborated that informal dialogue helped raise ?the comfort level and sophistication of the analysis of Chinese scholars?and made them much more savvy about how to think about these kinds of policy issues.? That participant contended that networks like NEACD helped Chinese analysts and scholars break out of their previous ideological constraints from the Communist system and ?I think the dialogue exposed them so they could see the difference between what they were doing and what good, independent policy scholars were doing.? 594 The goal or hope, of course, is that North Korea will follow in China?s footsteps. One question immediately raised is why North Korea began to participate in NEACD in 2002, ten years after it was one of the founding members of the organization. According to Shirk, the answer probably stems from two sources. First, she reached out to Han Sung Ryol, who had returned to the UN in the late summer of 2002. He had hosted Shirk in Pyongyang in 1992 when NEACD was initially established, and later visited Shirk in San Diego in the mid-1990s, when he was initially posted at the UN. Shirk said she called him in 2002 to appeal for his help to get North Korea to participate. With his ties in Pyongyang still strong, he agreed to try. A few weeks later he officially committed North Korean participation, as long as Shirk was willing to come to Pyongyang first as part of an unofficial bilateral discussion. The other reason, she cited, was that North Korea was seeking to establish better relations around the region, with Kim Jong Il?s summit meeting with Japanese Prime Minister Junichiro Koizumi in September 2002 the most obvious example, as well as 594 Interview B40 202 with the United States. A few months earlier, in January 2002, President Bush had labeled North Korea part of the ?axis of evil,? and the United States was in the midst of building its case at the UN and elsewhere to change the regime of another member of the so-called axis, Iraq, which it did a few months later in the spring of 2003. Through some combination of timing and personal connections, North Korea agreed to participate in NEACD in 2002. 595 How much of an effect the three meetings from 2002-04 have had on North Korea is unclear, as discussed earlier. North Korean participants apparently travel in pairs, presumably to monitor one another even on the margins of meetings, inhibiting the potential advantages of informality. 596 Yet such restrictions apparently don?t completely prohibit informal opportunities. Amazingly, a couple of younger North Korean participants at the 2004 meeting in San Diego reportedly went to a party in a graduate student?s dorm after the formal session, but it?s hard to know what effect, if any, something like that may have. 597 Whatever the effect may be on North Koreans, their participation affects the character of the NEACD dialogue among all other members. When they do attend, participants agree that discussions have been more formal, often enduring a series of North Korean talking points, and inhibited participants from speaking more frankly about how to deal with the DPRK. 598 Most recently, North Korea did not attend the 2005 NEACD meeting in Seoul, declaring it would not send its Foreign Ministry-affiliated think tanks and officials to South Korea which it treated as domestic territory. The 595 Interview with author, Washington, D.C., May 17, 2005 596 Electronic message from B4 597 Interview B37 598 Interviews A4 as well as B30 203 discussions at that meeting contrasted with the previous three years. The most recent NEACD meeting included a general, and of course unclassified, discussion of whether countries believed that North Korea had an HEU program as well as whether they had assembled nuclear weapons, and whether countries were willing to accept a nuclear North Korea; whether and when countries might be willing to refer North Korea to the UN Security Council or impose sanctions; the potential effects of pressure on North Korea; what leverage countries might wield as well as what signals they might send to encourage North Korea back to the six-party talks; what energy assistance the U.S. might give North Korea as part of any potential agreement; the effect of the new Rice/Hill team at the U.S. Department of State; and the state of progress of North Korean economic reform as well as the political stability of the North Korean regime. 599 On the other hand, even if the most recent discussions were more frank, a central purpose of NEACD has always been to open a dialogue channel with Pyongyang and, at least from the U.S. government?s perspective, to be able to reinforce formal negotiations in a more informal and unofficial, but still multilateral, setting. 600 At the last 2004 meeting that North Korea attended, discussions included debate about whether the United States had a hostile policy toward North Korea; participants? assurance that a Kerry administration would not pursue a different policy toward the DPRK; debate about the desirability of the CVID formula for regional disarmament; and inquiries about national positions on PSI. 601 In the end, one nongovernmental participant concluded that ?maybe there is a certain advantage in the North Koreans showing up every other time.? 602 If 599 Author?s notes from NEACD-16, Seoul, Korea, April 12-13, 2005 600 Interviews B34 and B301 601 Author?s notes from NEACD-15, La Jolla, California, April 5-6, 2004 602 Interview B30 204 North Korean participants were socialized over time, the discussions would presumably get increasingly frank and informal. NEACD?s Effects in Other Countries Whatever the effect on North Korea, NEACD arguably has had effects on other, non- American participants. For example, NEACD staff members argue that the network has facilitated interagency cooperation in other countries. At least Chinese participants, as well as possibly Japanese and South Korean members, reportedly meet as a delegation to review the presentation their Foreign Ministry representative will deliver. 603 Even if nongovernmental participants from China are often from the Foreign Ministry-affiliated think tank, it is unclear how much other communication exists between the Foreign Ministry official and the nongovernmental members that attend, not to mention between those participants and the Defense Ministry and military representatives. NEACD funders and staff also cited the evolution in Chinese participation as well as the high level of interest and participation from at least China, Japan, and South Korea as evidence that those countries see value and potential in the network. In contrast, NEACD has had to fight to get anyone from the Pentagon, and at times a sufficiently high representative from State or the Russian government, to attend. 604 Sustainability Although North Korean NEACD participation and rising regional tensions over the nuclear issue have boosted the network?s recent utility, two concerns about its 603 Interviews B31 and B41 604 Interviews B3, B4, and B31 205 sustainability were raised. First, some had bureaucratic concerns about the stability of DoE funding for the network. The initial funder, Ed Fei, had come over from the State Department and, as he said, ?implemented a State agenda with DoE money.? 605 The idea has been to build confidence among the countries that participate and reduce tensions that can cause proliferation. 606 According to one official previously involved in funding decisions, DoE was not historically involved in regional security. As it expanded into nonproliferation, it initially used NEACD to help understand the region and how to implement its programs more effectively. The question now is whether institutional support would continue. Some within DoE, according to one former employee, seemed to be beginning to question why the Department of State was not funding the program, although that person was not aware of any plans to discontinue funding for NEACD. 607 In a broader sense, as one former NEACD staffer argued, a program that relied on intangible results such as socialization was vulnerable to criticism from those who did not intuitively believe its value, including funders seeking to scientifically evaluate what to do with their money and whether their investments were getting sufficient returns. 608 Beyond the financial issues, a strategic question was also raised about the network?s sustainability. Because NEACD seems recently to be so focused on North Korea and its participation, participants wondered what the network?s value might be if North Korea becomes integrated into the region as China has, or at least if the six-party talks become more formally institutionalized. 609 In other words, if NEACD is successful in socializing North Korea or establishing permanent six-party talks, would it have 605 Interview with author, Washington, D.C., July 18, 2005 606 Interview B38 607 Interview B33 608 Interview B41 609 Interviews B2, B3, B36, and B40 206 planted the seeds for its own irrelevance? One possibility, as one former government policymaker mentioned, is that nongovernmental NEACD participation would become increasingly important if an official dialogue existed. 610 A former funder speculated that NEACD might become ?a think tank-type exercise that serves a formal structure in Northeast Asia, the way CSCAP has sought to do that for ARF.? 611 Another government policymaker also mentioned that network participants might be different. 612 Shirk herself felt it was important to have ?more frequent interaction among the director generals of the Asia departments of all of these Foreign Ministries?the guys who have real authority and are handling the regional policy, dealing with problems?otherwise it?s just a photo- op.? 613 If any formal six-party talks were only held at the foreign minister or head of state level, NEACD might provide a forum for more substantive, working level talks. Shirk also acknowledged that, even now, NEACD was considering evolving since the six-party talks have tenuously become more regular, at least on the North Korean nuclear issue. Evolving into a parallel organization that put more emphasis on nongovernmental participation was certainly possible, if not likely. But Shirk herself said that ultimately: I?m all but prepared to go out of business?.I think we might become more purely academic and maybe we could work on particular projects that they want to pursue as cooperative undertakings. I don?t really know?.I think it would be a mistake to have this all planned out. I also really don?t want to be like the March of Dimes, you know, polio goes away and the organization stays on. In other words, I?m not into organizational maintenance. We have plenty of other things to work on. If we create a real, ongoing official process in Northeast Asia, I will feel that 610 Interview B39 611 Interview B33 612 Interview B301 613 Interview with author, Washington, D.C., May 17, 2005 207 we contributed somehow?.If there?s no demand side, we can go back to other things. 614 Conclusions Although the network seems to have had little impact on the U.S. nonproliferation policy process?its agenda, options, or decisions?that has not been its purpose. Rather, NEACD has sought to build relationships among influential mid-level officials, who might rise to even greater heights, and to facilitate information sharing among them, increasing understanding about each other?s interests, motivations, policies, and political pressures shaping those policies. It has been more of a tool of diplomacy than of policymaking itself, although the two are certainly related. The network even served as an institutional model and trial for the six-party talks themselves. For U.S. government participants, NEACD has not necessarily helped establish many valuable new contacts, but it has provided the potential for individuals to learn from catalytic nongovernmental members, who might reshape policy debates in certain ways; from the informal nature of dialogue, allowing participants to go behind and beyond formal talking points; and from being able to observe the interaction of other participants in a multilateral setting. NEACD has also provided a forum for U.S. government participants to conduct diplomacy, reinforcing the six-party talks with North Korea, with other participants, and with the foreign policy elite. With North Korea, the network has provided a forum enabling U.S. government participants to preview significant issues at upcoming meetings; to follow up on recent formal meetings, particularly by clarifying media reports and reinforcing messages; and potentially to have former officials elaborate on U.S. policy in order to demonstrate both its bipartisan 614 Interview with author, Washington, D.C., May 17, 2005 208 support and unofficially explain it further than government officials are comfortable doing. With other parties in the six-party talks, NEACD has also provided a forum to conduct multilateral diplomacy, reinforcing messages in front of all the parties simultaneously, and clarifying media misperceptions. Finally, U.S. government participants have been able to conduct public diplomacy, generally signaling a willingness to engage North Korea, and more specifically enabling nongovernmental participants to see what dealing with North Korea is really like, as well as for U.S. officials to brief a few nongovernmental experts from other countries about current U.S. policy in an off-the-record setting. For U.S. nongovernmental participants, NEACD has provided an opportunity to establish contacts with Northeast Asian government officials, and to get a better understanding of North Korea and of current policies from all six parties. Some legitimate questions exist, given the network?s experience and its evolving utility, about whether the very limited number of seats around NEACD?s table is allocated most effectively. Nevertheless, it seems fair to say that the network as a whole, to this point, has been focused on maximizing benefits for government officials from six countries that in some cases have had only nascent or even no contacts with one another. 209 CHAPTER FIVE: THE PROGRAM ON NEW APPROACHES TO RUSSIAN SECURITY The Program on New Approaches to Russian Security (PONARS) ?is a small network of scholars who analyze Russian and Eurasian security issues and the region?s role in international affairs.? 615 First convened in 1997, PONARS has two objectives according to its founder, Celeste Wallander: ?to build a scholarly network of the leading social scientists in the United States and Russia (and other post-Soviet countries) in order to allow members to produce the best possible scholarly work, and to make that scholarship and those scholars known and useful to the policy world.? 616 PONARS began as just twelve young American scholars when Wallander started the network while she was at Harvard. It has been entirely funded by foundations, initially and primarily by the Carnegie Corporation of New York. According to interviews with both Wallander and David Speedie, who was the program chair in Preventing Deadly Conflict at Carnegie when PONARS was founded, the idea was to better understand Russia as a country in its own right, then more than five years after the Soviet Union had collapsed. These junior faculty members brought both a post-Soviet perspective on Russia, with some having recently lived in the country, and an interdisciplinary expertise in politics, economics, demographics, society, or culture, not just the defense issues which had dominated the Cold War. 617 615 From web description of PONARS, http://www.csis.org/ruseura/ponars/index.php?option=com_csis_progj&task=view&id=202 (accessed October 26, 2005), hereafter referred to as PONARS web description. 616 Celeste Wallander, ?Borderlands and the Value of Academic Research for Policy: A Case Study,? Asia Policy 1 (January 2006): 37. 617 Author?s interviews with David Speedie (telephone interview, November 14, 2005) and Celeste Wallander (Washington, D.C., November 17, 2005) 210 Over the next nine years, the network grew slowly and selectively. PONARS sought young members, primarily academics not think tank analysts. 618 It sought geographical diversity, not just the standard suspects from Ivy League and West coast institutions. 619 Members had to be Russia-area specialists who spoke Russian, not generalists who happened to work on Russia. 620 They are also supposed to be cordial, as working well with others was deemed important to the network (although some questioned whether the selection process had successfully screened for affability). 621 Beginning in 1999, the Macarthur Foundation gave additional money to help support the network?s expansion to admit first Russian members, and later Ukrainians and eventually Belarusians. 622 Partially to help manage this growth, in 1999 Wallander established an Executive Committee (ExCom), which also helped make other administrative decisions. 623 About seven members?Russians or other members from the former Soviet Union as well as Americans?make up ExCom with Wallander serving in an ex officio capacity. The committee makes membership decisions by majority vote, and in 2002 began both to suggest an agenda for PONARS conferences and to decide among members? applications to present at them. 624 In July 2001, PONARS moved its base of operations to the Center for Strategic and International Studies (CSIS) in Washington, D.C., where Wallander became the 618 Author?s interview with Wallander as well as C23 619 Author?s interviews with Speedie and Wallander 620 Author?s interview with Wallander as well as C26 621 Interviews C43 and C481 622 Author?s interview with Wallander 623 Celeste Wallander, principal investigator, PONARS Final Report on Grant B7308 (July 1, 2001-June 30, 2003), submitted to the International Peace and Security Program, Carnegie Corporation of New York (September 30, 2003): 7, hereafter referred to as 2003 Final Grant Report. 624 Wallander, PONARS 2003 Final Grant Report, pp. i, 12 211 director of its Russia and Eurasia program. By the end of 2005, PONARS had grown to 80 members?with just over forty percent from Russia, Ukraine, or Belarus. 625 Exclusively for those members, the network hosts a listserv that typically produces about 10-20 e-mail messages a day, 626 and convened a handful of academic conferences, which were funded until 2003. For non-members, PONARS produces no consensus documents but distributes products of its individual members. Two publication series are sponsored: scholarly working papers?about two dozen 20+-page academic essays?and over 400 about 2000- 2500-word policy memos, geared for government analysts. 627 Most memos are distributed for and discussed at PONARS?s premier public event: an annual day-long conference for the Washington policy community, including government officials and employees, journalists, businesspeople, other university scholars and think tank analysts. Prior to a typical policy conference, a book of about 25-30 policy memos authored by PONARS members was circulated to those who RSVPd. In recent years, more than 200 people replied to attend the event, generally held on a Friday in December, at which most memo authors would present their work for about 10 minutes each on a series of panels of 3-5 authors, and subsequently engage in 45-60 minute discussions with the audience. For this case study, primary written materials analyzed include the 26 PONARS memos on nonproliferation issues, and summaries from seven of the nine policy conferences held through the summer of 2005 (1997 and 1999 were not available). In addition, 28 original interviews were conducted with staff, funders, members, 625 Celeste Wallander, principal investigator, PONARS Final Report on Grant B7408.R01 (July 1, 2003- June 30, 2005), submitted to the Carnegie Corporation of New York (September 30, 2005): 5, hereafter referred to as 2005 Final Grant Report. 626 Wallander, ?Borderlands,? p. 40 627 PONARS web description 212 government attendees at the policy conferences, and senior U.S. government policymakers. PONARS members were sought for interviews if they specialized on nonproliferation or traditional security affairs, and attended three or more of the five policy conferences for which participant lists were available in the fall of 2005. Government attendees at the policy conference were also sought if they attended three or more of those five policy conferences. Table 5.1 summarizes the interview candidates, categorized by their relationship with PONARS and whether they were primarily NGO or government participants. Some interviews were not conducted either because candidates had retired, deceased, or otherwise could not be found, declined to be interviewed, an interview could not be arranged, or did not respond to requests to be interviewed. Table 5.1 PONARS Interview Candidates NGO USG Total PONARS Staff 3 - 3 PONARS Funders 4 - 4 PONARS Policy Conference Participants 16 24 40 Attended 3 or more policy conferences 14 20 35 DASs, NSC, or Hill staff attending twice 628 2 629 4 5 Senior government policymakers 630 - 4 4 Referrals from other interviews 1 5 6 Total 24 33 57 Fifteen of the 28 interviews conducted were with U.S. government officials, three of whom are members of the revolving door but were in government for significant periods of time or in senior policymaking positions. The other thirteen interviews were 628 Because of their positions in the policymaking chain, Deputy Assistant Secretaries were sought for interviews even if they only attended two meetings. Because of the theoretical evidence in chapter two, Hill staffers were sought for interviews even if they only attended two meetings. 629 Two current or former members of PONARS served on the NSC staff during fellowships, and are therefore classified as both PONARS members and NSC staff. 630 Senior government policymakers are defined as Assistant Secretaries of State for European Affairs, their equivalent for Russia, or senior directors of the NSC staff. 213 principally with PONARS members, staff, or funders, although three of those thirteen were also members of the revolving door who had government experience before or after their PONARS participation. In other words, of the 28 interviews conducted, twelve were with career government employees, six were with members of the revolving door, and ten were with nongovernmental employees who had not been employed in government. Categorizations of interviews actually conducted are listed in table 5.2. Throughout this chapter, opinions expressed in only a single interview, whether cited anonymously or specifically, will be noted as one person?s opinion. Table 5.2 PONARS Interviews Conducted NGO USG Total PONARS Staff 2 - 2 PONARS Funders 1 - 1 PONARS Policy Conference Participants 10 10 20 Attended 3 or more policy conferences 8 9 17 DASs, NSC, or Hill staff attending twice 2 1 3 Senior government policymakers - 2 2 Referrals from other interviews 0 3 3 Total 13 15 28 Having introduced the network?s infrastructure as well as the primary materials and interviews utilized, the subsequent three parts of this chapter will contain descriptions of the issues PONARS has covered; an evaluation of the network based on the four-part framework presented in the introductory chapter; and finally a series of observations and conclusions about PONARS. Chapter six will include comparisons of PONARS with the other two case studies. This chapter will be limited to PONARS itself, starting with the issues the network has addressed. 214 Issues PONARS Has Addressed PONARS has always sought to address not just traditional security challenges like proliferation, but also the new, internal challenges that Russia and the region face as they transition from their former Soviet union to independent statehood. More than any topic focusing on Russia (see table 5.3), especially in recent years, PONARS members published memos on trends in and policy toward the surrounding former Soviet region, such as the Caucasus, Central Asia, 631 and Ukraine 632 (including the 2004 Orange Revolution). 633 Among the topics that focused just on Russia, the economy and related socioeconomic challenges garnered more network activity than any other. A panel on some aspect of the Russian economy was held in all but one of the eight conferences for which agenda are available. In 1998, the ruble collapsed, becoming a major topic of network conversation focusing on the causes of the financial crisis and whether assistance should be provided to Moscow. 634 Just five years later, the economic outlook had changed dramatically with the central government reasserting some control and network debate shifting to potential Russian accession to the World Trade Organization (WTO) 635 and government relations with business, particularly after the 2003 YUKOS affair when Mikhail Khodorkovsky, the principal owner of Russia?s largest private oil company, was 631 Jacqueline M. Miller, Report of the January 2002 PONARS Policy Meeting (Washington, D.C., 2002): 2; and Miller, principal author, Report of the December 2003 PONARS Policy Conference (Washington, D.C., February 2004): 6. Hereafter, after the first full citation of each conference report, all will be titled after their date such as ?December 2003 Policy Conference report.? 632 See, for example, December 2003 Policy Conference report, p. 3 633 Susan Gasper, principal author, Report of the February 2005 PONARS Policy Conference (Washington, D.C., June 2005): 9-10. 634 Celeste A. Wallander, Conference Report on December 1998 PONARS Policy Meeting (Washington, D.C., 1999): 1-3. 635 January 2002 Policy Meeting report, p. 7 215 imprisoned for tax evasion and embezzlement. 636 PONARS also tackled socioeconomic afflictions such as the spread of HIV/AIDS, sparking articles in The Boston Globe and The San Francisco Chronicle, 637 and the declining health, life expectancy, and size of the Russian population. 638 Table 5.3 PONARS Memo Topics (1997-2005) 639 # Memos % total Other former Soviet states (Ukraine, CAsia, etc.) 64 15.9 Economic/socioeconomic 51 12.6 Russian regional issues (Federalism, Chechnya, etc.) 45 11.2 Nuclear issues 39 9.7 (Nonproliferation Issues 26 6.5) Russian politics (Putin) 39 9.7 Democratization 35 8.7 Russia-US bilateral issues 33 8.2 (U.S. Policy toward Russia 16 4.0) Relations with Europe/NATO 26 6.4 Russian military issues 24 5.9 General foreign policy and security issues 23 5.7 Terrorism 16 4.0 Relations with Asia (China) 8 2.0 Total 403 100.0 The second most common Russian topic has been relations between Moscow and the regions, including the war in Chechnya. Reflecting common concerns in the 1990s, the network?s early years considered whether Russia itself might fracture into smaller pieces. 640 Although that fear faded with time, the war in Chechnya dragged on, was 636 December 2003 Policy Conference report, p. 5, and February 2005 Policy Conference report, p. 8 637 Wallander, 2003 Final Grant Report, pp. 8-9 638 February 2005 Policy Conference report, p. 10 639 Nonproliferation issues and U.S. policy toward Russia are each a subset of nuclear issues and Russia- U.S. bilateral issues, respectively. They are accordingly marked in italics and flush right so as not to be double-counted in the totals. 640 See, for example, December 1998 Policy Meeting report, pp. 6-7. 216 reshaped by the September 11 attacks, and continued to be a topic at the annual policy conference. 641 Politics was another common feature of PONARS policy panels, including analyzing Putin?s rise to power in Yeltsin?s wake, the meaning of changes in election laws, 642 and results from the 2003 Duma contest and other political events. 643 More broadly, the status of Russian democratization and civil society was a focus of attention. Reflecting the chaos of the later Yeltsin years, panelists in the late 1990s expressed the need to stabilize Russian government and protect civil society against the emerging Russian mafia, or prevent an autocratic crackdown. 644 Over time, Putin asserted his power over the Russian media, 645 and used the fear of terrorism to justify additional central powers, leading PONARS members to reconsider strategies to assist Russian democracy. 646 Members of course also discussed Russian relations with the United States, particularly in the aftermath of the September 11 attacks 647 amidst a growing sense of nationalistic anti-Americanism, 648 and Russia?s relations with Europe as it sought to develop good relations with the rest of the West. 649 641 Jacqueline M. Miller, editor, Report of the December 2002 PONARS Policy Meeting (Washington, D.C., 2003): 1, and February 2005 Policy Conference report, p. 7. 642 December 2002 Policy Meeting report, p. 3 643 December 2003 Policy Conference report, pp. 1-3 644 December 1998 Policy Meeting report, pp. 4-8, and Celeste A. Wallander, Conference Report on December 2000 PONARS Policy Meeting (Washington, D.C., 2001): 1. 645 January 2002 Policy Meeting report, p. 5 646 February 2005 Policy Conference report, pp. 3, 6 647 January 2002 Policy Meeting report, p. 6 648 December 2002 Policy Meeting report, p. 6 649 December 2000 Policy Meeting report, p. 6; January 2002 Policy Meeting report, p. 6; December 2003 Policy Conference report, p. 4; and February 2005 Policy Conference report, p. 3 217 Proliferation Issues As table 5.3 indicates, nuclear proliferation was a common but not dominant network issue. Among its 80 members, only six (2 Americans and 4 Russians) claimed an expertise in nonproliferation on the PONARS website and wrote more than one memo on the subject. One PONARS staff member was not entirely surprised since, they argued, the network was initially shaped to focus on new, internal dimensions of Russian security rather than traditional challenges like proliferation. 650 Two other PONARS members argued that nonproliferation tended to be dominated in the United States by technical experts often based in think tanks rather than regional experts in universities. 651 Because PONARS recruited from the latter group, nonproliferation was not a focus for many members. Nevertheless, PONARS did discuss various aspects of nonproliferation over the network?s first nine years, particularly focusing on three: securing Russian fissile materials and scientific expertise (commonly known as the ?loose nukes? problem), missile defense, and Russian nonproliferation policy toward Iran, Iraq, and to a lesser extent North Korea. Loose Nukes Of the 26 nonproliferation memos, seven were on securing fissile materials or scientific expertise. One relatively early one by Nikolai Sokov, a Russian based at the Monterey Institute of International Studies, discussed the prospect that Russia might politically fracture and lose control over its nuclear weapons, concluding that ?the possibility that Russia might break apart?is extraordinarily small, and for all practical purposes 650 Interview C43 651 Interviews C24 and C25 218 nonexistent?.Nuclear regionalism, however, [the possibility that Russia?s regional leaders might establish de factor control over various nuclear assets on their territories] is a reality.? 652 The dominant PONARS analyst on loose nukes, writing or cowriting five of the seven memos, has been Deborah Yarsike Ball, a postdoctoral fellow when PONARS was first founded who subsequently was hired by the Lawrence Livermore National Laboratory. After writing an initial memo about a 1998 Department of Energy (DOE) anti-smuggling initiative called ?the Second Line of Defense,? 653 Ball focused on the state of Russian science, 654 including efforts to help develop the ten Russian so-called closed cities, the core of their nuclear weapons industry, 655 and the fear of brain drain, or that Russian scientists might sell their knowledge and services to aspiring proliferant states, given the poor state of the Russian economy. Ball?s most recent memo in 2004, cowritten with fellow PONARS member Theodore Gerber from the University of Wisconsin who is an expert in survey methods, reported the findings of ?an unprecedented survey of 602 Russian physicists, biologists, and chemists suggest[ing] that the threat of WMD brain drain from Russia should still be at the forefront of our attention.? The memo more optimistically reported that ?the data reveal that U.S. and Western nonproliferation assistance programs work. They significantly reduce the 652 Nikolai Sokov, ?The Reality and Myths of Nuclear Regionalism in Russia,? PONARS Policy Memo 133 (May 2000): 1. 653 Deborah Yarsike Ball, ?The US Second Line of Defense: Preventing Nuclear Smuggling Across Russia?s Borders,? PONARS Policy Memo 50, November 1998. 654 Deborah Yarsike Ball, ?The Security of Russia?s Nuclear Arsenal: The Human Factor,? PONARS Policy Memo 91, October 1999; as well as Theodore P. Gerber and Deborah Yarsike Ball, ?From Crisis to Transition: The State of Russian Science Based on Focus Groups with Nuclear Physicists,? PONARS Policy Memo 220, December 2001. 655 Deborah Yarsike Ball, ?Accessing the Inaccessible: The Case for Opening Up Russia?s Closed Cities,? PONARS Policy Memo 194, March 2001. For a more recent discussion by a PONARS member of a risk presented by the closed cities, see Robert Orttung and Louise Shelley, ?Linkages Between Terrorist and Organized Crime Groups in Nuclear Smuggling,? PONARS Policy Memo 392, December 2005. 219 likelihood that Russian scientists would consider working in [North Korea, Iran, Syria, or Iraq]. These findings have clear policy implications: the U.S. and its allies must continue to adequately fund nonproliferation assistance programs rather than hastily declare victory.? 656 Subsequently, the data was expanded into an article for the spring 2005 issue of International Security. 657 Missile Defense Ten memos were written on missile defense issues, splitting into two distinct phases. In the network?s early years, discussion focused on the Anti-Ballistic Missile (ABM) Treaty and Russian attitudes toward building U.S. momentum to withdraw from it. As early as 1999, Nikolai Sokov contended that the Russian military was actually looking forward to U.S. withdrawal so it could use that as leverage for its own modernization interests. 658 A number of PONARS members, however, pointed out that Russia was concerned that U.S. withdrawal would lead China to modernize its own nuclear forces, which might threaten Russia as well. 659 Within a year, according to Alexander Pikayev, a Russian scholar from the Carnegie Moscow Center, rising anti-Americanism within Russia and the uncertainty of a transition to a post-Yeltsin government made any sort of compromise to enable the U.S. to agreeably withdraw from the treaty at least temporarily difficult. 660 After Putin was elected in March 2000, Wallander wrote a memo outlining the potential incentives 656 Deborah Yarsike Ball and Theodore P. Gerber, ?Will Russian Scientists Go Rogue? A Survey on the Threat and the Impact of Western Assistance,? PONARS Policy Memo 357 (November 2004): 1. 657 Deborah Yarsike Ball and Theodore P. Gerber, ?Russian Scientists and Rogue States: Does Western Assistance Reduce the Proliferation Threat?? International Security 29:4 (Spring 2005): 50-77. 658 Nikolai Sokov, ?The Death of Arms Control: Popular Myths About What Russia Can and Will Do,? PONARS Policy Memo 93 (1999): 3-4. 659 See, among others, December 2000 Policy Conference report, p. 8 660 Alexander Pikayev, ?The Prospects for ABM Treaty Modification,? PONARS Policy Memo 108 (2000) 220 for Moscow?s new leader to reach an ABM agreement with Washington. 661 By the end of the following year, the Bush administration, which sought to withdraw from the ABM Treaty, had been elected, the September 11 attacks had completely changed the framework of the U.S. national missile defense debate, and the United States announced its intention to withdraw from the ABM Treaty, which took effect six months later, in June 2002. The second phase of the missile defense debate then commenced, focusing on potential missile defense cooperation with the West. Pavel Podvig, a Russian scholar primarily based at the Center for Arms Control of the Moscow Institute of Physics and Technology, recalled that as early as 1992, while Russia was ardently opposing U.S. withdrawal from the ABM Treaty, Yeltsin himself had publicly expressed Moscow?s readiness to develop and even jointly operate global early warning or missile defense systems. 662 Podvig and other PONARS members subsequently elaborated on the emerging technical, political, and bureaucratic obstacles to bilateral cooperation. 663 Although Pikayev also acknowledged some tactical and even fundamental obstacles to some forms of bilateral cooperation, in 2003 he outlined political and technical reasons why the United States was probably interested in some form of cooperation along with the potential areas for preliminary joint research and development. 664 Network members 661 Celeste A. Wallander, ?Russian Policy and the Potential for Agreement on Revising the ABM Treaty,? PONARS Policy Memo 134 (2000) 662 Pavel Podvig, ?A History of the ABM Treaty in Russia,? PONARS Policy Memo 109 (2000): 3. 663 Pavel Podvig, ?U.S.-Russian Cooperation in Missile Defense: Is It Really Possible?,? PONARS Policy Memo 316 (November 2003). Also see Alla Kassianova, ?Missile Defense Cooperation in the U.S.- European-Russian Triangle,? PONARS Policy Memo 313 (November 2003). Hereafter, after their first citation, memos will be cited as author, memo number such as Kassianova, PONARS Policy Memo 313. 664 Alexander Pikayev, ?U.S.-Russian Missile Defense Cooperation: Limits of the Possible,? PONARS Policy Memo 315 (November 2003). 221 have also recently explored incentives and obstacles to Russia?s quest to cooperate with Europeans as an alternative to the United States. 665 Russian Nonproliferation Policy Nine memos focused on Russian nonproliferation policy, particularly on Putin?s policy toward Iraq and Iran. In October 2002, during the heat of the international debate preceding the invasion of Iraq, network members contended that Russia?s relations with Saddam Hussein were driven primarily by its economic interests in Iraq, not by some geopolitical rivalry with the United States. 666 On Iran, Wallander argued that the Russian Ministry of Atomic Energy, or Minatom, had particular interests in developing Iran?s nuclear industry and was running Moscow?s policy up until at least 2002. 667 Subsequently, Vladimir Orlov, the founder of the Center for Policy Studies in Russia, or PIR Center, explained that Russia did prioritize nonproliferation, but was trying to engage Iran just as the United States had with North Korea when it reached the 1994 Agreed Framework. He even maintained that Russia had previously trusted the United States to look after Russia?s own economic interests in the earlier North Korea agreement. When they weren?t addressed, Moscow felt it learned the lesson that it had to take the lead to ensure its own interests. 668 Most recently, Wallander concurred that 665 Pavel Podvig, ?Putin?s Boost-Phase Defense: The Offer that Wasn?t,? PONARS Policy Memo 180 (2000); Kassianova, PONARS Policy Memo 313; and December 2003 Policy Conference report, p. 7 666 Celeste A. Wallander, ?Russia?s Interest in Trading with the ?Axis of Evil?,? PONARS Policy Memo 248 (October 2002); and Ekaterina Stepanova, ?Russia and U.S. Policy toward Iraq: Conflict of Interests and Limits of Dissent,? PONARS Policy Memo 249 (October 2002). 667 Wallander, PONARS Policy Memo 248, pp. 4-5 668 Vladimir Orlov, ?Nuclear Programs in North Korea and Iran: Assessing Russia?s Position,? PONARS Policy Memo 178 (November 2000): 3. Also see Orlov, ?Iran and Russia: U.S. Nonproliferation Dialogue,? PONARS Policy Memo 314 (November 2003); and ?Iran?s Nuclear Program Implications for the Dialogue Among Russia, the United States, and Europe,? PONARS Policy Memo 358 (November 2004). Orlov?s 222 Russia believed Iran was motivated by insecurity and that engagement was more likely to address the problem. 669 PONARS members also assessed Russia?s evolving global nonproliferation policy under Putin. 670 Both Pikayev and Wallander argued that although Russia values nonproliferation, it is a higher priority for the United States with states like Iran and North Korea. 671 Results Contacts and Perceptions Whether PONARS membership, or participation by non-members, established valuable networking connections or changed perceptions often depended on the previous experience of individual participants. Given the more frequent interaction among scholars within the network, members were apt to find different benefits from PONARS participation. PONARS Members Many U.S. scholars cited the value of meeting their American peers from universities scattered throughout the country. 672 Usually, the benefit was to know whom to contact, whether through the listserv or other means, to ask questions related to their research. The academic conferences also helped facilitate substantive feedback from arguments were later published in Orlov and Alexander Vinnikov, ?The Great Guessing Game: Russia and the Iranian Nuclear Issue,? The Washington Quarterly 28:2 (Spring 2005): 49-66, which this author edited. 669 Celeste A. Wallander, ?Geopolitics and Neo-containment, as well as Common Security,? PONARS Policy Memo 371 (December 2005): 26-27. 670 See, for example, Orlov, PONARS Policy Memo 178; and Wallander, PONARS Policy Memo 248, pp. 1- 2, 6; as well as Wallander, PONARS Policy Memo 371. 671 Alexander Pikayev, ?The U.S.-EU-Russian Nonproliferation Triangle,? PONARS Policy Memo 359 (November 2004), and Wallander, PONARS Policy Memo 371 672 Interviews C24, C26, C43, C481, C55, and C58 223 a larger group. 673 PONARS contacts even catalyzed collaborative research, such as coupling Ted Gerber?s methodological expertise on surveys with Debbie Ball on Russian nuclear science or with Sarah Mendelson on democratization trends and attitudes. Several U.S. members also mentioned that they met valuable young Russian scholars whom they would not otherwise have known. 674 When American scholars traveled abroad, these connections helped set up meetings or seminars in Moscow, St. Petersburg, or even more importantly in remote regions of Russia. 675 One aspect of membership which generated a lot of discussion during interviews is the members-only listserv. Conducted in English and mostly used by Americans, the most active participants have reportedly been about five U.S.-based academics along with Wallander. 676 Other members gave the listserv mixed reviews. On one hand, members based in Washington other than Wallander herself almost universally confessed that they did not have much time to read the listserv, 677 often filtering the stream of e-mails by deleting or only reading certain contributors unless they themselves had posed a question. 678 Some were reluctant to post any messages because they were not frequent participants. 679 On the other hand, many members valued the listserv as a way to ask their colleagues for research advice, recommendations on what to read on a particular topic, or what readings to include in a syllabus for an upcoming course. 680 Particularly to those based in Washington, the listserv was also praised for facilitating information- and 673 Interviews C24 and C26 674 Interviews C24, C42, C43, C46, and C47 675 Interviews C1, C24, C43, and C46 676 Interviews C24 and C43 677 Interviews C42, C46, C481, and C55 678 Interviews C481 and C55 679 Interviews C46 and C55 680 Wallander, 2003 Final Grant Report, p. 9; as well as interviews C1, C23, C26, C46, and C55 224 opinion-sharing during crises such as the 1998 financial crisis, the 2004 Orange Revolution in Ukraine, or the 2005 terrorist attack in Nalchik. 681 Beyond networking with and learning from other scholars, PONARS members lauded the opportunity, particularly early in their career, to draw the attention of media members or government officials to their memos or policy conference presentations. 682 This was less useful for those based in Washington who had similar opportunities elsewhere. 683 Harvard?s Mark Kramer mentioned that different memos he wrote had been cited in The Financial Times, The New Republic, and The Wall Street Journal. 684 He, among others, was subsequently asked to return to Washington to brief executive branch officials, 685 or to attend a summer retreat for Congressional members called the Aspen Congressional conference and brief them on their research. 686 In at least one case, a PONARS member may have even received funding for further research on their topic. 687 For some members, the conferences provided a reason to write about certain topics which they would not otherwise have addressed. 688 Some used PONARS memos as trials or first drafts for longer articles, such as Ball?s International Security piece on Russian brain drain, or Orlov?s on Russian policy toward Iran?s nuclear program which became a Washington Quarterly article, that this author edited. 689 681 Wallander, 2003 Final Grant Report, p. 9; as well as interviews C24, C46, C481, and C55 682 Interview C24, C481, and C54 683 Interview C43, C46, and C55 684 Interview with author, Washington, D.C., February 3, 2005 685 Ibid. and Wallander, 2003 Final Grant Report, p. 5 686 Interview with author, Washington, D.C., February 3, 2005 687 Interview C481 688 Interviews C1 and C23 689 Also interview C55 225 PONARS members also used the policy conferences to learn about current events, 690 other aspects of Russia or the region beyond their specific area of expertise, 691 or about U.S. policy. Although many members who were based in Washington or had spent significant time there did not find much value in conference feedback, 692 others found that presenting at conferences brought out audience members with expertise on a particular area, such as U.S. food aid or the Russian electricity industry, to approach them either during the plenary session or on its margins. 693 Just the questions from government officials gave some members a sense either of what issues those officials thought were interesting, or of the bureaucratic politics on topics such as Chechnya. 694 U.S. Government Employees U.S. government officials found different benefits from attending the annual policy conference or reading PONARS memos. Some used the policy conferences simply as a place to catch up with other Americans who studied Russia whom they already knew but did not see regularly. 695 Younger government officials seemed to actually meet new people. 696 For many, however, the value was to meet Russian participants. Some helped cement previous contacts, 697 while other government officials were introduced to some of the Russians for the first time. 698 One government employee specifically found it valuable to have the Russians come to Washington given that ?it?s not that easy to get 690 Wallander, 2003 Final Grant Report, p. 7 691 Interviews C1, C23, C25, and C58 692 Interviews C46, C55, and C58 693 Interviews C24 and C25 694 Interviews C23, C24, and C26 695 Interviews C50, C57, and C59 696 Interviews C51, C56, and C60 697 Interviews C59 and C62 698 Interviews C2, C50, and C59 226 over to Moscow.? 699 Although the policy conference may have initiated these contacts, it is not clear how valuable the interchange at the conference itself was perceived to be. In some cases, government officials mentioned that it was worthwhile to meet the Russians so they could follow up with them in greater depth on the margins of the conference or at a later time, 700 possibly to invite them back to speak at a government-sponsored conference, 701 or, in one case, simply to help determine whose work was worth reading later. 702 Many specifically valued the chance to meet the younger Russians. 703 Even if new contacts have been made, that still does not address whether government employees were actually learning valuable information from PONARS. The first benefit government officials cited was to hear the Russian perspectives. Although one retired diplomat discounted them as generally biased, 704 other government employees felt that these younger academics gave different, less ideological perspectives than what they heard from Russian diplomats. 705 One valued the stronger social science training and policy analysis skills compared to the older Russian generation. 706 Another government official mentioned they valued the chance to hear someone from some place other than Moscow, in this case from Volgograd. 707 Some government officials seemed to have debunked the perception that a younger Russian generation was universally more pro-Western than their elders. Although analysts certainly recalled Russian participants with more pro-Western views? 699 Interview C50 700 Interviews C47 and C56 701 Interviews C54, C59, C60, and C61 702 Interview C61 703 Interviews C2, C45, C47, C50, and C57 704 Interview C44 705 Interviews C2, C45, and C57 706 Interview C2 707 Interview C47 227 such as one Russian participant very critical of the Chechen war 708 or another who sympathized with Western fears of China 709 ?others, after hearing the Russian views, concluded that, in the words of one government staffer, ?we were deluding ourselves into thinking a certain generation likes us.? 710 In one particular case, a government policy analyst who focused on missile defense was surprised to hear traditional Russian arguments about the value of the ABM Treaty. ?To just hear from a young contemporary Russian that she believed this really had a lot of resonance with me.? 711 Unrelated to the nationality of the speakers, government officials valued PONARS perspectives for other reasons. Particularly for desk officers at lower levels of the government hierarchy, the network gave an opportunity to think about broader aspects of Russia beyond the specific ones, such as arms control, for which they were responsible in their job portfolio. ?I liked the fact that those things got me out of the more narrow focus of what I was doing. It broadened my horizons?Usually we didn?t get that broader picture at work?? 712 Another member of the revolving door explained that ?there can be a real tendency for the bureaus and offices to develop tunnel vision. We fail sometimes to see that there is a bigger picture out there.? 713 In at least one case, PONARS also helped a policy analyst look longer term beyond the various crises or events that drove their job from day-to-day. When asked, that staffer argued that even their interaction with the diplomatic and intelligence communities was ?very much event-driven reporting in both instances, whereas with the 708 Interview C57 709 Interview C51 710 Interview C57 711 Interview C61 712 Interview C45 as well as C61 713 Interview C62 228 academic community, the event is more than a week long, it?s more than a month long, it?s really looking at things over the longer range to try and draw more of a perspective on it.? 714 These themes of supplementing what the intelligence community provides the policy community with broader, longer-term analysis, particularly for political trends, recurred in a number of interviews. 715 In other words, U.S. government analysts valued PONARS members because they helped explain Russia. As one member of the revolving door, who had experience in both the intelligence community and policy planning, argued: ?when one thinks through different policy options, I think it gives you a better sense of what the Russian response might be.? 716 Another policy analyst went so far as to say, ?if you want to know about where Russia is heading and its political trajectory, you don?t go to the CIA or DIA?? 717 In other cases, the benefit was simply to have another informed opinion, as a sanity check or to protect against ?groupthink? within government. 718 It is an opportunity to hear what smart people have to say about Russia. One policymaker, who is a member of the revolving door, succinctly stated: ?there?s no monopoly on brilliant thinking inside the government?We close the door at our peril to thinking on the outside.? 719 One member of the intelligence community even admitted, ?my own experience, over thirty years, is that I have found very, very few issues where over time a good academic, a good scholar, or a really good journalist?could not uncover the story and get pretty much the 714 Interview C61 715 Interviews C45, C47, C50, C61, and C62 716 Interview C59 717 Interview C47 718 Interviews C47, C50, and C60 719 Interview C62 229 picture?? 720 Yet some government officials argued that the advantage went even further: nongovernmental analysts were able to talk to a broader range of contacts within Russia, either because government officials and diplomats stayed within their formal circles of contacts, 721 or because intelligence officials could not travel within Russia or were restricted by the need to maintain their cover. 722 When pressed for specific examples on the spot, most government employees were unable to come up with any. One member of the intelligence community explained, We spend our days doing preciously little but absorbing information in this job and, eventually, you start to lose track of where you got the information. There?s a little classified cabinet in your head?but other than that, I rarely remember exactly who told me what. But I do know that I keep going to these [PONARS] things so?I must be getting something out of it. 723 A policy analyst similarly argued that what information struck anyone depended in part on what any individual was thinking about at the time: ?In this job, you are inundated with information. The thing that comes along that scratches a particular itch of yours at a particular time [is what is interesting].? 724 That logic would explain the variety of specific examples that some subjects did recall, including how free and fair the 2003 Parliamentary elections were, 725 or what the Russian reaction might be if the Baltic states were to join the EU. 726 On nuclear and nonproliferation issues, recollections included when Russia might have to modernize their tactical nuclear weapons 727 as well as 720 Interview C60 721 Interviews C42 and C50 722 Interviews C47 and C50 723 Interview C51 724 Interview C47 725 Interview C51 726 Interview C59 727 Interview C47 230 Moscow?s policy toward Iran?s nuclear program, including one former Clinton administration White House official who remembered first hearing the argument from PONARS that pushing Iran too hard might lead them to quit the nonproliferation regime entirely. 728 Overall, U.S. government analysts were able to meet young Russian academics and learn about their perspective on their country, the region, and how it was changing. They also were able to use all PONARS members to help them understand broad, long- term changes, beyond the specific responsibility of their job portfolio, and supplement the analysis they received from the intelligence community about developments in Russian politics, or projections about Russian reactions to potential policy changes from smart, nongovernmental experts who might even have unique insights from a wider array of Russian experts in and out of the government. U.S. Government Access and Interaction Who from the U.S. government benefits from PONARS? Who goes to the annual conference? Who reads the memos? According to the PONARS Instructions and Notes on Policy Memos, ?Our primary audience and consumer of the policy memo series are policymakers who know a great deal about the issues facing Russia and Eurasia and U.S. interests in the region, but who have little time for long-term thinking, synthesis, self- criticism, and other luxuries we academics are permitted to indulge freely.? 729 The guidelines further advised that memos should be 2000-2500 words long. From its inception, at least Jim Goldgeier, a PONARS member who had experience in 728 Interview C2 as well as C59 729 PONARS Instructions and Notes on Policy Memos, p. 3 231 government, was skeptical of their value, arguing that the network should either stay focused on longer, analytical articles that academics were generally good at or should try to produce shorter op-eds (generally 600-800 words) that government officials might read. 730 Within government intelligence, research, and policy analysis communities, it turns out that some do seem to read the memos. Subjects interviewed value the discipline the shorter format imposes on scholars, forcing them to present their ideas in 2-3 pages instead of traditional, longer academic articles, 731 although one analyst demurred that they were too brief to ?shed that much light on things I was already thinking about.? 732 While the memos are circulated a few weeks in advance of the annual conference to be read prior to it, one analyst said they used the conference as a way to screen the memos, going back and reading ones written by speakers who interested them. 733 In contrast, the skeptical view turned out to be accurate for policymakers, or those at the deputy secretary level and higher, who claimed that they had not read more than one or two, if any, memos. 734 One former senior policymaker specifically criticized the quality of the memos and their ?plain brown wrapper? presentation, concluding that they are not presented in a way that makes them very inviting to readers who don?t know the writer or who are not looking for something on a very specific topic. So, I tended to toss them. I guess the ones that I looked at struck me as, I don?t want to be too snotty about this, commendable if somewhat amateurish work of younger, not super-informed, scholars. 735 730 Interview with author, Washington, D.C., October 18, 2005 731 Interviews C2, C56, C59, C60 732 Interview C50 733 Interview C51 734 Interviews C48, C52, and C62 735 Interview C52 232 One problem for policymakers seems to be that even 2000-2500 words does turn out to be too long. Another former official argued that Brevity is hard but it is how you communicate in government?.I think the people you probably want to affect are deputy assistant secretaries at State, office directors, NSC directors. The problem?is that when I was at NSC, I started at 7:20 in the morning, and if I left at 8 or 8:30 at night, it was a good night. When I was [a deputy assistant secretary] at State, I?d get in at 7 in the morning and get out by 7, 7:30, and Saturdays would be a half to a full day. The problem is just when you get home?the last thing [you] want to do is pick up a serious policy paper on Russia and then read it?[I]t has to be a 700-word op-ed piece. Beyond that, it?s going to be hard to have any certainty that it?s getting to the right target. 736 Compounding that problem, particularly in recent years, almost all the memos come out simultaneously in a 200-page bound book, right before the annual conference, overwhelming potential readers. 737 The same former official argued that ?If I get a book like this, it will go in my inbox right away, to the Russia desk. If it was a small paper, it at least had a chance to be read on a light Saturday. The trade-off is you have to decide which of the papers you are going to share.? 738 One analyst argued that they preferred memos targeted at key events, such as a summit, rather than the PONARS conference. 739 Events drove greater demand for memos, whereas the supply was driven by the network?s conference. To triage these memos, policymakers might wait for their staff to recommend anything worth reading, essentially having them serve as an additional layer of quality control. PONARS members interviewed for this study mentioned that they had heard of 736 Interview C48 737 Interviews C44, C45, C48, and C51 738 Interview C48 739 Interview C51 233 such cases, including a piece on AIDS in the former Soviet Union that was sent to Undersecretary of State for Global Affairs Paula Dobriansky. 740 In another case, a member was asked to testify before Congress because of a memo that they wrote. 741 Conversely, John Harvey, a policy adviser at the Department of Energy, mentioned that the Deputy Secretary of Energy for Nuclear Security, Linton Brooks, would actually occasionally recommend to his own staff that they read a PONARS memo and look into an issue further. 742 Usually, however, it seemed that analysts and staffers were more likely to read the memos than policymakers. A second way that PONARS sought to access government was by inviting them to attend the policy conference. Overall, government turnout at the conferences has been quite high, surprising even some PONARS members. 743 Based on the participants lists available since 2000 (5 of 6 sessions with 2001 unavailable), the conferences have averaged over 50 RSVPs to attend from Capitol Hill or traditional security agencies (mostly from Defense, State, and the intelligence community but also from Energy or the National Nuclear Security Administration, NSC, or the Office of the Vice President), with a high of 60 in 2000, a low of 43 in 2003, and 49 and 50 in the last two conferences in 2005. Additional members also came from the U.S. Agency for International Development (USAID), the Department of Justice, and the Department of Treasury, reflecting the network?s emphasis on new security issues. Interestingly, almost 2 out of every 3 employees from traditional security agencies who RSVP?d to one PONARS conference did not RSVP again, with 1 in 10 attending four or all five sessions (see table 740 Interview C24 741 Interview C481 742 Interview with author, Washington, D.C., November 1, 2005 743 Interviews C42 and C55 234 5.4). (Although this may indicate that many attendees did not perceive PONARS valuable after attending once, it is also possible that it simply indicates that government officials do not necessarily stay in one job, on the same portfolio, or even in Washington for more than 2 or 3 years at a time.) Table 5.4 Repeated Government RSVPs to PONARS (2000-05) Number of Sessions RSVPd (of 5) 1 2 3 4 5 Number of Attendees 106 34 8 13 3 % of Total Government Attendees 65% 21% 5% 8% 2% Although the participants lists only note the agency at which an official works, not their title or level within the government, Wallander herself acknowledged that most of the attendees were desk officers, department heads, military officers, and senior intelligence analysts, not policymakers. 744 Current and former policymakers interviewed for this study confirmed that they were either unable to or uninterested in attending the annual conference. Deputy assistant secretaries were at least interested in attending a panel or two, although none could fathom attending for a full day, but stated they were unable to attend until after they were no longer the DAS. 745 As one argued, an event like PONARS hosted by a nongovernmental organization can be useful but it ?tends to be the first thing cut from your schedule when you are pressed for time.? 746 For the assistant secretary-level, neither interviewed for this study could even consider attending unless they were asked to be the speaker (PONARS does not host government speakers). 747 More generally, these policymakers, some of whom are even 744 Interview with author, Washington, D.C., November 17, 2005 745 Interviews C47, C48, and C62 746 Interview C48 747 Interviews C49 and C52 235 members of the revolving door, felt that they rarely learned anything useful during their time in office from any event hosted outside government. 748 One stated: Very rarely [are outside groups useful]. So rarely that I?m going to have to pause to see if I can dredge up an exception for you. It?s less that you hear from an outsider an idea that nobody ever thought of, and more that hearing from outside groups gives you kind of a lay of the land in the world of opinion, and a sense of what the distribution of views is among outsiders. The government, more commonly, is interested in knowing what the traffic will bear, what kinds of things people are saying rather than getting authentically new ideas. There are a lot of people in government who have the job of generating new ideas. In one way or another, they usually come up with pretty much the same menu of possibilities that outsiders do. It?s more of a process of validation, confirmation, even merely reporting than actually hearing something new. 749 Two of these policymakers with DAS or assistant secretary-level experience also argued, however, that outside groups were often useful to help provide public evidence to either Russian interlocutors or even U.S. interagency counterparts that outsiders cared about certain issues, such as democratization, which might otherwise be overlooked by strategic considerations. 750 In other words, outside groups could be useful leverage in interagency or international negotiations. One perception that emerged from interviews is that the set of policymakers responsible for Russia has declined drastically, both after the Cold War and again after the September 11 attacks, further straining the already immense time pressure on that handful of officials. 751 One PONARS member even perceived that the shrinking number of true policymakers within the government was one, but not necessarily the only, reason 748 Interviews C49 and C52 749 Interview C52 750 Interview C49 and C62 751 Interviews C2, C481, and C55 236 why they felt ?the quality of the people who come to the meetings has declined both on the Russian and American side.? 752 Unlike policymakers, desk officers do attend the conference, but desk officers are only an indirect way to influence the declining number of policymakers responsible for Russia. Many arguably were responsible for implementing policy, but not making it. As Goldgeier pointed out, ?I?ve worked in the State Department, but there are a lot of people in the State Department who aren?t policy-relevant either.? 753 Some members of the policy planning staff also attended, but questions were raised about that staff?s relevance and its recent focus on Russia. 754 For the Pentagon, a desk officer who frequently attended network conferences (one of the 10% of recidivists) confessed that their counterparts were skeptical about the value of PONARS because they felt that even they didn?t have the time, they perceived the event to have a pro-Russian bias, and they felt that it was a ?luxury? to be able to attend the event but that ?it?s not really pertinent to my day-to-day work.? 755 Nevertheless, the RSVPs and testimony from PONARS members reveal that some desk officers attended, even if their policy influence is unclear. The main government consumers that find PONARS useful, however, seem to be intelligence analysts. It is hard to say for certain for a number of reasons. Attendees only are listed by the department in which they work, so an analyst from the Bureau of Intelligence and Research (INR) appears as ?Department of State,? while one from the Defense Intelligence Agency may register as ?Department of Defense.? Others might not be able to say where they work, potentially asking not to be included on the participant?s 752 Interview C481 753 Interview with author, Washington, D.C., October 18, 2005 754 Interviews C59 and C62 as well as C54 755 Interview C61 237 lists or just putting ?U.S. government.? To compound these problems, some members of the intelligence community resisted being interviewed for this study or were prohibited by their office. Nevertheless, a significant indicator is that of the 16 government attendees who RSVP?d to attend either 4 or all 5 of the PONARS conferences since 2000, at least 10 of them are from the intelligence community?either the CIA, DIA, INR, or National Intelligence Council (NIC). (Of the other six, one was a military officer, another from the Pentagon, another a desk officer from the Pentagon, a desk officer from the State Department, a Congressional Research Service staff member, and a recently retired State Department official.) Interviews with policymakers, PONARS staff, PONARS members, and intelligence community members themselves all reinforced this evidence. 756 Goldgeier, previously cited for being initially skeptical about the network?s value, admitted that ?it did not occur to me [that PONARS might be relevant to the intelligence community]. All I was thinking at that first [brainstorming] meeting was: why would someone who was in the position I just left at NSC going to read a policy memo?? He explained why the relationship was natural: ?[Academics are] not very good at prescribing policy-relevant things to do. They don?t know much about what to do. They don?t know what the possibilities are on any given issue. They don?t know what?s going on, on the inside, but they?re very good at analyzing a situation.? 757 As mentioned earlier, PONARS members were not just perceived to be smart, but also useful because they talked to experts from Russia outside of traditional government contacts. To the intelligence community, those connections appear to be particularly useful because of 756 Interview C1, C42, C43, C47, C59, and C60 757 Interview with author, Washington, D.C., October 18, 2005 238 travel restrictions, possibly because Russia may not give them visas. 758 Goldgeier also argued that academics may be able to get different insight from Russians because ?they?re not threatening because they?re just academics. They?re not from the U.S. government.? 759 That could lead to more frank conversations. Intelligence community members could also invite PONARS experts, both Americans and Russians, to conferences in the United States for further discussion. Within the intelligence community, members specifically said that the National Intelligence Council (NIC) and the State Department?s Intelligence and Research (INR) bureau found PONARS most useful because they were the intelligence community?s most public parts, unlike the more sheltered and distant CIA or DIA. 760 Especially INR and the NIC were not like the 20 th -century intelligence community seeking state secrets and supporting covert operations, but part of the 21 st -century?s emphasis on seeking more transparent and open-source analysis. One INR analyst explained that ?though INR is part of the intelligence community, we are part of the diplomatic community: the main institution that?s supposed to be reaching out.? 761 In addition, another analyst explained, ?Most of us here, at least at INR, are only one deep on most issues?.If there is a network of people out there looking at the same questions, there is certainly going to be an interest here in hearing what they have to say.? 762 Similarly, with the recent changes in the intelligence community after the 9/11 Commission recommendations, one analyst from the NIC explained that ?the NIC?s role is changing now too: we?re now getting more into the policy realm since we now work for the director of national intelligence?.The line 758 Interview C59 759 Interview with author, Washington, D.C., October 18, 2005 760 Interviews C51, C59, and C60 as well as C46 761 Interview C60 762 Interview C51 239 between analysis and policy has become somewhat fuzzier, if you like, more fluid in the NIC. Some of the memos address directly policy issues.? 763 These analysts argued that as the intelligence community continues to put greater emphasis on open-source analysis, at least the NIC and INR will find nongovernmental experts, including PONARS members, even more valuable. 764 Any policy impact that PONARS members might indirectly have through the intelligence community is, however, even more difficult to trace. In interviews for this study, intelligence community members were not forthcoming with specific examples of things they may have learned. Even if they did, the IC finds it hard to trace its own impact. When they do interact with policymakers, as one of its members admitted, ?I?m not going to give [PONARS members] credit for any arguments, but I may well pass the argument along. Obviously, the people who I write for don?t care who said it.? 765 While executive branch attendance seemed to be concentrated in the intelligence community and desk officers, about 5-10 Congressional employees were also among the roughly 50 government RSVPs each year. PONARS members had some contact with committee staffers but generally felt Congressional participation was much smaller than the executive branch, possibly because the sessions were not held on the Hill. 766 One former House committee staff member who did attend explained that their colleagues did not for a variety of reasons. Most personal staffers, in this staffer?s opinion, were not interested because they were more political than analytical. Although it helped that 763 Interview C59 764 Interviews C59 and C60. For the increasing emphasis on open-source analysis , see Scott Shane, ?A T- Shirt-and-Dagger Operation,? New York Times, November 13, 2005; and Susan B. Glasser, ?Probing Galaxies of Data for Nuggets: FBIS Is Overhauled and Rolled Out to Mine the Web?s Open-Source Information Lode,? Washington Post, November 25, 2005, p. A35. 765 Interview C51 766 Interviews C24 and C43 240 Congress was usually out of session (conferences are typically held on a Friday in December, when Representatives and Senators were often traveling to their home districts) and they personally found PONARS useful because they did not often get to travel to Russia, getting an elected official to focus on Russia recently was not a high priority. ?As much as I want to delve into detail about Russian political life, a member of Congress really doesn?t care?. The bottom line is that the things I learned there didn?t go anywhere because there was just no place for an outlet.? 767 Network members might improve their chances of being heard on the Hill when they were asked to testify 768 or through a Congressional Briefing Series that Wallander runs at CSIS. 769 The Carnegie Foundation also runs another program during the summer recess called the Aspen Congressional program, a unique event at which 20-25 members of Congress attend a week-long retreat with parliamentarians and scholars from Europe, Russia, and the United States. Carnegie?s David Speedie explained ?for that, we drew very much on the PONARS group?.I could probably rattle off a dozen PONARS members who have been called on as scholars to have this really intimate, in-depth dialogue with members of the Congress.? 770 What policy impact they may have had is ultimately unclear. A final consumer from the Hill is the Congressional Research Service. 771 As mentioned earlier, one of the 16 government employees who attended almost all of the recent conferences is a CRS analyst named Stuart Goldman who succinctly explained, 767 Interview C57 768 Interviews C43, C481, and C53 769 Wallander, 2005 Final Grant Report, p. 3 770 Telephone interview with author, November 14, 2005 771 Interviews C24 and C57 241 ?I?m a researcher and so it?s one of the many fora in which I acquire information and contacts that I use in my work.? 772 Overall, although some government analysts generally found the conferences useful, 773 others complained that they had become too big for meaningful dialogue, 774 or were too infrequent. 775 One policy advisor felt it was just unrealistic to think that one meeting in December could have that much of an impact. ?The timing is important and the quality of the idea is important. You don?t get that from simply having one meeting a year in which folks come to Washington?.[E]ven if there is a good idea, it may not be the right time in December when the PONARS meeting hits.? 776 Even if the immediate impact of the conference itself was limited, government attendance served a positive networking role, making initial contacts between PONARS and government officials, particularly from the intelligence community, which could be drawn upon in subsequent meetings or conferences. Beyond memos and government attendance, PONARS might also theoretically be able to influence the U.S. government through the revolving door. At this stage, however, there are few members?all of whom are young academics?with government experience. It may be that academics are generally less interested in going into government or it may just be their age. One exception worked at the National Security Council but reported that even though they actually did talk to former network colleagues, 772 Interview with author, Washington, D.C., December 1, 2005 773 Interview C51 and C56 774 Interviews C50 and C58 775 Interview C60 776 Interview C47 242 I also hasten to add that I can count on the fingers of one hand the occasions in which I had?policy influence in any meaningful sense of the word, even as an NSC staffer. I didn?t matter very much?. It has to do with the limitations of even middle-level officials in government service. There?s very little real influence. The only issues that were the sort of thing that would be interesting for an academic to think about were issues that were decided at the very highest levels. The kind of things that I might have had some influence on were very, very, low level, down in the weeds, implementation issues. The question, in a way, is the wrong question because even if I learned something somehow from PONARS, it would be almost completely irrelevant to any actual policy decision. 777 Over time, the question remains whether these young academics might become future government officials. 778 Another member with government experience told what is now a familiar story: ?I don?t know how many of those folks in PONARS have gone into government, but in general to my mind, that?s where think tanks have the biggest role to play, which is in helping people while they are out think about policies they want to pursue when they are in?Once they go in, they?re not reading.? 779 Another PONARS member speculated ?you assume people in this network will someday be in a position of office of policymaking, and that?s when it will matter. Condi [Rice] told me this once a long time ago: it?s very hard to renew your intellectual capital while you?re in government office?.You do it when you?re on the outside.? The uncertainty, of course, is ?that it?s a big role of the die, and you never know who is going to be in power.? 780 Agenda Setting The PONARS conference agenda is established through a bottom-up process where members submit memo ideas to the Executive Committee (or Wallander herself prior to 777 Interview C58 778 Interview C2 779 Interview C42 780 Interview C55 243 its establishment in 1999) and the committee decides which ones to host at the conference. In theory, this process might increase the chance that a young scholar would highlight an agenda item that the U.S. government was not currently considering. Wallander acknowledges that the process has risks because ?PONARS does not assign topics, research, or areas of expertise?.[I]n the end we shape our conferences and publications around what the members do best, not on what the policy community might deem important. That is a potential weakness in making PONARS useful to the U.S. policy community, but in practice it has never been a problem.? 781 Since 2002, ExCom has sent out nonbinding guidance to members suggesting some potential topics that were hot in Washington, but the agenda was ultimately a mix of those issues and others that members suggested. 782 Despite the potential novelty provided by this process, interviews did not reveal any sense that PONARS had raised any new issues on the agenda. The proliferation topics covered?loose nukes, missile defense, and Russian policy toward Iraq and Iran? were certainly already there. Government analysts more frequently replied that when PONARS was useful, it provided different, often but not always Russian, perspectives on subjects that they were already thinking about. 783 One former policymaker endorsed the utility of the results, arguing that the U.S. agenda with Russia was fairly static, making it difficult for any organization to change the government agenda: ?I think the way the bureaucracy works, it will be harder to put a new issue on the agenda?. you are more 781 Wallander, ?Borderlands,? p. 38 782 Interviews C43 and C54 783 Interviews C50, C51, and C60 244 likely to get somebody to really pay attention if it addresses one of the first-tier agenda issues with a new idea or new way to think about it.? 784 Policy Options Similarly, PONARS members raised few if any policy options. PONARS staff and members explicitly acknowledged that the network often steers clear of policy advice or developing policy options because it is not the strength of its academic members, most of whom have had no experience in government. 785 From its inception during the 1996 brainstorming meeting, Wallander explained There was a lot of resistance to the idea that young professors could or should necessarily end everything they wrote for a policy audience saying U.S. policy should be x. We decided early on that PONARS policy memos can be about: here?s something you haven?t thought about before, or here?s a different take on something you think you understood and you haven?t got it right and let me tell you why?.Because the point of PONARS is to make whatever valuable comes out of academic research accessible to a policy audience, it?s not necessarily the case to do policy advocacy. Actually, I think that?s one of the reasons why people like coming to the conference: because it?s not all about what U.S. policy should be. Policymakers or analysts get that all the time. 786 The network?s instructions for writing policy memos embodied this early consensus, guiding ?we have received feedback in the past not to think in terms of writing action memos, and not to try to duplicate the work of government analysts. Our memos will be most effective and useful if, unlike the usual government analyses, they address the long term, with a synergistic perspective and greater breadth.? 787 784 Interview C48 785 Interviews C43 and C54 as well as C23 and C25 786 Interview with author, Washington, D.C., November 17, 2005 787 PONARS Instructions and Notes on Policy Memos, pp. 1-2 245 Two former policymakers expressed some frustration with this approach, arguing it left them with little practical advice. One former senior former policymaker summarized that, in general, ?My frustration with think tanks has been, for a very long time, that it is very eager to say what is going wrong and very little effort to provide useful alternatives?.Leaving it there isn?t an option that I have.? 788 Conversely, an intelligence analyst explained that the option-free approach made it even more useful for them as, ?It?s illegal for me to actually recommend policy?.I think the commentary of PONARS tends to be on Russian policy and trying to understand how the Russians are developing.? 789 Even if PONARS organizers sought to develop policy options, its structure and membership are not ideally suited to have much of a direct and immediate impact. For one thing, as a policy analyst pointed out, the policy conferences are not focused on one particular issue but are an annual survey of Russian and regional security issues, making it too diffuse to have much of an effect on any individual subject. 790 More fundamentally, PONARS members resist even trying to make policy recommendations. 791 At least one sensed that policymakers were simply saturated with suggestions and that PONARS should do something different. 792 Others contend that academics are just not well suited to give specific policy recommendations anyway. One scholar was resigned that ?I don?t see my views standing much of a chance of being implemented,? so they were content to give policy-relevant advice and leave it to policymakers to ?draw their own policy 788 Interview C49 as well as C47 789 Interview C51 790 Interview C45 791 Interviews C23, C24, C25, C26, and C55 792 Interview C24 246 conclusions if there are any.? 793 Another stated, ?I?m not really in the business myself of offering policy recommendations myself so I?ve always been kind of uncomfortable [doing it].? 794 That does not mean it was not important to network organizers to influence the policy community, just that some individual members were not as interested or did not seek to get specific policy recommendations adopted. Some network members felt that having an impact in the policy world was clearly important to Wallander herself. 795 Other members thought the funders, particularly Carnegie, wanted to see the network?s policy impact. 796 Wallander confirmed that, for Carnegie in particular, policy impact ?is key; that?s the core.? When funding for PONARS was reduced, she said, Carnegie wanted to keep the policy conferences and eliminate the academic conferences, much to the chagrin of many network members, because ?the policy view was more valuable and more important? to Carnegie. 797 Carnegie?s Speedie confirmed that policy impact ?was very important for a couple of reasons:? because Carnegie generally defined itself ?as operating in the intersection of scholarship and policy? and because, at the time, ?as this unruly period following the collapse of the Soviet Union was evolving?the need, at least to us, was clearly evident for there to be good input into policy recommendations.? Generally, Speedie acknowledged that how you measure policy impact is ?a lot more difficult to answer.? Nevertheless, Carnegie continued to support and fund the network?s policy conferences. Speedie stated that he had been impressed at how well attended they were. 793 Interview C26 as well as C55 794 Interview C24 795 Interviews C1, C25, and C58 796 Interviews C26, C42, and C481 797 Interview with author, Washington, D.C., November 17, 2005 247 ?In other words, the Washington insiders were paying attention. Where that ends up in terms of legislation or policy directions, I?m not qualified to judge. All I know is that people were spending time and effort to come to these meetings.? 798 Similarly, one PONARS staff member said that they measured their impact by access to government officials. The key was not to change policy, because that was too unlikely and at least too difficult to measure, but ?I would say [it was] important to inform policy, which I think is different?.I think we at least made people stop and think.? 799 Government policymakers participating in this study were less optimistic that access made that much of a difference. One member of the revolving door said, ?It?s true that some lower or mid-level folks show up at the policy conferences, but I don?t think it means anything.? 800 Other government officials who thought PONARS was trying to influence government policy felt it was unsuccessful. One former senior policymaker criticized, ?It?s a problem when you?re a young scholar and you?re just learning how to do it. The presumption that you?re going to be tooting your plain brown-wrapper essay, along with seven or so others under the same cover over the transom, and have somebody pay attention is a little bit unrealistic.? 801 One desk officer specifically highlighted their frustration with how little contact PONARS members had with the policy community, and how uninformed they felt much of the network?s recommendations and discussions were about current policy. 798 Telephone interview with author, November 14, 2005 799 Interview C43 800 Interview C58 801 Interview C52 248 I think the way PONARS is structured right now, with a once a year conference with the policy community, it?s inherently difficult for them to really try to influence the policy community because they have less interaction?.For academics, without that kind of feedback loop where they are aware of what we?re working on and what we?re interested in, they?re inherently likely to be coming at it from a distance?.[I]t?s not really well informed by what?s truly going on in the government and the interagency right now. 802 That does not mean the network is useless, they concluded, but it is not about influencing policy. ?When I travel, it tends to be interacting with officials and not having as much time to get out into the community and talk to the general public ?.What I really find more valuable is more the in-depth analysis of what?s going on inside the country with the populous?getting some historical perspectives to that.? 803 Because of this focus on informing not recommending policy, an intelligence analyst praised the network?s approach, highlighting how often people outside government did not realize that their proposals had been considered. I?m sure there are many that will disagree with me, but I usually find relatively little value in outsiders, academics particularly, making policy recommendations. I usually get to the point where, when they get to the policy implications, I start to cringe and think I can leave. It?s not because they?re not thoughtful; it?s not because they?re not bright people doing it. They?re not there. They?re not in the heat of the battle. They don?t really know what already has been done, and what is possible that you can do in the building?.[G]enerally what happens, I find, when we get to policy advice and implications [is] that it is at a level of generality that is not particularly useful and probably well over half the time, probably two- thirds or three-quarters of the time, they?re talking about things that are already being done, because other bright people?already came to that. So, in that sense, I kind of have appreciated PONARS self-imposed restraint. 804 802 Interview C61 803 Interview C61 804 Interview C60 249 The network?s decision, therefore, not to develop many policy options seems to be justified. Although many lower-level policy analysts and members of the intelligence community valued the analysis PONARS members provide, developing policy options was not the network?s focus. Observations Beyond the structured part of this study, a number of other issues related to the network?s operations emerged during interviews. Rather than ignore them entirely or leave them unreported, they will be briefly mentioned here for others who may examine networks. Although some of these issues have already been touched on earlier, six will be discussed further here: the network?s potential drawbacks, its membership, how it might be structured to improve policy impact, its effects during a crisis, the potential socialization of academic members for policy impact, and any effects in Russia. Drawbacks? Although every subject interviewed was directly asked whether PONARS was harmful in any way, not one expressed any examples or even suspicions that the network was harmful. The overwhelming consensus of members and government participants is that PONARS is not harmful in any way. 805 Many, both in and out of the network, specifically said that the network is not influential enough to be harmful. 806 Two different current or 805 Interviews C2, C23, C25, C42, C48, C49, C50, C52, C54, C59 C60, C61, and C62 806 Interviews C26, C44, C46, C47, C51, and C57 250 former government officials specifically pre-empted any idea that network criticism might complicate their lives, saying it was just part of living in a democracy. 807 The only serious suspicions raised were harms to members themselves. Two network members complained that the listserv could get ?nasty? or ?acrimonious,? particularly as the network had gotten larger and its intimacy dissipated. 808 More generally, one member of the revolving door thought the network?s quest for policy influence was counterproductive, going so far as saying, ?My recommendation would be, especially in the current climate, to give up on policy relevance and focus on being an academic network. That is what it is good at. That is what it should do.? 809 The most common suspicion, however, was whether PONARS membership ran any risks for Russian members. 810 The implicit concern was simply that critical comments by Russian network members would alarm hard-line members of their government or intelligence services. As one U.S. participant put it, ?there are times when I think to myself ?holy crap? I wonder if the Russians realize this is getting said.? 811 One Russian member, however, dismissed those concerns stating simply ?I can watch a number of programs or read newspapers which are much more critical to Putin.? 812 Government Participation and Network Membership Who attends the annual conference is important to the network?s influence. While the primary government consumers seem to be the intelligence community and, to a lesser 807 Interviews C45 and C49 808 Interviews C54 and C55, respectively 809 Interview C58 as well as C60 810 Interviews C24, C43, C51, C54, and C59 811 Interview C51 812 Interview C25 251 extent, desk officers, chapter two highlighted Congress as another promising avenue. What happened to the legislature? Their interest and actual results was minimal in practice. As discussed earlier, staff members did not predominantly seem interested enough to leave the Hill. The network?s attention seemed too detailed for a committee staff member likely responsible for an entire region, much less a personal staffer covering the whole world. PONARS also is not focused enough on policy options to serve Congress, which focuses on budget allocations and U.S. policy. Wallander explained that from her experience, ?I think the fit is better with the executive branch? The demand is greater. They need to know more, they need to think a little more long term?.The U.S. Congress is supposed to think about what program they like, how are they going to fund it, and whether it will get implemented right.? 813 Within the network itself, a few questions were raised about membership. Two PONARS members were disappointed by the quality of the Russians, particularly the recent emphasis that seemed to be a ?near-random? quest for ?warm bodies who will come to meetings.? 814 Two different government participants complained either that the Russian members tended to be more pro-Western, 815 or that, more broadly, nongovernmental Russian analysts had no influence or contacts with their government today. 816 On the U.S. side, the primary question was the network?s utility to members as they became more experienced, particularly for those who moved to Washington think 813 Interview with author, Washington, D.C., November 17, 2005, as well as interview C57 814 Interviews C55 and C481 815 Interview C59 816 Interview C60 252 tanks. 817 These members seemed to ?outgrow? PONARS as their own contacts developed, and the PONARS conference became ?trivial for them.? One member, now based in a Washington think tank, was eternally grateful for the government and broader policy community contacts that PONARS initially established, but over time found that once these contacts were established ?instead of [the policy conference] being the one time a year you?re seeing people, you?re in constant dialogue with them.? As a result, the people that remained active in PONARS were, in their opinion, ?the residence of a bizarre set of personalities that are having trouble finding outlets elsewhere, as opposed to a gathering of the best and brightest.? 818 One possibility they suggested was to have a type of ?term membership,? similar to the Council on Foreign Relations, that would admit new members on a trial basis, after which they would either be promoted to full membership or no longer be members. Wallander and other network organizers had previously considered whether to ?graduate people from PONARS? once young scholars got tenure or some similar standard, but ultimately decided that ?maintaining the network was more important than sticking to some kind of age or tenure criteria.? 819 In Wallander?s defense, building those connections over time is arguably how you build trust, collaboration among members, and ultimately how younger members might learn from older ones (if the latter participate). Unfortunately, some government participants also found declining utility in hearing the ?same people? who ?after a while, you know what [they are] going to say on an issue?.At the moment, at least, I don?t really see enough of a variety of voices on the 817 Interviews C55 and C481 818 Interview C481 819 Author?s interview with Celeste Wallander, Washington, D.C., November 17, 2005 253 questions.? 820 Another similarly recommended ?there should be turnover, younger folks coming in all the time, and people graduating and going off on their careers?and then maybe their professional connections will carry them through in the future, as opposed to having the same people over 5-10 years meet all the time.? 821 Even Speedie, who has since moved within Carnegie and is no longer involved in reviewing PONARS, advised ?by definition of the title of the organization itself, new approaches would mean that some people would move on, and others are brought into the fold?.Let me put it this way: if the membership does not evolve in that way, then that would be something of a criticism.? 822 Maximizing Policy Influence Although PONARS is not exclusively designed to influence policymakers, some issues raised during interviews recommended four ways that the network could increase it. The first is another critique on the membership itself. The network could be more influential if it allowed more members with government experience or even government officials themselves. Some current members?including Fiona Hill, Mike McFaul, Sarah Mendelson, and Wallander herself?have subsequently moved to Washington. That should theoretically increase the network?s influence. But these are the members who have arguably found the network less useful as they have developed their own contacts. They also do not have government experience, at least not yet. Wallander cautioned that changing the membership criteria would be a radical step and ?it?s really important to 820 Interviews C57 and C51, respectively 821 Interview C47 822 Telephone interview with author, November 14, 2005 254 remember what you?re trying to do.? 823 If the purpose is to bring scholarship to government, then such a change would dilute scholarship. If, on the other hand, the primary purpose is to influence government, interviews reaffirmed the theoretical findings of chapter two and the other case studies: the revolving door helps. One current member advocated opening membership to ?young people working the Russia desk or the guy working at policy planning on Russia right now?.I think one needs to revisit the selection process for who is in and who is out.? 824 A member of the revolving door stressed that government membership could facilitate relationships that are often the key to influencing, or at least informing, policy from outside government. 825 From government?s perspective, it is a part of ?a public diplomacy strategy?. We ignore at our peril what folks on the outside are saying, thinking, and writing about issues that we deal with on a regular basis.? Outside government, more intimate exposure to government could help develop ?an appreciation of how difficult it is to be in the government,? particularly ?the difficulties of achieving consensus? and how, at the end of the clearance process, ?you can wind up with something that is dumbed down [to its] lowest common denominator.? 826 A government desk officer stressed that incorporating government members would enable government employees to better inform nongovernmental members about what steps the United States had taken, considered, or rejected, and why. 827 A second, less radical, suggestion was to have smaller working groups on discrete issues supplementing the annual policy conference. In this way, membership would 823 Interview with author, Washington, D.C., November 17, 2005 824 Interview C55 825 Interview C62 as well as C55 826 Interview C62 827 Interview C61 255 remain limited to academics, but dialogue with the government would be enhanced. Two current members with Washington experience separately suggested the idea. 828 Independently, a number of government participants expressed interest in smaller, more focused groups in addition to or even instead of the annual conference, which one analyst called ?a circus.? 829 PONARS actually has run and received funding from Carnegie for occasional working groups called PONARS Plus. Among six that were held from July 2003-July 2005, one in February 2005 focused on nonproliferation, although its content was private. The problem with doing too many of them, explained Wallander, is the amount of time and staff work required to run them. 830 A third suggestion was an even more intimate interaction: more briefings for individual government officials or other ways to interact beyond the annual conference. At least one former PONARS staff member seemed to take great pride that PONARS primarily distributed even memos and conference invitations only to people who requested them. ?I got stuff in my e-mail all the time from other think tanks that I haven?t signed up for. We never did that. People had to come to us to get the memos, the working papers. If they signed up for them, it is because they wanted them.? Although they acknowledged in retrospect that they might have been more aggressive about marketing PONARS, ultimately the staff member felt, ?It?s really not up to us. We put out a product.? 831 828 Interviews C55 and C481 829 Interviews C50 as well as C44, C59, and C60 830 Wallander, 2003 Final Grant Report, p. 2, and 2005 Final Grant Report, p. 7, as well as interview with author, Washington, D.C., November 17, 2005 831 Interview C43 256 One government analyst who frequently attended the conference, while complimenting the organization for providing a valuable opportunity to listen to ?very erudite and informed and useful? discussions once a year, simultaneously criticized that It has always struck me that this is a club of insiders?I have never felt like anybody reached out to pull me in?.I don?t mean any of this disparagingly, and I am very grateful that this organization exists, but the idea that it exists, in part, to act as a transmission belt between the policy community and the academic community, and that they try to network, I think, is something that while I won?t say is laughable, I will simply say I have never seen it. 832 In addition to reaching out to brief officials, a number of interview subjects recommended consulting officials more widely to help determine the policy conference agenda. 833 Such consultations might give PONARS members greater insight into what issues were really dominating government policymakers? attention. Even if they didn?t, they could still help build the relationships that might make PONARS more influential. The key is to get more than annual interaction with the policy community. As one government official advised, ?you can?t trust the fact that the guy will realize it?s a good idea the first time that he hears it. You have to be persistent and timing is really critical. There comes a time when people are searching for new ideas, and if you?re around at that time and you got it, then you?re in. It?s really amazing how that works.? On the bright side, they continued, since the network?s headquarters and some of its members had moved from Harvard to Washington over time, Now you have a mechanism at CSIS and elsewhere to get ideas into the system. I think what that means is that, for the young American scholars, it might mean a more concerted effort by Celeste and her team to follow up 832 Interview C51 833 Interviews C48, C56, and C61 257 on a good idea that gets identified and maybe get some resonance at the December meeting, and then push it in January and February, or create opportunities to re-engage. You know how it is in Washington, you can?t get a policy guy to come spend all day in a meeting. Maybe you?ll get half an hour of his time, targeted. It may well be that part of PONARS has to evolve into something where the Washington PONARS folks help identify the good ideas that could play, and then have a more focused campaign to engage the right people in State and Defense and Energy to try to advance the idea, as opposed to being more passively briefing a meeting where you may not even get the people showing up that need to hear about it. 834 None of those interviewed expressed any intention to criticize Wallander or the job that she has done in creating PONARS. On the contrary, interviews with both members and government officials revealed an overwhelming and nearly universal admiration for an ?outstanding job of leading the organization,? even among some skeptics who initially didn?t think PONARS would work. 835 It turns out that Wallander had actually put together at least one of these briefings for about ten analysts in 2000, at the request of a White House official at the time who claimed to have gotten ?unique and interesting? insights from it. The session was held just prior to the annual policy conference, focusing on nonproliferation policy toward Iran and the priorities of the new Russian leadership under Putin, who had just been elected president earlier in the year. 836 The problem is not necessarily the desire to run these smaller sessions, according to Wallander, but the limits on staff and resources to be able to consult more widely, and more effectively track ?where U.S. policy on Russia is so we can find those nodal points?. Honestly, I think we would be more useful if we had our own independent office and capacity to organize meetings more regularly, and to 834 Interview C47 835 Interviews C1, C25, C42, C45, C47, C56, and C57 836 Interviews C2 and C54 258 shape the discussion in Washington more regularly instead of just a yearly conference?but that costs money.? 837 The fourth and final suggestion to maximize policy influence revolved around other opportunities to have government officials address network members, not simply listen to them or read their memos. Some government analysts seemed to perceive a sort of arrogance that the network would only have something to tell government and not also be able to learn from them, 838 potentially reducing the network?s influence just by its appearance. More substantively, earlier discussion consistently highlighted the perception that PONARS analysts are minimally informed about internal U.S. policy debates. Addresses by government officials could help fill that gap. Assistant-secretary level officials, who would not consider attending unless they were a speaker, might be invited. At least their deputies and NSC directors would also presumably be more likely to attend if they could address the group. An early attempt at including a government speaker on a PONARS panel was ?a complete disaster,? 839 according to one member. Wallander also expressed concerns about ?the amount of work involved in trying to get these folks? as well as a potential quelling effect on discussion because a high-level administration official was there. More specifically, she said ?I want it to be a discussion about what is going on in Russia, and if you have a U.S. policy person standing up there, automatically you are going to start arguing about defending or not defending U.S. policy?.that?s not our niche.? 840 Carnegie?s Speedie added, ?I?ll be very honest, I?d never thought of [government 837 Interview with author, Washington, D.C., November 17, 2005, as well as C55 838 Interviews C51 and C61 839 Interview C54 840 Interview with author, Washington, D.C., November 17, 2005 259 speakers]. All I knew was that in the beginning, it was clearly designed as an opportunity for PONARS people to strut their stuff.? 841 Network members expressed mixed support for such an idea with one, more involved in policy on a regular basis, saying ?I think it could have been brilliant if you could have, I don?t know, Bob Joseph [now the Undersecretary of State for Arms Control and International Security Affairs and formerly an NSC director for Proliferation Strategy, Counterproliferation, and Homeland Defense] or someone like that? give the administration?s view about these issues. 842 A prominent academic member was more lukewarm, ?It wouldn?t be useful so much for me professionally because my work these days is not all that policy-oriented anyway, but it would be intellectually interesting for me to get more insight into how the government feels about these sorts of issues.? 843 Ultimately, a government official need not speak on one of the panels. They might be invited to exclusively address network members on the margins of the conference, either the day before or in a dinner afterward, or even to be a keynote lunch speaker (they would be unlikely to attend the rest of the conference based on these interviews). Effects in a Crisis? Although the theory discussed in chapter one predicted a role during a crisis for these networks, the reality is that any impact has been constrained because the conference is almost always held at the same time every year, preventing any ability for the network to respond in real-time. What a crisis does do is concentrate the focus of network members, resulting in a surge in listserv discussion or an occasional memo detached from the 841 Telephone interview with author, November 14, 2005 842 Interview C25 843 Interview C26 260 conference. Conferences did reflect on crises, elections, or even upcoming summits. Examples include a December 2005 preview of Russia?s July 2006 hosting of the G-8 summit; February 2005 discussions of the implications of Bush?s and Putin?s 2004 reelections, the September 2004 Beslan school hostage tragedy, and the November 2004 Orange Revolution in Ukraine; 844 December 2003 panels on the 2003 Duma elections and the October Yukos scandal; a December 2002 lunch session on the October 2002 Nord- Ost theater hostage crisis; 845 a January 2002 panel on the implications of the September 11 attacks; discussions in May 2001 after Bush first took office in January 2001 and in December 2000 on Putin?s initial March 2000 electoral victory; and a December 1998 focus on the implications of the 1998 financial crisis. The influence of these sessions on government is unclear, but they demonstrate an interest and focus by network members, at least in interpreting their long-term implications. Socialization, With a Twist Unlike the other cases that focused on socializing China or North Korea into the practice of multilateralism or regional or global norms, interviews for PONARS focused on whether the network socialized academics to think about policy issues. Two interviews suggested that the editorial process could help train members to write more concise policy memos, even if a more intrusive process might not be popular among academics traditionally used to lighter editorial touches. 846 PONARS staff also reported that memos are being used in classrooms in 21 U.S. universities?including Cornell, Columbia, 844 February 2005 Policy Conference report, pp. 1,4 845 December 2002 Policy Meeting report, pp. 5-6 846 Interviews C481 and C52, as well as Wallander, ?Borderlands,? p. 39 261 Dartmouth, Harvard, MIT, Princeton, and SAIS?and 12 institutions in the former Soviet Union including St. Petersburg State, integrating policy issues into teaching. 847 The problem, however, is more fundamental in the academic system. Although there may be some exceptions, members complained that universities traditionally dissuade junior faculty both from networking or collaborating with others, emphasizing individual achievement, and from seeking policy relevance, preferring to emphasize theory. 848 Wallander explained that those who focused on Russia or the region risked being branded as area studies experts, rather than social scientists, and thus ?as lowlife academics who aren?t real scholars because they?re not really theoreticians.? 849 Wallander herself was denied tenure at Harvard, prompting her move to Washington, and revealed that ?one of her senior colleagues referred to how she had ?wasted? her time providing for the ?public good? of PONARS.? 850 Ironically, Wallander had heard anecdotal evidence that PONARS membership was ?noted as a positive factor in the successful tenure cases of members? because it was ?selective.? It was too hard to tell how widespread that effect might have been or, in other words, whether PONARS had a transformative effect on academic standards and interests, although it seemed unlikely. It appeared to be sporadic. 851 Interviews also revealed that PONARS had a positive, bonding effect on members who were coping with a decline in Russian studies after the Soviet collapse. Some of the effect was simply social by boosting morale among a group of young academics facing 847 Wallander, 2005 Final Grant Report, p. 4 848 Interviews C26, C43, and C47, as well as Wallander, ?Borderlands,? p. 36, and 2003 Final Grant Report, pp. 2-3 849 Interview with author, Washington, D.C., November 17, 2005 850 Wallander, 2003 Final Grant Report, p. 2 851 Wallander, ?Borderlands,? p. 36; and interview with author, Washington, D.C., November 17, 2005, as well as interview C43 262 job or research money scarcity. Part of the effect was also practical, building collaboration among members, sharing research and information about job openings or grant money. 852 More specifically, members felt PONARS improved studies not just by making memos available for teachers but because, as one member said, ?it puts students also in contact with a network of experts,? 853 and because faculty members could more easily consult other members for their own research or to ask for suggestions for what to put on a syllabus for a new course. 854 One member of the revolving door broadly warned that ?this is a field that is in jeopardy.? 855 Wallander herself was surprised that PONARS ?actually ended up helping preserve something of a post-Soviet studies field in the universities, which was completely unanticipated and I honestly had nothing to do with it. It just kind of happened. I think that a lot of PONARS members, to my surprise, value it more for that than almost anything else.? 856 The other postulated socializing effect was on raising academic standards of the Russian members: ?how you do research, how you attribute your sources, what is considered valid scholarship, and what is not,? and even ?to understand what peer review is like.? 857 One specific advantage, one government analyst raised, was to give scholars from generally underfunded institutions money to travel so they could air their ideas and ?to listen to the responses that they get from their presentations?It helps enrich the field.? 858 In one exceptional case, PONARS member Fiona Hill took members of Brookings?s Board of Trustees to The Humanities University at St. Petersburg, where 852 Interviews C42, C55, C59, and C60 853 Interview C26 854 Interviews C26 and C54 855 Interview C59 856 Interview with author, Washington, D.C., November 17, 2005 857 Interviews C43 and C47, respectively, as well as C2 and C42 858 Interview C56 263 ?several? PONARS members were employed, to take a look at one of the dynamic examples of Russian scholarship. ?Some of the trustees were so taken with the university that they actually made a massive donation to it?.That, I have to say for me, was one of the most gratifying experiences that has come out of this?? 859 Influence in Russia? Beyond potentially raising academic standards among Russian members, interviews revealed no confidence that PONARS made a difference in that country. 860 The network didn?t really make policy recommendations; Russians were not members until 1999; only two policy conferences were held in Moscow; and think tanks in Moscow are not generally considered influential. In his 1999 book, Unarmed Forces, discussed in chapter one, Matt Evangelista found that nongovernmental analysts did have influence in Russia at the end of the Cold War, but that was because Gorbachev gave them influence. There is no similar evidence that Yeltsin or Putin went outside government to seek advice, much less to junior Russian scholars. One Russian PONARS member concluded, ?If you ask me how PONARS is known in Moscow, not among its members, but in policymaking circles, the answer would be zero?.I am not sure there were any significant attempts to bring attention [to] the Russian decisionmaking community.? 861 Conclusions Ultimately, the policy influence of PONARS has been to provide young, rising academics with access to government analysts to facilitate longer-term, open-source analysis for the 859 Interview with author, Washington, D.C., October 27, 2005 860 Interviews C43 and C51 861 Interview C25 264 intelligence community and desk officers. From the government?s perspective, this has been useful but its format has not been ideal. Memos are too long for policymakers; they are usually distributed all at once, overwhelming some analysts and marginalizing them to become reference materials for future use; and the network?s members, generally academics with little other contact with government officials, are not perceived to be particularly well-informed about current policy. Some government participants also felt the conferences were too big, although others thought it was an advantage to get a comprehensive annual survey of Russian and regional security issues. Because it was only held once a year, with little interaction beyond that session and few memos sent at other times during the year, the relationships that might be developed to facilitate further influence were blunted. Some in government resented that PONARS appeared to lecture them without an opportunity to give feedback to network members. In the end, it seems that smaller, focused roundtable sessions, also known as PONARS Plus sessions, were favorably viewed as one way to focus discussion and facilitate dialogue, at least to supplement the annual conference. Nevertheless, the members of government that kept coming back to PONARS, particularly analysts from the State Department?s Intelligence and Research (INR) bureau and the National Intelligence Council (NIC)?the analytical branches of the intelligence community that delved into open-source analysis?and some policy analysts served by those entities valued PONARS for five reasons: direct access to Russian nongovernmental analysts; 265 indirect access through American members to a deeper community of informed Russian opinion, particularly in think tanks and universities, beyond traditional government contacts; wider perspectives on Russia and the region beyond specific issues on which they worked; longer-term regional trends beyond the crisis of the day; and ultimately just a smart set of Russian and American scholars who could provide a sanity check on some of the issues about which they were thinking. The value was not from specific policy recommendations, which were usually not even made, but to provide policy-relevant analysis particularly on the long-term implications of crises, potential Russian or regional reactions to current or proposed U.S. policies, regional political trends, and the implications of political changes either in Russia, the region, or in the United States for policy toward the region. If the post-9/11 emphasis on open-source analysis in government and the intelligence community continues, networks like PONARS providing access to nongovernmental experts should have even greater potential influence. Outside government, among network members, PONARS has had an even greater and much more fundamental impact. Interviews revealed that it has knitted together a community of academics over a wide area throughout the United States, to Russia, within 266 Russia beyond Moscow and St. Petersburg, and more recently to other parts of the former Soviet space like Ukraine, at a time when attention to and funding for the field has withered after the Soviet Union collapsed. Through its listserv, academic conferences, the policy conferences, and relationships subsequently built from participation, PONARS has helped members with their own research, keep up with events throughout Russia beyond specific issues they were researching, and for those members that were either fundamentally or casually interested, on the interests of the Washington policy community as well as connections to that community, particularly early in a member?s career. 267 CHAPTER SIX: CONCLUSIONS Do transnational security policy networks matter? If so, why? Do they make a difference in U.S. policymaking? How should similar networks be structured to maximize their value? Whether contemporary transnational security policy networks have changed U.S. nuclear nonproliferation policies or the perceptions that shape them has been evaluated in this study to help gain insight on some of these questions. Whether the networks have succeeded at meeting their own goals, such as encouraging multilateralism in Asia, is not assessed here because of my own language limitations and inquiry into U.S. policy. None of these networks focused exclusively on nonproliferation, and none sought solely or even primarily to change U.S. policy. Other researchers might seek to ask different questions, including their effects on other security issues. Inquiries might be made about network effects in other countries, particularly given preliminary evidence on some South Korean perceptions of the Proliferation Security Initiative, and their apparent use by China and South Korea to persuade North Korea to enter into the six-party format. Many other benefits or drawbacks may exist. The answers in this study seek to shed just a sliver of initial light on the larger questions of why these networks matter, starting with their effects on the nuclear nonproliferation policymaking process in the United States. On occasion, references in this concluding chapter will be made back to individual theories that helped form the hybrid framework for analysis constructed in chapter one. However, since no single existing theory sought to explain the emerging phenomenon of contemporary transnational security policy networks, no section of this concluding chapter is dedicated to confirming or denying those individual theories which 268 informed but did not define this analysis. Instead, this concluding chapter contains an evaluation of the four benchmarks for analysis?any contacts and perceptions facilitated, U.S. government interaction, agendas set, and policy options generated?and recommendations before concluding. Findings One initial finding is that these networks serve a cooperative security role. The pioneering work on track-two diplomacy has examined their role as conflict resolution tools. The findings here show that there is some tension between the two roles. Including North Korea facilitated a conflict resolution exercise, but changed the agenda of network activities, making them broader, and the tenor of discussion, making them more formal and strained. When North Korea did not participate, other members of the six-party talks more freely discussed cooperative strategies and disagreements among them about the threat and how to manage it. Despite the tensions, these networks managed to do both. Contacts and Perceptions These transnational security policy networks have helped introduce participants to valuable contacts. The significance of those contacts varies from person to person and depends on the network. Most of CSCAP?s U.S. nongovernmental members knew each other, but the network linked those U.S. experts on East Asia with functional experts on nonproliferation, export controls, nuclear energy, and other security issues as well as with experts from East Asia and with government officials from all countries, including the United States itself. Most of NEACD?s governmental members knew each other, but not 269 necessarily the nongovernmental or military participants. PONARS?s generally younger academic members did not know other members from their own country, much less other countries, the media, or government. The concept of Congressional enterprises?or the ability of individual U.S. Congressional members to build their own pool of nongovernmental experts from which they could draw?was introduced in chapter two. Transnational networks similarly empowered individual governmental and nongovernmental participants with their own personal enterprises, particularly early in their careers. Because these networks have all been founded within the last thirteen years, the long-term effect of these ties remains to be seen. Early returns such as Jim Kelly?s rise from directing CSCAP to become the U.S. Assistant Secretary of State for East Asia, or Fu Ying?s ascent from participating in NEACD to become a principal Chinese participant in the 6-party talks are promising. Although it is difficult to prove, or in some cases even imagine, that these networks may have improved relations over time between the United States and countries like Russia, China, or North Korea, such initial contacts may at least enable stronger future personal working relationships if relations get better for their own reasons. In just their first few years, however, these networks have been particularly useful when they strengthened weak ties, as sociologist Mark Ganovetter advised in the 1970s. They helped introduce and maintain relations among experts from different countries (or even those dispersed within countries), professions (such as policymakers, professors, and analysts), and generations who worked on different issues (regional security, proliferation, or politics) or simply might have very little time to otherwise get together. 270 For its primarily nongovernmental members, these contacts and network activities facilitated the exchange of ideas and information from U.S. members? recent travel (such as after the bombing of the Chinese embassy in Belgrade), with officials and analysts from the region (as demonstrated by a Chinese member?s explanation of the Shanghai Cooperation Organization), or simply among smart people from any country. The exchange of such ideas and information provides further verification of both Joe Montville?s track-two diplomacy concept and Matt Evangelista?s transnational networks. Networks also enhanced contact with and the flow of information from U.S. government officials, helping addressing one of the most common criticisms of nongovernmental analysts? work expressed in this study: that they often didn?t know that their recommendations had already been considered in government, much less why they had been rejected. A theoretical exaggeration of government participation in networks is that it allows officials to express personal disagreement with their own government?s policies. Although there was no such evidence in these cases, off-the-record network sessions did allow government officials to go beyond what they were willing to tell the media, and explain to domestic experts, many of whom they know personally, details of recent events like what had happened at the latest round of six-party talks. They provided an intermediate option offering more information than what would be shared at public briefings, but less than what officials would pass along in one-on-one meetings with trusted colleagues. Conversely, government officials used network activities such as USCSCAP lunches to conduct public diplomacy, eliminating the media middlemen who had incentives to sell copies by highlighting disagreements with policy rather than elaborating 271 on government rationale. These networks enabled governments to be more candid and get their side of the story out to nongovernmental opinion leaders and experts. Government officials also learned from participating in, or briefings on, network activities because they were multilateral (particularly initially in Asia at a time when such gatherings were uncommon), informal, and because nongovernmental members brought different perspectives to ongoing events. Informality enabled conversations among government participants to wander beyond confined talking points, and try to better understand the sources of disagreement rather than having to conclude with negotiated settlements or talking points for the media that formal, official talks required. Meanwhile, nongovernmental participants enhanced that informality by asking questions that diplomatic protocol wouldn?t allow and ?getting in people?s faces,? such as the NEACD discussion directly asking North Korean participants whether it was even possible, and if so how, for the U.S. government to reassure Pyongyang. Government officials also received human intelligence from nongovernmental network members. This is not the 20 th -century style of intelligence from spies gathering and protecting state secrets, but a 21 st -century strain emphasizing open-source analysis about politics, perceptions, and policies overseas in a more transparent, globalizing security environment. As Stanford Professor Alexander George advised, networks often sought to diagnose the root causes of security crises such as North Korea?s pursuit of nuclear weapons, Chinese policy toward Pyongyang, or perceived Russian support for Iran or Iraq before Saddam Hussein was ousted. These nongovernmental experts got deeper into societies by interacting with their counterparts overseas with greater agility since they were not burdened with the inertia of only meeting in diplomatic channels or 272 the transaction costs of filing out clearance forms, receiving official approval for meetings, and filing reporting cables with these nontraditional contacts. They and their nongovernmental counterparts abroad more widely integrated issues, such as Russian nonproliferation policy with political trends, and countries throughout a region, particularly in East Asia, which bureaucratic stovepipes or the time pressures of governance often impeded. They looked longer term at either the implications of ongoing crises, such as the downing of the U.S. EP-3 plane over China in the spring of 2001, or at the historical roots of and lessons for current crises, like the perspective that Robert Scalapino was praised by U.S. government participants for bringing to NEACD discussions on North Korea. There is no question that recollections expressed in interviews were often frustratingly vague, describing general types of information learned without specific examples. Although some are cited here and in earlier chapters, only a few participants or government officials could recall specific pieces of information, new perspectives, or policy recommendations from networks or individual members when interviews pressed for further examples. Some officials attributed this to the deluge of information to which they were subjected, while others thought it might simply be because events or crises which were hot at the time changed so much from year-to-year, leaving more general residual impressions of a country?s objectives and strategies. Nevertheless, the types of information and perspectives that networks and their members brought to those current events were cited as valuable by government officials themselves. As one official frankly put it, ?I rarely remember exactly who told me what. But I do know that I keep going to these things, so?I must be getting something out of it.? The key is that claims are not 273 just being made by nongovernmental members tooting their own horns, but by government officials themselves. It would be inaccurate to suggest that the effects of these nuggets of information or distinct perspectives were earth-shattering or paradigm forming. One might best think of them as threads that were woven into a broader tapestry of individual official perspectives. U.S. government officials could draw on policy networks or their nongovernmental participants for the acumen of its members and their access to different sources of information for more honest assessments as well as deeper, wider, and longer- term perspectives on how other countries are politically changing and their foreign policy implications, how they might perceive current or future U.S. policies, and how to diagnose the root causes and implications of crises that unfold over time. U.S. Government Access and Interaction The preceding section still leaves questions of whom in government, how, and when they might be most likely to gain any advantages from networks. Because of time pressure on government officials, particularly policymakers at the deputy assistant secretary (DAS) level and up, it turns out to be very difficult to get their participation in transnational networks. After the inaugural session or two, DASs did not usually travel to participate in NEACD until North Korea joined, or observe PONARS conferences for even part of a December Friday a few blocks from their office. If these officials were invited to speak at a luncheon in town, such as the USCSCAP sessions, they were much more likely to come, although even then they might cancel at the last minute. At lower levels of the 274 government?such as office directors or desk officers?officials were more likely to participate or travel, particularly if they were new to the region. A more successful target other than policymakers, in at least one of the networks, has been the intelligence community (IC) that informs the executive branch. Since all three case studies included China and/or Russia, residual restrictions and suspicion about the community interacting with these countries may have made the results of these networks even more valuable to the IC. Their participation with some members from other countries, however, may still be tainted by the stigma of perceived spying especially on sensitive discussions, such as on North Korea or that some feared might evolve on Russia. Nevertheless, the more transparent branches of the community such as the National Intelligence Council (NIC) and the State Department?s Intelligence and Research Bureau (INR) have been interested consumers of and even contributors to the open-source analysis that network activities facilitate. Outside of the executive branch, the behavior of domestic think tanks predicted that Congress might be targeted by nongovernmental actors to raise attention to certain problems or propose solutions. In these networks, however, the level of discussion on foreign policy issues was usually too detailed for Congressional interest. Legislative participation turned out to be mostly a dead end. The challenge for all of these networks has been how they or their individual members should interact with government officials that do not participate. Written materials might be read, but it was much more likely if they were one-page briefs such as the PACNET series. Longer materials might be read by desk officers or intelligence analysts, but even then, longer individual memos or the volume of a large collection was 275 more likely to overwhelm already overworked officials. NEACD took an entirely different approach, recently hosting an annual briefing in Washington but almost entirely avoiding written materials. Another channel?attempts to virtually interact through listservs?worked among academic members in PONARS but was less successful even with members based in Washington think tanks, while NEACD?s attempts to virtually include government officials did not catch on. The most successful means of reaching out to policymakers has been the simple labor-intensive briefing of principal officials. Evidence such as former CSCAP director and Assistant Secretary of State Jim Kelly?s connections to Ralph Cossa and other network members demonstrates the value of personal connections and the revolving door to gain access to government policymakers. Other individual members may have connections or pursue such meetings on their own, but network organizers are the most knowledgeable about the full range of network activities, often making it incumbent upon them to do the hard work. In some senses, this seems fair because the directors gain knowledge and notoriety from running them. On the other hand, directors and their staff are already doing all the heavy lifting to organize the networks, their membership, meetings, agenda, and activities in the first place. Given that networks have not received the resources to make running them the sole or even principal responsibility of any of these three directors, it is a very time-intensive but arguably necessary responsibility to add briefing a wide range of government officials. Determining when those briefings might be most effective, and therefore when to seek them, is one way to limit the burden. Evidence from these cases reveals that summits, political changes, and recent crises drove many but not all network activities 276 (working groups in all three cases are notable exceptions). These activities could at least inform and crystallize the messages of participants that might be passed along to governments. Sometimes, governments even sought advice from networks or individual participants at these times. Summits like Clinton?s 1998 and Bush?s 2002 trips to China, Jiang Zemin?s 2002 trip to Crawford as well as Wen Jiabao?s 2003 visit to the United States, the 2000 intra-Korean summit, and Russia?s 2006 hosting of the G-8 summit were often the focus of network sessions. Summits involving the United States also spurred the U.S. government to seek outside advice, or at least be more willing to accept it, because officials invariably look for what are called ?deliverables,? or an agreement, statement, or other form of tangible evidence, that can be announced after a summit to demonstrate its success. Summits that did not involve Washington still provided opportunities for analysts to predict their long-term implications. Political changes in the United States or overseas also opened these policy windows, as John Kingdon called them, into government. An opportunity predicted by Donald Abelson and Christine Carberry?U.S. presidential election campaigns?did not yield results in these three cases. Each network included some participants who played a part in presidential campaigns at some level or sent materials from their activities to senior campaign advisors. 862 For one reason or another, however, in no cases did the participants or networks feel they had any noticeable influence. One explanation might be the nature of the three presidential campaigns run since these networks were created. Abelson and Carberry argued in 1997 that whether campaigns were likely to be influenced by nongovernmental experts would depend on two characteristics of candidates: their experience in Washington and the strength of their ideological views. 862 Interviews A13, A16, A21, B32, B37, and C42 277 Insiders were less likely to seek advisers; ideologues more so. The reality is that the absence of campaign influence may say more about the campaigns than networks, reinforcing Abelson and Carberry?s earlier findings. Clinton v. Dole featured an incumbent against a Washington insider; Gore v. Bush saw an incumbent square off against a candidate who some would consider an ideologue but had his own set of advisers and sought to distinguish himself from his father and the Republican establishment; 863 and Bush v. Kerry pitted an incumbent against another insider. At least for now, the implications of presidential campaigns for networks are inconclusive. Turnover in U.S. government personnel was more promising. Ralph Cossa?s opportunities to separately brief a new director of the regional security policy office in the State Department?s East Asia bureau and a new ambassador to South Korea, Chris Hill who has since replaced Jim Kelly as the Assistant Secretary of State for East Asian and Pacific affairs, provided new avenues into government. Discussions about the implications for U.S. policy toward East Asia when Hill replaced Kelly and Condoleezza Rice replaced Colin Powell as Secretary of State in the current administration?s second term also piqued network interest. Political changes in other countries provided another set of opportunities to open the U.S. government to outsiders. The former White House official who requested and praised a PONARS-organized session in 2000 on Putin?s policy toward Iran?s nuclear program, just after he replaced Yeltsin, is one example. All three network activities are littered with attention to these political transitions including Hu Jintao?s rise in China and Putin?s in Russia. Although it is possible that changes in China or Russia may have 863 For an excellent look at the advisers to George W. Bush, see James Mann, Rise of the Vulcans: the History of Bush?s War Cabinet (New York: Viking Press), 2004. 278 increased network opportunities because of their dynamic changes and unique relationships with the United States in the last fifteen years, networks also focused on the implications of political changes in allies such as South Korea after Kim Dae Jung and then Roh Moo-Hyun were elected. Crises, on the other hand, had a different effect than summits and political changes on government interaction. Specific crises such as the 1998 North Korean missile test, the 1999 Chinese embassy bombing in Belgrade, and the September 11 terrorist attacks certainly focused network activities. They even shaped networks by leading NEACD to shift the morning sessions of their second day from non-security issues to the host country?s perspective on recent events, and CSCAP to form a working group on export controls after the A.Q. Khan network was discovered. But little evidence exists from these three cases of any access to government immediately after these crises, when they are hottest. On one hand, this seems to negate John Kingdon?s prediction that, along with political changes, crises are a catalyst to open policy windows and Peter Haas?s conclusion that epistemic communities can help explain why crises emerged, identify policies to alleviate them, or help states define their interests in a crisis. Alternatively, it may refine these findings in the security field, which neither Kingdon nor Haas focused on exclusively. In fact, as discussed in chapter one, both John Ikenberry and Alexandra Carter raised doubts about the permeability of the U.S. security policymaking system, given the national interests at stake and at least the perceived need to have access to classified information. That system appears to be least accessible in the heat of a crisis. 279 Quite simply, the demands on government officials become overwhelming during these times. Networks whose members are scattered around the world are also not often agile enough to organize any immediate responses. 864 On the other hand, networks either virtually or physically organize opportunities to exchange information, helping inform individual members who are often asked to comment on crises in the media. After the immediate crisis has settled down, they may open windows for members to provide their opinions to government officials on the root causes or long-term implications of crises. In the near-term though, networks and their members are better off if crises provoked them to focus on information gathering and public education. Although the heat of a crisis is not usually the best time to approach government officials, experiences from these networks yielded better results before summits and around political and personnel transitions. A related lesson here is that, while it is natural for well-positioned network members and their organizers to focus on supplying government with findings from network activities, demand in government was not greatest at that time. Some government interest might be generated from network activities, particularly from working groups that focused on over-the-horizon challenges. But governments often opened policy windows and were more interested in opinions and analysis from outsiders based on their agenda and events, not network activities. Even the most determined network member efforts are not necessarily sufficient to get the attention of government officials?particularly because transnational policy networks did not seek to leverage public attention through the media. Governments must be willing to open policy windows for networks to inform them. In these three cases, that 864 Also see Wolfgang H. Reinicke, ?The Other World Wide Web: Global Public Policy Networks,? Foreign Policy (Winter 1999-2000): 47. 280 was more likely when officials had themselves been members of the revolving door, or when they knew and trusted the individual organizers or members who sought to brief them. Ultimately, although at least lower level officials interviewed for this study were arguably a sample biased in favor of networks by their continued participation, the results precisely demonstrate the point about policy windows: those officials who did participate in networks or at least interact with their members believed they benefited. It is in their interest to open these windows and accept briefings, at least at key times. The effort here is to better define the circumstances and times most likely to benefit officials. But as one DAS who is a member of the revolving door said, ?We close the door at our peril to thinking on the outside.? Agenda Setting Despite the findings of John Kingdon, Andrew Rich, and others that U.S. think tanks were often most successful at setting the government agenda, there is little evidence that these networks were, or even sought to do so. Instead, networks often sought to bring a different?longer range, more integrated?perspective to issues already on the agenda. The one exception might have been the work on nuclear transparency through PACATOM that appeared to stumble into an area that slipped through U.S. bureaucratic cracks, effectively serving more like a safety net for the U.S. government. By design, though, this study focuses on transnational policy networks, as opposed to transnational advocacy networks like the International Campaign to Ban Landmines. In other words, it effectively screened out those networks that would primarily use the media to pressure the government and raise an issue on its agenda. 281 Generally, transnational networks have two avenues to seek to influence government: behind the scenes, as policy networks sought, or publicly, as advocacy networks do. It is possible that some members of policy networks, often populated by experts who had served in government before, were more sympathetic to the demands of governance and not inclined to pressure current officials through the media (or, as former military officers might say, to go outside the chain of command). The only network that allowed media to participate and report on their activities was PONARS, whose members were academics who almost universally had not served in government and did not feature any government speakers anyway. Conversely, NEACD published very few written documents at all, relying on confidentiality to maximize the candor of proceedings for government participants, and some in CSCAP even speculated that Cossa might have been hesitant at times to get his friend and former boss Jim Kelly in trouble by calling him out in public. It is possible that other transnational networks might sequentially try both strategies: working with government officials first and, if that yields no results, shifting to a more public strategy to pressure them. On the other hand, it is also possible that differences in membership (with policy networks comprised of more experts) and the subject matter (with advocacy networks generally focusing on more principled or humanitarian concerns) lend networks to choose either a behind-the-scenes (policy) or public (advocacy) strategy. There is a trade-off, however, between having influence with current government officials behind the scenes, and seeking to influence them through the media. 865 While policy networks might have access to government officials, their autonomy could be 865 This point is also made in Herman Joseph S. Kraft, ?The Autonomy Dilemma of Track Two Diplomacy in Southeast Asia,? Security Dialogue 31:3 (September 2000): 343-356. 282 compromised. In other words, not only might policy networks not seek to set the government agenda, governments may actually seek some influence on network agenda. The trade-off is that governments would by definition be interested in the results of network activities. In two cases, this dilemma manifested itself in an even more direct way: the government funded CSCAP and NEACD. The effect may best be described by recalling Kent Weaver?s categorization of think tanks from chapter two: government-funded networks were more likely to act like consulting tanks, while other sources of funding enabled what Weaver called ?universities without students? or more autonomous enterprises, such as PONARS whose members are academics and funding is from foundations. Because of the effects of CSCAP and NEACD, it is worth a one-paragraph digression to comment on the overlooked impact of consulting tanks. The existing literature on think tanks, for obvious methodological reasons, has been limited to a public study of the impact of think tanks on public discourse, the media, or Congress. This effectively has confined study to look at two of Weaver?s three categories, but has excluded an analysis of the behind-the-scenes consulting role that think tanks might play. Greater study on think tanks that directly interviews executive branch officials and investigates their consulting role is warranted. Unfortunately for networks, the playing field to perform either of these two different roles is not even. Foundations generally do not fund projects indefinitely, seeking instead to start them and move on to new things, confronting networks with a dilemma. Governments, with greater resources at their disposal and their own priorities, 283 appear to be willing to fund projects for longer periods of time. That provides an incentive for networks to seek government funding which enhances their consulting, rather than agenda-setting, role. It is tempting to conclude that such government influence nefariously co-opts networks. But that is not necessarily the only reason networks seek a consulting role. Their members are frequently experts who formerly served in government, some of whom may prefer the greater access and nuance behind the scenes. While there is surely a role to be played by autonomous organizations that bring public pressure and scrutiny on governments, the reality is that these more tightly linked policy networks play a different, still useful and not necessarily sinister, role. The absence of any evidence that overseas participants or governments dismissed policy networks funded by the U.S. government demonstrates this value. While it undoubtedly was essential to be transparent about U.S. government funding sources, these linkages may have assured overseas participants of Washington?s interest in network activities. Participants seemed to relish the opportunity to use networks to explain their own country?s policies. The fact that the U.S. government funded a network made it no more likely, for example, for U.S. participants to convince Japan to give up using plutonium as an energy source than it did for Japanese participants to convince the United States why it should pursue it. Although these three networks rarely sought to set the U.S. agenda, U.S. participation in at least CSCAP and NEACD was more focused on agenda setting overseas. In other words, they facilitated the ability of U.S. participants to conduct unofficial diplomacy. These findings reinforce the three stages that Matt Evangelista observed in these networks? Cold War predecessors: transnational networks can help 284 share ideas and information, transfer norms, and even coordinate policies. Contemporary examples demonstrated these functions were being performed by both nongovernmental and governmental actors, either simply out of coincidental interests or even explicitly as partners. Just as these networks had helped share information, explanations, and ideas between U.S. officials and nongovernmental experts as a domestic public diplomacy forum, they played a similar international public diplomacy role. U.S. Secretary of Defense Cohen?s 1997 briefing on the Quadrennial Defense Review for the first foreign audience, Lew Dunn as well as, separately, former Assistant Secretary of State for Nonproliferation Robert Einhorn on the Proliferation Security Initiative (PSI), and a team from the University of Georgia on export controls were all examples of either official, unofficial former government, or nongovernmental experts explaining U.S. policy to overseas participants. From CSCAP?s inception, its U.S. nongovernmental members sought to prioritize a nonproliferation norm with Asian governments and strategic thinkers, even after facing regional resistance, reaffirming Evangelista?s second stage. This general nonproliferation work was followed by specific efforts to embed a nuclear energy transparency norm through PACATOM. Other initatives, such as CSCAP?s support of a regional arms registry and defense white papers as well as NEACD?s principles of cooperation and defense information sharing, sought to more generally embed transparency as a norm in security affairs. The third stage, policy coordination, was also predicted by Joe Montville?s final stage of eventual joint strategies in track-two diplomacy. U.S. nongovernmental member 285 efforts to get Asian governments to accede to and indefinitely extend the NPT in 1995 demonstrate policy coordination, as do similar accession efforts to the CWC, BWC, MTCR, and the Additional Protocol to the IAEA Safeguards agreement. European participation helped convey the potential for EURATOM to serve as a model for PACATOM, and for an EU Action Plan for Nonproliferation to provide a template for an Asian effort. U.S. government clarifications of media misperceptions of the initial round of the six-party talks, and efforts to explain and gain support for the Proliferation Security Initiative do the same. Other examples outside nonproliferation such as preventive diplomacy and coordinating maritime policy demonstrate similar unofficial policy coordination. Did PONARS include similar diplomatic efforts, or agenda setting overseas? Evangelista himself is a member of PONARS, and was kind enough to be interviewed for this study. Although he thought the network was analogous to the Cold War predecessors he had studied, Evangelista predicted that PONARS?s members, including himself, were too young and did not have the access to influence the Russian government. He also felt that the political context was so different because the shared anxiety to end the nuclear arms race did not bind a network today as tightly as it had during the Cold War. 866 In Unarmed Forces, he had also concluded that Gorbachev had turned outside the government for ideas and allies to implement them in the Soviet system. Now, that system was too disorganized under Yeltsin or, more recently, more suspicious of nongovernmental actors under Putin as the recent controversial legislation restricting NGO activities and increasing their reporting requirements to Moscow shows. In this case, the network did not have the right members, the Russian government was not 866 Interview with author, Washington, D.C., February 4, 2005 286 permeable to outsiders, and the common cause did not exist to have the same purpose and effect. In the other two cases, diplomacy was conducted by more senior nongovernmental and sometimes even official participants in their private capacity, like former Principal Deputy Assistant Secretary Don Keyser. As previously discussed, their multilateral, informal format and the catalytic presence of nongovernmental actors could spark discussions to explore behind talking points, formal agenda, and the search for negotiated press statements and agreements that governed official negotiations. The ability to get beyond these talking points and understand the interests that shaped them, not just in North Korea but in China or even South Korea, through transnational networks were what Roger Fisher and William Ury advised could help eventually reach a negotiated agreement or, as they put it, to get to ?yes.? The methodological problem, of course, is that it is difficult to trace how U.S. or overseas officials might take these unwritten insights from networks and put them into agreements. Rather, it is more likely to affect the understanding and norms that diffusely help shape them, as well as their coordination. In these cases, embryonic official relations with China in networks? initial stages, and with North Korea recently, provided additional diplomatic value from transnational networks. Although some theoretical concerns were expressed that these networks provided a platform for North Korea to conduct their own public diplomacy and potentially could relieve their isolation as well as reduce their incentive to participate in the six-party talks, the networks appeared to facilitate those official talks. At a March 2003 CSCAP working group, South Korean and Chinese delegations may have 287 encouraged North Korean participants for the first time to abandon Pyongyang?s demand for bilateral talks with the United States and participate in the multilateral format. After the first round of those talks were held five months later, a September 2003 NEACD session in Qingdao allowed Keyser to follow-up on that first round, reiterate what Washington felt was most important, preview to Pyongyang what the United States perceived was crucial to the success of the second round, clarify media misperceptions that the United States demanded North Korea make all concessions before receiving any benefits, and elaborate on the benefits of an executive agreement as opposed to a treaty which required Congressional ratification. An April 2004 NEACD meeting enabled Einhorn, who had formerly served in Democratic administrations, to try to dissuade North Korean participants from stalling official talks for the subsequent seven months if they were hoping that they could get a better deal if John Kerry won the U.S. election. Ultimately, the primary agenda setting role that transnational policy networks played was not in the United States. The primary effort was either to provide a new perspective on an already existing agenda item or to set the agenda in other countries by acting as a forum for unofficial public or private diplomacy. What these networks often sought to do, therefore, was to increase understanding and, where possible, agreement. In other words, whether in cooperation with government or by independent nongovernmental initiatives, these networks have served as a forum for soft power. In their most fundamental role, simply ?information is power,? according to the opening page of Joseph Nye?s influential book, 867 and its exchange across countries and a variety of other weak ties is facilitated by transnational policy networks. But even 867 Joseph S. Nye, Jr., Soft Power: the Means to Success in World Politics (New York: Public Affairs, 2004): 1. 288 further, the diplomatic initiatives they enable has the potential to build a consensus among participants ?to influence the behavior of others to get the outcomes one wants.? 868 As the world globalizes, Nye emphasizes, this power to reach agreements does not rest exclusively in governments but is increasingly devolving to nongovernmental actors. 869 The soft power of networks will undoubtedly vary depending on who participates in, or are briefed by, networks as well as how other participants wield influence in their own country?s governments and societies; each country?s context will be different. Nevertheless, transnational networks exchange information and build consensus among opinion leaders, policy planners, government participants, and others within the foreign policy elite on the seriousness of threats such as proliferation itself, diagnose their root causes, seek norms, explore strategies, and even potentially coordinate international policies to respond. Policy Options With two possible exceptions, these networks did not develop policy options for Washington to consider adopting. For the most part, they did not even try. Some CSCAP efforts focused on the ASEAN Regional Forum or coordinating regional policies in countries overseas. Others, such as the North Pacific working group where much of CSCAP?s early work on North Korea was conducted, expressed concern ?about its capacity to produce policy recommendations? from the beginning and never even issued a memo. PONARS memo instructions advised members ?not to think in terms of writing action memos, and not to try to duplicate the work of government analysts.? NEACD also 868 Ibid., p. 2 869 Ibid., especially pp. 14-15, 17, 32, 99 289 avoided trying to reach policy conclusions. In fact, an advantage cited by a senior government policymaker on North Korea was precisely that these sessions did not seek to reach policy conclusions, allowing greater opportunity to identify and explore differences that shaped negotiating positions. One of the two exceptions may have been the recommendation to synchronize comprehensive, verifiable, irreversible disarmament (CVID) of North Korea with similar assurances (CVIA) to them. Even this instance was not a network recommendation, but a suggestion by at least one Chinese CSCAP scholar, and a separate discussion in NEACD provoked by a nongovernmental member asking participants for their perspective on the CVID formula. Over time, that formula appears to have been publicly dropped from U.S. declarations (although its principles still shape U.S. thinking) and the September 2005 joint statement of the six-party talks in principle incorporated security assurances. These networks were two of a number of channels through which the United States was being discouraged to publicly drop the CVID formula and extend security assurances as part of any agreement with North Korea. By no means does that mean networks were irrelevant, but they were not unique. To borrow the words of one of the interviews from another context, it may have been more like another ?drop of water on the rock? eroding the CVID declaration. The second exception is a set of contingency planning options, which might also be called a form of trial balloons. As one government participant explained, networks were where you could ?imagine the possible? but not be held accountable for it as official U.S. policy. Work to explore what a verification regime with North Korea might look like and discussions about the merits of an executive agreement are both examples of 290 contingency planning if political conditions were to change and an agreement with North Korea were reached. More broadly, the entire NEACD network was a giant organizational experiment to see what six-party talks might look like, a decade before that format officially existed. The idea was not to make immediately implementable policy recommendations, but to relax practicality a bit and be able to provide a first draft of an agreement if politics changed. A number of reasons could explain why networks did not seek to change U.S. policy. It might be unique to the United States. Network behavior could confirm John Ikenberry?s hypothesis that the strength of the global hegemon made it less permeable to outsiders. It could corroborate Matt Evangelista?s thesis that democracies are too messy for nongovernmental actors to have much influence and that, ironically, centralized systems are more amenable. It might be unique to the Bush administration, which has been roundly criticized for limiting decisionmaking to a small inner circle and not listening to outsiders. The networks may have been too weak in their early stages to have much effect at all, with PONARS only moving its operations to Washington in 2001, North Korea just participating in NEACD since 2002, and CSCAP restructuring its working groups in 2004. Think tanks and the nongovernmental sector in other countries did not often effectively exist before the end of the Cold War, potentially leaving them too weak to seek to change U.S. policy through networks. American network directors may have been more focused on changing other countries. As explained before, U.S. government funding may have encouraged networks to look at other countries. All these reasons are possible, but they only explain why networks may not have effectively changed or focused on the United States. In reality, some network members 291 did make an impact in Washington, but they rarely sought to change policy immediately; they sought instead to inform it and shape it more broadly. The networks most successful at informing its members have been the ones that strengthen weak ties. As it turns out, the very weakness of these ties and diversity of opinions and experiences that makes networks the most useful are the same factors that make consensus among members, and therefore policy recommendations, difficult to build. Diversity impedes consensus. While individual members might have their own policy recommendations improved by what they learned at network events, networks themselves rarely reached joint statements. CSCAP memos were infrequent, NEACD principles for cooperation were bland, and PONARS did not even try. Tanja Borzel?s distinction between homogenous and heterogeneous networks, 870 or Hugh Heclo?s between issue networks and shared-action groups, could have predicted this result. There is a trade-off between heterogeneous networks, which are more effective at exchanging ideas and information, and homogenous networks, which are more likely to reach consensus recommendations. Each of the three networks created smaller units?CSCAP or NEACD working groups and PONARS Plus?to focus on more specific issues when they sought to make policy recommendations. Chinese participants were also drawn from CSCAP for an independently run 1999-2001 Wilton Park project on P-5 security cooperation. One way to think of this effect might be to draw on network theory. 871 The networks studied here have often served as idea and information exchange bases, from which specific policy- focused cells might be formed for discrete periods of time. Although individual members 870 Tanja A. Borzel, ?Organizing Babylon?on the Different Conceptions of Policy Networks,? Public Administration 76 (Summer 1998): 261, 266. 871 Also see James G. McGann, Think Tank Networks and Transnational Security Threats (Philadelphia: Foreign Policy Research Institute, December 2003): 7. 292 may improve their own recommendations, networks themselves rarely generated or advocated policy options, and that is the way it should be. Recommendations These findings lead to a series of recommendations for governments and for networks that seek to influence U.S. security policymakers, but possibly also for foundations that might fund networks. Foundations are not a primary subject analyzed in this study, principally funded only one of the cases, and clearly have a responsibility to assess the return on their investment of any project they fund. That is extraordinarily difficult when the principal product is an intangible idea, exchange of information, or principle from which a diffuse process that shapes policy might begin. Nevertheless, to the extent that foundations look for evidence that networks produced a policy to demonstrate their value, they are looking in the wrong place. Rarely do any nongovernmental actors, even U.S. think tanks, make a radical policy impact, almost always having a much more nuanced and incremental effect. Their role is to analyze or shape or, as one network staffer said, to ?inform not influence? or make policy, at least in part because there are too many factors in policymaking beyond their control. For policy networks that seek to work with government officials, the standard should be access to government officials, as David Speedie indicated Carnegie emphasized, not about actual policies themselves. It is simply not realistic. A concomitant set of lessons follow for the U.S. government. Interviews for this study revealed a consistent sense of information overload among officials at all levels, and often an awareness of the stovepipes or lanes within which officials were expected to 293 operate, confining themselves to single issues and countries, rarely having or allowing themselves the ?luxury? to think about the bigger picture. As a result, networks can be useful because nongovernmental analysts integrate issues across countries as well as time and are not burdened by the demands of governance which impede such thinking. Given 9/11 Commission recommendations and recent failures, it is no great insight to add that the U.S. intelligence system is broken, burdened by information overload, stovepiping, and Cold War-era stigma on interacting with some other countries. Unfortunately, diplomatic protocol is also too formal to get some messages to the United States that it needs to hear. Finally, attention has recently turned to improving public diplomacy, but understanding the challenges remains in its infancy. It is well beyond the scope of this paper to holistically advise how to fix intelligence, revolutionize diplomatic protocol, or reinvent public diplomacy. It does, however, mean that government officials can and should use transnational networks as modern tools to conduct diplomacy and to learn. Except in special circumstances, like recently with North Korea, it does not require that senior officials necessarily travel overseas to network events or attend day-long conferences, especially given demands on their time. It is beneficial for lower level officials, analysts, and certain members of the intelligence community to do so, particularly those new to a region, issue, or even to diplomacy. But it does mean that even senior officials benefit from candidly speaking off-the-record over a meal with select nongovernmental experts, and from accepting or even seeking briefings from directors and members of select networks to improve the open-source analysis available to them. 294 Which networks are the most useful to government officials? The final set of recommendations is for networks that seek to inform U.S. security policymakers or provide a forum for governmental and nongovernmental actors to conduct public or private diplomacy. It is tempting to compare, contrast, and where they fall short criticize, the three pioneers studied here. But that is simply not fair to the networks as they did not primarily seek these roles in all cases. Nevertheless, they yield lessons?do?s and don?ts?for others that aspire to those roles. First, membership matters. Diversity powers the exchange of information and ideas through transnational security policy networks, strengthening weak ties across geography, expertise, generations, and professional backgrounds. Inertia will settle networks to work with current members, risking devolving into a group think mentality as new ties grow old and familiar. 872 New members can continually revitalize networks. Among their participants, transnational security policy networks should include members of the revolving door to maximize network value to other members and the U.S. government. Adding members who have recently left government is one way to keep membership fresh, but that is just the tip of the iceberg. Those with government experience better understand the challenges of governance and the realities of policymaking from their own experience. They are often able to get information from former government colleagues. Perhaps most importantly, they are more likely to go back into government, taking the training and knowledge they have received from network 872 Also see Wolfgang H. Reinicke, ?The Other World Wide Web: Global Public Policy Networks,? Foreign Policy (Winter 1999-2000): 56; Diane Stone, ?Introduction: Global Knowledge and Advocacy Networks,? Global Networks 2:1 (2002), p. 8; and Jan Martin Witte, Wolfgang H. Reinicke, and Thorsten Benner, ?Beyond Multilateralism: Global Public Policy Networks,? International Politics and Society (Politik und Gesellschaft Online) 2/2000 (www.fes.de/IPG/ipg2_2000/artwitte.html (accessed April 16, 2004)). 295 participation with them. As was the case with Jim Kelly, it is not that network contacts will change policy when those officials are in government, but network membership prepares them before they go into government, and enhances scarce access to exchange unclassified information with them when they are serving. Second, networks should seek multiple channels to influence governments. For starters, they should conceptually consider informing multiple governments at various layers: to multilateral institutions (such as ARF, the Gore-Chernomyrdin Commission, or even the UN or G-8), to national governments (such as the members of the six-party talks), and through individual policymakers or members who may be in government now, a member of the revolving door, or even nongovernmental members who may seek to influence governments publicly or privately on their own. When they do seek to interact with the U.S. government, networks should do so with strategies tailored to different levels of seniority. Particularly for policymakers at the DAS level and up, networks should seek to have them speak at special sessions for a limited period of time and do not require travel, otherwise it is unlikely they will participate. At lower levels, networks should seek, where appropriate, to get office director and desk officer participation particularly if they are new to an issue. Today?s office directors might even be tomorrow?s assistant secretaries. Networks should also strongly consider incorporating intelligence analysts, particularly from INR and the NIC, while being aware of the potential for an intelligence presence to quell discussion on sensitive issues or with participants from certain countries overseas. But when even the greenest government analysts participate, networks would be wise not to treat them exclusively like notetakers to take advice back to government. These participants are not 296 only sources of information to improve nongovernmental understanding of government perceptions, policy debates, and negotiations, but governmental analysts are more likely to learn and participate if they engage in a discussion than when they attend a lecture. They need not, and sometimes can not, address a large network conference, but they can engage in roundtable discussions on equal footing with network members, at least after a member or official delivers short initial remarks. If they do not participate in person, some officials will read publications, most likely one-page op-eds, but are unlikely to engage in any virtual diplomacy or online discussions. It is incumbent upon network directors and interested individual members to seek to brief officials at various levels both when networks can supply new findings or contingency recommendations, and when government demand is greater such as well in advance of summits or around political or principal transitions overseas or in the United States. Finally, transnational security policy networks should focus on facilitating the exchange of information and ideas, or the perceptions that shape policies. As such, they should remain vigilant against developing organizational bureaucracy and preserve flexibility to constantly reinvent new meeting formats and membership to adjust to the rapidly changing security challenges that globalization presents. Networks should envision themselves primarily as an expert base, to facilitate information and idea exchange among its members and with officials, while also providing members for working groups, or cells, which can delve into specific policy issues for a discrete period of time. Networks should also be endowed with resources from diverse sources to pursue these goals with the independence, energy, and creativity required. 297 Conclusions As security challenges have globalized in the wake of the Cold War, the ideas industry has been globalizing with it in the form of transnational security policy networks. Based on these three case studies, some preliminary hypotheses about the effects of transnational security policy networks on U.S. nonproliferation policies or the perceptions that shaped them have been proposed in this concluding chapter. Strictly speaking, the findings of this chapter are only based on three case studies and only apply to nuclear nonproliferation. But their implications are potentially greater. They are preliminary hypotheses which can be further tested not just by other transnational security policy networks on nonproliferation, but on other cooperative security issues as well. As defined in the introductory chapter, these preliminary hypotheses can be further tested by other transnational security policy networks (those involving regular international interactions including non-state agents not operating on behalf of a national government or intergovernmental organization) that deal with security threats which take a relatively long time to become acute, have a normative framework to address, and are a common interest of countries to cooperate to do so. One immediately obvious other potential security issue to which these findings may apply is terrorism. Networks other than CSCAP, NEACD, and PONARS may provide additional values screened out by case selection. Some may have members, such as the scientists that formed Pugwash, who may be sought for technical expertise not available in government. Others may be led by more politically seasoned directors such as the Aspen Strategy Group?s tandem of Brent Scowcroft and Joseph Nye. But these three case 298 studies provide a base from which we can begin to better understand the emerging phenomenon and value of transnational security policy networks. Assuming that developing improved capabilities in open-source analysis and public diplomacy remains a high priority in the wake of the 9/11 attacks and rising anti-Americanism, these networks can help. Their value is not in having an immediate and catalytic effect on policy or even in changing the government agenda or developing policy options for consideration. Some potential exists, but a big bang or magic bullet can rarely be traced to networks or even individual members. These networks are not primarily about changing policies immediately, they are about changing the perceptions that shape them. To better understand the value of these networks to U.S. security policymaking, analysts and officials should take a step back to look at their potential contributions to intelligence and diplomacy. Both as a sanity check on and to the intelligence community, particularly the NIC and INR, nongovernmental network members combine their acumen and perspective with blunter feedback from overseas to produce three-dimensional improvements in open-source analysis?deeper into societies, more widely integrated across issues and regions, across longer periods of time?to help diagnose the root causes and understand the implications of perceptions, politics, policies, and threats abroad particularly around summits, political and principal transitions overseas or in the United States, and the longer-term aspects of crises. Diplomatically, transnational security policy networks provide a forum to conduct public diplomacy within the United States as well as public and private diplomacy abroad to better get behind talking points, particularly but not only with nascent relationships such as China in the 1990s or North Korea today. 299 The benefits of these networks transcend their potential contributions to government security policy, even when they are structured and utilized most effectively for those purposes. They build the capacity for individual members and participants to do their jobs better by empowering them with ideas, information, and personal enterprises, drawing from a community of experts across national boundaries and stovepiped issue areas to strengthen the otherwise weak ties among them. They can even enable their members, participants, and interlocutors among the foreign policy elite to better understand and seek agreement on how to define and diagnose contemporary threats, on the political pressures shaping national responses, and on the merits and drawbacks of strategies to respond, serving as instruments of soft power for contemporary global security. These transnational security policy networks improve intelligence and private as well as public diplomacy, enhancing the analytical capacity and soft power of their participants and interlocutors. 300 BIBLIOGRAPHY Networks and Theory Adler, Emanuel, ?The Emergence of Cooperation: National Epistemic Communities and the International Evolution of the Idea of Nuclear Arms Control,? International Organization 46:1 (Winter 1992): 101-145. Berman, Maureen R. and Joseph E. Johnson, eds. Unofficial Diplomats (New York: Columbia University Press, 1977). Borzel, Tanja A. ?Organizing Babylon?on the Different Conceptions of Policy Networks,? Public Administration 76 (Summer 1998): 253-273. Carter, Alexandra G. Politics and Physics: Epistemic Communities and the Origins of United States Nuclear Nonproliferation Policy. PhD Dissertation, Georgetown University, 2000. Denning, Dorothy. ?Activism, Hacktivism, and Cyberterrorism: The Internet as a Tool for Influencing Foreign Policy,? ch. 8 in Networks and Netwars: The Future of Terror, Crime, and Militancy (Santa Monica, CA: RAND), 2001, ed. by John Arquilla and David Ronfeldt, pp. 239-288. Diamond, Louise and John McDonald. Multi-Track Diplomacy: A System Approach to Peace (West Hartford, CT: Kumarian Press, 1996), 3 rd edition. Evangelista, Matthew. Unarmed Forces: The Transnational Movement to End the Cold War (Ithaca, NY: Cornell University Press, 1999). Evans, Paul M. ?The Dialogue Process on Asia Pacific Security Issues: Inventory and Analysis,? in Studying Asia Pacific Security (Toronto: University of Toronto- York University Press, 1994), ed. by Evans: 297-318. Evans, Peter B., Harold K. Jacobson, and Robert D. Putnam, eds. Double-Edged Diplomacy: International Bargaining and Domestic Politics (Berkeley, CA: University of California Press, 1993). Fisher, Cathleen S. Reformation and Resistance: Nongovernmental Organizations and the Future of Nuclear Weapons (Washington, DC: The Henry L. Stimson Center, May 1999), Report No. 29. Fisher, Roger and William Ury. Getting to Yes: Negotiating Agreement without Giving In (New York: Penguin Books, 1981). 301 Florini, Ann M. ?Lessons Learned,? ch. 8 in The Third Force: The Rise of Transnational Civil Society (Washington, DC: Carnegie Endowment for International Peace, 2000), ed. by Ann M. Florini, pp. 211-240. ----- and P.J. Simmons, ?What the World Needs Now?? ch. 1 in The Third Force: The Rise of Transnational Civil Society (Washington, DC: Carnegie Endowment for International Peace, 2000), ed. by Ann M. Florini, pp. 1-15. Gelman, Lisa H. Talking Politics: The Role of Conference Diplomacy and Non-State Actors in the Global Regime Process. PhD Dissertation, University of Pennsylvania, 2000. George, Alexander L. and Andrew Bennett, ?Case Studies and Policy Relevant Theory,? ch. 12 in Case Study and Theory Development in the Social Sciences (MIT Press: Cambridge, MA), 2005: 263-285. -----. ?Process Tracing and Historical Explanation,? ch. 10 in Case Study and Theory Development in the Social Sciences (MIT Press: Cambridge, MA), 2005: 205-232. Goldstein, Judith and Robert O. Keohane, ?Ideas and Foreign Policy: An Analytical Framework,? in Ideas and Foreign Policy: Beliefs, Institutions, and Political Change (Ithaca, NY: Cornell University Press, 1993): 3-30. Granovetter, Mark S. ?The Strength of Weak Ties,? The American Journal of Sociology 78:6 (May 1973): 1360-1380. Haas, Peter M. ?Introduction: Epistemic Communities and International Policy Coordination,? in International Organization 46:1 (Winter 1992): 1-36. Heclo, Hugh. ?Issue Networks and the Executive Establishment,? ch. 3 in Anthony King, ed. The New American Political System (Washington: AEI for Public Policy Research, 1978), 1 st ed., pp. 87-124. Johnasson, Frans. The Medici Effect: Breakthrough Insights at the Intersection of Ideas, Concepts & Cultures (Boston, MA: Harvard Business School Press), 2004. Johnson, Rebecca. ?Advocates and Activists: Conflicting Approaches on Nonproliferation and the Test Ban Treaty,? ch. 3 in The Third Force: The Rise of Transnational Civil Society (Washington, DC: Carnegie Endowment for International Peace, 2000), ed. by Ann M. Florini, pp. 49-81. Keck, Margaret E. and Kathryn Sikkink. Activists Beyond Borders: Advocacy Networks in International Politics (Ithaca, NY: Cornell University Press, 1998). Keohane, Robert O. and Joseph S. Nye, Jr., eds. Transnational Relations and World Politics (Cambridge, MA: Harvard University Press, 1972). 302 Kingdon, John W. Agendas, Alternatives, and Public Policies (New York: Harper Collins Publishers, 1995), 2 nd edition. Knopf, Jeffrey W. ?A Framework for Assessing Activism?s Influence,? in Domestic Society and International Cooperation: The Impact of Protest on U.S. Arms Control Policy (Cambridge, UK: Cambridge University Press, 1998): 49-78. Kraft, Herman Joseph S. ?The Autonomy Dilemma of Track Two Diplomacy in Southeast Asia,? Security Dialogue 31:3 (September 2000): 343-356. Mendelson, Sarah E. Changing Course: Ideas, Politics and the Soviet Withdrawal from Afghanistan (Princeton, NJ: Princeton University Press, 1998). -----. ?Internal Battles and External Wars: Politics, Learning, and the Soviet Withdrawal from Afghanistan,? World Politics 45:3 (April 1993): 327-360. Montville, Joseph V. ?The Arrow and the Olive Branch: A Case for Track Two Diplomacy,? in The Psychodynamics of International Relationships: Unofficial Diplomacy at Work (Lexington, MA: Lexington Books, 1991), eds. Vamik D. Volkan, Joseph V. Montville, and Demetrios A. Julius: 161-175. Parmar, Inderjeet, ?American Foundations and the Development of International Knowledge Networks,? Global Networks 2:1 (January 2002): 13-30. Reinicke, Wolfgang H. ?The Other World Wide Web: Global Public Policy Networks,? Foreign Policy (Winter 1999-2000): 44-57. Risse-Kappen, Thomas, ed. ?Introduction,? in Bringing Transnational Relations Back In: Non-State Actors, Domestic Structures and International Institutions (Cambridge, UK: Cambridge University Press, 1995): 3-32. Rochefort, David A. and Roger W. Cobb, ?Problem Definition: An Emerging Perspective,? ch. 1 in The Politics of Problem Definition: Shaping the Policy Agenda (Lawrence, KS: University Press of Kansas, 1994), eds. by Rochefort and Cobb: 1-31. Ronfeldt, David and John Arquilla, ?What Next for Networks and Netwars?? ch. 10 in Networks and Netwars: The Future of Terror, Crime, and Militancy (Santa Monica, CA: RAND), 2001, ed. by John Arquilla and David Ronfeldt, pp. 311-361. Rutherford, Kenneth Robin. NGOs and the Ban on Antipersonnel Landmines. PhD Dissertation, Georgetown University, 2000. Shambaugh, David. ?Letter of Transmittal: US-China Track II Security Dialogue Survey,? to Thomas Graham, Rockefeller Foundation, April 22, 1998. 303 Stone, Diane. ?Introduction: Global Knowledge and Advocacy Networks,? Global Networks 2:1 (January 2002): 1-11. Struyk, Raymond J. ?Transnational Think-Tank Networks: Purpose, Membership, and Cohesion,? Global Networks 2:1 (January 2002): 83-90. Witte, Jan Martin, Wolfgang H. Reinicke, and Thorsten Benner, ?Beyond Multilateralism: Global Public Policy Networks,? International Politics and Society (Politik und Gesellschaft Online) 2/2000 (www.fes.de/IPG/ipg2_2000/artwitte.html (accessed April 16, 2004)). Think Tanks in the United States Abelson, Donald E. American Think Tanks and their Role in U.S. Foreign Policy (New York: St. Martin?s Press), 1996. -----. ?The Business of Ideas,? ch. 13 in Think Tank Traditions: Policy Research and the Politics of Ideas, ed. by Diane Stone and Andrew Denham (Manchester: Manchester University Press, 2004): 215-231. Do Think Tanks Matter? Assessing the Impact of Public Policy Institutes (Ithaca, NY: McGill-Queens University Press), 2002. -----. ?Policy Experts and Political Pundits: American Think Tanks and the News Media,? NIRA Review (Summer 1998). -----. ?Think Tanks and U.S. Foreign Policy: An Historical View,? U.S. Foreign Policy Agenda 7:3 (November 2002): 9-12. -----. ?Think Tanks in the United States,? ch. 5 in Diane Stone, Andrew Denham, and Mark Garnett, eds. Think Tanks Across Nations: A Comparative Approach (Manchester: Manchester University Press, 1998): 107-126. -----. ?Trends in Search of Policy Influence: The Strategies of American Think Tanks,? NIRA Review (Spring 1998). ----- and Christine M. Carberry, ?Policy Experts in Presidential Campaigns: A Model of Think Tank Recruitment,? Presidential Studies Quarterly 27:4 (Fall 1997): 679-697. ----- and Evert A. Lindquist, ?Think Tanks in North America,? ch. 2 in James G. McGann and R. Kent Weaver, editors. Think Tanks & Civil Societies: Catalysts for Ideas and Action (London: Transaction Publishers, 2002): 37-66. ?The Carousels of Power,? The Economist, May 25, 1991, pp. 23-??. 304 ?The Charge of the Think Tanks,? The Economist, February 13, 2003. Clark, Ann Marie. ?Non-Governmental Organizations and their Influence on International Society,? Journal of International Affairs 48:2 (Winter 1995): 507-525. Easterbrook, Gregg. ?Ideas Move Nations: How Conservative Think Tanks Have Helped to Transform the Terms of Political Debate,? The Atlantic (January 1986): 66-??. Gellner, Winand. ?The Politics of Policy: ?Political Think Tanks? and Their Markets in the U.S.-Institutional Environment,? Presidential Studies Quarterly 25:3 (Summer 1995): 497-510. George, Alexander L. Bridging the Gap: Theory and Practice in Foreign Policy (Washington, D.C.: U.S. Institute of Peace Press), 1993. ?The Good Think-Tank Guide,? The Economist, January 3, 1992, pp. 49-??. Goodwin, Craufurd D. ?The Fifth Estate: Research for Informed Debate in Democratic Society,? ch. 3 in Think Tanks in a Democratic Society: An Alternative Voice (Washington, D.C.: The Urban Institute, 1996), ed. by Jeffrey Telgarsky and Makiko Ueno, pp. 23-34. Haass, Richard N. ?Think Tanks and U.S. Foreign Policy: A Policy-maker?s Perspective,? U.S. Foreign Policy Agenda 7:3 (November 2002): 5-8. Heclo, Hugh. ?The In-and-Outer System: A Critical Assessment,? Political Science Quarterly 103:1 (Spring 1988): 37-56. Hicks, Sallie M., Theodore A. Couloumbis, and Eloise M. Forgette, ?Influencing the Prince: A Role for Academicians?? Polity (Winter 1982): 279-294. Higgott, Richard and Diane Stone. ?The Limits of Influence: Foreign Policy Think Tanks in Britain and the USA,? Review of International Studies 20 (1994): 15-34. Linden, Patricia. ?A Guide to America?s Think Tanks: Powerhouses of Policy,? Town & Country (January 1987): 99-106, 170, 174-175, 178-179. Matlack, Carol. ?Marketing Ideas,? The National Journal 23:25 (June 22, 1991): 1552-. McGann, James G. The Competition for Dollars, Scholars and Influence in the Public Policy Research Industry (Lanham, MD: University Press of America), 1995. -----. ?How Think Tanks Are Coping with the Future,? The Futurist 34:6 (November 2000): 16. 305 -----. Responding to 9/11: Are Think Tanks Thinking Outside the Box? (Philadelphia, PA: Foreign Policy Research Institute): July 31, 2003. -----. Scholars, Dollars, and Policy Advice (Philadelphia, PA: Foreign Policy Research Institute), August 2004. -----. Think Tank Networks and Transnational Security Threats (Philadelphia: Foreign Policy Research Institute), December 2003. -----. Think Tanks and Policy Advice in the U.S. (Philadelphia: Foreign Policy Research Institute), August 2005. -----. ?Think Tanks and the Transnationalization of Foreign Policy,? U.S. Foreign Policy Agenda 7:3 (November 2002): 13-18. -----. Think Tanks: Catalysts for Ideas in Action: An International Survey (Philadelphia: Foreign Policy Research Institute), October 1999. Reiss, Mitchell. ?The Role of Foundations in the Making of Foreign Policy,? Remarks to the German Marshall Fund, Washington, D.C., December 6, 2004. Available at http://www.state.gov/s/p/rem/39495.htm (accessed February 17, 2005). Ricci, David M. The Transformation of American Politics: The New Washington and the Rise of Think Tanks (New Haven: Yale University Press), 1993. Rich, Andrew. Think Tanks, Public Policy, and the Politics of Expertise. PhD Dissertation, Yale University, 1999. -----, Think Tanks, Public Policy, and the Politics of Expertise (Cambridge, England: Cambridge University Press), 2004. ----- and R. Kent Weaver. ?Advocates and Analysts: Think Tanks and the Politicization of Expertise,? ch. 11 in Allan J. Cigler and Burdett A. Loomis, eds. Interest Group Politics (Washington, D.C.: Congressional Quarterly), 5 th edition, 1998: 235-253. Roberts, Brad, Stanton H. Burnett, and Murray Weidenbaum, ?Think Tanks in a New World,? The Washington Quarterly 16:1 (Winter 1993): 169-??. Shulock, Nancy. ?The Paradox of Policy Analysis: If It Is Not Used, Why do We Produce So Much of It?? Journal of Policy Analysis and Management 18:2 (1999): 226-244. Smith, James Allen. The Power Brokers: Think Tanks and the Rise of the New Policy Elite (New York: Free Press, 1993). 306 Snow, Donald M. and Eugene Brown. ?Outside Influence I: Interest Groups and Think Tanks,? ch. 6 in Puzzle Palaces and Foggy Bottom: U.S. Foreign and Defense Policy- Making in the 1990s (New York: St. Martin?s Press, 1994): 179-205. Stanfield, Rochelle L. ?The Golden Rolodex,? The National Journal 22:10 (March 10, 1990): 552-. -----. ?The New Establishment,? The National Journal 21:3 (January 21, 1989): 132-. Stone, Diane. Capturing the Political Imagination: Think Tanks and the Policy Process (London: Frank Cass), 1996. ----- ?Private Authority, Scholarly Legitimacy, and Political Credibility: Think Tanks and Informal Diplomacy,? in Non-State Actors and Authority in the Global System, ed. by Richard A. Higgott, Geoffrey R.D. Underhill and Andreas Bieler (London: Routledge), 2000: 211-225. -----. ?Recycling Bins, Garbage Cans, or Think Tanks? Three Myths Regarding Policy Analysis Institutes,? unpublished manuscript. ----- ?Think Tanks Across Nations: The New Networks of Knowledge,? NIRA Review (Winter 2000): 34-39. ----- and Mark Garnett. ?Introduction: Think Tanks, Policy Advice and Governance,? in Diane Stone, Andrew Denham, and Mark Garnett, eds. Think Tanks Across Nations: A Comparative Approach (Manchester: Manchester University Press, 1998): 1-20. Wallace, William. ?Afterword: Soft Power, Global Agendas,? in Think Tank Traditions: Policy Research and the Politics of Ideas, ed. by Diane Stone and Andrew Denham (Manchester: Manchester University Press, 2004): 281-290. ?Between Two Worlds: Think-tanks and Foreign Policy,? ch. 8 in Christopher Hill and Pamela Beshoff, eds., Two Worlds of International Relations: Academics, Practitioners and the Trade in Ideas (London: Routledge, 1994): 139-163. -----. ?Conclusion: Ideas and Influence,? in Diane Stone, Andrew Denham, and Mark Garnett, eds. Think Tanks Across Nations: A Comparative Approach (Manchester: Manchester University Press, 1998): 223-230. Weaver, R. Kent. ?The Changing World of Think Tanks,? PS: Political Science and Politics 22:3 (September 1989): 563-578. ----- and James G. McGann. ?Think Tanks and Civil Societies in a Time of Change,? ch. 1 in James G. McGann and R. Kent Weaver, editors. Think Tanks & Civil Societies: Catalysts for Ideas and Action (London: Transaction Publishers, 2002): 1-36. 307 Weiss, Carol H. ?Congressional Committees as Users of Analysis,? Journal of Policy Analysis and Management 8:3 (1989): 411-431. -----. ?Helping Government Think: Functions and Consequences of Policy Analysis Organizations,? ch. 1 in Organizations for Policy Analysis: Helping Government Think (London: Sage Publications, 1992), ed. by Weiss: 1-20. Whiteman, David. ?The External Context of Communication: Issue Networks,? ch. 3 in Communication in Congress: Members, Staff, and the Search for Information (Lawrence, KS: University Press of Kansas, 1995): 39-50. Wiarda, Howard J. ?New Actors: Think Tanks and Foreign Policy,? ch. 8 in American Foreign Policy: Actors and Processes (New York: HarperCollins, 1996): 137-156. Yee, Albert S. ?The Causal Effects of Ideas on Policies,? International Organization 50:1 (Winter 1996): 69-108. CSCAP Partial Bibliography ?Asia?s Informal Diplomacy,? Harvard International Review 23:1 (Spring 2001): 38-41. Ball, Desmond. The Council for Security Cooperation in the Asia Pacific (CSCAP): Its Record and Prospects, Canberra Papers on Strategy and Defense 139 (Canberra, Australian National University), 2000. -----, ?CSCAP: Its Future Place in the Regional Security Architecture,? in Managing Peace and Security in the Asia-Pacific (Kuala Lumpur: Institute for International and Strategic Studies, 1996), ed. by Bunn Nagara and Cheah Siew Ean, pp. 289-325. Chalk, Peter. ?Research Note: CSCAP Study Group on Transnational Crime,? Studies in Conflict and Terrorism, vol. 21, 1998, pp. 229-231. Cossa, Ralph A., ed. Asia Pacific Confidence and Security Building Measures, Significant Issues Series 17:3 (Washington D.C.: Center for Strategic and International Studies), 1995. -----, ed. Asia Pacific Multilateral Nuclear Safety and Non-Proliferation: Exploring the Possibilities (Honolulu, HI: Pacific Forum Occasional Papers), December 1996. -----. PACATOM: Building Confidence and Enhancing Nuclear Transparency (Honolulu, HI: Pacific Forum Occasional Papers), October 1998. -----. ?Proliferation and Regional Security,? ch. 7 in Keeping Proliferation at Bay (Jakarta: Center for Strategic and International Studies, 1998), ed. by Ramesh Thakur: 109-120. 308 -----. ?Track Two Diplomacy: Promoting Regional Peace, Stability,? U.S. Foreign Policy Agenda 3:1 (January 1998), http://usinfo.state.gov/journals/itps/0198/ijpe/pj18coss.htm (accessed May 16, 2002). CSCAP International Working Group on Confidence and Security Building Measures. Countering the Spread of Weapons of Mass Destruction: the Role of the Proliferation Security Initiative, Issues and Insights 4:5 (Honolulu: Pacific Forum/CSIS): July 2004. CSCAP Working Group Special Report, Asia Pacific Multilateral Nuclear Safety and Non-Proliferation: Exploring the Possibilities (Pacific Forum Occasional Papers, December 1996). Evans, Paul M. ?Assessing the ARF and CSCAP,? ch. 10 in The Security Environment in the Asia-Pacific, ed. by Hung-mao Tien and Tun-jen Cheng (London: M.E. Sharpe, 2000): 154-172. -----. ?Building Security: The Council for Security Cooperation in the Asia Pacific (CSCAP),? The Pacfic Review 7:2 (Summer 1994): 125-139. Hamada, Kazuko. CSCAP?s NEEG: Exploring Nuclear Energy Transparency as a Regional Confidence and Security Building Measure, Issues and Insights 3:7 (Honolulu: Pacific Forum/CSIS): September 2003. Johnston, Alastair Iain. ?Socialization in International Institutions: the ASEAN Way and International Relations Theory,? ch. 3 in International Relations Theory and the Asia- Pacific, ed. by G. John Ikenberry and Michael Mastanduno (New York: Columbia University Press, 2003): 107-162. Roberts, Brad. ?Evaluating Global Non-Proliferation Instruments,? ch. 3 in Keeping Proliferation at Bay (Jakarta: Center for Strategic and International Studies, 1998), ed. by Ramesh Thakur: 60-74. Simon, Sheldon W. ?The ASEAN Regional Forum Views the Councils for Security Cooperation in the Asia Pacific: How Track II Assists Track I,? Managing Security Challenges in Southeast Asia, National Bureau of Asian Research Analysis 13:4 (July 2002): 5-23. -----. ?Evaluating Track II Approaches to Security Diplomacy in the Asia-Pacific: the CSCAP Experience,? The Pacific Review 15:2 (2002): 167-200. Wanandi, Jusuf, ?The Future of the ARF and CSCAP in the Regional Security Architecture,? in Managing Peace and Security in the Asia-Pacific (Kuala Lumpur: Institute for International and Strategic Studies, 1996), ed. by Bunn Nagara and Cheah Siew Ean, pp. 281-288. 309 Woods, Lawrence T. ?Learning from NGO Proponents of Asia-Pacific Regionalism: Success and its Lessons,? Asian Survey 35:9 (September 1995): 812-827. 310 LIST OF INTERVIEW SUBJECTS In addition to four subjects who requested or were required to have complete anonymity to be interviewed, the following list of subjects (in alphabetical order by last name) were interviewed for this study, identified by name and most relevant institution, either where they were employed when the interview was conducted or where they used to work. Name of Subject Deborah Yarsike Ball David Brown Kent Calder Kurt Campbell Ralph Cossa Scott Davis Joseph DeTrani Lewis Dunn Robert Einhorn Matthew Evangelista Ed Fei Evan Feigenbaum Bonnie Glaser Brad Glosserman James Goldgeier Stuart Goldman James Goodby Stephen Haggard John Harvey Fiona Hill Steedman Hinckley David Hitchcock Thomas Hubbard Duc Hyunh Frank Januzzi Beth Jones Stuart Kaufman Jessica Kehl James A. Kelly Don Keyser David Kramer Mark Kramer Wayne Limberg Robert Manning Tanya Mazin Michael McDevitt Most Relevant Institution (at interview or former) Lawrence Livermore National Laboratory School of Advanced International Studies School of Advanced International Studies Former Deputy Assistant Secretary of Defense Pacific Forum/CSIS U.S. Department of Energy U.S. Department of State (at interview) Science Applications International Corporation Former Assistant Secretary of State Cornell University U.S. Department of Energy U.S. Department of State, Policy Planning staff Center for Strategic and International Studies Pacific Forum/CSIS George Washington University; former Dept. of State Congressional Research Service U.S. Department of State (retired) University of California at San Diego U.S. Department of Energy; former DASD Brookings Institution U.S. government (various) U.S. Department of State (retired) Former Deputy Assistant Secretary of State University of California at San Diego graduate U.S. Congressional staff Former Assistant Secretary of State Delaware University; former NSC staff U.S. Department of Defense Assistant Secretary of State (at interview) Former Principal Deputy Assistant Secretary of State Deputy Assistant Secretary of State Harvard University U.S. Department of State U.S. Department of State, Policy planning staff U.S. Congressional staff (former) Center for Naval Analyses 311 Michael McFaul Sarah Mendelson Jackie Miller Mark Mohr John Olsen Vladimir Orlov Matthew Ouimet Torkel Patterson T. J. Pempel Steven Pifer Pavel Podvig Brad Roberts Alan Romberg Robert Scalapino Randy Schriver Kathryn Schultz Stephen Sestanovich Susan Shirk Pamela Slutz Scott Snyder David Speedie James Steinberg Steven E. Steiner Angela Stent William Sullivan Celeste A. Wallander Stanley Weeks Stanford University and Hoover Institution Center for Strategic and International Studies Center for Strategic and International Studies (former) U.S. Department of Energy (retired) Sandia National Laboratories PIR Center U.S. Department of State Deputy Assistant Secretary of State; former NSC University of California at Berkeley Former Deputy Assistant Secretary of State Moscow Institute of Physics and Technology Institute for Defense Analyses Stimson Center University of California at Berkeley Deputy Assistant Secretary of State (at interview) U.S. Department of State Former Special Advisor to the Secretary of State University of California at San Diego; former DAS U.S. Department of State Asia Foundation Carnegie Corporation of New York Former Dept of State and deputy national security advisor U.S. Department of State Georgetown University and former U.S. govt (various) U.S. Department of Defense Center for Strategic and International Studies Science Applications International Corporation