University of Southern California Dissertations and Theses

Things fall apart: the unraveling of international institutions through withdrawal

(USC Thesis Other)

Things fall apart: the unraveling of international institutions through withdrawal

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THINGS FALL APART: THE UNRAVELING OF INTERNATIONAL INSTITUTIONS THROUGH WITHDRAWAL by Taylor Reeves Dalton A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulllment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (POLITICAL SCIENCE AND INTERNATIONAL RELATIONS) August 2022 Copyright 2022 Taylor Reeves Dalton ii Declaration of Authorship I, Taylor Reeves Dalton, declare that this thesis titled, `Things Fall Apart: The Unraveling of International Institutions through Withdrawal' and the work presented in it is my own. I conrm that this work submitted for assessment is my own and is expressed in my own words. Any uses made within it of the works of other authors in any form (e.g., ideas, equations, gures, text, tables, programs) are properly acknowledged at any point of their use. A list of the references employed is included. Signed: Date: iii \Turning and turning in the widening gyre The falcon cannot hear the falconer; Things fall apart; the center cannot hold; Mere anarchy is loosed upon the world." William Butler Yeats, The Second Coming \He has put a knife on the things that held us together and we have fallen apart." Chinua Achebe, Things Fall Apart iv I dedicate this work to my parents, Rick and Tanya, for allowing me to be myself and encouraging me unconditionally, and my sons, Emmett and Titus, who inspire me and ll me with hope. v Acknowledgements \The reward of a thing well done is to have done it." | Ralph Waldo Emerson The sprawling nature of navigating a doctorate program and writing a dis- sertation is a community endeavor. I am thankful for the support, en- couragement, and guidance from so many people. Key among them is my advisor, Wayne Sandholtz, and my committee members (in alphabetical or- der): Andrew Coe, Benjamin Graham, Patrick James, and Abby Wood. All of them have been tremendously generous with their time and advice over the years. Despite all of their other commitments, professional goals, and personal lives, they each made time to review manuscripts or funding appli- cations; write letters of recommendation; talk at length about research, the profession, and life; make connections; and invite me into the profession as a junior scholar. Beyond my formal committee, other faculty members also contributed invaluable lessons about research and professional development, including: James Lo, Morris Levy, Pablo Barbera, Barbara Koremenos, Stephen Chaudoin, Ange-Marie Hancock Alfaro, Brian Rathbun, and Jens David Ohlin. The research in this dissertation was funded by the generous support of the Michael R. Young Memorial Fund Fellowship; the USC Graduate School Russell Endowed Fellowship (Competitive), including the Russell Endowed Fellowship Travel/Research Award; USC School of International Relations and Political Science and International Relations (POIR) funding; and mul- tiple grants from the USC Center for International Studies (CIS). CIS has been instrumental to this research project. Because of the generous funding from CIS, I was about to fund terric work from two undergraduate research assistants: Jamie Schlegel and Kaili Ganigan. Both contributed to vi the empirical research. Matt Stevens, Chandra Caldwell, and Cort Brinker- ho at CIS also provided invaluable support and friendship over the years. Additionally, completing the dissertation was only possible with the support of the POIR sta, including Danielle De Rosa Ballard, Aurora Ramierz, Cathy Ballard, Linda Cole, and Veridiana Chavarin. They all helped me navigate the process and were patient with me when I popped in with a question and to chat. Veri was especially key, particularly when my mother passed away and during the birth of my sons. So many friends and colleagues helped me work on this project. Although I risk missing someone, I want to acknowledge Suzie Caldwell Mulesky, Therese Anders, Kelebogile Zvobgo, Stephanie Kang, Xinru Ma, Ellen Kim, Miriam Barnum, Kyle Rapp, Anne van Wijk, Adam Feldman, James Clegg, Joey Huddleston, Simon Radford, Beau Squires, and Jesse Matlock for the conversations, advice, and friendship over the years. Beyond friends, I re- ceived insightful comments on drafts and papers from discussants and par- ticipants in numerous conferences and workshops. Finally, my family and friends supported my crazy idea to move away from practicing law to jumping back into becoming a student. Professionally, Howard Hall, Jody Browning, and Ira Rivin supported my goals. Personally, my parents, Rick and Tanya, always encouraged me to pursue my interests and do so to the best of my ability. Although my mother will not be able to see the completion of this work, I hope I made her proud. My wife, Sarah, has been a gracious partner and dealt with more uncertainty than deserved. Thank you. Table of Contents vii Table of Contents Declaration of Authorship ii Epigraph iii Dedication iv Acknowledgements v List of Figures xii List of Tables xiv Abstract xv 1 Introduction 1 1.1 Research Question: Why Does Treaty Exit Occur? . . . . . . 4 1.2 International Law's Aect on State Behavior . . . . . . . . . 10 1.3 Jurisprudence of Treaty Withdrawal . . . . . . . . . . . . . . 16 1.4 Existing Theories . . . . . . . . . . . . . . . . . . . . . . . . . 20 1.4.1 Are Entry and Exit Simply Two Sides of the Same Coin? . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 1.4.2 Why Do States Exit? What We Know . . . . . . . . . 22 1.5 A Bargaining Model of Treaty Exit . . . . . . . . . . . . . . . 29 1.6 Roadmap of Dissertation . . . . . . . . . . . . . . . . . . . . . 34 1.7 Limitations and Lingering Questions . . . . . . . . . . . . . . 35 2 Bargaining, Bending, and Breaking: A Theory of Treaty Exit 39 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Table of Contents viii 2.2 Types of Treaty Exit . . . . . . . . . . . . . . . . . . . . . . . 41 2.3 Building Blocks of a Theory: Flexibility and Obligation . . . 46 2.3.1 Flexibility . . . . . . . . . . . . . . . . . . . . . . . . . 46 2.3.2 What is obligation? . . . . . . . . . . . . . . . . . . . 48 2.4 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 2.4.1 Theoretical Model: . . . . . . . . . . . . . . . . . . . . 58 2.4.2 Analysis of Game . . . . . . . . . . . . . . . . . . . . . 62 2.5 Alternative Explanations . . . . . . . . . . . . . . . . . . . . . 66 2.6 Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 2.6.1 International Convention for the Regulation of Whaling 72 2.6.2 Cascading Episodes of Exit from UNIDO . . . . . . . 82 2.6.3 North Korea from NPT . . . . . . . . . . . . . . . . . 86 2.6.4 Universal Postal Union . . . . . . . . . . . . . . . . . 90 2.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 3 Broad Analysis: Treaty Exit Over the Last 80 Years 95 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 3.2 Data Collection from UNTS . . . . . . . . . . . . . . . . . . . 97 3.3 General Descriptive Statistics . . . . . . . . . . . . . . . . . . 99 3.4 Data Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 3.5 Measuring Flexibility and Obligation . . . . . . . . . . . . . . 111 3.5.1 Collecting Actions . . . . . . . . . . . . . . . . . . . . 111 3.5.2 Collecting Treaty Texts . . . . . . . . . . . . . . . . . 114 3.5.3 Building Measure of Obligation and Flexibility from Treaty Text . . . . . . . . . . . . . . . . . . . . . . . . 115 3.5.4 Other Treaty Variables . . . . . . . . . . . . . . . . . . 118 3.5.5 Summary of Variables and Expectations . . . . . . . . 120 3.6 Analysis and Results . . . . . . . . . . . . . . . . . . . . . . . 124 3.7 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 3.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 4 ICC: Institutional Flexibility and Relative Importance 134 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 4.2 Background: Legacy, Creation, and Early Cases . . . . . . . . 138 4.2.1 Brief History of International Criminal Tribunals . . . 138 4.2.2 The Rome Statute . . . . . . . . . . . . . . . . . . . . 145 4.2.3 Early Cases . . . . . . . . . . . . . . . . . . . . . . . . 151 Table of Contents ix 4.3 Legitimacy, Eectiveness, and Growing Critiques . . . . . . . 154 4.4 Bargains and Exits: Reforming the Institution for the Right Member . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 4.4.1 Kenya . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 4.4.2 South Africa . . . . . . . . . . . . . . . . . . . . . . . 171 4.4.3 Burundi and The Philippines . . . . . . . . . . . . . . 175 4.5 Assessing the Theory . . . . . . . . . . . . . . . . . . . . . . . 179 4.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 5 Arms Control: Breach or Exit 185 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 5.2 Summary of Arms Control between the Superpowers . . . . . 188 5.2.1 ABM Treaty . . . . . . . . . . . . . . . . . . . . . . . 194 5.2.2 INF Treaty . . . . . . . . . . . . . . . . . . . . . . . . 196 5.3 Stresses on Arms Control Regime . . . . . . . . . . . . . . . . 198 5.3.1 Star Wars and the ABM Treaty Reinterpretation De- bate . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 5.3.2 Extending Arms Control after the USSR Collapse . . 202 5.3.3 Changing Security Concerns in the 21st Century . . . 208 5.4 Cooperation and Withdrawal of Superpowers . . . . . . . . . 218 5.4.1 US Approach to Disengagement . . . . . . . . . . . . . 218 5.4.2 Russian Approach to Disengagement . . . . . . . . . . 229 5.5 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 5.5.1 ABM Treaty . . . . . . . . . . . . . . . . . . . . . . . 237 5.5.2 INF Treaty . . . . . . . . . . . . . . . . . . . . . . . . 244 5.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 6 Brexit: Bargaining Failure 250 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 6.2 Long Road to Brexit: A History . . . . . . . . . . . . . . . . 254 6.2.1 UK-EU Relationship . . . . . . . . . . . . . . . . . . . 254 6.2.2 2008 Recession and Eurozone Crisis . . . . . . . . . . 258 6.2.3 Promised Negotiations on Reform . . . . . . . . . . . 262 6.2.4 Campaign . . . . . . . . . . . . . . . . . . . . . . . . . 266 6.2.5 The Vote . . . . . . . . . . . . . . . . . . . . . . . . . 271 6.2.6 Aftermath . . . . . . . . . . . . . . . . . . . . . . . . . 272 6.3 Application of Theory . . . . . . . . . . . . . . . . . . . . . . 274 Table of Contents x 6.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288 7 Conclusion 291 7.1 Contribution, Implications, and Remaining Questions . . . . 291 7.2 Where do we go from here? . . . . . . . . . . . . . . . . . . . 297 Bibliography 302 Appendices 343 A Formal Model Analysis and Proofs 343 A.1 Analyzing the Bargaining Model of Cooperation, Breach, and Withdrawal . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343 A.1.1 Making a Deal or Exit . . . . . . . . . . . . . . . . . . 351 A.2 Analyzing Punish Breach Game . . . . . . . . . . . . . . . . . 353 A.2.1 Basic Model Setup . . . . . . . . . . . . . . . . . . . . 354 A.3 Model Version 1: Uncertainty with Costless Punishment . . . 356 A.4 Model Version 2: Uncertainty with Costly Punishment . . . . 359 A.5 Model Version 3: Removing Uncertainty from Costly Punish- ment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360 A.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361 B Case Vignettes 370 B.1 Treaty of Wuchale, Ethiopia (1893) . . . . . . . . . . . . . . . 370 B.2 Unraveling of the League of Nations (1925{1942) . . . . . . . 373 B.3 American Convention on Human Rights (ACHR), Trinidad & Tobago (1998) and Venezuela (2012) . . . . . . . . . . . . . 375 B.4 Collective Security Treaty Organization (CSTO), Uzbekistan (1999, 2012) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378 C Data Appendix 381 C.1 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . 381 C.1.1 Collecting Data from UN Treaty Series . . . . . . . . 381 C.1.2 Filtering Actions . . . . . . . . . . . . . . . . . . . . . 382 C.2 Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 C.3 Text Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 C.4 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 Table of Contents xi C.5 Case Study Treaties . . . . . . . . . . . . . . . . . . . . . . . 388 D RA Instructions for Coding Treaty Obligation 392 D.1 Summary of Approach . . . . . . . . . . . . . . . . . . . . . . 392 D.2 Coding Instrument and Instructions . . . . . . . . . . . . . . 394 List of Figures xii List of Figures 2.1 Types of Treaty Exit . . . . . . . . . . . . . . . . . . . . . . . 44 2.2 Intuitions About Flexibility and Obligation . . . . . . . . . . 55 2.3 Game Tree and Payos . . . . . . . . . . . . . . . . . . . . . . 59 3.1 Treaties Concluded and Registered since 1851 . . . . . . . . . 100 3.2 Treaties Concluded and Registered since 1945 . . . . . . . . . 101 3.3 All Exits Over Time (Annual) . . . . . . . . . . . . . . . . . . 102 3.4 Exits Percent by Treaty Type . . . . . . . . . . . . . . . . . . 104 3.5 Exits [Restricted] Over Time (Annual) . . . . . . . . . . . . . 106 3.6 Exits per State . . . . . . . . . . . . . . . . . . . . . . . . . . 107 3.7 Exits per Treaty . . . . . . . . . . . . . . . . . . . . . . . . . 108 3.8 Top 20 Exiting States . . . . . . . . . . . . . . . . . . . . . . 109 3.9 Average Permissive-Obligatory Score of Treaties Overtime . . 117 3.10 Marginal Eects of Permissive-Obligatory Score on Predicted Treaty Exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 4.1 Results of Major Int'l Criminal Tribunals . . . . . . . . . . . 144 4.2 The ICC at The Hague . . . . . . . . . . . . . . . . . . . . . 146 4.3 Number of States Ratifying or Acceding Over Time . . . . . 149 4.4 Length of Documents Establishing Int'l Criminal Tribunals . 151 5.1 Treaty Exits by the US . . . . . . . . . . . . . . . . . . . . . 225 5.2 Congressional Actions re Treaty Exit . . . . . . . . . . . . . . 228 5.3 Treaty Exits by the Russia (Soviet Union) . . . . . . . . . . . 233 6.1 Uncertain Future for the U.K. . . . . . . . . . . . . . . . . . . 272 6.2 Net EU Contributions . . . . . . . . . . . . . . . . . . . . . . 277 6.3 Complaints Against the U.K. re EU Law . . . . . . . . . . . . 281 6.4 Infringement Cases Against the U.K. re EU Law . . . . . . . 282 List of Figures xiii 6.5 Continued Opposition to Brexit . . . . . . . . . . . . . . . . . 287 A.1 Cooperation, Breach, Punish Game . . . . . . . . . . . . . . . 354 A.2 Uncertainty re Costless Punishment . . . . . . . . . . . . . . 357 A.3 Uncertainty with Costly Punishment . . . . . . . . . . . . . . 359 A.4 Removing Uncertainty from Costly Punishment . . . . . . . . 361 A.5 Equilibria Diagram . . . . . . . . . . . . . . . . . . . . . . . . 366 C.1 Distribution of Treaty Conclusion Dates in Sample of No-Exit Treaties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384 C.2 Distribution of Permissive-Obligatory Word Scores . . . . . . 385 C.3 Correlation Plot of Cooperation Problems and Permissive- Obligatory Word Scores . . . . . . . . . . . . . . . . . . . . . 387 List of Tables xiv List of Tables 2.1 Summary of Case Studies . . . . . . . . . . . . . . . . . . . . 70 3.1 Select Treaty-Level Variables . . . . . . . . . . . . . . . . . . 120 3.2 Measuring Independent Variables . . . . . . . . . . . . . . . . 122 3.3 Eect of Treaty Attributes on Treaty Exit . . . . . . . . . . . 125 3.4 Eect of Treaty Attributes on Number of Treaty Exits . . . . 127 4.1 Situations under Investigation at the ICC . . . . . . . . . . . 152 5.1 Signicant Arms Control Agreements between US and Russia (USSR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 C.1 Types of Treaties in No Exit Sample . . . . . . . . . . . . . . 384 C.2 Dictionary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386 C.3 Treaty Attributes Aect on Number of Treaty Exits . . . . . 389 C.4 Eect of Treaty Attributes on Number of Treaty Exits . . . . 390 C.5 Case Study Treaties . . . . . . . . . . . . . . . . . . . . . . . 391 xv Abstract Britain's decision to withdraw from the E.U., or \Brexit;" members of the African Union calling for mass withdrawal from the International Crimi- nal Court (ICC); and the United States' announced withdrawal from the Paris Climate Accord, are all examples of an increasing number of inter- national institutions seemingly under threat by members exercising their power to formally withdraw. Why do states choose to withdraw rather than bargain for accommodation or simply breach to meet their needs? This dissertation develops and presents a bargaining model of treaty exit that incorporates treaty features of obligation and exibility to understand how the distributional shift in treaty costs and the eect of treaty precision and enforcement shape a member state's preference to withdraw overreaching a bargain, maintaining the status quo, or breaching. Since withdrawing is costly to international cooperation|foreclosing a forum within which to address interstate problems|it is puzzling why a state would choose to withdraw over other options. The theory is empirically tested using quan- titative, large-N statistical analyses and qualitative, small-N case studies in a multi-method research design. Quantitatively, the texts and meta-data of hundreds of treaties was collected and used test aspects of the theory regard- ing exibility and obligation using automated text analysis. In addition, the high-prole withdraws from the International Criminal Court (ICC); arms control agreements between the United States and Russia; and Brexit are examined in detail to uncover the mechanisms working in the states' de- cisions to exit. The dissertation concludes with a summary of the theory and results, implications of the ndings; and re ections on the prospects for future cooperation in an increasingly competitive world. Keywords: treaties; international cooperation; international law; formal theory; text analysis 1 Chapter 1 Introduction \Bargaining has neither friends nor relations." | Benjamin Franklin 1 \Con ating these two types of conduct|the failure to subscribe to international agreements, and the decision to withdraw from international agreements|with a disregard for international law, as critics often do, is confused. No one doubts that states enjoy the capacity to decide for themselves which international agree- ments are in their national interest, and failing to endorse certain agreements due to national interests has no bearing on a state's reputation for abiding by international law." | John B. Bellinger, III, U.S. Dept. of State Legal Adviser 2 1 Poor Richard's Almanack, 1736. 2 Remarks to Atlantic Council Workshop entitled \Transatlantic Approaches to the International Legal Regime in an Age of Globalization and Terrorism? on Nov. 1, 2005, in Digest of U.S. Practice in International Law 2005, Oce of the Legal Adviser, U.S. States Dept. of State (Oxford UP, ed. Sally J. Cummins), pp. 215{219, available at https://2009-2017.state.gov/documents/organization/138677.pdf 2 On a sunny summer day in the White House Rose Garden, Vice Presi- dent Mike Pence introduced President Donald Trump to sustained applause. Wearing a red tie in his trademarked fashion, President Trump took to the podium to describe his Administration's eorts to fulll its \America First" policies both domestically and internationally. As he nished describing his accomplishments, the President continued, being interrupted mid-way through by enthusiastic applause from the audience: On these issues and so many more, we're following through on our commitments. And I don't want anything to get in our way. I am ghting every day for the great people of this country. Therefore, in order to fulll my solemn duty to protect America and its citizens, the United States will withdraw from the Paris Climate Accord|(applause)|thank you, thank you|but begin negotiations to reenter either the Paris Accord or a really entirely new transaction on terms that are fair to the United States, its businesses, its workers, its people, its taxpayers. So we're getting out. But we will start to negotiate, and we will see if we can make a deal that's fair. And if we can, that's great. And if we can't, that's ne. (Applause.) 3 As his sunlit, blond hair bounced in the breeze, President Trump went on to describe the unacceptable burdens the treaty placed on the U.S., describing the costs of \[c]ompliance with the terms of the Paris Accord and the onerous energy restrictions" 4 citing to loss of economic activity, production, and jobs. He also pointed to actions of others members of the Accord, like China and India, who he felt could unfairly pollute to the disadvantage 3 \Statement by President Trump on the Paris Climate Accord." The White House. (June 1, 2017), available at https://trumpwhitehouse.archives.gov/briengs- statements/statement-president-trump-paris-climate-accord/. 4 Id. 3 of the U.S. Channeling his nationalist view of international cooperation, he memorably stated: \I was elected to represent the citizens of Pittsburgh, not Paris." 5 This would be one of many announcements by the President and his Administration that the U.S. would be exiting an international agreement causing excitement for his supporters and consternation for his opponents and U.S. partners abroad. The United Kingdom's decision to withdraw from the E.U., better known as \Brexit;" members of the African Union calling for mass withdrawal from the International Criminal Court (ICC); and the United States' withdrawal from the Paris Climate Accord described above, 6 are all recent examples of an increasing number of institutions under threat by members formally withdrawing. Treaty withdrawal has even made its way to becoming part of partisan politics in America. 7 Treaty withdrawals include a number of other pending withdrawals from treaties and states with less media coverage than 5 Id. 6 Article 28 of the Accord allows states to formally withdraw from the agreement three years after it has been in force, withdrawal being eective one year after formal notice. The Accord went into force on November 4, 2016. Therefore, despite President Trump announcing the U.S.'s withdrawal from the agreement on June 1, 2017, formal notice was not given until November 4, 2019 and withdrawal was not eective until November 4, 2020|a day after the next presidential election. Upon taking oce as the forty-sixth president, Joe Biden signed an executive order reversing the Trump policy and accepting the agreement. The U.S. was fully re-admitted on February 19, 2021 after a 30-day waiting period (See Paris Climate Accord, art. 21.3). 7 The Republican Party adopted a platform that called for restrictions on the U.S. President's power to enter into executive agreements and other instruments without rat- ication by the U.S. Senate. \The United States will withdraw from all agreements and arrangements failing those standards." \Republican Party Platforms, 2016 Repub- lican Party Platform Online by Gerhard Peters and John T. Woolley, The American Presidency Project https://www.presidency.ucsb.edu/node/318311. In 2020, the Repub- lican Party essentially readopted this platform. Reid J. Epstein. The G.O.P. Delivers Its 2020 Platform. It's From 2016." The New York Times (Aug. 25, 2020) available at https://www.nytimes.com/2020/08/25/us/politics/republicans-platform.html . The Democratic Party's platform repudiates this position and calls for expanding arms 4 Donald Trump's withdrawals. For example, media coverage of Guatemala's recent announcement to withdraw from the International Coee Agreement 8 has been subdued|as is the case with most instances of treaty withdrawal. Lesser known instances of withdrawal, like Mauritania's 2006 withdrawal from the Agreement for the Establishment of a Commission for Controlling the Desert Locust in North-West Africa or Canada's 2011 withdrawal from the World Tourism Organization, received far less attention than many oth- ers. Based on media coverage, treaty withdrawal appears to be happening more frequently and a preferred action for states to meet their needs at the expense of international cooperation. Yet, it is not immediately clear that it is happening more frequently or for new or distinct reasons. 1.1 Research Question: Why Does Treaty Exit Occur? Why do countries like the U.S. decide to end a treaty relationship? More- over, why does this seem to occur so rare historically even when exit pro- visions or the ability to withdraw are ubiquitous in most treaties and insti- tutions? Given that a well-designed agreement will make continued coop- eration equilibrium-path behavior (Koremenos 2016), why do states choose to withdraw rather than bargain for an accommodation or simply breach control eorts. Democratic Party Platforms, 2020 Democratic Party Platform On- line by Gerhard Peters and John T. Woolley, The American Presidency Project https://www.presidency.ucsb.edu/node/342177. 8 Nick Brown. \Guatemala to Exit the International Coee Agreement." Daily Coee News. (July 7, 2020). 5 to meet their needs? In brief, this project presents a theory of institutional bargaining to answer these questions and to understand why members to an international agreement or institution 9 decide, after generally extensive cooperation, to exercise their option to withdraw their membership in the institution. In this project, I present a bargaining model of treaty withdrawal to inves- tigate the distributional con ict that may arise between one state and the other members of a treaty given an exogenous shock that shifts the weight of obligations or benets of the treaty. Additionally, a treaty's precision and enforcement mechanisms regulate the value of breaching, which in uences the preference for bargaining, withdrawing or maintaining the status quo. Being that withdrawing is costly to international cooperation|foreclosing a forum within which to address interstate problems|it is puzzling why a state would ever choose to withdraw. At rst blush, a bargaining model seems applicable to a number of historical and recent cases of withdrawal. After bargaining over an interpretation of the Treaty of Wuchale failed, King Menelik II of Shewa (and later Emperor of Ethiopia) denounced the treaty in 1893. 10 This lead the Kingdom of Italy to attempt to forcibly impose a protectorate over Ethiopia in the First Italo- Ethiopian War. \We are not coming back" were the words of the Yosuke 9 I often use the two terms interchangeably, even though they may not mean the same thing, broadly speaking. An institution is a set of rules, which is also the case for an agreement. However, this project primarily investigates formal agreements, generally manifested in treaties which, for the most part, contain explicit provisions for withdrawal. 10 \Treaty of Wichale." Encyclopaedia Britannica (April 25, 2020) available at https://www.britannica.com/event/Treaty-of-Wichale 6 Matsuoka, the head of the Japanese delegation to the League of Nations, said before walking out of the League's assembly hall. 11 Japan's dramatic withdrawal from the League|one of the most dramatic withdrawals of the twentieth century|was preceded by eorts by Japan and members of the League to address Japanese military expansion into China. In the end, Japan's demands were unacceptable to the League and, thus, no bargain was possible. More recently, the United Kingdom's decades long discomfort within the European Union gave way to multiple rounds of negotiations between it and the other member countries. Faced with a rising new wave of anti-EU sentiment, exacerbated by the global nancial crises of the early part of the 21st century, 12 David Cameron gambled that he could bargain his way out of the problem|as the Thatcher government had before him|only to realize that what his constituents wanted was far beyond what Europe was willing to give. 13 Even President Trump's decisions to withdraw from international agreements were often framed in terms of a negotiation with other treaty members. 14 In each of these three withdrawals from three dierent centuries, the withdrawal was preceded by eorts to bargain to avoid the exit. 11 Stewart Brown. \Japan stuns world, withdraws from league" UPI (Feb. 24, 1933) available at https://www.upi.com/Archives/1933/02/24/Japan-stuns-world-withdraws- from-league/2231840119817/. 12 Tooze (2018). 13 In his memoir, Cameron explained: \It came in two parts: a renegotiation and a referendum. Both were necessary, and both were long overdue. . . . I didn't see them as separate endeavours; they were two sides of the same coin." (Cameron 2019). 14 U.N. Ambassador Nikki Haley's notice to the U.N. Secretary-General noted that the U.S. intended to exercise its \right" to withdraw \[u]nless the United States identies suitable terms for reengagement" (U.S.A.: Commu- nication. C.N.464.2017.TREATIES-XXVII.7.d (Aug. 4, 2017) available at https://treaties.un.org/doc/Publication/CN/2017/CN.464.2017-Eng.pdf.) 7 To be clear, this project is aimed at examining the phenomenon of states exercising the option to withdraw, 15 not simply choosing to breach the agree- ment and renege on their promise. Breaching a treaty is a defection from a state's obligations, and those disfavored defections are disincentivized by various costs, as discussed below. Even after breach, a state remains a party to the institution and can come back into compliance. Arguably, withdrawal from an institution is almost always permissible under international law, given certain criteria, even if the specic treaty does not include a provi- sion for termination, denunciation, or withdrawal (Helfer 2012, 636{641). 16 Thus, with the option to simply breach, it is a puzzle as to why states ever withdraw from membership in an institution|forfeiting the opportunity of continued cooperation in the future. Also, it is generally assumed that withdrawal is rare, being characterized as o-equilibrium-path behavior. This may be one reason for the dearth of theory and data on the subject. Yet, the relative rarity of withdrawal does not negate its importance. Withdrawal is a signicant blow to international cooperation and may endanger the very existence of an international insti- tution. Similarly, studying war has been the focus of much of the research 15 As Helfer (2012, 635{636, fn.6) explains, the terms \denunciation" and \withdrawal" are used interchangeably; however, withdrawal is likely the better term. Some also use \exit." \Termination" results when all or all but one member withdraws from an insti- tution. Here, I use \withdrawal" unless another term was used in an empirical case or explanation. See Section 2 for a more in-depth discussion of terminology and jurispru- dence. 16 The 1969 Vienna Convention on the Law of Treaties (VCLT), articles 42{45, 54{ 56, 65{68, 70{71, contain rules that apply when agreements do not include termination, denunciation, or withdrawal rules. The \Treaty Handbook" prepared by the U.N. Treaty Section of the Oce of Legal Aairs provides guidance on withdrawal, denunciation, and termination events at sections 4.5 and 4.6. (See also Voeten 2001; Johns 2007). 8 in international relations based on its perceived cost to society, despite its rarity (Gartzke 1999; King and Zeng 2001; Collier and Sambanis 2002). Given con ict's relative rarity but high costs, it too could be considered o-equilibrium path behavior worthy of study. 17 A systematic empirical analysis of the phenomenon of withdrawal is still useful in developing our understanding of this category of treaty behavior beyond conjecture and anecdotes. In short, exercising one's right to withdraw from an institution ends coop- eration; however, it does so in a dierent way than breaching. Withdrawal threatens the very existence of the institution and removes a state from a forum where shared interests could develop. Understanding the cases of treaty withdrawal{and, later, its eects|is important in understanding the durability of international cooperation. Institutional design is crucial in cre- ating a treaty; however, conditions change and states make decisions based on those changed conditions. The ability to withdraw promotes interna- tional cooperation in giving states enough exibility at the time a treaty is created for a deal to be reached. Without an option to exit, international institutions could be too rigid. Deciding to breach a treaty given a change in circumstances undermines the legitimacy of international law. Withdrawal promotes exibility and legitimacy but can undermine cooperation. As is well known, the breakdown of a bargain within a treaty framework 17 von Borzyskowski and Vabulas (2019) have a similar assessment. 9 can have signicant consequences, potentially leading to a decrease in hu- man welfare and at worst suering and war. 18 Ending a treaty relationship can have profound costs for the states and their citizens. With the recent withdrawal of the U.S. from the Intermediate-Range Nuclear Forces (INF) Treaty and Open Skies Treaty, 19 some commentators fear a total collapse of the arms control regime between the U.S. and Russia leading to increasing arms spending and little transparency. 20 Some also argue for the U.S. to withdraw from the 1967 Outer Space Treaty so that the nascent U.S. Space Force can operate unencumbered. 21 This may prompt other treaty members to exit and accelerate the further militarization of outer space activities. At the extreme, withdrawal may precipitate war. 22 18 The fact that the collapse of some treaties leads to these consequences shows the power of formal international institutions to control or shape state behavior. These treaties may be acting as dams for the unrestrained ood of preferences of a state. 19 Michael R. Gordon and Vivian Salama. \Trump Moves Closer to Ending An- other Post-Cold War Treaty." The Wall Street Journal (Oct. 27, 2019) avail- able at https://www.wsj.com/articles/trump-moves-closer-to-ending-another-post-cold- war-treaty-11572177600; \Sens. Cruz, Cotton, Burr Call on President Trump to Withdraw the United States From Open Skies Treaty" (March 12, 2020) available at https://www.cruz.senate.gov/?p=press release&id=4984. 20 Lara Seligman and Robbie Gramer. \What Does the Demise of the INF Treaty Mean for Nuclear Arms Control?" Foreign Policy (August 2, 2019) available at https://foreignpolicy.com/2019/08/02/what-does-the-demise-of-the-i-n-f-treaty-mean- for-nuclear-arms-control-intermediate-nuclear-forces-new-start-strategic-arms-limitation- nonproliferation-trump-russia-arms-control-explained/; Steven Pifer. \As US-Russian arms control faces expiration, sides face tough choices." Brookings Institution (March 23, 2020) available at https://www.brookings.edu/blog/order-from-chaos/2020/03/23/as-us- russian-arms-control-faces-expiration-sides-face-tough-choices/. 21 Schmitt (2020a). 22 After Poland rejected Nazi Germany's demand for Polish territory, Hitler unilaterally rescinded the German-Polish Non-Aggression Pact in a speech before the Reichstag on April 28, 1939. Four months later, Nazi Germany invaded Poland. 10 1.2 International Law's Aect on State Behavior Treaty withdrawal is a political act of a state that has been formalized and reied in international legal doctrine. International law prescribes what is a proper method for a state to disengage from its formal commitments with other states. By following the proper procedure, or in a more primitive sense reciting the proper incantation, a state can break its prior commitment to cooperate and subject itself to the obligations of the relationship. Interna- tional law also governs other state behavior through norms and rules that overlap and connect, including rules on treaty making, treaty interpretation, and treaty breaking. The law serves as the banks and levees of rules and norms constructed to channel the ow of state interests and conduct in their relations with each other in the absence of a global sovereign. These banks and levees are created by the states themselves over decades and centuries, evolving overtime (Sandholtz 2008; Sandholtz and Stiles 2009). Interna- tional law as contest, dialectic, or compromise of political positions is one way to understand how it substantively develops overtime (Koh 1996; Rapp 2020). From the Peace of Westphalia (1648) to the post-World War II era, states have crafted the modern contours of international law to govern their relations between each other and, increasingly, within their borders. 23 23 Scholarship on the history and theory of international law is vast (See generally, Fassbender et al. 2012). Koskenniemi (2012) provides a summary of the history of inter- national histories from early Christianity through natural law during the Enlightenment to the professionalization in the 19th century on through to modern times. Most histories of international law mark the Peace of Westphalia as the birth of the modern form in international law. This is not to say a system of norms or laws did not exist between nations before 1648 (Preiser and Bernhardt 2008); those standards existed and are impor- tant to the heritage international law as we know it. Our current form of international law can trace its birth to the order established after the costly Thirty Years War in Europe. 11 How does international law do this? From the outset, international coopera- tion through institutions requires states to cede some sovereignty by impos- ing some limitations on their agency, primarily by granting the others rights that can be exercised against the state (Stein 1982; Keohane 1984; Abbott and Snidal 2000). International law is the set of rules binding states to one another as legally binding (Bull 1977, 127). In describing the \soveignty costs" states incur when entering an international agreement, Abbott and Snidal (2000, 436) explain: \[a]ccepting a binding legal obligation, espe- cially when it entails delegating authority to a supranational body, is costly to states." International law is a tool for states to use to govern behavior both substantively and procedurally. Substantively, the question of whether and to what extent international law aects state behavior is often examined in both legal and political scholarship under the heading of compliance. 24 Compliance is often examined from the perspective of addressing non-compliance. Scholars have often pointed to the broken promises of states, noting: But, after all allowances and reservations have been made for obsolete or imperfect or doubtful obligations, I am afraid it must be admitted that a very lax morality exists among nations, as to the observance of treaties, even where there is no reasonable ground for calling in question their substantial validity. The From 1648 through the 19th century, European colonial powers engaged with each other and with states outside Europe, primarily in Asia, under a set of norms and rules that recognized states as sovereign and relatively as equals in law (if not in power or racial or ethnic terms) (Verosta 2007). Modern international grew out of the Westphalian system, yet also was in uenced by other systems, including older systems developed in East Asia (Kang 2010; Coe and Wolford 2020) and communist systems like the Soviet Union and People's Republic of China (Grzybowski 1970; Hanqin 2012). 24 Compliance can be understood as distinct from eectiveness, although the concepts may overlap. 12 highway of history is strewn with the d ebris of broken treaties (Richard 1877-1878, 93). Yet, international lawyers echo Henkin (1979, 47) in stating \almost all na- tions observe almost all principles of international law and almost all of their obligations almost all of the time." Starting from this view, why do we see states violate their commitments? Where violations of international law occurs, compliance can be understood as a managerial problem where uncertainty about the contours of compliance leads to violation (Chayes and Chayes 1993). Critics take a dierent view, seeing wide-spread \compliance" as the result of shallow commitments and states simply pursuing their in- terests without any obligation from international law (Downs, Rocke and Barsoom 1996; Goldsmith and Posner 2005). The argument of Goldsmith and Posner (2005) has served as a persistent foil to scholars of international law, prompting signicant work aimed at pushing back on their argument and those like it (See, e.g., Ohlin 2015). Compliance can be explained, at least in part, by the communal dynamics of international relations, where the community can police behavior through social interactions. Outcasting is a mechanism for the community of states to force compliance (Hathaway and Shapiro 2017; Mantilla 2020). \Outcasting occurs when a group denies those who break its rules the benets available to the rest of the group" (Hathaway and Shapiro 2017, 375). As described by Mantilla (2020, 8{9), international law is a \collective political achieve- ment" of a socially pressured compromise driven by reputation and face saving. Outcasting is in part a function of collective reciprocity. Moreover, 13 the communal management of its members' reputations for being a good member of the community can be used to shame norm violators (Hafner- Burton 2008). 25 Returning to Henkin's maxim above, we might understand widespread compliance as the result of threatened or potential outcasting. \Outcasting, then, is a victim of its own success. It is so ubiquitous and so often eective that it is usually invisible" (Hathaway and Shapiro 2017, 377). International law carries with it normative power that can shape interests and behavior (Klotz 1995; Finnemore and Sikkink 1998; Checkel 2001; Reus- Smit 2003). Through interactions in the international society, states develop a \culture of compliance" as international law is seen as a legitimate source of authority (Franck 1990; Henkin 1995). States comply, or rather are \obe- dient," with international norms and law when they internalize those norms into their own value system through a \transnational legal process" (Koh 1996). Through the transnational legal process of interaction, interpretation and internalization, states comply more or less depending on their level of in- ternationalization. Full internationalization results in proactive motivation for obedience rather than compliance out of fear of punishment. Obedience springs from an internal desire to do what is right. Norms can diuse across the international community through entrepreneurs, transnational advocacy networks, and socialization (Finnemore and Sikkink 1998; Keck and Sikkink 1998; Johnston 2001). Although, international law in uences state behavior through its normative power, it can also do so through economic or material 25 Greif (1993) describes how 11th-century traders could enforce contracts through methods of group accountability and punishment. 14 channels (See, e.g., Sandholtz and Gray 2003). Moreover, scholars also nd evidence that states comply even with \soft law" (Shelton 1997, 2000). Using a rational choice paradigm, states comply with the terms of a treaty for various reasons, including considerations of reputation, retaliation, reci- procity, and future cooperation and other assessments of costs and benets as described in the large compliance literature (Keohane 1984; Fearon 1997; Simmons 1998; Guzman 2008; Simmons 2010). The prospects of future ben- ets or punishments, namely repeated interactions among states, allows in- ternational institutions to have in uence without an overarching sovereign (Keohane 1984). These reasons for compliance appear to tap into ways in which many societies have developed systems of dispute resolution and justice where justice is imbued with some level of equal treatment for like situated individuals and that a breach of societal norms by one is punishable by a reciprocal breach by the injured party or the society as a whole. 26 In incorporating normative ideas, Guzman (2002, 65) understands legal com- mitments as inherently raising ex post costs, writing that treaties \represent the complete pledge of a nation's reputational capital." Since enforcement depends largely on reciprocity, this framework is useful for explaining stable trade agreements (Goldstein, Rivers and Tomz 2007) and some aspects of the laws of war, where militaries risk retaliation in kind 26 Harari (2014) discusses early human societies' ability to cooperate through the use of ctional institutions, like law or religion, to harmonize the functioning of ever increasing groups of people. In his estimation, homo sapiens would not have been able to cooperate on the grand, civilization-creating scale without these unifying ctional institutions. This may correspond to theories of justice as fairness as laid out by Rawls (1971) (See generally D'Amato 1975; Franck and Franck 1995). 15 (Morrow 2007). It is also suitable for analyzing obligations whose violation might provoke negative market reactions, as is plausible in the area of mon- etary aairs and investment (Simmons 2000). The compliance literature has largely focused on the negative inducements of compliance; however, states are also motivated by positive incentives for complying with international law (Schimmelfennig 2005; van Aaken and Simsek 2021). 27 Also, obtaining a balance in the scope of compliance incentives across powerful and weaker states is crucial in designing an international institution (Stone 2008) There- fore, a properly balanced carrot and stick understanding of international law, in part based on reciprocity, seems best to motivate states. Aside from pressure from the community of states, domestic political pres- sure, including domestic interest groups, can be leveraged to ensure compli- ance with international agreements (Dai 2005; Hafner-Burton and Tsutsui 2005; Simmons 2009; Conrad and Ritter 2013). International institutions provide information about leaders' actions to the public (Manseld, Milner and Rosendor 2000, 2002; Johns and Rosendor 2009; Manseld and Milner 2012). Domestic groups can also persuade states to violate international law (Pollack and Shaer 2009; Rosendor and Milner 2001; Rosendor 2005). Again, these groups can can leverage transnational advocacy networks to pressure governments to circumvent domestic blocks to advocacy (Keck and Sikkink 1998). 27 The American Society of International Law recently devoted a symposium to the topic. See AJIL Unbound, Volume 115, 2021. 16 In understanding state behavior, compliance is only part of the picture. Empirically studying state behavior must account for other outcomes po- tentially in uenced or caused by international law (Martin 2013). There are other outcomes or potential outcomes that are of interest and potentially give power to an empirical investigation. International law also prescribe how states interact and are bound by inter- national obligation broadly. Studying treaty exit allows for an investigation of international law's broad eect, or lack thereof, on state behavior, by fo- cusing on a formalized and relatively easily observable behavior. Moreover, this work furthers the growing trend between scholars in political science and law to bridge our understandings of both domains and how they are intertwined (Slaughter, Tulumello and Wood 1998; Abbott 1999; Raustiala and Slaughter 2002; Ohlin 2015). Legal doctrine, examined in law schools, is aected by the politics of the time, and politics are shaped by the le- gal landscape of acceptable actions. Understanding the contours of one is informative in understanding the other. 1.3 Jurisprudence of Treaty Withdrawal Treaty withdrawal is a legally dened action that comes with various le- gal justications and parameters. To start, a number of terms are used to describe the end of a treaty or participation in a treaty. The terms \ab- rogation," \denunciation," \termination," and \withdrawal" are all used to describe the act of ending a state's consent to be bound by or adhere to a 17 treaty. These terms are often used interchangeably (Helfer 2012, 635{636, fn.6). \Termination" may encompass withdrawal but it is broader and can also describe the end of the application of the treaty to all participants by formal act, by change in state of the world, or by its terms. Treaties may also be \invalidated" whereby a state terminates a treaty on grounds of fraud, corruption, coercion, error, and violation of a domestic law of funda- mental importance. A treaty may also be invalidated if its terms violate a peremptory rule of international law or jus cogens|examples being a treaty encouraging maritime piracy, starting a war, or committing genocide. In this project, I use the terms \withdraw" and \exit" to refer to a state's in- tentional and pro-active decision to end its consent to be bound by or adhere to a treaty. To start, a treaty is dened as \an international agreement concluded be- tween States in written form and governed by international law, whether embodied in a single instrument or in two or more related instruments and whatever its particular designation" in Article 2 of the 1969 Vienna Con- vention on the Law of Treaties (VCLT) or \treaty on treaties." 28 Treaty withdrawal is legally permissible based on a number of grounds as described in the VCLT. Generally, a state may withdraw by the terms of the treaty or by the consent of all parties to the treaty (VCLT, art. 54). If a 28 Although the United States signed the VCLT in 1970, the U.S. Senate has yet to give its \advice and consent" ratifying the treaty. Nonetheless, the U.S. abides by many of its terms, considering some to be codied customary international law on the law of treaties. See U.S. State Dept. \Vienna Convention on the Law of Treaties." available at https://2009-2017.state.gov/s/l/treaty/faqs/70139.htm The Restatement of the Law (Third) on the Foreign Relations Law of the United States tracks many of the legal prin- ciples in the VCLT. 18 treaty is silent on withdrawal, the state must not violate the treaty based on pacta sunt servanda (VCLT, art. 26). However, there are exceptions to this rule: withdrawal is permissible if implied by the nature of the agreement or other circumstances (VCLT, art. 56). Beyond general reasons for withdrawal, a treaty may end because (i) it is invalid under jus cogens, as discussed above, (VCLT, art. 64); (ii) of the impossibility of performance where a permanent disappearance or destruc- tion of an object indispensable for the execution of the treaty has occurred (VCLT, art. 61); or (iii) a material breach (VCLT, art. 60). A material breach of a treaty is the unjustied repudiation of the treaty or a violation of a term that is essential to the accomplishment of any object or purpose of the treaty (VCLT, art. 60). Here, I use the term \breach" to refer to material breach, which is generally opposed or undesired (if not ignored) by the other parties to a treaty. Also, withdrawal is permissible under the legal maxim rebus sic stantibus, which refers to a fundamental change in circumstances that aects the per- formance of a state's obligations under the treaty that were unforeseen at the time the treaty was concluded, providing grounds for termination of a treaty (VCLT, art. 62). Relatedly, a state may \suspend" the application of a treaty by agreement or unilaterally, especially in cases of crisis (VCLT, art. 57, 58, 60, 61, 62). This legal option is exceptional and rarely used. The United States used it once in 1941 when the Roosevelt Administra- tion suspended the country's participation in the International Load Line 19 Convention based on the ground that World War II constituted changed circumstances not anticipated by the parties. 29 States often use these legal rationales in explaining their decision to with- draw from a treaty. President Trump explained the rationale for why the U.S. was planning to withdraw from the Open Skies Treaty 30 with Russia: But Russia didn't adhere to the treaty, so until they adhere, we will pull out.. . . But whenever there's an agreement that another party doesn't agree to|you know, we have many of those agreements around the world, where it's a two-party agreement, but they don't ad- here to it, and we do. When we have things like that, we pull out also. 31 President Trump points to Russia's prior material breach as the reason for the U.S. withdraw|as was the case with a number of earlier instances of the U.S. withdrawing from arms control treaties with Russia. 32 Moreover, states may rescind a notice of withdrawal before it takes eect. Although courts at the international level have provided only a few rulings on the parameters of withdrawal, the ability of a state to rescind its notice was the subject of a recent ruling of the European Court of Justice (ECJ) in the context 29 Congressional Research Service. \Treaties and Other International Agreements: The Role of the United States Senate" (January 2001), p. 195. 30 The Treaty on Open Skies was signed in 1992 and went into force in 2002. Negotiated by then NATO members and the Warsaw Pact, the treaty allows for aerial, unarmed surveillance ights of member territories. President Dwight Eisenhower proposed the concept of \mutual aerial observation" in 1955; however, it was not accepted until the renewed initiative of President George H. W. Bush in the nal days of the Cold War. 31 Remarks and an Exchange With Reporters Prior to Departure for Ypsilanti, Michigan (May 21, 2020) https://www.presidency.ucsb.edu/node/341977 32 Author interviews with former State Dept. ocials and staers for Congressional committees. 20 of Brexit. 33 There, the ECJ stated that the decision to revoke a notice of withdrawal \is for that member state alone to take, in accordance with its constitutional requirements." Now having laid out the legal justications for treaty withdrawal, the question remains as to when does a state invoke these reasons to end its participation in a treaty. 1.4 Existing Theories 1.4.1 Are Entry and Exit Simply Two Sides of the Same Coin? States create or join treaties to solve problems and create value; 34 states leave treaties, in part, when the reason for joining is no long valuable (See, e.g., Leeds and Savun 2007). Can we simply explain exit as the other side of the same coin? Put another way, do we need a theory of treaty exit? There are a number of reasons we might think of exit as a phenomenon with enough distinguishing features to warrant independent study. To start, exit diers from entry in that exit occurs not in the vacuum of anarchy but in the context of an existing relationship. Before a state enters into a treaty, its relationship with other potential partners can be essentially neutral. 35 Once they enter into a treaty, that commitment will color any future decision 33 C-621/18, Andy Wightman and Others v. Secretary of State for Exiting the European Union, 2018. 34 For example, state economic preferences are strong predictors of treaty commitment preferences (Lupu 2016). 35 Of course, states have histories and long-memories of interactions. 21 to continue in a treaty, breach, or exit. Treaty exit is akin to divorce. 36 The decision to exercise an option for divorce is made in the context of an existing relationship with obligations and revealed information. This is distinguished from entering into the marriage relatively unencumbered, with limited information, and with uncertainty about future behavior. The gains or losses of trust in the relationship are much more signicant than outside the relationship. Exit also has the unique feature that it is largely governed by rules either set out by the state parties or by international custom. But more explicit and involved rules exist for entering a treaty or international agreement, requir- ing multiple steps to become binding, including signature and ratication. Relatedly, how each state exercises its option to withdraw is generally less dened and heterogenous across states. When entering a treaty, a state gen- erally signs through an executive and then raties through some legislative body. However, when leaving a treaty, it is less clear who within a state has power to do so and if that actor can do so unilaterally. In South Africa, the Supreme Court determined that Jacob Zuma, then the President, could not unilaterally withdraw from the Rome Statute without the approval of the legislature (Woolaver 2017). In the U.S., this has been the subject of legal ghts and political tug-of-war between the President and Congress, especially in the age of the imperial presidency. Thus, in domestic systems, 36 In his recent work on withdrawal from multi-lateral treaties, Morelli (2021, 3) also compares withdrawal to divorce. 22 who can leave an international agreement has fewer explicit rules leaving room for more political contests. 1.4.2 Why Do States Exit? What We Know The literature on institution and treaty formation is fairly robust and ad- dresses the design and inclusion of \exit" clauses. According to Koremenos (2016), 70 percent of her random sample of agreements contain withdrawal clauses. Helfer (2005, 2010, 2012) argues that \exit" provisions help states to mitigate the uncertainties within the international environment by allowing for lawful, public mechanisms for a state to terminate treaty obligations and withdraw completely from membership in an institution. Generally, treaties include an \escape clause" to preserve the idea of state sovereignty (Helfer 2005). At the core of these theories is the basic idea that international agreements are rationally designed to solve various problems like cooper- ation, distribution, or coordination problems between states. (Stein 1982; Morrow 1994; Fearon 1998; Koremenos, Lipson and Snidal 2001; Rosendor and Milner 2001; Koremenos 2016). Once cooperating within a rationally designed institution, the state generally will not defect from cooperation without some signicant, motivating change. Koremenos (2005, 2016) ar- gues that withdrawal from an agreement occurs when there is a signicant shift in the underlying \bedrock" preferences of the withdrawing party away from cooperation. For example, a state abrogates its military alliances when it experiences a substantial change to circumstances like a change in interna- tional power or creation of a new, outside alliance (Leeds and Savun 2007). 23 Although the design of these provisions is important to understanding state behavior, it does not by itself answer the question as to why and when a state will choose to exercise this option to exit. In the literature on institutional design, treaty withdrawal has been linked to treaty breach in passing and without deep analysis. Koremenos and Nau (2010) assert that withdrawing has the same eect, strategically, as defecting from international cooperation. However, this seems unsatisfying in explaining the dierence between defecting by breaching and \defecting" by choosing to exercise a bargained-for exit. The substantive dierence between the two behaviors, as well as the strategic calculations leading to each, is under-theorized. 37 The foundation of a theoretical explanation for treaty withdrawal comes from work in economics and organizational theory. Hirschman (1970) pro- vides a theory to explain the choice a member of an organization|whether a consumer, business, or citizen|faces when the quality of the organization and benet to the member are perceived to be decreasing. Members can choose to either exit the organization, an economic action; voice a reform the organization, a political action; or remain quiet. The choices interact and are impacted by members' loyalty to the organization. In economics and business, research using formal models has considered the eect of in- cluding the option to exit on partner matching (Hauk and Nagel 2001) and 37 Pevehouse et al. (2020) note that \the work on the determinants of IGO membership (including accession, withdrawal, or expulsion) is relatively sparse, compared to mem- bership in other international institutions such as alliances (Mattes, 2012) or preferential trade agreements (Manseld & Milner, 2015)." 24 rm alliances (Arend and Seale 2005). In a unilateral-choice setup of the partner-matching game, intending defectors are more likely to exit than in- tending cooperators (Hauk and Nagel 2001). This lends support to the idea that treaty exit and breach are related, as mentioned earlier; however, the underlying theory driving these decisions in the international treaty context needs additional exploration. Helfer (2005)'s seminal work on treaty withdrawal bridges the early theo- ries from Hirschman (1970), international relations, and law and then draws from empirical data to expand on them. In discussing international relations theories on collaboration, coordination, and defection, he notes that schol- ars have often con ated withdrawal and breach and presents a new way to conceptualize the traditional collaboration and coordination games in inter- national relations theory. This work has in uenced much of the theoretical and empirical research, limited as it is, on treaty exit over the last decade and a half (Helfer 2019). Much of the work on treaty withdrawal and in- ternational cooperation has focused on subsets of agreements, institutions, and organizations. One growing line of research regarding treaty exit focuses on the eect in- ternational courts and tribunals have on a state's preference to leave the court's jurisdiction by withdrawal. Pauwelyn and Hamilton (2018) provide two sets of changes that drive exit from international tribunals: tribunal- centered based on judicial activism and tribunal bias, and state-centered based on sovereignty costs and policy change. Johns (2015) provides a the- ory of how the strength of international courts aects dispute settlement, 25 compliance, and the stability of the international institutions. As she ex- plains: \[C]ourts must promote short-term compliance with the law and the long-term stability of the regime" (21). Using a formal model and investigat- ing the International Court of Justice and the Dispute Settlement Body of the World Trade Organization (WTO), she nds that a strong court is ideal when laws are precise and the court is embedded within a robust political structure, for example the European Union. These strong courts increase compliance when it requires less eort and exit when compliance requires substantial eort. A weak court is ideal when law is imprecise and treaty exit has minimal political and economic repercussions. Johns (2015) gives us a rigorous theory by which we can understand state withdrawal from an international adjudicative body's jurisdiction. However, Johns (2015, 60)'s theory collapses a state's choices into cooperation or violation, where a violation can be resolved by dispute resolution or exit. Doctrinally, there is legal debate about the domestic political and legal pa- rameters of treaty exit. The American Society of International Law's (ASIL) recent symposium collected a number of articles that focused on the domes- tic constraints on the power to withdraw (Helfer 2017). McLachlan (2019) posits that international law places checks on withdrawal, to protect the collective interests of the member states|namely the peaceful resolution of disputes. Thus, he argues that national courts should constrain the unilat- eral acts of an executive to withdraw from an international court's juris- diction. Such was the case in South Africa (Woolaver 2019). Specically, Bradley and Helfer (2017) examine how the courts in the United Kingdom 26 and South Africa have constrained the executive's authority to withdraw from treaties in those countries, with implications for the U.S. In the U.S., the United States Supreme Court has determined that assessing whether the President has the power to withdraw from a treaty unilaterally, i.e. without the \advise and consent" of the Senate, is a nonjusticiable politi- cal question. 38 Where the courts decline to provide guidance, legal scholars have posited various analyses of the historical practice of treaty termina- tion (Schnitzer 2012; Bradley 2013) and with regard to the constitutional issues regarding separation of powers between the executive and legislative branches of government (Eichensehr 2012). Interest in research on the determinants and covariates of treaty withdrawal has increased in recent years, possibly in light of the number of high prole withdrawals from signicant and to date durable international institutions. As pointed out in this dissertation, Gray (2020) notes that concerns over high prole withdrawals in recent years may skew perceptions about its frequency or abnormality. International organizations can follow dierent paths including life, death, inertia, and change, each relevant to understand- ing international cooperation (Gray 2020, 35{37). An international organi- zation's existence does not evidence meaningful on-going cooperation, some organizations may persist as lifeless or as \zombies" Gray (2018). Nev- ertheless, states appear to prefer ending organizations and creating new 38 Goldwater v. Carter, 444 U.S. 996 (1979); see also Kucinich v. Bush, 236 F.Supp.2d 1 (D.D.C. 2002). Scholars have critiqued the application of the political question doctrine, especially on issues of foreign policy. See, e.g., Adler, David Gray. \The Law: Termination of the ABM Treaty and the Political Question Doctrine: Judicial Succor for Presidential Power." Presidential Studies Quarterly 34, no. 1 (2004): 156-166. 27 ones rather than to reform or perpetuate dysfunctional or ineective ones (Eilstrup-Sangiovanni 2020). Continuing on this path, much of the recent work focuses on a subset of treaty withdrawal, namely from international organizations. von Borzyskowski and Vabulas (2019) use theories on why states join international organiza- tions, as explained above, to understand why states leave them, testing these explanations against data on international organizations from 1945. They nd that geo-political factors including preference divergence and contagion are the principle drivers of exit from international organizations. Despite commentary regarding the recent withdrawals in America and Europe, they nd that nationalism is not a driver of withdrawal from international orga- nizations (von Borzyskowski and Vabulas 2019, 2022). Moving back to a broad understanding of treaty withdrawal, Morelli (2021) examines the legal and conceptual framework and development of with- drawal from mutli-lateral treaties. Similar to the conception used here, Morelli (2021) understands withdrawal to be related but distinct from breach, conceptualizing considerations by states as a balance between \commitment and compliance." Unilateral treaty withdrawal can undermine the trust be- tween states, requiring a balance between too permissive withdrawal and too restrictive (Crampin 2020). Morelli (2021, 147{164) categorizes state reasons from withdrawing mutli-lateral treaties as falling into two broad cat- egories: reactions to domestic politics and as an act of state power. Those reasons included increased economic costs, \mission creep," perceptions of bias or unfairness, change in domestic leadership, leverage in amending a 28 treaty, a defense strategy, and obtaining an alternative consensus. His ty- pology is useful in understanding the direct rationale for a state's withdrawal from a mutli-lateral treaty, but does not move to the underlying theory un- derpinning each stated rationale. New work on treaty exit looks at the eects of withdrawal on future coop- eration (von Borzyskowski and Vabulas 2018), how states address concerns of withdrawal in the case of important exits, like Brexit (Larik 2020), and investigates the overlap of the scope of international organizations in in- uencing a state's preference to withdraw from an organization or treaty (Clark 2020). Burgeoning scholarship also investigates pending threats of treaty exit from specic regimes, like the Nuclear Non-Proliferation Treaty (Pretorius and Sauer 2022). Rigorous and deep work on individual cases is useful in understanding how general mechanisms driving treaty withdrawal are utilized in each case. The eect of network structures and states' indi- rect and parallel connections may also aect decisions to exit a treaty exit where states can substitute the benets of one treaty for that of another (Dalton 2021). A similar logic appears to apply to termination of bilat- eral investment treaties (BITs) where integration into global value chains leads to higher probability of terminating a BIT (Ge 2022). However, al- though empirical work on treaty withdrawal continues to expand, questions regarding the theoretical underpinnings of exit remain open. 29 1.5 A Bargaining Model of Treaty Exit This project seeks to understand how the design of an international legal institution in uences whether a state can bargain with other members to address changes to circumstances in light of the options to breach the agree- ment or exit. Building on and complementing the growing body of work on treaty withdrawal (Helfer 2005; von Borzyskowski and Vabulas 2019; Helfer 2019; Morelli 2021; von Borzyskowski and Vabulas 2022), this project focuses on the institutional features that frame the on-going bargaining between states that cooperate in a formal treaty. Here, I move beyond looking at treaty withdrawal from international orga- nizations or courts, and expand our understanding to all types of formalized treaty relationships between states. Within a random sample of treaties, only 44 percent of treaties create an international organization (Koremenos 2016). 39 We might think that many of the considerations of states in those institutions that have created a treaty organization would apply to those treaty relationships without an organization. However, treaty relationships are prior to and perhaps more primitive to those that create an international organization. International relationships that produce an international or- ganization may dier in systematic ways from other relationships. 39 Koremenos (2016) denes \organization" broadly, including bodies that include a secretariat and commissions created by bilateral arms control agreements, going beyond the denition of the oft-used COW IGO data set. Because of this broad denition, we might expect the proportion of treaties that create organizations that meet the COW IGO denition is even lower. 30 Further, through research on international institutions (Barnett and Finnemore 2012), we know that the delegation of authority and accretion of power through moral claims and expertise provide organizations with their own authority and agency. This may complicate the bargaining of the member states when changes to circumstances arise, if for no other reason than in- troducing another interested party who may or may not be interested in preserving the status quo. This is not to ascribe any normative view on the organization's involvement in the bargaining process. It simply raises the specter that studying only those relationships between states that resulted in an organization may only give a partial glimpse into how states cooperate within formal international relationship. As explained in Chapter 3, most treaties are bilateral treaties, with 94% of treaties concluded on a bilateral basis from 1800 to 2020. 40 Widening our scope to all types of treaties allows us to incorporate our understanding of both well-known, multi-lateral treaty organizations as well as obscure and discrete bilateral treaties. 41 As theorized here, international cooperation within the bounds of interna- tional law is an on-going negotiation between states. States come together with some desire to address a common problem or issue, but the bargaining does not stop once the treaty is signed and ratied. As the states move for- ward with the relationship formalized in a treaty, they are confronted with waves of changing circumstances. Much like a sail boat cutting through the 40 For treaties concluded and registered with the United Nations Treaty Series (UNTS), 94% are bilateral treaties, 3% are closed multi-lateral treaties, and 3 % are open multilat- eral treaties. 41 Morelli (2021, 2{3) chooses to limit his analysis to multi-lateral treaties for feasibility. 31 water, once the sails are properly trimmed and a bearing set, the boat con- tinues to its destination despite minor waves that cause slight deviations in its course. The boat navigates those minor changes with ease and small ad- justments to the rudder. However, every so often, a large wave or changing winds necessitate a more signicant change. This is where the skipper, like the state, faces a choice on how to proceed. When a state is confronted with a signicant shock or change to its circum- stances, which can occur quickly or over time, it can bargain to account for those changes. Given that states have already entered into a formal treaty relationship, bargaining to reform the treaty is the most straightforward way to account for the change without a major shift in relationship between the states. Part of the bargaining at the start includes the design of the treaty, as explained above, but it also includes the making of declarations and reservations to adjust the purview of the treaty to the state's future interaction. 42 States engage in this behavior to adjust the terms and ap- plicability of the treaty, sometimes called \treaty options" (See Galbraith 2012). Some of these concessions are bargained for from the start but oth- ers come later. States may also bargain to completely reform or amend the treaty or seek to add an addendum or protocol to the treaty. Bargaining might also simply result in a shared understanding regarding a deviation 42 Declarations explain a state's understanding of some matter or interpretation of a provision of a treaty, usually made at time of deposit of the treaty instrument or at the time of signing. Declarations generally do not purport to modify the legal eect of the treaty but carry weight as to the state's interpretation of the treaty. Reservations are declarations by states that exclude or alter the legal eect of a provision of the treaty in its application to that state as long as the reservation is not incompatible with the object and the purpose of the treaty (VCLT, art. 2(1)(d), 19{23). Other states may object to a reservation (VCLT, art. 20{23). 32 from some treaty obligation that is less formal. The bargaining options allow states wide latitude to solve problems within the treaty framework. However, a state unsatised with the status quo also has other options avail- able to it. A state could forego continued bargaining with the other member states|or see bargaining as futile in certain situations|and seek to redress the changed circumstances on its own. It could do so by either breaching its obligations under the treaty or shedding its future obligations altogether by exiting. 43 The states' choice between options is largely in uenced by two fundamental parameters of the treaty relationship, namely the treaty's exi- bility and level of obligation it puts on the state. These two dimensions may not be the only dimensions of a treaty's design that in uence behavior, but they constitute the core of a state's cost and benet analysis of its continued participation in a treaty relationship. States incorporate exibility into treaties in various ways in light of the uncertainty of the state of the world in the future and with distribution problems (Downs and Rocke 1990; Lipson 1991; Koremenos 2001; Kore- menos, Lipson and Snidal 2001; Koremenos 2016). States may incorporate exibility into the design of their treaties in various ways, including setting a nite duration for the life of the treaty and adding \escape clauses," as mentioned above. These treaty provisions allow states to adjust to shocks to their cost and benet analysis. States also incorporate exibility by using imprecise language and adding reservations when joining. Finite duration 43 Recent work on treaty withdrawal from mutli-lateral treaties also nds these three choices as foundational for states (Morelli 2021). 33 provisions are useful to allow for later renegotiation after the parties have gained more information about the costs and benets of a treaty. However, the costs of renegotiation increases with the more treaty partners trying to negotiate (Koremenos 2001). Escape clauses are calibrated by the treaty partners to avoid opportunistic use by adjusting notice and waiting periods (Koremenos 2016). As in the case of the U.S. announcement of its intention to exercise its right to withdrawal for the Paris Climate Accord, the U.S. was required to wait a certain number of years before it could do so. In that case, the parties incorporated the wait period to give the other parties a chance|if they so desired|to try to prevent the withdrawal. International agreements create obligations on the member states. Some obligations are legal, while others may be moral or political. Raustiala (2005, 586) argues that there exists a great variety in the \legality" of in- ternational law. Some of that heterogeneity is described by Abbott and Snidal (2000) in distinguishing covenants and contracts in the underlying obligations motivating state behavior. Importantly, obligations created by international law and norms come in two broad forms based on their un- derlying function. The rst and most commonly understood version is as a primary or substantive obligation to do or refrain from some behavior. The second is procedural, namely the rules about making rules. Here, obligations regarding how to withdraw from a treaty would be secondary obligations on a state. Again, when the U.S. announced its withdrawal from the Paris Accord, it did not immediate leave the treaty but abided by the procedural rules for proper withdrawal. Empirical work on treaty obligation and state 34 behavior (Almeida and Pereira 2013; Mulesky, Sandholtz and Zvobgo 2020), highlights how obligation and exibility function in tandem and interact. In the context of a bargain, a state will be more likely to choose to exit when a treaty framework is too rigid to make an opportunistic breach valu- able and the shock to the state increases the burden of the treaty obligation to levels that are not capable of a bargained-for solution. The threat of exit (or breach) is used by a state when bargaining to increase in its demands. This allows for larger adjustments to larger shocks. Understanding the insti- tutional bargaining within the context of obligation and exibility produces a framework by which to understand treaty behavior. 1.6 Roadmap of Dissertation What follows is an examination of the phenomenon of treaty exit, develop- ing an institutional theory of bargaining and testing it against a variety of evidence. Chapter 2 presents a theory of treaty exit based on a bargaining model that describes the institutional factors that lead to withdrawal. The formal model presented builds on previous models of bargaining, dispute settlement, and arms control. From the formal model, I derive propositions and expectations that lead to hypotheses to be tested empiricially. In that chapter, I present short case studies that help inform the model and test its plausibility. In Chapter 3, the expectations developed in the theory from Chapter 2 are tested using quantitative analysis across many treaties from the seven 35 decades following World War II. First, I explore the contours of treaty- making and exiting for the relevant time period. I then introduce a dataset of treaties collected from the UNTS and a number of variables I construct using automated text analysis to test the hypotheses, primarily based on the concepts of treaty exibility and obligation. In Chapters 4, 5, and 6, I examine treaty exit from three dierent institutions or sets of institutions, exploring the factors that lead to the exit and the bargaining process|to the extent any exist|that accompanied the decision to exit. First, I examine the history of the Rome Statute establishing the International Criminal Court and how the institution has grappled with threats of withdraw and actual exit from a number of states. Here, we can see how bargaining is used to avoid exit in some cases and fails in others. In the security context, I next look at the U.S. withdrawal from two arms control treaties with Russia. Here, we focus on the in uence of the changing security environment and the presence of breaching behavior on withdraw. Finally, I examine Brexit and the United Kingdom's international and internal bargaining that lead to withdraw. Each case takes an important case of withdrawal and seeks to understand that event in the context of the theory. 1.7 Limitations and Lingering Questions This project provides a framework for understanding how an institution's design in uences the continued bargaining of the states to lead to various 36 outcomes, including continued cooperation, breach, or exit. This is by no means the only framework that can be used to understand treaty exit. Al- though domestic factors can be incorporated into this model, they are taken as simply a part of the institutional bargaining process. A deeper investi- gation at how those domestic features in uence the bargain is warranted. Moreover, questions about the eect treaty exit has on future cooperation (see, e.g., von Borzyskowski and Vabulas 2018) or how treaty exit may have indirect eects across treaty relationships are only touched on here. Where treaty relationships create a network of connections between states, treaty exit does not occur in a vacuum and may be aected by or cause changes in the network structure of the states's relationships (Dalton 2021). International cooperation, generally, appears to be under strain. Interstate war in Europe between Russia and Ukraine has shattered a sense of set- tled norms and behavior. Various commentators have noted the illiberal underpinnings of recent high prole episodes of treaty exit (Tooze 2018; Applebaum 2020). 44 In describing the work of Karen Stenner, Applebaum (2020, 16) notes that \[a]uthoritarianism appeals, simply, to people who cannot tolerate complexity." Perhaps the complexity of a globalized world addressing numerous challenges has lead to the appeal of authoritarian- ism, especially within democratic societies. Resentment about changing social and economic structures, the movement of economic and political power to metropolitan hubs, and envy over ascendant new power in once 44 Reid-Henry (2020) provides a history of the \reimagining" of the liberal order after the post-War World Golden Age and how the anti-liberal movements grew out of that history. 37 weaker groups can fuel these desires. Is the story of treaty exit one of ris- ing illiberalism, retrenched nationalism, and a rejection of complexity? von Borzyskowski and Vabulas (2019) nd evidence that treaty exist is not driven by high levels of nationalism; however, it does seem to at least be used as a narrative for those advocating for a withdraw from internationalism. 45 Although broad cooperation may be going through the crucible of renewed great power competition, challenges that face countries across the globe may allow space for cooperation to ourish on certain issues. If international co- operation is facilitated by powerful states (Kasner 1995; Gilpin 2001), then a disengagement by those powerful states{like the U.S.|will have signicant impacts on future governance of pressing global issues. Arguably, coopera- tion on eorts to mitigate the eects of climate change are a concern to all, and states|should they choose to address the climate crisis|will have to cooperate. On all these substantive challenges, international law will guide states in their eorts to cooperate. As explained in this project, treaty exit is one mechanism by which international law can guide behavior. Even after periods of disengagement, states can come together under the inter- national framework to repair and create new relationships. Nonetheless, understanding how this framework channels political goals is important to understanding cooperation for our past, present, and future. This seems to be both a timely and important area for continued investiga- tion. Although notices of withdrawal are \generally short, stylized letters of 45 von Borzyskowski and Vabulas (2022) revisit their ndings and address critiques of their earlier study to test \leader nationalism" as proposed by Choi (2021). They conrm their original ndings that international cooperation and nationalism can coexist. 38 two or three paragraphs" (Helfer 2012, 643), they have signicant impact on international cooperation. As Snyder (2018, 13) insightfully notes: \[A] his- tory of disintegration can be a guide to repair. Erosion reveals what resists, what can be reinforced, what can be reconstructed, and what must be recon- ceived." Designing exible institutions from the start is useful in beginning the cooperative relationship; yet, it may not always adapt to changing pref- erences and circumstances. Understanding the changing dynamics that can lead to withdrawal could be useful in on-going bargaining eorts to preserve institutions and avoid the costs of institutions falling apart. 39 Chapter 2 Bargaining, Bending, and Breaking: A Theory of Treaty Exit \[S]ince we do not belong to a world of law but only of the jungle law, the eect of [breach] would be the same as withdrawal" from the Treaty \because nobody is going to be able to enforce the [T]reaty against us." | US Senator J. William Fulbright, Chairman of the Senate Foreign Relations Committee 1 1 Quoted in Republic of the Marsh. Islands v. United States, 865 F.3d 1187, 1197{1198 (9th Cir. 2017). 40 2.1 Introduction Scholarship on treaty behavior recognizes a link between exit and breach (Helfer 2005; Morelli 2021), the aect of design on continued cooperation (Morrow 1994; Fearon 1998; Koremenos, Lipson and Snidal 2001; Rosendor and Milner 2001; Koremenos 2005, 2016), and the bargaining nature of co- operation in international institutions (Fearon 1998). Yet, many theories are ambiguous about their parameters and mechanisms or focus on subsets of institutions like international organizations (von Borzyskowski and Vab- ulas 2019; Gray 2020; Clark 2020), alliances (Leeds and Savun 2007), or courts (Johns 2015; McLachlan 2019). Moreover, our understanding is still in its early stages as empirical work gives us some foundation on which to build and rene our theories (Helfer 2005; von Borzyskowski and Vabulas 2019; Helfer 2019; Morelli 2021; von Borzyskowski and Vabulas 2022). Tak- ing an institutional-level view, here I provide a theory that synthesizes our understanding of \how international law works" in a fundamental sense to explain broad treaty behavior, including exit. The theory presented is pur- posefully general, seeking to be applicable to dierent institutions and fora of international cooperation. This theory distills aspects of international co- operation to their essential components so that they are able to stand in for any number of costs or benets that may be of interest in further research. This chapter proceeds by rst considering, in general terms, dierent types of treaty exit and how eorts to reform through bargaining have wide ap- plicability to dierent rationales. Next the chapter turns to describing the 41 concepts of exibility and obligation in terms of international cooperation and law before then presenting the institutional-level bargaining theory of treaty exit. In describing the theory, I highlight the interactions of exibil- ity and obligation for treaty behavior and then present a formal model to further rene the theory and make parameters explicit. After setting up the model, key aspects of the game are analyzed and propositions of interest are presented. Using the propositions and analysis to provide expectations, a series of case studies are used to understand the underlying mechanisms at work in the model and to assess the model's plausibility. The chapter con- cludes with an assessment of the model and expectations for more rigorous testing. 2.2 Types of Treaty Exit It is worth taking a moment to consider the range of possible reasons a state may choose to withdrawal from a treaty, at least in general terms. The range of possible reasons for treaty withdrawal is one of the appealing, and also daunting, aspects of this research. It touches on various interconnected aspects of international politics and law, opening many avenues to explore. Here, I consider a number of potential justications for treaty exit. 2 Obsolescence: An existing treaty becomes obsolete because the original purpose no longer exists, it has expired, or is simply not relevant. In short, 2 Morelli (2021) proposes a typology of treaty withdrawals that is similar but not identical. The rationales sketched here may provide a starting point for further research. 42 the treaty is no longer useful. 3 This may also capture treaty replacement, whereby a new treaty or treaty regime is created that abrogated or super- sedes an old treaty. In these cases, exit might be required (see GATT and WTO). Domestic pressure: A state may withdraw from a treaty if the treaty has become unpopular with an in uential part of the electorate; a treaty's unpopularity can take many forms including unpopular economic eects or feelings of imposition by the international community (possibly tied to nationalism). Treaties likely have fans and detractors from various segments of domestic society. Those cleavages may provide fascinating paths for future research. Bargaining failure: States may use the threat of exit when bargaining with other treaty members. As explained by Hirschman (1970), actors can use exit and voice to in uence decisions or policies in an organization. If the threat of withdrawal is successful, the state should be able to obtain policy concessions from other treaty members. If the state is unable to obtain a satisfactory level of concessions, then it must withdraw or lose all credibility. However, the exit itself will likely prompt the organization to action{thus, still achieving some eect. This reason should capture any scenario where the treaty members desire to continue cooperating and have the ability to negotiate. Rhetoric, diplomacy, and economic incentives are often used in parallel in cases of bargaining. 3 \Zombie" organizations, as described by Gray (2018, 2020) may fall into this category as well. 43 Fundamentally altered international environment: Treaties are cre- ated in certain international environments or orders. Because of this, they are often crafted to function within these orders and support interests based on this order. Yet, when the environment fundamentally changes (e.g. end of the Cold War or entry of China into the World Trade Organization) treaties may not properly t in the new environment given fundamental changes in interests, power, and alliances. These changes are generally exogenous shocks to the international system prompting some response. Principal-agentMisalignment: When treaties create treaty bodies, those treaty bodies, most importantly organizations, generally take on a life of their own. Organizations beget bureaucracies that develop expertise and competence, empowering the organizations in relation to their principals, i.e., the member states (Barnett and Finnemore 2012). Institutions may also suer from mission creep. When there is misalignment between princi- pal and agent, the principal may decide to leave the institution. Treaty inhibits other goals: A state may have a goal that is hindered or prohibited by the treaty. It is dicult to determine whether this is a distinct reason for treaty exit. If a state has some policy goal that is thwarted by a treaty obligation (e.g. missile defense), then there may be increased pressure on the state to withdraw from the treaty to free itself to pursue this goal. Punishment of other members: A state may withdraw from a treaty to punish other members or a treaty organization. This may be especially true 44 when the exiting state is a signicant contributor to the treaty institution or organization in funding, material, or moral weight. Figure 2.1: Types of Treaty Exit The general rationales can be synthesized into three main categories, illus- trated in Figure 2.2. First, and most simply, treaty exit can be ministerial or clerical, functioning to carry out the terms of a new or related agreement. Second, treaty exit may also be adversarial and used as a means to signal displeasure with another treaty member. Third, treaty exit may be a part of a state's eort to reform an international institution. Reforming an in- stitution may be purely to amend the institution to change the members' changed preferences. Yet, it also includes adversarial aspects of protest and clerical aspects of substituting institutions. Therefore, eorts to reform a treaty overlap with the other two motivations for withdraw, as shown in gure 2.2 where adversarial reform is labeled as protest and clerical reform is labeled as substitute. Thus, focusing on how treaty exit ts into a state's 45 eorts to reform an institution should provide an interesting and useful focus that touches on multiple rationales at once. Categorizing the types of treaty exit in this way also sheds light on whether an exit from a treaty will be publicized generally. On one hand, a protest exit is done in an adversarial manner, in part to pressure a response from the remaining treaty members. In this case, we should expect that the with- drawing state makes their intention to exit known widely. Diplomats may hold press conferences and leaders highlight the exit in speeches. Especially in cases where the state leaving the treaty accuses another of cheating or views the treaty as unfair, the global press and public will likely pick up on that message and amplify it. In the case of a clerical or substitute exit, we should expect much less publicity and drama. On the other hand, if the states are replacing a treaty, they may announce the replacement but the exits are less important|especially where future cooperation is assured. Other clerical treaty exits, either by operation of another treaty provision or some change in the law are generally less contentious and, thus, there is less for diplomats, politicians, journalists, and activists to comment on. These types of exit likely happen without much awareness among the public or even those in government not directly involved in the treaty relationship. Therefore, when the treaty exit is adversarial, it is made loudly. When treaty exit is clerical, it is made quietly. From this starting point, we can best understand treaty members' eorts to reform as part of a bargaining process. In an institutional context, a state's actions and preferences within a treaty institution would be most in uenced 46 by the obligations imposed by the treaty and the exibility of the state to change the status quo or deviate from the obligations under the treaty. The two dimensions of exibility and obligation capture crucial elements of a treaty regime and potentially in uence how a state will respond to changes in preferences. 2.3 Building Blocks of a Theory: Flexibility and Obligation The theory developed here relies on two aspects of international law and institutions: exibility and obligation. Both are high-level concepts that encompass a wide array of mechanisms related to treaty design and state behavior. Below, each concept is dened and explained, before moving to describing the theory. 2.3.1 Flexibility In the context of treaties and international law, exibility is the ability of a state or institution to deviate from some requirement or mandate to address changes in the environment or preferences. States incorporate exibility into treaties in various ways in light of the uncertainty of the state of the world in the future and with distribution problems (Downs and Rocke 1990; Lipson 1991; Koremenos 2001; Koremenos, Lipson and Snidal 2001; Koremenos 2016). The use of exibility tools in treaty making increases as states face 47 more dire distribution problems (Koremenos 2016). However, coordination problems require low levels of exibility, or high levels of rigidity, to be solved. States may incorporate exibility into the design of their treaties in various ways, including setting a nite duration for the life of the treaty and adding \escape clauses." These treaty provisions allow states to adjust to shocks to their cost and benet analysis. States also incorporate exibility by us- ing imprecise language and adding reservations when joining. When treaty language is imprecise or general, states are more free to oer up varying interpretations of acceptable behavior. \States like other legal actors, take advantage of the indeterminacy of legal language to justify indulging their preferred course of action" (Chayes, Chayes and Mitchell 1995, 12). And as (Simmons 2010, 277) notes: \[P]recision reduces the plausible deniability of violation by narrowing the range of reasonable interpretations." Addition- ally, imprecise language generally leads to more delegation (as is the case in domestic settings by leaving interpretation to a judge or jury) to foster compliance. Finite duration provisions are useful to allow for later renegotiation after the parties have gained more information about the costs and benets of a treaty. However, the costs of renegotiation increases with the more treaty partners trying to negotiate (Koremenos 2001). Escape clauses are calibrated by the treaty partners to avoid opportunistic use by adjusting notice and waiting periods (Koremenos 2016). As in the case of the U.S. announcement of its intention to exercise its right to withdraw, the U.S. was required to wait 48 a certain number of years before it could do so. In that case, the parties incorporated the wait period to give the other parties a chance|if they so desired|to try to prevent the withdrawal. Also, exibility is essential in managing a state's obligations in and behavior mandated by a treaty commitment. Along with the other methods used to increase exibility, informality is also a tool states can use to shape behav- ior exibly. Informality in punishment is preferable to states when there is a wide range of preferences and economic power asymmetry among the members (Koremenos 2016, 228, 239{248). Informality as a supplement to the explicit treaty instruments and design elements increases exibility. In- formality is no doubt a part of how states regulate treaty behavior, but it can be dicult to observe. Nonetheless, exibility is needed because treaties create obligations for states. 2.3.2 What is obligation? In law, an obligation is \a legal or moral duty to do or not do something" (Black et al. 2004, 1104). Put another way, an obligation is specic conduct for a specic matter. As one commentator noted in the late nineteenth century: There can be no doubt as to the general rule that should be ap- plied to Treaties. They are the solemn promises of nations, and must therefore be held binding, unless the contracting parties are released by mutual consent. We cannot in the interests of morality and peace insist with too much emphasis upon this. All 49 jurists are of one mind in laying down the general principle on this question (Richard 1877-1878, 91). As explained by Cassese (2005), there are two, overlapping conceptions of obligation in international law. One is traditional and one more modern, arising out of the post-World War global order. Traditionally, obligations are owed as between two states, and that obliga- tion is based on reciprocity. This is sometimes described as \synallagmatic". This traditional conception of obligation comports with the horizontal na- ture of the international system, whereby states have equal sovereignty| albeit unequal power|and aorded equal treatment by the law. \Interna- tional rules, even though they address themselves to all states (in the case of customs) or group of states (in the case of multilateral treaties) confer rights or impose obligations on pairs of states only. As a result, each state has a right or an obligation in relation to one other State only" (Cassese 2005, 14). Under the traditional view, states who are members of multilateral treaties are thought to owe obligations, and have rights against, each member. In a way, a multilateral treaty is a network of many bilateral relationships. Breach and punishment under the traditional view of obligation is made by the state to which an obligation was owed; however, this is largely based on power dynamics. Of course, this is where the international system most starkly diers from the domestic system, where a member of the community (the \people" or state) can bring punishment. Exceptions existed for cer- tain violations like piracy or riparian freedom of navigation (see universal 50 jurisdiction). Moreover, treaty obligation likely was motivated by various factors in earlier periods of international relations (Hertz 1990-1991). Beginning in the aftermath of World War I, and blossoming after World War II, a new conception of community obligations has begun to take root in the international system. These obligations are those that are owed by all to all without the condition of reciprocity. These community obliga- tions primarily are based in certain issue areas of paramount importance to the international community, including human rights, prohibitions against genocide and aggression, and some labor protections. Community obliga- tions generally apply when the following conditions are met: (1) obligations are for the protection of fundamental values (e.g. peace, human rights, self-determination); (2) obligations are erga omnes, i.e. owed to all (or to contracting states for multilateral treaties|erga omnes contractantes); (3) correlative right that belongs to any state (or contracting states for multilateral treaties); (4) the right may be exercised by any other (contract- ing) state, whether or not it has been materially or morally injured; and (5) the right is exercised on behalf of the whole international community (or community of contracting states) to safeguard fundamental values of this community (Cassese 2005). The concept of community obligations and rights is a cognate concept of jus cogens. However, community obligations are still relatively rare and this newer conception of obligation overlaps the more widely used traditional conception. Obligations in treaties, and norms more broadly, also consist of at least two 51 types. The rst is a primary rule, as explored by Hart (1961), where le- gal subjects are commanded to do or refrain from doing something. These can be thought of as the substantive rules. The second is a secondary rule, which is the modalities or means of fulllment of the obligations contained in primary rules. One can think of these as the procedural rules of the agree- ment or institution. These may be voting rules or other conditions places on the state members to carry out the purpose of the treaty agreement. An example of a secondary rule, which Cassese (2005) notes highlights the prob- lematic interaction of these types of obligations in the international system, is common Article 1 of the Geneva Conventions (1949). Common Article 1, found in the four Geneva Conventions, states: \The High Contracting Par- ties undertake to respect and to ensure respect for the present Convention in all circumstances." Thus, state parties must not only respect the rules of the Geneva Conventions but also must ensure other states respect them as well. The distinction between the types of obligations is sometimes referred to \substance" and \form" of an agreement (Guzman 2008). Obligation is a central part of our understanding of how international law functions. Obligation is one leg of the tripod of international legalization as described by Abbott et al. (2000). Scholars of international law and institu- tions have also recognized that international agreements sometimes create legally binding obligations and other times merely memorialize suggestions or aspirations. Raustiala (2005) argues that there is variation in the \legal- ity" of rules in international law, and Abbott and Snidal (2000) distinguishes between covenants and contracts. Johns (2015) recognizes the variation of 52 obligation, in particular with regard to how international court rulings can put normative and instrumental pressure on a member state. We are also obtaining a better understanding of how treaty obligation af- fects state behavior and commitments to treaty terms. Almeida and Pereira (2013) look at the degree and importances of obligations of treaties and ex- ecutive agreements, particularly where states sign agreements with no inten- tion to ratify. Using content analysis, they nd simply signing a treaty has been growing in importance in a state's obligations of \good faith" with re- gard to the treaty and an obligation of full responsibility for complying with executive agreements. Their ndings suggest obligation under international law is growing in both primary and secondary ways. Mulesky, Sandholtz and Zvobgo (2020) nd that demanding obligations in human rights treaties lead to a decrease in the likelihood that a state will ratify the treaty. States appear to be able to assess the level of commitment required of them before entering a treaty, 4 which corresponds with our understanding of how states use methods of exibility to adjust those commitments over the life of the treaty. Obligation and exibility interact and function in tandem. 2.4 Theory Given the almost ubiquitous appearance of clauses allowing withdrawal from treaties, and the other exibility provisions, to avoid breakdown of cooper- ation (Koremenos 2016), why do states ever choose withdrawal? Why do 4 An interesting extension of this research is to expand beyond ratication to determine how perceptions of obligation change over the life of the treaty. 53 states choose to withdraw rather than bargain for an accommodation or simply breach to meet their needs? To start, we must recognize that this is a phenomenon in uenced by many factors and circumstances. It is likely the case that characteristics of the member states, the goods provided, costs levied, exibility of the institution, geo-politics, and network of relationships in the international community all aect the likelihood of an institution un- raveling. In fact, as other authors have shown, it is exactly the job of well-designed institutions to prevent cooperation breakdowns. Nonetheless, the decision to withdraw is a strategic one made based on a process lled with an array of in uencing factors. The theory presented here focuses on the institutional factors and the bargaining process of the member states in providing an explanation of treaty exit. At the outset, it is important to note that this is a theory that seeks to apply to all types of treaties: bilateral and multilateral, economic- and security- focused, of great importance and insignicant. Treaties are agreements or \contracts" that function in the same, fundamental way regardless of their subject matter or consequence. What changes is the diering weights the parties to the agreement put on the benets and costs of the agreement. It is from these preferences that the importance of a treaty springs, not from the subject matter of the treaty itself. Bargaining models, like the bargaining model of war (Fearon 1995; Wagner 2000), are premised on the idea that some outcome, e.g. war, is costly and the relevant parties should bargain to avoid that outcome. As explained here, this logic applies to treaty withdrawal. Here, some shock increases the 54 costs or changes the value of a treaty for a state. The party bearing the costs from the shock can choose to bear the costs or negotiate an accommodation from the other party. If the cost is great enough and an agreement cannot be reached, then the state can withdraw from the treaty to avoid that added cost or decreased benet. However, withdrawing, although largely permissible and legal, is costly. Withdrawing terminates a forum for the parties to address issues and pro- vide goods. By withdrawing, the state removes itself from the table, is not able to set the agenda in the future, and receives no club benets from the treaty. Moreover, the treaty's overall value could decrease from losing a contributing member who can bear some of the load or from losing a partic- ipant in a regime. For example, this can be as direct as the loss of funding or other material support for the institution's mission. Also, withdrawing likely comes with some reputational costs in that states who withdraw often or without good cause may be viewed suspiciously at the next negotiation. This has been exemplied by the reaction of members of the Paris Climate Accord after the U.S. announced its withdrawal. States appear to antici- pate this reputational cost when they routinely give some justication for the withdrawal, e.g. another party's breach or court's overreach, even when one is not required. There is political value in pointing to another party's breach in announcing withdrawal, rationales sanctioned by international law (See, e.g., VCLT). Trachtman (2014) argues that withdrawal cannot be costless if delegation to an international organization is to have any real meaning. If withdrawal is costless, then the binding nature of international law appears 55 compromised. In an institutional context, a state's actions within a treaty institution are most in uenced by the obligations imposed by the treaty and the exibility of the treaty to allow a state to change the status quo or deviate from the obligations under the treaty. The two dimensions of exibility and obli- gation seem to capture crucial elements of a treaty regime and potentially in uence how a state will respond to changes in preferences. Figure 2.2 il- lustrates intuitions regarding exibility and obligation and their interaction with regard to treaty actions. Figure 2.2: Intuitions About Flexibility and Obligation Inflexible Flexible Low High OBLIGA TION FLEXIBILIT Y Status Quo Negotiate Exit Status Quo Breach Negotiate Breach Negotiate When exibility is low, a state should choose to exit. Where the treaty is too rigid, the state cannot benet from minor deviations or negotiate to address its changing preferences. If the treaty is highly exible, then the state would 56 likely take advantage of that exibility and breach its obligations. However, examining exibility on one dimension provides an incomplete picture. With obligation, a state would choose to exit where the burden is exceedingly high. In such a situation, remaining in the treaty would be too costly. Where obligations are low, the state generally would not have any incentive to change the status quo. As the obligation increases, the state would rst prefer to negotiate, until the point where the obligation is severe enough for breaching to be more attractive. When these dimensions are combined, as shown in Figure 2.2, various inter- actions become visible. Where obligations are low, a state prefers the status quo. In such a case, a treaty may fall into obsolescence or become a \zom- bie" (Gray 2018, 2020). That is likely true for a range of exibility because the stakes are so low. As those stakes increase, a state should be incen- tivized to take action to change its position. Where the treaty is in exible, breaching would be more attractive. When exible, the state would attempt to negotiate for accommodations. When obligations are high and exibility low, the state will choose to leave the treaty by exit. In this case, the burden of remaining is intolerable and the success of bargaining or breach is either improbable or impossible. To be clear, negotiation includes a wide range of behavior. It includes formal dialogues with other treaty members to address concerns, as well as other methods of addressing concerns. States can make declarations or reserva- tions, as well as withdraw the same. Specically, reservations can be used 57 to mitigate the obligations of a treaty (See Galbraith 2012). These are often done unilaterally, so this is a strong version of negotiation close in nature to exit but less drastic. Here, I examine bargaining broadly dened as negoti- ations to adjust the status quo to meet a change in party preferences. To build a theory as to what factors are most important in driving a de- cision to withdraw, incorporating the intuitions regarding obligation and exibility, I use formal modeling and qualitative case studies to provide guidance on what variables are key. Formal modeling is useful here where the decision is inherently strategic and builds on the international politics literature which focused on this outcome in limited ways. Generally, it is reasonable to model state decisions to cooperate, breach, or exit as rational because these decisions are generally made after substantial consultation and debate by policymakers. 5 Moreover, in considering actions like whether to exit, policymakers consider the responses of other actors, namely other treaty members to the contemplated action. Modeling a theory in this way will allow for assumptions and variables to be explicitly presented and their relationships understood. In addition, using case studies will ground the abstract theory in empirical evidence to rene and assess the plausibility of the mechanisms at work in the theory. 5 In keeping with the general practice in the literature, this model assumes that the state risk-adverse. 58 2.4.1 Theoretical Model: This model focuses on a bargain between two states in light of some exoge- nous shock within a treaty framework. The institutional parameters I am interested in are what I broadly call treaty obligation and exibility, which are incorporated into the model in the net value of the treaty to each player and the likelihood that a breach would be punished. Model Set-up: This is a game between two players, P 1 and P 2, which are both in a treaty. 6 Although this is a two player game, a player here can represent a number of members of a treaty as a unied group. 7 Each player's preference of actions is simply based on weighing the benets and costs of remaining in the treaty. The benets may include arms reduction, an alliance, new trade opportunities, or reputational benets. The players also bear various costs of the treaty, which may include the direct monetary costs in sustaining a treaty organization, being subject to an international court's jurisdiction, injury to a domestic business sector, adjustment costs to legal obligations, or opportunity costs. That balancing of benets and costs give the players a net value of being a member of the treaty. The total net value of a treaty (V ) is composed of the individual net value of the treaty to each player, i.e. V =v 1 +v 2 . The order of play, illustrated in Figure 2.3, is as follows. Nature chooses some shock (e), 0e, to the net values of the players where Player 1's (P 1) 6 This models an interaction within a multilateral treaty as noted below. 7 Of course, it is possible for factions to form with varying preferences. 59 net value decreases and Player 2's (P 2) net value increases. 8 This shock can be thought of as changes in a domestic environment, electoral pressures, eco- nomic distress, a change in the security environment, or any other change of circumstances that stresses the obligations of Player 1 (P 1). Here, the shock is modeled as a long-term shift in value, building on the general assumption that fundamental shifts in preferences drive state decisions to leave a treaty (Koremenos 2016). If the shock was temporary, which is plausible, the state would face more complicated tradeos. Figure 2.3: Game Tree and Payos v 1 − e v 2 + e + δ(v 2 + e − I) v 1 − e + x + δ(v 1 − e + x) v 2 + e − x + δ(v 2 + e − x) v 1 − e + δ(v 1 − e) v 2 + e + δ(v 2 + e) v 1 − e + x(1−q) − k(q) + δ(v 1 − e) v 2 + e − x(1−q) − s + δ(v 2 + e) v 1 − e + x + δ(v 1 − e) v 2 + e − x + δ(v 2 + e) N e x Accept Reject Breach Withdraw Cooperate Punish No Punish P1 P1 P2 P2 Each player receives its net value of the treaty and the shift in costs now and in the future. This is modeled as a two-period utility for simplicity. To 8 This could be expanded to allow the total value of the treaty (V ) increase or decrease as the result of a shock. The shock modeled here is a redistribution of value among members. 60 account for the state's weighting of future value of the treaty, that future value is weighted by a factor (), where 0<. This factor can approximate whether the states value the treaty more now than in the future. Where < 1, the state is fairly short-sighted, viewing the pay o today more than the future. Where > 1, the state sees the payo as expanding into the future and having long-term value. In that case, the state might believe that a treaty will be around in the future and still providing benets. There are reasons we might think states are short-sighted, focusing on pressing and immediate issues, like being reelected, economic development, appeasing an unruly domestic faction, or on-going con ict. States have long time horizons, but the individuals who make policies have more immediate goals on the whole. In light of the exogenous shock (e), Player 1 (P 1) makes an oer (x), where 0 x, to Player 2 (P 2) to shift Player 1's share of the treaty's obligation. Player 1 has a constraint on how much it can demand ( x), in one sense being limited by the total value of the treaty (V ). Player 2 (P 2) chooses to accept or reject the oer. If it accepts, a bargain is made and the oered shift (x) is subtracted from Player 1's total costs and added to Player 2's costs into the future. 9 If Player 2 (P 2) rejects, then Player 1 (P 1) chooses one of three options: to continue to cooperate in the treaty notwithstanding the added costs or lower benets from the shock; to withdraw from the agreement avoiding costs but losing the treaty benets; and to breach the agreement to unilaterally shift 9 Again, this models a shift in the distribution of the treaty's value between the players. 61 costs (x) without a bargain. If Player 1 cooperates, the game ends and the players receive their payos with the shift from the shock (e). If Player 1 withdraws, the game ends and the players receive payos without the bene- ts and costs of the treaty in the future. However, Player 2's net value of the treaty|which remains if it is a multilateral institution 10 |is decreased by Player 1's \importance" (I), 0 0, and the proposal is less than Player 1's discounted importance to the treaty,x 0, and the proposal is not greater than Player 1's discounted importance, x < I (1+) , and the costs of uncertain punishment to Player 2, x< s q+ . As punishment becomes more certain, q! 1, the less 11 In extensions of the game, this may be relaxed and long-term eects of breach will be modeled. 63 room there is for a bargain because s q+ decreases as q increases. Player 1 makes a proposal, x, that is greater than the discounted value of the treaty with the shock,x> (ev 2 ) 1+ , and greater than the cost of being punished, x> kq (+q) . Because, as dened, any positive oer will be greater than the cost of being punished (negative), the second condition will be satised. Thus, the parties will reach a bargain as long as Player 1 proposes a shift in costs that accounts for the future costs of the shock and is less than the importance of Player 1 to the treaty and costs of punishing a breach. Having established that a bargain is possible under the right conditions, we can now look to circumstances that will prevent a bargain and lead to treaty exit. The most straightforward scenario is one in which the shock shifts the costs to Player 1 in such a way that those new burdens outweigh the present value of the treaty to Player 1. Proposition 2. If e > v 1 , the bargain will fail where the shock is greater than Player 1's treaty importance e > I +v 1 and successful punishment is so certain to restrict the range of acceptable proposals to Player 1, i.e., q! 1, s q+ decreases. When the shock is greater than the value of the treaty to Player 1, e>v 1 , Player 1 will need to make a proposal to account for the shock. However, for Player 1 to be satised with a new deal, the proposal will exceed Player 1's importance to the treaty and Player 2's costs to successfully punish a breach. In that case, Player 2 will reject the oer, and Player 1 will choose between breaching or withdrawing, contingent on the probability of successful pun- ishment by Player 2. As the likelihood that punishment will be successful 64 increases, q! 1, the more costs Player 1 would suer for breach and the more Player 2 prefers to withdraw as opposed to breach. These propositions lead to the following broad hypotheses regarding when we should expect to see bargain fail and a treaty to experience a treaty exit. First, the probability of a successful punishment (q) is a form of treaty exibility, where a treaty is more exible when the risk of a costly and successful punishment is low. In that case, Player 1 adjust to the shock unilaterally without the bargain. Inversely, an in exible treaty can make unilateral adjustments infeasible. Thus, increasing the probability of suc- cessful punishment may decrease the bargaining range and push a state to favor exit. Claim 1. A treaty that is less exible to allow for a breach will be more likely to experience an exit, all else being equal. Another aspect of the bargaining process highlighted in the model is how the importance of the treaty member experiencing the negative eects of the shock sets the boundary of the bargaining range between the treaty members. Where the member is important, we should expect the other member to be willing to accept a larger proposal in the bargain. Thus, more drastic reforms are possible to address more substantial shocks. Claim 2. A bargain to account for a shock should be more feasible where the treaty member is important to the value of the treaty, all else being equal. Although this hypotheses is intuitive to our understanding of bargaining and collective action, it is useful to make this expectation explicitly. It also 65 captures variation of outcomes among dierent treaty members within a single treaty framework. Where a member has less importance, the other members may not be willing to accept even a modest reform proposal. Finally, the distributional costs and benets generated by the shock will likely lead to treaty exit being a preferred option. This aspect of the game captures perceived imbalances of fairness among the treaty members. When the costs shift and fall more heavily on one member, that member might resent being stuck with a worse deal based on no or little fault of that treaty member. As mentioned in the introduction, U.S. President Trump highlighted the perceived unfairness of the the Paris Climate Accord based on obligations on other countries like China and India. Claim 3. A treaty that becomes relatively more obligatory compared to other members will be more likely to experience an exit, all else being equal. In this model, the distribution of costs among the players is the focus. Yet, as mentioned above, the shock could aect the total treaty value (V = v 1 +v 2 ). In doing so, both player may experience an increase or decrease in value, which would certainly aect their preferences to bargain, breach, or exit. Further, although not modeled here, we can posit that the baseline obligation of a treaty will likely determine how sensitive the treaty members are to the shocks. If the treaty is already obligatory and onerous, a small shift may be enough to prompt bargaining or exit. 66 2.5 Alternative Explanations Are other explanations possible? Yes, there are other possible reasons a state decides to withdraw from a treaty (See Helfer 2005; von Borzyskowski and Vabulas 2019; Helfer 2019; Morelli 2021; von Borzyskowski and Vabulas 2022). Many explanations can be incorporated or explained as either subsets of the bargaining model presented here or some variation on the model. The institutional-level theory here is capable of incorporating a number of mechanisms into its parameters. Nonetheless, some explanations may not t neatly into this theory. There are numerous possible explanations but a few of note are presented here. One explanation for treaty withdrawal could simply be individual leader preference. It could be the case that each U.S. president simply approached international cooperation in a unique fashion. One might argue that Donald Trump was an aberration or so unique that the U.S. exits under his watch should be excluded from analysis. This explanation con icts with research on state behavior and common sense. Although each leader brings dierent background, preferences, and style to their decisions regarding a treaty, those decisions are still made in the context of a deliberative domestic institution. In domestic politics, it could be that the spread of illiberalism is under- mining eorts at international cooperation. von Borzyskowski and Vabulas (2019, 2022)'s ndings challenge the assertion that nationalism is driving treaty exit. However, the networks of advocates pushing illiberal policies 67 (Applebaum 2020) may be having an eect in a way we have yet to mea- sure and test. There may be normative shifts away from internationalism and cosmopolitanism, which may be distinct from nationalism per se. This might be captured by the model through changes in how states value a treaty; however, further theorizing and analysis is needed. In addition, a state's web of connections with the international community might also drive decisions to leave a treaty relationship. On one level, a state's economic integration with the international community likely has a signicant eect on its desire to remain in the treaty (See Lupu 2016). In addition, a state's position within the network of treaty relationships, along with the parallel and indirect connections to other states, might in uence a state's decision to decouple one or more relationships (Dalton 2021). Where a state is well-connected in the network, possibly being more central to the community, it may have many options to substitute the benets of a treaty. In such a case, treaty exit may be, at least marginally, more likely. These network based theories could be incorporated or extend the bargaining model. Pure power dynamics may also play a role in a state's decision to with- draw from institutional membership. Power dynamics play a key role in agreement formation (Koremenos 2016). If a state has signicant power in the international system, for example the United States or China, it may be more willing to withdraw from an institution being that it is more self- reliant and assumes it will not suer great consequences. Great powers tend to have outsized in uence in international aairs and may not see the rules 68 as applying to them (See generally Simpson 2004; Baron 2014). Similarly, pariah states may make a similar calculation regarding the consequences of its withdrawal, believing it already sits on the outside of the interna- tional system (Simpson 2004; Lawal 2012; Crampin 2020). A pariah state, like North Korea, may believe that it is already shunned by the world and thus would not have much more to lose by withdrawing from an institution. Moreover, its audience costs are likely lower for doing so, having to only appease its small constituency and a limited range of benefactor states. For these states, reputations will be aected to a lesser degree as being tem- pered by their importance or lack thereof to the international community. Additionally, goods provided by the agreement may be more or less crucial for these states on the extremes. In describing great powers and \outlaw" states, Simpson (2004, 5) explains \states can be dierentiated in law ac- cording to their moral nature, material and intellectual power, ideological disposition or cultural attributes." This is essentially a state's standing in the international community. This may manifest itself in the calculus pre- sented herein in a variety of ways but further extensions can model it more explicitly. Finally, it is possible that there is no general, systematic reason for treaty withdrawal. It could be that subsets of treaty exit share similar reasons but there is no overarching theory that can explain them all, or even most of them. This position seems dubious, but the plausibility of the bargaining model presented here can be assessed by examining a few episodes of treaty withdrawal. 69 2.6 Cases To understand the mechanisms at work and assess the plausibility the the- ory, four short case studies are used as preliminary examinations of the parameters of the theory. Table 2.1 summarizes the short case studies. Us- ing these case studies as plausibility probes allows me to determine if the key mechanisms of the model match the key mechanisms at play in the examined cases. It will also allow for potential omitted variables and hypotheses to be uncovered and generalization made from the examined case to a potential class of cases within which the examined case belongs (George and Bennett 2005). The cases here include most-likely cases of withdrawal. The rst case pre- sented is Japan's failed attempts to rst bargain for reform of the Interna- tional Whaling Commission, its breach of is obligations, and its ultimate exit when the regime proved too rigid and burdensome for continued mem- bership. The second case presented covers the two episodes of cascading exit from United Nations Industrial Development Organization (UNIDO) by a number of western democracies after the end of the Cold War and also leadership concerns in the organization. Third, the case reviewed is that of North Korea's withdrawal|or attempted withdrawal{from the Nu- clear Non-Proliferation Treaty (NPT). Again, these cases would be most likely cases where it was widely understood that the costs of remaining in the treaty increased substantially given new circumstances for the member states. The cases are selected as events of withdrawal that have variation 70 in a number of independent variables of interest in the literature, namely regime type, treaty subject matter, and regional variation. With each case, I review the background and purpose of the treaty; set out the key provi- sions and obligations of the treaty as provided in the treaty text; highlight the state's characteristics, like regime type, economy, and standing in the international community; and explain the circumstances leading to a state's decision to withdraw from the treaty. A nal short summary of the announced and then rescinded withdrawal from the Universal Postal Union by the United States is informative where the bargaining process over reform was a signicant feature of the episode and was ultimately successful, from the standpoint of preserving the U.S. membership in the UPU. Although the bargaining occurred in earnest after the U.S. announced its withdrawal, it may still provide insights into the bargaining framework. Table 2.1: Summary of Case Studies Withdrawal Event Treaty Entered into Force Withdrawal Date Japan from ICRW ICRW Nov. 10, 1948 December 26, 2018 Canada, U.S., and Australia from UNIDO UNIDO June 21, 1985 1993{1997 Cascading Exit from UNIDO UNIDO June 21, 1985 2012{2016 North Korea from the NPT NPT March 5, 1970 January 10, 2003 U.S. from Universal Postal Union UPU Constitution October 9, 1874 October 15, 2018 These preliminary case studies do not serve as a true test of the theory where the theory was informed by a sample of well-known cases of treaty withdrawal (including these cases), i.e. cases selected on the dependent variable. Nonetheless, these preliminary studies can be used to identify common patterns and determine whether the theory is plausible, without 71 the outcome variable needing to vary (George and Bennett 2005). This exploratory exercise looks at whether the variables of interest \contributed to or added weight in favor of a specic outcome in a particular case" (Goertz and Mahoney 2012, 109). This diverse set of cases can illuminate the range of explanatory variables that may aect withdrawal, likely being minimally representative of the variation in the larger population of withdrawal events (Seawright and Gerring 2008). Given the general intuition that withdrawal is rare, it is fully acknowledged that these may be outlier cases, unrepresentative of a broader population of withdrawal case studies. It may be that exiting the Rome Statute is fundamentally dierent from Mauritania's 2006 withdrawal withdrawal from the Agreement for the Establishment of a Commission for Controlling the Desert Locust in North-West Africa or Canada's 2011 withdrawal from the World Tourism Organization. Perhaps the notoriety or importance of an institution signicantly weighs on a state's decision to withdrawal, such that lesser know institutions are easier to exit because there is less backlash for doing so. Also, given the ease of nding some recent cases, it may be the case that withdrawal occurs more often than expected|albeit still being rare overall. These theory-building case studies primarily involve relatively well-known, signicant treaties. However, until a more rigorous examination of the prevalence of the phenomenon can be accomplished, these types of case studies can assist in building out the theory. 72 2.6.1 International Convention for the Regulation of Whal- ing The International Convention for the Regulation of Whaling (ICRW) is a treaty that was signed in Washington D.C. on December 2, 1946 to \provide for the proper conservation of whale stocks and thus make possible the orderly development of the whaling industry." 12 The ICRW established the International Whaling Commission (IWC) (ICRW, art. III). As the IWC describes itself: \The IWC is the global body charged with the conservation of whales and the management of whaling." 13 Member states also implement domestic legislation to restrict whaling. 14 Peoples have practiced whaling since as early as 2000 BC. The modern industry of whaling developed in the 19th century. That industry was pri- marily driven by the collection of oil from the animals. Over the course of time, the interests of states with whaling industry has changed over the years, often changing with economic demands and changing social attitudes. The United Kingdom and United States had robust whaling industries in the 1800s. Yet, those industries became unprotable with new technolo- gies. Further, concerns of overshing, driven by technological advancements, added to concerns regarding the practice. As Fitzmaurice and Fitzmaurice 12 International Convention for the Regulation of Whaling, with Schedule of Whaling Regulations, 2 December 1946, 62 Stat. 1716, 161 UNTS 72. 13 https://iwc.int/home 14 For example, in the United States, whaling is restricted by the Marine Mammal Protection Act of 1972 (MMPA), the Endangered Species Act of 1973 (ESA), and the Whaling Convention Act of 1949 (WCA). 73 (2015) explain, technology was and has been a driving factor in the reg- ulation of whaling. Whaling was originally regulated through land-based operations when the processing of the oil was done at port. However, with technological advances that allowed for the processing to occur on ships, regulation became more dicult and the speed of whaling increased. The ICRW consists of eleven articles and originally included a protocol of three articles to extend the application of the convention to helicopters and other aircraft and includes additional inspection methods. Article IX man- dates: \Each Contracting Government shall take appropriate measures to ensure the application of the provisions of this Convention and the punish- ment of infractions against the said provisions in operations carried out by persons or by vessels under its jurisdiction." State parties can withdraw from the agreement on June 30th of any year by giving notice before January 1 of the same year|a six month wait period (ICRW, art. XI). Interestingly, any other party may also withdraw in response within one month of receiving notice of the rst state's withdrawal, taking eect also on June 30 (Id.). There has been consistent disagreement over the scope of the ICRW. The 1946 Convention does not dene a `whale' along with other terms (See ICRW, art. II). Some members of the IWC claim that it has the legal competence to regulate catches of only great whales (the baleen whales and the sperm whale). Others believe that all cetaceans, including the smaller dolphins and porpoises, fall within IWC jurisdiction. The IWC's eectiveness has oft been criticized on a number of grounds. 74 First, any member countries are free to simply leave the organization and declare themselves not bound by it if they so wish. Second, any member state may opt out of any specic IWC regulation by lodging a formal objec- tion to it within 90 days of the regulation coming into force (such provisions are common in international agreements, on the logic that it is preferable to have parties remain within the agreements rather than opt out altogether| oering exibility). 15 Third, the IWC has no ability to enforce any of its decisions through penalty imposition. Despite these potential shortcomings, an analysis by the Carnegie Council determined that while the ICRW has had \ambiguous success" owing to its internal divisions, it has nonetheless \successfully managed the historical transition from open whale hunting to highly restricted hunting. It has stopped all but the most highly motivated whale-hunting countries. This success has made its life more dicult, since it has left the hardest part of the problem for last." 16 On July 23, 1982, members of the IWC voted by the necessary three-quarters majority to implement a pause on commercial whaling. The 1982 Mora- torium amended the Schedules calling for \catch limits for the killing for commercial purposes of whales from all stocks for the 1986 coastal and the 15 \[Any] amendment [to the ICRW] shall become eective with respect to all Con- tracting Governments which have not presented objection but shall not become eective with respect to any Government which has so objected until such date as the objection is withdrawn." Id., para. 3. Contracting states also may withdraw from the organization. Id., Art. XI. 16 Ian Hurd. \Almost Saving Whales: The Ambiguity of Success at the International Whaling Commission [Full Text]." Ethics & International Aairs. (March 2012), available at https://ethicsandinternationalaairs.org/2012/almost-saving-whales-the-ambiguity-of- success-at-the-international-whaling-commission-full-text/ 75 1985/86 pelagic seasons and thereafter shall be zero." 17 The measure passed by 25 votes to seven, with ve abstentions. The seven countries voting against the measure were Brazil, Iceland, Japan, Norway, Peru, South Ko- rea and the USSR. As the moratorium applies only to commercial whaling, whaling under the scientic-research and aboriginal-subsistence provisions of the ICRW is still allowed. Further, the IWC has, in recent decades, consisted of two blocks of states: pro-whaling and anti-whaling states. Pro-whaling view the purpose of the ICRW as creating a framework and protections for responsible whaling. Anti-whaling states view the current purpose of the regime as protecting whales and restricting hunting altogether. These two factions of states have lobbied to shape the institution to their interests. In the lead up to the 1982 Moratorium, anti-whaling states recruited non-whaling, sometimes land-locked, states to join the ICRW and increase the vote share. States like Austria made clear that it was joining the institution to abolish commercial whaling and transform the ICRW into a preservation agreement." 18 The IWRC and IWC are interesting in the context of treaty exit because the institution has existed for a signicant period of time, undergone intense negotiations about its reform by the members, and has seen numerous states \breach" their commitments or withdraw. As such, the institution should provide insights into how bargaining between members of an international 17 International Whaling Commission Schedule, para. 10(e). 18 (Kobayashi 2005, 208) 76 institution choose between reform, breach, and exit in the context of the institutions exibility and obligations. Eight states have withdrawn from the Convention since its ratication: Canada, Egypt, Greece, Jamaica, Mauritius, Philippines, the Seychelles and Venezuela. Belize, Brazil, Dominica, Ecuador, Iceland, Japan, New Zealand, and Panama have all withdrawn from the convention temporarily but ratied it a second time. The Netherlands, Norway, and Sweden have each withdrawn from the convention twice, only to have accepted it a third time. 19 Japan is the most recent member to depart, in January 2019, so as to resume commercial whaling. The case of Japan's recent exit will be reviewed here. Japan adhered to the ICRW on April 21, 1951, the same year the U.S. occupation ended after World War II. Japan's willingness to join the ICRW was done in part to rehabilitate the country's international reputation and signal to the international community that it was a reliable partner (Holm 2019, 5{6). Thereafter, Japan expanded its whaling capabilities, mostly destroyed during the war, to possess one of the largest whaling industries in the world (Holm 2019, 6). Even though Japan sought to grow and protect its whaling industry, it was willing to cooperate to implement monitoring and conservation measures. In 1975, Japan and the United States entered into an agreement concerning an international observer scheme for whaling 19 https://2009-2017.state.gov/documents/organization/191051.pdf 77 operations from land stations in the North Pacic Ocean. 20 As explained above, the anti-whaling block of states was successful in imple- menting a moratorium on commercial whaling. After the 1986 Moratorium, Japan pursued the practice of whaling under the exception for scientic re- search. 21 The Institute of Cetacean Research (ICR) in Japan was formed in 1987 to facilitate these scientic endeavors. Japan had been a target of pressure by the other treaty members. Those members engaged in pre-moratorium \soft" and post-mortatorium \hard" shaming of Japan's behavior and policies (Kolma s 2021). In particular, Aus- tralia had sent sheries and customs patrol ships to the Southern Ocean to collect information of Japan's whaling activities there for use as evidence. 22 Australia later brought a case against Japan at the International Court of Justice (ICJ) on March 31, 2010, challenging Japan's scientic whaling aci- tivites which Australia argued were in breach of the ICRW. In its March 31, 2014 judgment, the ICJ ruled against Japan and ordered it to halt its whaling. The Court rejected Japan's argument that its whaling activities were for \scientic research," allowable under Article VIII of the ICRW. Because it found Japan to be in breach of its treaty obligations, the Court ordered Japan to revoke any authorization or permit to kill whales and refrain from granting any permits under Article VIII. The ICJ ruling 20 Agreement between the United States of America and Japan concerning an inter- national observer scheme for whaling operations from land stations in the North Pacic Ocean, Jp-U.S., May 2, 1975, 998 U.N.T.S. 313. 21 ICRW, Art. VIII(1). 22 Majority of Japanese support whaling -poll." Reuters (Feb. 8, 2008) available at https://www.reuters.com/article/idUST286484 78 against Japan conrmed to Japan that the moratorium would not be lifted (Holm 2019). Japan announced that it would resume whaling in 2015. 23 Prime Minister Shinzo Abe explained to a committee of the National Diet that he intended to resume commercial whaling, and \[t]o that end, I will step up eorts further to get understanding from the international community." 24 Yoshi- masa Hayashi, Japan's sheries minister, even promoted a \whale week" campaign to raise awareness among the Japanese people about this aspect of their culture. 25 In January 2017, a helicopter photographed the Japanese whaling vessel Nisshin Maru engaged in whaling activities in the Australian Whale Sanc- tuary. Upon seeing the helicopter, the crew members attempted to cover up the carcass of a freshly-killed minke whale on its deck, a seeming acknowl- edgement of violating the treaty obligations. On September 14, 2018, the Japanese delegation at the 67th meeting of the IWC in Florian opolis Brazil, proposed a reform to the organization, which would allow the IWC to set future catch quotas for non-endangered whale species. Japan also sought a rule change regarding the required vote share for IWC decisions, moving from a three-quarters majority to a simple majority. Doing so would allow for easier changes to the treaty obligations 23 \Japan to resume whaling in Antarctic despite court ruling." BBC News (Nov. 28, 2015) https://www.bbc.com/news/world-asia-34952538 24 Agence France-Presse in Tokyo. \Japanese PM admits seeking resump- tion of commercial whaling." The Guardian (Jun. 9. 2014) available at https://www.theguardian.com/environment/2014/jun/09/japan-pm-commerical- whaling-shinzo-abe-antartic-hunt 25 Id. 79 and redistributing power in the organization between the two blocs of states. The Japanese proposal was roundly rejected (41 in favor, 27 against) by the other members. In response, Japan's Vice Minister for Fisheries, Masaaki Taniai, ruminated on Japan's future participation in the organization. Japan notied the treaty members of its withdrawal from the ICRW on De- cember 26, 2018, little more than two months later. One rationale given for the timing of the withdrawal was the view that the Trump Administration in the United States would not likely sanction or criticize Japan for the with- drawal (Holm 2019). With the Trump Administration's own withdrawals from numerous high prole international agreements and organizations, for example with Paris Climate Accord among others, the American adminis- tration may view other withdrawals as justiable or not a part of American interests. This is likely a misunderstanding of the U.S. view of treaty exit, which was focused on benets to the U.S. rather than a wholesale rejection of international cooperation generally. The reaction to Japan's withdrawal from other member states and other organizations was generally negative and condemnatory. Then U.K. En- vironment Secretary, Michael Gove, tweeted that he was \[e]xtremely dis- appointed" about Japan's exit and that the United Kingdom would \con- tinue to ght for the protection and welfare of these majestic mammals." 26 Australia's Foreign Minister, Marise Payne, also stated that the Australian 26 @michaelgove (Dec 26, 2018 5:33 AM) 80 government was \extremely disappointed" about the \regrettable" with- drawal. 27 After the withdrawal became eective, Marise Payne and Sussan Ley, Minister for the Environment, welcomed the end to whaling in the Southern Ocean but lamented Japan's withdraw and resumed commercial whaling, stating: \We continue to urge Japan to return to the Convention and the Commission as a matter of priority. The Australian Government's position on whaling has not changed. We remain opposed to all forms of commercial and so-called `scientic' whaling." 28 At the time, the New York Times Editorial Board wrote: \Withdrawing from the whaling commission for short-term political gain is a dangerous and foolish move, especially for an advanced country like Japan that has generally supported multilateral eorts on the environment." 29 It went on to criticize Japan's action as a \politically motivated decision," which, as ar- gued here, is accurate of any decision to exit a treaty. The Sydney Morning Herald's Editorial Board took a similar position of disapproval, compar- ing whaling to trophy hunting on elephants and rhinoceroses. 30 The Sea 27 Senator the Hon Marise Payne. \Japan to exit from International Whaling Com- mission and cease Southern Ocean whaling." Joint Press Release. (Dec. 26, 2018) avail- able at https://www.foreignminister.gov.au/minister/marise-payne/media-release/japan- exit-international-whaling-commission-and-cease-southern-ocean-whaling 28 Senator the Hon Marise Payne and Senator the Hon Sussan Ley MP \Japan withdraws from the International Whaling Convention" Joint Press Release. (July 2, 2019) available at https://www.foreignminister.gov.au/minister/marise-payne/media- release/japan-withdraws-international-whaling-convention 29 The Editorial Board. \Japan: Stop Slaughtering Whales." The New York Times (Dec. 31, 2018), available at https://www.nytimes.com/2018/12/31/opinion/editorials/japan-whale-hunt- whaling.html 30 The Editorial Board. \Australia must lead ght to save whales." The Sydney Morning Herald (Dec. 28., 2018), available at https://www.smh.com.au/environment/conservation/australia-must-lead-ght-to-save- whales-20181227-p50oco.html 81 Shepard, an environmental organization that worked to thwart Japanese whaling, 31 welcomed the end of Japan's \scheme to pose as researchers" and Japan's exit \means there can be absolutely no justication for hunting whales in an internationally established whale sanctuary." 32 Captain Paul Watson characterized Japan, along with Iceland and Norway, as \pirate whaling nations." However, some argued that by leaving the international agreement and giv- ing up the pretext of whaling for \scientic research" would be benecial to whale populations because the demand for whale meat was low and shrink- ing. 33 In July 2019, Japan resumed commercial whaling in its territorial waters. Japan's domestic laws were amended as it decided to pull out of the IWC. The revised laws supported whale consumption, including whale meat in school lunches, as a way to pass on culinary culture to future generations. 34 In the case of Japan, its eorts to bargain for a reform were rejected leaving it the option to live with the status quo, breach, or exit. Here, Japan rst attempted to skirt its obligations by using a \loophole" under Article VIII. 31 Sea Shepard was founded in 1977 by Captain Paul Watson, an original member of Greenpeace, in Vancouver, Canada. (\Who We Are: History" Sea Shepard, available at https://www.seashepherdglobal.org/who-we-are/history/) 32 https://seashepherd.org/2018/12/26/sea-shepherd-welcomes-the-end-of-whaling-in- the-southern-ocean/ 33 Dennis Normile. \Why Japan's exit from international whaling treaty may actually benet whales." Science. (Jan. 10, 2019), available at https://www.science.org/content/article/why-japan-s-exit-international-whaling-treaty- may-actually-benet-whales 34 The Sankei Shimbun. \Japan Moves Toward Revision of Whaling Law." (Nov. 20, 2019), available at https://japan-forward.com/japan-moves-toward-revision-of-whaling- law/ 82 Moreover, when that was unsuccessful, Japan moved to publicly breach the agreement until it later decided to exit. At that point, the ICRW was impos- ing too great of obligations on Japan|namely completely shutting down its whaling industry|and was not exible enough for Japan to cheat without repercussions indenitely. The ICJ ruling also removed any ambiguity in Japan's actions, clearly labeling Japan's actions a violation, and reducing the exibility of the regime. Therefore, exiting the institution became the preferred choice. 2.6.2 Cascading Episodes of Exit from UNIDO The U.N. Industrial Development Organization (UNIDO) was created in 1966 to address problems of industrialization in the developing world. Broadly, it promotes industrial development for poverty reduction, inclusive global- ization and environmental sustainability. UNIDO grew out of the work of the Industry Section of the U.N. Secretariat in the late 1950s and even- tually converted to a specialized agency of the U.N. in 1985. Its primary purpose \is to promote and accelerate inclusive and sustainable industrial development (ISID) in Member States." 35 The institution is governed by the UNIDO Constitution, adopted by 1984 after years of wrangling between the member states. The UNIDO Consti- tution entered into force in 1985. The agreement includes 29 articles and three annexes. Article 2 lays out in detail the functions of the organization, 35 \Who we are." UNIDO (2018), available at https://www.unido.org/who-we- are/unido-brief. 83 which is primarily to provide assistance and services to promote develop- ment. Contributions to UNIDO by members are apportioned as they are in the U.N., with a cap that prohibits any apportionment on a member that exceeds twenty ve percent. (Art. 15.) The Constitution extensively lays out the governance structure of the organization, which has been reformed a number of times based on criticism from members. Article 6 allows member states to withdraw from the organization, taking eect one year after noti- cation. The provision requires a state to continue to make contributions to the organization before the withdrawal becomes eective. As explained in article 5, a member state may be suspended by the organization, including for falling in arrears for two years on its scal contributions. The history of the institution has been characterized by intermittent strug- gles between developed and developing countries. These struggles may have lead to two signicant episodes of withdrawal from the institution. The rst episode occurred in the 1990s, shortly after the collapse of the Soviet Union and capitalist optimism at the end of the twentieth century. With a growing preference that development assistance should come from the private sector along with views loosely summarized in the Washington Consensus, a num- ber of developed states decided to withdraw from the institution. The rst state to withdraw being Canada in 1993, followed by the U.S. in 1996, and Australia the following year. The primary reasons for the withdraws being criticism of the governance of the institution. The Australian legislature held a number of public hearings on the issue of withdrawal from UNIDO 84 in 1996 and 1997. 36 As explained by the legislative committees, budgetary and governance concerns dominated the reasons underpinning its decision to withdraw. 37 Moreover, Australia viewed its own foreign aid eorts as more eective. Australia also cited to the earlier Canadian and U.S. withdrawals as supporting its decision. In Canada, the results of the 1993 federal election led to a major shift in Canadian politics with the Liberals, led by Jean Chr etien, winning a ma- jority and soundly defeating Progressive Conservatives, who never regained their former political power. Chr etien, the new Prime Minister of Canada had promised to renegotiate trade treaties like the North American Free Trade Agreement (NAFTA). The U.K. (with Germany close behind) also moved to withdraw from UNIDO, citing criticism of the U.N. and its gover- nance structure as a reason for its decision to withdraw. 38 It was reported that the Clinton Administration withdrew the U.S. from UNIDO based on growing anti-U.N. sentiment among conservatives in the U.S. Congress. 39 At the time, the U.S. and the U.K. contributed approximately one third of UNIDO's budget, making the withdrawals a direct threat to the survival of the institution. 36 \Australia's Withdrawal from UNIDO and Treaties Tabled on 11 February 1997." (7th Report) Joint Standing Committee on Treaties, the Parliament of the Commonwealth of Australia. (Mar. 1997). 37 Australia had withdrawn from UNIDO in 1988 for similar reasons, only rejoining in 1992 once it determined UNIDO governance was being reformed. 38 Raymond Bonner. \British Pullout Threatens U.N. Development Agency's Future." The New York Times (Dec. 15, 1996), available at https://www.nytimes.com/1996/12/15/world/british-pullout-threatens-un-development- agency-s-future.html. 39 Id. 85 The second episode of cascading withdrawal occurred in recent years with many European states, including the U.K., Belgium, Denmark, France, Por- tugal, Greece, Lithuania, and Slovakia, as well as New Zealand deciding to withdraw. Again, the rationale for these decisions focused on criticism of the governance in the institution. Yet, some reporting suggests these countries were engaged in a process of reconsidering their international commitments because of budgetary concerns|highlighting possible increase in costs. 40 There does not appear to be signicant shifts in domestic politics, like those seen in Canada and the U.S. the previous decade, that led to the cascading exit. Being that each of these states shares strong ties with other developed, Western states, it is likely that there was a contagion aect occurring, at least in part. The international community connections here are strong as many of the states in both episodes of withdrawal are central and key play- ers in the international system and many great powers. Additionally, most of the withdrawing states are developed states substantially integrated into the global economy. Thus, these community connections are likely at play, although in an interesting direction|leading to contagion of withdrawal. Moreover, these connections and strong preference for rule of law in the countries would make any breach particularly costly. Therefore, in light of the shock of the changes in governance, and potentially based on a lack of 40 \France quits U.N. industry agency, in continuing Western exodus." Reuters. (Apr. 23, 2013), available at https://www.reuters.com/article/us-un-france/france-quits-u-n- industry-agency-in-continuing-western-exodus-idUSBRE93M0YN20130423. 86 exibility in UNIDO, these states could not bargain to restructure the gov- ernance framework of the institutions. The costs of remaining were simply too high. 2.6.3 North Korea from NPT The Treaty on the Non-Proliferation of Nuclear Weapons (NPT) is a multi- lateral, international treaty with the primary objective to prevent the spread of nuclear weapons and weapons technology. The NPT opened for signature in 1968 and came into force on March 5, 1970. Evidencing its breadth, 190 states are now parties to the NPT. The NPT provides a public good in avoid- ing the proliferation of nuclear weapons, something from which all benet. One need not be a member of the NPT to enjoy this benet, and members can benet from obtaining assistance with peaceful nuclear activities from member states. The NPT contains eleven articles, which seek to promote non-proliferation, disarmament, and the right to peacefully use nuclear technology. Like other arms control treaties the NPT uses the term \undertake," as opposed to \shall," in describing member state's obligations under the treaty (NPT, arts. I-VI). The treaty delegates verication of compliance to the Interna- tional Atomic Energy Agency (IAEA) (NPT, art. III). In article X, the same article allowing withdrawal from the treaty, the parties are to reconvene 25 years after the NPT entered into force to decide whether it should continue in force indenitely or for some xed period. In 1995, the parties agreed to the 87 NPT's indenite duration, decreasing its exibility. In practice, the treaty may be more exible for some states, like nuclear weapons states, than non- nuclear-weapon states. Being that North Korea was a non-nuclear-weapon state and a pariah in the international community, its latitude under the NPT was particularly constrained. On January 10, 2003, North Korea announced its withdrawal from the NPT after a previous attempt in 1993 was suspended. Its withdrawal became eective ninety days after the announcement pursuant to Article X of the treaty. There continues to be disagreement as to whether North Korea's withdrawal conformed to the provisions of the treaty. Months before, the U.S. had accused North Korea of violating the \Agreed Framework," which had kept North Korea within the NPT regime. The Agreed Framework was a 1994 agreement between the U.S. and North Korea to avoid its earlier attempt to withdraw, whereby North Korea agreed to remain in the NPT and not develop nuclear weapons in exchange for proliferation-resistant nu- clear power stations and fuel oil during the transition. 41 However, disputes over the validity of North Korea's withdrawal essentially became moot when the September 19, 2005 Joint Statement at the end of the Fourth Round of the Six-Party Talks called for North Korea's return to the NPT, implicitly recognizing its withdrawal. Pinpointing the shock that prompted North Korea's withdrawal is somewhat dicult. As Huntley (2004) explains, North Korea had nuclear ambitions 41 Bas and Coe (2018) give a theory of when a deal to prevent proliferation will be successful, occuring either early or late in the development process and backed up by a credible threat of con ict. 88 for some time. It appears that North Korea's shift toward developing nu- clear weapons with haste and withdrawing from the NPT was prompted by the U.S.'s more bellicose and hawkish foreign policy under the Bush Admin- istration (Huntley 2004, 91{95). The Administration had taken a hardline position, including North Korea in the \Axis of Evil" along with Iraq and Iran. Linkage to the war in Afghanistan prompted the North Korean leader- ship to reassess its vulnerability to U.S. aggression, believing that a nuclear deterrent was necessary to avoid the same fate of Afghanistan, and later Iraq and Libya. This raised North Korea's perceived costs in not having a deterrent and the potential costs of violating the treaty. The public bene- t, i.e. a world with less nuclear proliferation, remained roughly the same. Thus, the relevant shock can be described by the shift in U.S. foreign policy to being more hawkish in light of the \war on terror." In the wake of North Korea's exit from the NPT, there was substantial analysis of how the parties to the NPT should handle an exiting violator. Congressional testimony detailed the decade-long eorts by the U.S. to nd a bargain with North Korea to avoid its breach or exit from the NPT. 42 At the rst session of the preparatory committee for the 2020 Review Conference of the Parties to the NPT in Vienna, on May 2, 2017, U.S. Ambassador Robert A. Wood emphasized the importance of maintaining accountability for withdrawing parties while preserving the right to withdraw: 42 Testimony by James A. Kelly, Assistant Secretary of State for East Asian and Pacic Aairs, delivered on July 15, 2004, to the Senate Foreign Relations Committee, in Digest of US Practice in International Law (2004), 1149{1155 89 Allowing States Parties to violate the Treaty and then withdraw without consequence if caught in a violation ignores this basic principle of international law. To vigorously enforce this princi- ple is not to curtail the right but rather to vindicate the interests of the remaining NPT Parties and preserve the integrity of the Treaty itself. We must also ensure that a withdrawing state can- not escape its obligations to other Parties not to misuse the fruits of peaceful nuclear cooperation. 43 This had been a repeated concern raised by U.S. representatives at the 2015 NPT Review Conference; 44 the June 5, 2014 G7 Summit; 45 and the Second Session of the Preparatory Committee for the 2010 NPT Review Confer- ence. 46 Ultimately, the U.N. Security Council passed Resolution 1887 in 2009, with President Barack Obama presiding, to develop means by which NPT members could collectively respond to any notice of withdrawal from the treaty to ensure compliance of the withdrawing state under interna- tional law. 47 In essence, the global community reacted to North Korea's withdrawal from the NPT by working to make withdrawal more costly. The North Korean case, and failure of the related Six-Party Talks, again highlights the central role of bargaining to the outcome of treaty withdrawal. In the end, the insecurities that North Korea had without nuclear weapons made its demands too large for the rest of the member states to accept. Looked at from another angle, the other NPT states were not willing to accommodate North Korea's concerns to the extent necessary to avoid the 43 Digest of US Practice in International Law (2017), 767. 44 Digest of US Practice in International Law (2015), 812, 816. 45 Digest of US Practice in International Law (2014), 818 46 Digest of US Practice in International Law (2008), 936{938. 47 UN Docs. S/RES/1887 (2009). 90 withdrawal. Moreover, breaching was not feasible given the ability of other member states to punish North Korea through sanctions. 48 2.6.4 Universal Postal Union The Universal Postal Union (UPU) was established in 1874, making it the second oldest international organization still in operation. 49 The UPU is a constituent organization of the United Nations, and any member of the United Nations may become a member to the UPU. The UPU serves as a forum for its 192 members to cooperate regarding issues concerning the postal sector. According to the UPU Constitution, the aims of the Union is: \to secure the organization and improvement of the postal services and to promote in this sphere of development of international collaboration." 50 The UPU was also created to \give developing countries access to global markets." 51 The UPU has regulated international mail service for over a century, solving the coordination problem by avoiding the mix of rates and regulations of the 19th century. On August 23, 2018, U.S. President Trump issued a memorandum setting out the priorities of his administration in its negotiations with the UPU. 52 48 North Korea was still sanctioned for its activities by states like the United States even after withdrawal. 49 \About UPU." https://www.upu.int/en/Universal-Postal-Union. 50 UPU Const., art. 1. 51 Peter Yeo. \Why the Universal Postal Union Matters." UN Foundation (Oct. 19, 2018) available at https://unfoundation.org/blog/post/universal-postal-union-matters/. 52 Modernizing the Monetary Reimbursement Model for the Delivery of Goods Through the International Postal System and Enhancing the Security and Safety of International Mail, 83 Fed. Reg. 183 (Aug. 23, 2018) [hereinafter Presidential Memorandum] 91 Under the then existing incarnation of the UPU framework from 1969, termi- nal due rates were based on a country classication system where developing nations paid less than developed states, like the United States. 53 This had the eect of increasing costs on these developed states. Among the United States' concerns was the contention that the costs of delivering items from developing countries, namely China, put U.S. mailers at a disadvantage. The U.S. accused the UPU of unfair discounts for China. Trump warned that if these issues were not addressed, the United States would \consider taking appropriate actions." 54 The White House Director of trade and manufacturing policy, Peter Navarro, later wrote: \[U]nder the [UPU]'s antiquated `terminal dues' system, the United States Postal Service was being forced to subsidize a ood of small packages, primarily from China, at an annual cost in the neighborhood of $500 million. This forced subsidy gave China an unfair advantage against American manufacturers and workers . . . ." 55 The United States' concerns prompted the UPU to hold is Second Extraordinary Congress to attempt to address those issues and avoid withdrawal. However, those negotiations were unsuccessful. Like many treaties, the UPU Constitution includes a withdrawal provision. 53 Universal Postal Union, About Terminal Dues and Transit Charges, at http://www.upu.int/en/activities/terminal-dues-and-transit-charges/about-terminal- dues-and-transit-charges.html. 54 Presidential Memorandum, supra. 55 Peter Navarro. \The Trump Guide to Diplomacy." The New York Time (Oct. 15, 2019), available at https://www.nytimes.com/2019/10/15/opinion/trump-universal- postal-union.html. 92 Article 12 of the UPU Constitution gives \[e]ach member country may with- draw from the Union by notice of denunciation of the Constitution . . . [and] becomes eective one year after the day on which the notice of denuncia- tion . . . is received . . . ." On October 17, 2018, the Trump Administration announced that it would pull the United States out of the UPU, exercising this option based on a lack of \sucient progress . . . on reforming terms of the Acts of the Universal Postal Union (UPU)." 56 Some feared the unilateral action would set o \postal wars" where countries set their own rates and create complicating barriers to global ows of mail. 57 However, after the announcement of the withdrawal, the UPU parties began to negotiate to change the status quo and meet the United States's demands in the Third Extraordinary Congress, convened from September 23 to 26, 2019. The U.S. advocated a number of options, including for allowing self- declared rates. The U.S,'s preference to immediately move to self-declared rates was opposed by China and rejected by the UPU Congress. However, they quickly and unanimously adopted the U.S.'s less drastic proposal on September 25, 2019. China expressed support for \ `a positive solution and compromise in the spirit of UPU.' " 58 Shortly thereafter, the United States rescinded its notice of withdrawal. 56 Statement from the White House Press Secretary (Oct. 17, 2018) available at https://www.whitehouse.gov/briengs-statements/statement-press-secretary-38/. 57 Yeo, supra. 58 Stephanie Nebehay, \U.N. Postal Union Clinches Deal to Keep U.S. in Club." Reuters (Sept. 25, 2019), available at https://www.reuters.com/article/us-un-postal/un-postal- union-clinches-deal-to-keep-us-in-club-idUSKBN1WA247. 93 This case highlights the central role of the bargaining process in driving the United States' decision, and rescission of that decision, to withdraw from the UPU. The United States had suered the shock of increased costs of the global marketplace, especially relative to China. Based on the model, the value of the institution decreased for the U.S. and increased for China. This proved to be signicantly large that the U.S. demanded a change to the status quo or withdrawal. The UPU accepted the U.S. demands given the U.S.'s large importance in the international economy and mail system, being a founding member of the UPU. 2.7 Conclusion Treaty exit is a strategic action taken by states in the context of their on- going bargains with their treaty partners. Here, I have presented a theo- retical model to explain the relationship between an institution's level of obligation and exibility in uence a state's decision to successfully bargain to remain in a treaty, unilaterally breach to take more value (or lower cost) from the treaty, or end cooperation by leaving. The theoretical model re- veals that when an agreement is less exible such that breaching is not protable, a state is more likely to choose to leave in the face of large shocks or relatively small shocks to treaties with high initial commitments. In examining the four case studies, the theory and mechanisms at work appear plausible in explaining each of the states' decisions to withdraw. The bargaining framework ts comfortably in each case, where the states that 94 were bueted with a shock attempted to bargain to stay. In a counterfactual world, the states may have simply acted, namely by exiting the institution, without any bargain. At least in these cases, that does not appear to have occurred. In all four cases, the states attempted to bargain for concessions. The obliga- tions of the institutions in each case were particularly burdensome for these states. Japan felt that its culture, whether widely practiced or not, was under threat by the ICRW and the other member states. Australia, U.S., Canada, and European states protested UNIDO where it was not exible enough to respond to their displeasure and economic burden of support- ing the organization. The North Korean leadership, desperately fearful of being toppled, saw deterrence from external threats as essential but could not reach a bargain within the NPT to guarantee its security. Finally, the UPU is an example of how a suciently exible institution can respond to threats of withdrawal by reforming itself. However, examination of more cases would allow for a more conclusive assessment. With a theoretical framework in hand, we can examine other cases to test whether it explains the outcomes of interest: whether a state decided to remain in a treaty or exit. Ideally, we would want to examine every treaty and case of withdrawal to nd proof of the theories validity. In the following chapters, I use a mixture of methods to test aspects of this theory through a quantitative analysis of treaties over a 70 year period and additional more in-depth case studies of prominent cases of exit. 95 Chapter 3 Broad Analysis: Treaty Exit Over the Last 80 Years 3.1 Introduction As mentioned in Chapter 1, most are generally familiar with the high-prole treaty withdrawals of the Trump Administration from 2017 to 2020 and the 2016 Brexit referendum and long divorce between the U.K. and the E.U. However, these well-known treaty withdrawals only account for a fraction of the treaty withdrawals that have occurred in the modern era. Lesser known instances of withdrawal, like Mauritania's 2006 withdrawal from the Agreement for the Establishment of a Commission for Controlling the Desert Locust in North-West Africa or even Canada's 2011 withdrawal from the World Tourism Organization, received far less attention than those of the 96 Trump Presidency. As described in the prior chapter, the type of treaty exit will also likely correspond with the publicity of the exit, where adversarial or protest exits will be made loudly and clerical or substitute exits will be made quietly. As explained in Chapter 2, exit occurs when the member states cannot reach a bargain to reform the treaty and breaching the treaty after a shock is not feasible because it is not exible enough. Additionally, treaties with high levels of obligation are more sensitive to shocks, which make bargains less appealing. The interplay between exibility and obligation shape the preferences of the states over their desired outcome of interest. In this Chapter, I explore treaty exit from 1945 to 2020 across a broad range of treaties. In developing measures of a treaty's obligation and exibility, us- ing registered treaty actions and automated test analysis, I test the theory across hundreds of bilateral and multilateral treaties. This chapter pro- ceeds by rst describing the approach to data collection, namely using the United Nations Treaty Series (UNTS), as a repository of valuable data on treaties and state behavior. After describing the data collection approach, this chapter then presents descriptive statistics on the conclusion of treaties from the mid-nineteenth century to the twenty-rst century. Then, focus- ing on the post-World War II era, the chapter presents the make up of the types of treaties, i.e. bilateral or multilateral, and types of treaty exit over that time period. Included in this exploration are descriptive statistics re- garding the frequency of exit across time, states, and treaties. Then the chapter describes the sampling method for the analysis, the construction of 97 key variables, and provides a statistical analysis of the relationship between exit and measures of treaty obligation and exibility. After the analysis, the results are discussed and the expectations from the theory are assessed. The chapter concludes with implications of the results, areas for further data collection and testing, and the need to understand the mechanisms at work through case studies. 3.2 Data Collection from UNTS To empirically test the hypotheses derived in Chapter 2, I collected treaty information from the UNTS, which includes \more than 250; 000 treaties or treaty actions that have been registered and published by the Secretariat." The UNTS \is a publication produced by the Secretariat of the United Nations containing all treaties and international agreements registered or led and recorded by the Secretariat since 1945, pursuant to Article 102 of the Charter." 1 Article 102 incentivizes treaty registration in stating \No party to any such treaty or international agreement which has not been registered . . . may invoke that treaty or agreement before any organ of the United Nations." Thus, should a state desire to invoke the forum of the United Nations to address issues of treaty interpretation or compliance, those treaties need to be made public and registered with the UNTS. As such, the UNTS serves as a repository of numerous treaties of all types from 1 United Nations Oce of Legal Aairs Treaty Section, \Overview," United Nations Treaty Collection, available at: https://treaties.un.org/Pages/Overview.aspx?path=overview/overview/page1 en.xml 98 states around the globe. The UNTS also includes treaties from the League of Nations giving it a wide range of treaties over time. Of course, not all treaties are registered or led with the UNTS. Yet, of all the possible sources of treaty information, the UNTS provides the most comprehensive collection of treaties over the last two centuries and across most states. Moreover, the UNTS has catalogued the treaties, with meta-data for each, including information about the number of parties, the date of signature, the subject matter, and other useful information. Furthermore, the UNTS includes information on actions by state parties, which allows for an un- derstanding of how the member states have participated in the treaty over time. Included in those actions are the date of signature, date of ratication, any reservations or declarations made by the states, and date on which a state withdrew or denounced the treaty. This suite of information largely standardized across many treaties makes it an ideal source|acknowledging that the UNTS is not a perfect compilation of treaties concluded over the last hundred years. Any limitations of the UNTS in its incompleteness are outweighed by the utility of the information it does have. There is no reason to believe that there is systematic omission of treaties from the UNTS. Some classes of treaties, perhaps secret treaties between states 2 or agreements that 2 States do enter into \secret" agreements. Although a global norm of making treaties public can be traced to Woodrow Wilson and others in the aftermath of World War I (Hudson 1925), the use of secret treaties does persist (Azaria 2017; Donaldson 2017). In such a case, those treaties would not be made public or likely registered with the UNTS. Secret treaties may dierent from other treaties in that they seek to avoid confrontation with other states or domestic opposition. Nonetheless, the mechanisms at work in the bargaining model of treaty exit proposed here likely apply to secret treaties, just with dierent degrees of magnitude. Avoiding breach by the other member may be more dicult when a treaty is secret. 99 are not characterized as treaties per se, may not be included here. However, that omission should not aect the mechanisms of interest in this study. My approach is similar to Helfer (2005) and Koremenos (2016) in using the UNTS as a source of this information. Helfer (2005) collected similar in- formation without performing much analysis and reviewed a limited set of treaty actions. Koremenos (2016) used the UNTS to sample 243 treaties to code and test theories of treaty design. The variables of interest of the 243 treaties were hand-coded, which likely improved the measures but lim- ited the coverage. Moreover, it is dicult to replicate such an extensive coding scheme on a new random sample of treaties. The data used here was collected using automated web-scraping methods to build a large and replicable dataset. This process is labor-intensive, even if automated, given the volume of data to be collected, cleaned, and organized. 3.3 General Descriptive Statistics Before exploring treaty exit more broadly, we can examine the landscape of the treaties concluded and registered with the UNTS over the course of the last two centuries. Using the date of conclusion for a treaty as a guidepost, from 1800 to 2020, the UNTS includes 58,806 unique treaties with dates of conclusion. Of those treaties, 94% are bilateral treaties, 3% are closed multilateral treaties, and 3% are open multilateral treaties. This highlights the fact that almost all treaties for the last two centuries were bilateral. This is crucial in our understanding of state behavior where much of the focus of 100 the literature on treaty behavior is focused on the high-prole, multilateral treaties of the past few decades. Figure 3.1: Treaties Concluded and Registered since 1851 As shown in gure 3.1, the number of treaties concluded and registered with the UNTS is minimal in the later half of the 19th century and early 20th century, then explodes in the post-war era. This is likely due to the institutional nature of the United Nations where treaties concluded after 1945 when the U.N. was founded would be more likely to be registered with the UNTS. Yet, it is important to include treaties concluded before 1945 that were still in force after 1945 to capture later actions by the state members. As discussed in Chapter 2, the U.S. announced its plans to leave the Universal Postal Union (UPU), which was founded in the late 1980s, long before the creation of the U.N. 101 Focusing in on the post-World War II period, gure 3.2 shows the number of treaties concluded and registered from 1945 to 2020 by year. The number of treaties concluded each year increases in the 1970s dropping in the 1980s and again peaking in the mid-1990s. The year with the most treaties concluded is 1994 with 1,240 treaties concluded and registered. On average, 761 treaties were concluded and registered each year. Again, gure 3.2 highlights that since 1945 the overwhelming majority of treaties concluded by treaty type are bilateral treaties. Figure 3.2: Treaties Concluded and Registered since 1945 As mentioned, the UNTS also collects information of treaty actions, like signature, ratications, declarations, and terminations. To collect the data on treaty exits, I determined how states had registered their exit actions with the UNTS. States generally registered exits from treaties as either 102 withdrawals, denunciations, or terminations. 3 When a state withdraws from or denounces a treaty, it is a clear, pro-active action exiting the treaty. Treaty termination is a species of treaty exit; however, it includes a class of behavior that is distinct from the proactive step of treaty exit at the focus of this research. Treaty termination may include the end of a treaty simply by operation of its terms or for some other reason. This may complicate the analysis where terminations that occurred for reasons other than a state's considered decision to exit would be included. Figure 3.3: All Exits Over Time (Annual) 3 In collecting the data directly from the UNTS using search criteria, additional actions that included similar terms were captured, requiring ltering of the results for an accurate assessment of the number of exits. For example, the action \Withdrawal of objection to the accession of Albania" was captured in the broad search; however, this is not the type of withdrawal at issue in this study. Withdrawing an objection is not a treaty exit. Actions like these were ltered using hand coding of actions. That hand coding included a strict coding scheme, where only denitive treaty exits were included, and a exible coding scheme, where ambiguous actions were included. The main analyses conducted here used the strict coding scheme to ensure that only unambiguous treaty exits would be included. 103 Before moving on, we can examine all types of treaty exit since 1945 for gen- eral patterns and trends. Figure 3.3 shows the annual number of exits from 1945 to 2020 by type of exit. On average, 75 exits were registered per year; however, it is clear exits uctuated from year to year. Further, the type of exits uctuated from year to year as well, although terminations dominate treaty exit making up 76% of all treaty exits in this time frame, followed by denunciations at 18% and withdrawals 6%. As shown in gure 3.3, there was a large volume of terminations in the mid- to late-1970s. This mass of terminations is most likely the result of some realignment of treaty members from one set of treaties to another. On closer examination, many of the ter- minations in that time period involved treaties ceasing to have eect because of the adoption of a new treaty. For example, on March 16, 1974 Qatar and 14 United Nations organizations 4 terminated the Standard Agreement on Operational Assistance signed at New York on September 29, 1972. That agreement ceased to have eect on that specic date because the Agreement concerning assistance between Qatar and the United Nations Development Programme, signed at Doha on March 16, 1974, came into eect. 5 Breaking down treaty exit by treaty type, namely bilateral or multi-lateral, a lower percentage of bilateral treaties have registered exits compared to either closed or open multilateral treaties. Figure 3.4 shows the percent of treaties experiencing a treaty exit (withdrawal, denunciation, or termination) by treaty type. Of all treaties and exit actions registered with the UNTS, 9:6% 4 Those organizations include: United Nations; UNIDO; UNESCO; Universal Postal Union; World Health Organization; International Labour Organization; International Atomic Energy Agency; World Meterorological Organization, among others. 5 Pursuant to article XIII(l) of the 1974 Agreement. 104 Figure 3.4: Exits Percent by Treaty Type of bilateral treaties experience exit compared to 23% of closed and 23% open multilateral treaties. Assuming all (or at least most) exits are registered with the UNTS, the data suggests that multilateral treaties experience exit more than twice as much as bilateral treaties. Of course, we might question this supposition where states may not register the exit or end of a bilateral treaty. Yet, the fact that most treaties registered with the UNTS are bilateral, it suggests that states would also want to register related treaty actions, including exit, for the same reason. Examining treaty exit broadly highlights a few challenges with the data. First, the challenge of including terminations within the set of treaty exits 105 to study. If an agreement is completely terminated, it terminates for all member states such that all states (or organizations) would register a termi- nation action even if that was not a proactive decision by that each actor. Further, many terminations are the function of either the treaty terms, thus bargained for ends to the treaty relationship, or the result of a new bargain replacing an old one. Either situation would require a detailed investigation of the treaty text and termination notication for every individual termina- tion. With 4,465 terminations registered from 1945 to 2020, investigating each one is not feasible at this time. Excluding terminations avoids a num- ber of these challenges at the sacrice of losing information about termina- tion actions that may t the scope of interest here. Yet, denunciations and withdrawals seem less likely to suer from some of these challenges while providing information on proactive behavior of treaty participants. A second challenge as the circumstances described above highlights, is the inclusion of international organizations. International organizations have in- ternational legal personality and may enter and, by extension, exit treaties. 6 These organizations are active in treaty relations but are dierent from states as not being the primary focus of international law. Moreover, international organizations represent a collection of states in many cases complicating our understanding of their rationales for decisions. The theory to be tested here focuses on state calculations to exit. 6 The Vienna Convention on the Law of Treaties Between States and International Organizations or Between International Organizations of 1986 (VCLTIO), which has yet to enter into force as of 2022, extends the VCLT to agreements between states and IOs or between IOs themselves. 106 Figure 3.5: Exits [Restricted] Over Time (Annual) For a more restricted view of treaty exit, I collect all instances of with- drawal or denunciation registered with the UNTS from 1945 to 2020. This includes over 1; 500 exits (withdrawals and denunciations) from 398 treaties by 180 dierent exiting states. Excluding terminations, there were signif- icant spikes in denunciations and withdrawals in the late 1980s and late 1990s. This may be associated with the realignment of states after the col- lapse of the Soviet Union, lending plausibility to the proposition that shocks to a treaty's value precipitate its decision to exit. Figures 3.5, 3.6, 3.7, 3.8 illustrate descriptive statistics regarding the subset of exits registered with the UNTS after excluding terminations. Figure 3.5 shows all treaty withdrawals and denunciations from 1945 to 107 Figure 3.6: Exits per State 2020 registered with the UNTS. On average, 21 of these type of exists were registered each year, with some years experiencing signicant spikes in exit. The 1980s and 1990s were particularly active. Figures 3.6 and 3.7 show the numbers of exits per state and per treaty, respectively. The average number of exits per state is 10 and the average number of exits per treaty is 4. As gures 3.6 and 3.7 show, most states exit infrequently and most treaties suer from few exits (one exit is the median). Thus, a small number of states seem to be behind many exits and only a few treaties suer from more than a couple withdrawals. Figure 3.8 shows the top 20 states with the most exits registered with the UNTS. Interestingly, most of the states are western European states or North 108 Figure 3.7: Exits per Treaty American states. At rst glance, this may seem surprising to have so many of the advocates for international institutions appearing to be the most fre- quent exiters. This could be because these countries, mostly democracies, have frequently shifting preferences based on regular elections. Shocks may be more impactful in these states when the public has greater say in state policy. It could also be that these states are so integrated into the interna- tional network of treaties, with relatively more connections than most, that each simply has more opportunity to exit. Also, these states may have a higher respect for international law and the rule of law, which would make breaching particularly costly when compared to exit. 109 Figure 3.8: Top 20 Exiting States 3.4 Data Set Moving beyond examining treaty exit descriptively, I develop a strategy to test the theory over a large sample of treaties. Again, the data set used in this analysis comes from the UN Treaty Series. The data were generated in two phases, based on the recommendations of King and Zeng (2001). As King and Zeng (2001) explain, in international relations, some of the most consequential and interesting phenomena are rare events. This poses a challenge with broad statistical analyses where the rare event may be under- estimated and the absence of the event is overestimated. A core example of this is the study of war, which although hugely consequential to the parties 110 and individuals caught in its snare, is relatively rare across time and space. Thinking of the rare event as ones (1) and the absence of that event as zeros (0), the informational benet of the zeros decreases the more one has in their data. In this case, King and Zeng (2001) suggest working sequentially starting with an equal amount of ones and zeros, and then increasing to twice as many zeros than ones and so on to minimize standard errors and condence intervals. Returning to the data used in the analysis here, the data was collected in two phases where phase one captured all instances of treaty exit (the ones) and the second phase randomly sampled treaties with no exits. In the rst phase, all treaties that include a registered action of \withdrawal," \denunci- ation," or \termination" were collected as discussed above. In collecting the data, the actions had to be ltered to restrict them to only a \withdrawal," \denunciation," or \termination" from the agreement, as compared to the withdrawal of an objection or reservation. To do this required hand-coding the types of treaty exit. 7 This generated a dataset of 398 treaties that had at least one exit. In the second phase, I collected a random sample of treaties that do not include any registered actions \withdrawal," \denunciation," or \termina- tion." As shorthand, these are called the \no-exit" treaties. In line with King and Zeng (2001), double the number of treaties with exits (ones) is 796. To avoid the eects of likely attrition in the number of treaties to 7 The raw data from the UN Treaty Series included 467 variations of exit actions, some of which were not relevant to this project. For a discussion of the coding procedure, see Appendix C. 111 be used in the analysis below, I sampled 1; 102 \no-exit" treaties, giving a buer of 306 treaties above what is needed. In examining the random sample, it follows many of the overarching patterns and trends of the full set of treaties concluded and registered with the UNTS, as explained in Ap- pendix C. Combining these two datasets produces data from 1; 500 treaties to use for further analysis. 3.5 Measuring Flexibility and Obligation As explained above and in Chapter 2, a treaty's level of obligation and ex- ibility in uence the ability for the treaty members to bargain and avoid breaching or exit. To test the theory that the treaty design has an eect on whether a state decides to exit, we need to be able to assess, not only the treaty characteristics, including time of conclusion, subject matter, and number of parties, but also subsequent treaty actions and the treaty text. Text analysis of treaties is a laborious process for which international lawyers spend signicant time. For purposes of comparing treaties across time, sub- ject, and states, I leverage the use of automatic text analysis to build a number of measures to approximate the two variables of interest: exibility and obligation. 3.5.1 Collecting Actions On way to measure the exibility and obligation imposed by a treaty is to consider the subsequent actions of the member states. To do this, I collected 112 all registered actions for each treaty in the sample. There is a wide range of actions that states register with the UNTS. Of those, signature, ratications, accessions, and acceptances generally register a state's willingness to accept the obligations contained in the treaty but at dierent levels. Generally, ratications, accessions, and acceptances fully bind the state to the terms of the treaty (Vienna Convention on the Law of Treaties (1969) [VCLT], arts. 2(1)(b), 14(1), 14(2), 15, and 16). Signatures also attach obligations but to a lesser degree where the state at a minimum must act in good faith to avoid frustrating the purpose of the treaty (VCLT, arts. 10 and 18; see also Almeida and Pereira 2013). The number of each of these actions is coded as an integer in the variables Signatures, Ratifications, Accessions, and Acceptances for each treaty and treaty-year. Additionally, states may make other actions with regard to agreeing to bind themselves to the terms of a treaty. Importantly, states may make reser- vations to exclude or alter the legal eect of some portion of the treaty as applied to that particular state (VCLT, art. 2(1)(d) and 19{23). It is a device for states to join a treaty with a carve out for some particular provi- sion. Reservations are generally registered when a state sees some portion of the obligations of a treaty being particularly costly (Zvobgo, Sandholtz and Mulesky 2020). As applied in this study, we might expect treaties with high levels of obligation to attract more reservations. Put another way, if the treaty does not carrying particularly onerous obligations, a state will not need to lodge a reservation. This logic is limited by the fact that some 113 treaties may limit the subject matter open for reservation or prohibit reser- vations all together. The number of reservations is coded as Reservations for each treaty and treaty-year. States may also make declarations, which are statements of the state's under- standing of some matter of interpretation of a treaty provision. Declarations dier from reservations in that they merely express an understanding regard- ing interpretation rather than modifying or excluding some provision of the treaty. Declarations are also not always legally binding yet sometimes in- corporate customary international law. Additionally, the term \declaration" can be used for instruments themselves, for example the 1948 Universal Dec- laration of Human Rights. However, in this study, only declarations led as a comment on some referenced treaty are included. As related to obligation and exibility, we should expect treaties that impose a substantial amount of costs to elicit a declaration. For example, in joining the Rome Statute establishing the International Criminal Court (ICC), Australia registered a declaration (reservations are prohibited by the Rome Statute) to highlight that the court's admissibility jurisdiction. Specically, Australia noted that where a case is already being investigated by Australian authorities, that case will be inadmissible before the ICC. Australia's declaration also notes that, although it is not a reservation, it has \full eect in Australian law." 8 The Rome Statute imposes substantial obligations on its members, and Aus- tralia's declaration implicitly acknowledges this high level of costs. Like the 8 The Australian Parliament enacted the International Criminal Court Act 2002 creat- ing mechanisms for Australia to implement the terms of the Rome Statute (International Criminal Court Act 2002, No. 41, 2002). This understanding regarding inadmissibly was incorporated into the law. 114 other variables, the number of declarations is included in the variable Dec- larations for each treaty and treaty-year. The main dependent variable, treaty exit, was created using the UNTS in- formation on treaty actions. If any member state registered a \withdrawal" or \denunciation," the binary variable Exit was coded as one (1) for exit and zero (0) for no exit. In addition, the number of exits was computed for each treaty and treaty-year as the count variable Num Exit. 3.5.2 Collecting Treaty Texts In collecting treaty texts, the English version of the treaty text was ob- tained where available through the UNTS. The UNTS makes the treaty texts available in PDF format, which themselves come in variety of formats. Some PDFs contain only the English text of one treaty. Other PDFs contain the treaty text with various translations of the text either in a two-column format with translations side by side or where each alternating page includes a dierent translation (e.g. English, then French, then English, and so on). Additionally, many of the treaty texts are included in large volumes where several other treaties are included in one PDF. Finally, a portion of the cor- pus of treaty texts was either unavailable or only a Certicate of Registration by the U.N. Secretary General is available. Because of the heterogeneity of the availability and format of the treaty texts, signicant processing was re- quired to obtain a subset of treaty documents, even before converting them 115 to text les for analysis. 9 Because the UNTS stores PDF versions of the treaty documents, it is nec- essary to convert the treaty documents to a format that can be used in automated text analysis. To do this, each PDF version of the treaty doc- ument was processed using optical character recognition (OCR) in Adobe Acrobat and saved as a plain text le. Because the process involves the conversion of PDFs|sometimes old and unclear PDFs|some of the text is converted imperfectly. However, because the text analysis used here uses the \bag of words" approach, namely that each treaty text is considered a collection of words without punctuation or stop words, inaccuracies in the OCR process will not likely have a signicant impact on the outcome of interest in the aggregate. In spot-checking the accuracy of the conversion from PDF to text les, the English text remained intelligible on the whole. In weighing the trade-o here, any inaccuracy is balanced against increased coverage of dierent treaties. 3.5.3 Building Measure of Obligation and Flexibility from Treaty Text Using the treaty texts, I use a dictionary method to classify treaty texts as either obligatory or permissive. Dictionary methods are often used in this 9 The UNTS should consider storing the text of treaty instruments in a uniform text format in a manner similar to EDGAR lings with the U.S. Securities and Exchange Commission (SEC). Doing so would allow for more functionality for users of the UNTS database, including diplomats, attorneys, researchers, activists, and educators. Moreover, it would increase accessibility and transparency, especially where texts can sometimes be buried in long PDFs and contained in a format that is cumbersome to read. 116 manner to assess the negative or positive sentiment of texts, for example social media posts (Rathje, Van Bavel and van der Linden 2021). To build new measures of treaty obligation using automated text analysis, I take a multi-pronged approach and follow the guidance of Grimmer and Stewart (2013) and Barber a et al. (2021). To start, I must obtain a corpus of treaty text. As explained above, I collected the English treaty text of each treaty to be analyzed. Using the English text of each treaty in a text le, I use the Quanteda package in R to create a corpus of text and a dictionary to be applied to the corpus. In using a dictionary method, I categorize certain words as obligatory or permissive. Table C.2 in Appendix C gives the full dictionary used in this analysis. Words like \shall," \oblige," \verify," and \punish" are categorized as obligatory words. Words like \may," \undertake," \endeavor," \observe," and \aspire" are categorized as permissive. The dictionary categorizations were created in collaboration with two research assistants who reviewed over 200 treaty articles and coded them for various measures of obligation, precision, and strength (See Appendix D). 10 The dictionary is then used to create a score for the treaty based on the balance of obligatory and permissive words used in the treaty text. Simi- larly to sentiment analysis of text and social media posts, the two categories are set o from each other, where obligatory words are positive and permis- sive words are negative, such that a positive score indicates an obligatory 10 This work will be used to rene and expand the text analysis of treaty texts in later projects. 117 Figure 3.9: Average Permissive-Obligatory Score of Treaties Overtime treaty and a negative score indicates a permissive treaty. This permissive- obligatory score is stored in the variable Score. Based on scoring each treaty, the most permissive treaty is a bilateral treaty entitled the Project Grant Agreement for Housing and Community Upgrading for Low Income Egyptians entered into by the United States and Egypt in 1978. 11 The treaty with the highest permissive-obligatory score is the multilateral International Convention for the Safety of Life at Sea of 1960. 12 That treaty has had over 100 members and experienced at least four exits. The next highest score was assigned to the bilateral Settlement Treaty (1963) between the Netherlands and West Germany regarding the settlement of lingering border disputes. 13 It is not surprising that a treaty regarding border disputes includes a high 11 1169 UNTS 287. 12 536 UNTS 27. 13 508 UNTS 14. 118 number of obligatory terms. Neither party has registered an exit to that bilateral treaty. Figure 3.9 plots the annual average of the permissive-obligatory score for the treaties (623) in the data from 1890 to 2020. On the whole, most treaties create obligations and that is shown in the the overwhelming majority of years having positive (obligatory) permissive-obligatory scores. Interest- ingly, the average scores appear to become more permissive overtime, with many treaties in recent years using more permissive terms on balance. Ap- pendix C includes additional descriptions of the permissive-obligatory score. 3.5.4 Other Treaty Variables In addition to the key variables described above, a number of treaty-level variables are included. The variableMultilateral is an indicator variable for whether a treaty is bilateral or multilateral, coded as one (1) if it is multilateral and zero (0) otherwise. As shown in gure 3.4, multilateral treaties appear to see more exits on the whole. Having more than two members allows for more opportunities to exit. Yet, the higher rate of exits may also be a symptom of the diculty of reforming an institution with many members. The number of participants or members is also coded for each treaty and treaty-year in the variableNumParticipants. The number of participants and multilateral are distinct in that multilateral treaties can vary from a 119 handful of members to over one hundred members. Thus, Num Partici- pants should be able to highlight the issues of collective action, which may constrain the ability to devise a reform to the treaty obligations. The number of words in a treaty, namely the length of the treaty, is coded as the variable Words. Longer more wordy treaties likely have more pro- visions imposing obligations, are harder to interpret or reform, and might describe enforcement mechanisms. Therefore, a more wordy treaty would be a potentially less exible and more obligatory treaty. The permissive- obligatory score also captures length in part because a longer treaty will likely use more obligatory and permissive terms in total. However, having a separate variable disaggregates the measures to allow more ne-grained analysis. Finally, the variable Lifespan (Years) provides the number of years from date of conclusion to either the rst exit date or the end of 2020, whichever comes rst. All else being equal, we should expect older treaties to be more established and states to have grown accustomed to the obligations imposed by the treaty. Member states' preferences likely would become more aligned overtime (Stein 1982), and, therefore, treaty exit should be less prevalent from longer lived treaties. Note, this is a treaty-level measure of lifespan, not related to any particular member. Additional state-treaty-level data, can be coded to nd the length of membership of each state in each treaty, but that is not included here. 120 3.5.5 Summary of Variables and Expectations In collecting the meta-data and treaty texts, there was attrition in the data based on incomplete information, bad links, corrupted les, and other is- sues. As mentioned, this was anticipated to a degree. Of the 1,451 treaties included in the main sample, usable treaty texts were collected for 623 of those treaties (43% percent). A summary of select treaty-level variables for the 623 treaties that have useable treaty texts is described in Table 3.1. Of the 623 treaties, a third are multilateral treaties and the other two thirds are bilateral treaties. The number of signatories to each treaty ranges from two in bilateral treaties up to 202 signatories (sometimes including international organizations and non-state actors). The average lifespan of each treaty is approximately 38 years. Only a few treaties have registered reservations and declarations. Table 3.1: Select Treaty-Level Variables Statistic N Mean St. Dev. Min Pctl(25) Pctl(75) Max multilateral 623 0.3 0.5 0 0 1 1 num participants 621 14.4 30.1 2.0 2.0 8.0 202.0 words 623 4,465.1 13,158.4 212 735 4,377.5 280,279 score 623 31.5 74.2 50 1 35 1,247 reservations 623 0.2 3.8 0 0 0 93 declarations 623 0.6 2.6 0 0 0 34 exit 623 0.3 0.5 0 0 1 1 num exits 623 1.4 4.8 0 0 1 55 lifespan years 623 37.8 17.8 1.4 27.6 49.6 100.8 On average, the treaties are 4,465 words long, varying widely from the pithy 212 words to the voluminous 280,279 words. The average permissive- obligatory score is 31.5, which indicates that a treaty creates a minimal level 121 of obligations on the whole. This comports with our intuition that treaties are created and entered into to change state behavior in someway. To do that, the treaty drafters likely use some basic level of obligatory words to achieve that goal. Table 3.2 summarizes how each of the variables is linked to the concepts of obligation and exibility. The permissive-obligatory score captures the balance of the concepts of obligation and exibility in the treaty text. Pos- itive scores represent treaties that are more obligatory and negative scores represent treaties with more exible, less onerous, terms. The number of declarations captures the member states' concern over the interpretation of some provision in the treaty that will impose a cost on the state. There- fore, the more declarations registered with a treaty, the more obligatory we should expect the treaty to be. We cannot have the same expectation with exibility because the absence of declarations does not necessarily imply high levels of obligation or exibility. The number of reservations should provide insight regarding both concepts. As mentioned before, states register a reservation when they see some por- tion of the treaty as being too costly. However, once registered, the reserva- tion legally exempts the registering state from some portion of the obligation in the treaty. Where many states lodge reservations, the total obligations of the treaty will be diminished. Therefore, with more reservations registered, we should assume more exibility than where less are registered. Again, on the low end, it is dicult to draw too strong of a conclusion regarding the absence of reservations, where some treaties prohibit them explicitly. 122 Longer treaties have more provisions and leave less room for interpretation, thus they will have higher levels of obligation. The provisions of the treaty will be spelled out in more detail, memorializing the parties' understanding at the time. When the terms are made explicit in writing, it is more di- cult to take a dierent position in the future. Shorter treaties leave more room for interpretation and informal dealings, thus being more exible all else being equal. Short treaties can still create substantial obligations yet may leave more room for interpretation. Finally, when a treaty has a low number of participants it will be easier to negotiate an accommodation or reform of the treaty. A high number of participants may not necessarily rep- resent higher levels of obligation. In fact, high levels of treaty acceptance may indicate low levels of obligations, attracting many states to the lowest common denominator. Table 3.2: Measuring Independent Variables Obligation Flexibility Dictionary Score Obligatory (positive) Permissive (negative) No. of Declarations More - No. of Reservations Less More Length Long Short No. of Participants - Less Based on the theory developed in Chapter 2 and the mapping of the concepts of exibility and obligation onto the treaty measures, I devise the following hypotheses for each variable of interest. Hypothesis 1. A treaty that has a higher permissive-obligatory score will be more likely to experience an exit and more exits, all else being equal. 123 Hypothesis 2. A treaty that has more declarations registered with it will be more likely to experience an exit and more exits, all else being equal. Hypothesis 3. A treaty that has more reservations registered with it will be less likely to experience an exit and more exits, all else being equal. Hypothesis 4. A longer treaty (words) will be more likely to experience an exit and more exits, all else being equal. Hypothesis 5. A treaty with few participants will be less likely to experience an exit and more exits, all else being equal. The data is organized into three datasets. The rst contains the treaty-level variables of the 1,451 treaties sampled from the UNTS. The second is a subset of the rst (623 treaties), with additional variables derived from the treaty texts, as described above. The third dataset is a time series cross- sectional dataset organized by treaty-year (43,927 observations). The third dataset is useful in assessing treaty actions overtime and controlling for time and treaty xed eects. Variation in treaty texts is not included in this study, but is an area for further work to build on the analysis here. It theoretically should be possible to track the treaty text from original signature, additional protocols and annexes, amendments, reservations, declarations, and other subsequent actions. Those additional action change the interpretation, ex- ibility, and obligations of a treaty relationship in many dierent and likely nuanced ways. But for purposes of this initial study, the treaty text is taken as invariant over time and not included in the time series analysis. 124 3.6 Analysis and Results After the long process of collecting and processing the data, we can now analyze their relationships. In assessing the relationship between the vari- ous treaty attributes that are related to the two concepts of exibility and obligation, I estimate the variables on the occurrence of treaty exit (whether it occurs or not) and how many exits each treaty suers. In doing so, I use a logistic model to estimate the relationship of the variables to treaty exit, Exit, and use poisson and negative binomial models to estimate the rela- tionship of the variables to the number of exits per treaty, Num Exit . 14 Tables 3.3 and 3.4 report the results of the models. Table 3.3 shows that the permissive-obligatory score, Score, has a very small positive relationship with treaty exit. By exponentiating the coe- cients, we see that the odds of treaty exit increase by one percent for each increase in the permissive-obligatory score, holding all else equal (model 1). The hazard ratio (HR = exp()) for permissive-obligatory score is 1.001, such that a one unit increase in the score is associated with a one percent increase in the hazard of experiencing a treaty exit (model 2). While this a small eect, it can add up with signicant increases in the obligation score. The permissive-obligatory score is not statistically signicant when control- ling for treaty and year xed eects. This is sensible where the score did not vary by year, essentially being a static measure of the text at the start of the treaty relationship. Yet, when controlling for treaties and time, the 14 In addition, the results of quasi-poisson and zero-in ated models are reported in Appendix C. 125 Table 3.3: Eect of Treaty Attributes on Treaty Exit Dependent variable: Treaty Exit Lifespan (Years) Treaty Exit Lifespan (Years) Treaty Exit logistic Cox logistic Cox logistic prop. hazards prop. hazards (1) (2) (3) (4) (5) Obligation Score 0.009 0.001 0.0001 (0.004) (0.001) (0.001) Length (Words) 0.0001 0.00000 0.00000 (0.00004) (0.00001) (0.00001) Multilateral 3.634 2.331 3.406 2.382 2.453 (0.381) (0.194) (0.256) (0.150) (0.162) No. of Members 0.009 0.006 0.029 0.007 0.002 (0.008) (0.002) (0.009) (0.002) (0.002) No. Reservations 12.316 0.007 9.996 0.004 0.043 (696.650) (0.011) (429.476) (0.011) (0.379) No. of Declarations 0.075 0.045 0.061 0.011 0.472 (0.066) (0.021) (0.048) (0.007) (0.139) Lifespan (Years) 0.056 0.018 (0.008) (0.005) Year 0.006 (0.002) Constant 0.278 2.138 17.322 (0.312) (0.193) (4.193) Treaty Fixed Eects? No No No No Yes Observations 621 621 1,301 1,301 22,892 R 2 0.371 0.307 Max. Possible R 2 0.977 0.951 Log Likelihood 185.100 1,028.659 365.810 1,720.356 2,432.978 Akaike Inf. Crit. 386.199 743.620 4,883.955 Wald Test 238.400 (df = 6) 390.400 (df = 4) LR Test 287.573 (df = 6) 477.786 (df = 4) Score (Logrank) Test 397.133 (df = 6) 692.523 (df = 4) Note: p<0.1; p<0.05; p<0.01 Data from UN Treaty Series. 126 number of declarations registered with a treaty (Declarations) becomes signicant, and each additional declaration registered increases the odds of treaty exit by 60 percent. Additionally, multilateral treaties (Multilateral = 1) are by far more likely to have a treaty exit, increasing the odds by 3,690 percent. This ap- plies across models and corresponds to the preliminary exploration of the data described above. The overall number of treaty participants (NumPar- ticipants) has a small but positive relationship to treaty exit, where each additional member increases in the odds of exit by 3 percent. Finally, older treaties (Lifespan (Years)) are less likely to suer from treaty exit, where the longer a treaty is in eect, the odds of treaty exit decrease by 6 per- cent for each additional year the treaty exists. When controlling for treaty and year, an exit is slightly more likely in recent years, with a one percent increase in odds of treaty exit for each additional year. Table 3.4 shows the results of the poisson and negative binomial regression analyses on the relationship between the key variables of interest and the number of exits a treaty might suer (Num Exits). Again, multilateral treaties (Multilateral = 1) are likely to suer from many more treaty exits, as estimated across multiple models. The more members a treaty (Num Participants) also has a positive, albeit small, relationship with the number of treaty exits. In the rst model, the permissive-obligatory score (Score) and length of treaty (Words) have a small and negative relationship with the number 127 Table 3.4: Eect of Treaty Attributes on Number of Treaty Exits Dependent variable: Number of Exits Poisson Negative Poisson Negative Poisson Negative Binomial Binomial Binomial (1) (2) (3) (4) (5) (6) Obligation Score 0.002 0.0001 0.0005 0.0001 (0.001) (0.002) (0.001) (0.001) Length (Words) 0.00002 0.00002 0.00000 0.00001 (0.00001) (0.00002) (0.00000) (0.00001) Multilateral 3.440 3.133 3.673 3.428 2.540 2.488 (0.153) (0.206) (0.121) (0.159) (0.135) (0.154) No. of Members 0.007 0.010 0.007 0.013 0.003 0.007 (0.001) (0.002) (0.001) (0.002) (0.002) (0.002) No. Reservations 0.035 0.035 0.036 0.038 0.182 0.267 (0.015) (0.017) (0.015) (0.018) (0.390) (0.579) No. of Declarations 0.014 0.040 0.002 0.002 0.510 0.511 (0.007) (0.023) (0.002) (0.008) (0.056) (0.238) Lifespan (Years) 0.009 0.019 0.0001 0.004 (0.002) (0.004) (0.001) (0.003) Year 0.008 0.011 (0.002) (0.003) Constant 1.793 1.473 2.583 2.482 21.808 26.705 (0.164) (0.231) (0.124) (0.159) (3.259) (4.982) Treaty Fixed Eects? No No No No Yes Yes Observations 621 621 1,301 1,301 22,892 22,892 Log Likelihood 1,061.810 618.284 1,670.446 987.521 3,831.592 3,044.046 0.750 (0.087) 0.693 (0.066) 0.057 (0.004) Akaike Inf. Crit. 2,139.620 1,252.568 3,352.891 1,987.043 7,681.185 6,106.093 Note: p<0.1; p<0.05; p<0.01 Data from UN Treaty Series. 128 of treaty exits, where a one unit increase in the score reduces the rate of expected treaty exits by 0.2 percent, all else equal. Similarly, the length of a treaty in words has a very small and negative relationship to the num- ber of treaty exits. Those results do not appear in the other models at a statistically signicant level. The number of reservations (Reservations) and declarations (Declarations) have statistically signicant relationships to the number of treaty exits in many of the models, moving in opposite directions respectively. As the total number of reservations per treaty increases, the expected rate of treaty exits decreases by 3.4 percent, all else equal (model 1). When controlling for time and treaties, these results are no longer statistically signicant. The rela- tionship of the number of declarations and number of treaty exits is in the opposite direction, where for each additional declaration to a treaty, the rate of expected treaty exits increases by 1.5 percent, all else equal (model 1). When controlling for time and treaties, each additional declaration is asso- ciated with an increase of 66.5 percent in the rate of expected treaty exit, all else equal (model 5). Finally, the number of treaty exits (Num Exits) increases as we move for- ward in time. Similar to the models of treaty exit generally, the longer a treaty exists (Lifespan (Years)), the number of treaty exits decreases. 129 3.7 Discussion The results above appear to conrm some of the theoretical expectations described in the hypotheses laid out in section 3.5.5 above. Regarding the actions that imply obligation or exibility, there is evidence to suggest that an increase in the number of reservations decreases the number of exits a treaty experiences (Hypo 3). More declarations is associated with a higher likelihood of exit and a short time until an exit will occur (Hypo 4). As expected, the presence of more declarations is likely an indication that states are particularly concerned with the interpretation and implementation of treaty terms. The results for these two actions do not appear in all models, which may be a result of the low number of total registered reservations and declarations and the asymmetric nature of how they function. The number of participants is associated with an increased likelihood of exit and more exits per treaty (Hypo 5). Again, this is likely due to the fact that the collective action problem becomes more acute as the number of members increases. Additionally, it is also possible that the more members to a treaty the wider the range of preferences is among treaty members. It may be the case that treaty members become dissatised with the preferences of some other member or bloc of members. As explained in Chapter 2, the states that left UNIDO explained their exit, in part, based on the change in leadership of the organization and disagreement over the organization's operations. Nonetheless, if we accept that states would bargain to address concerns to avoid exit, then doing so would occur more easily with fewer members, 130 regardless of the wide range of preferences. Additional measurement and testing should be conducted to assess the mechanism behind this result. Finally, the results nd support|albeit small|for the treaty text measures, where both the permissive-obligatory score and words have very small eects when statistically signicant (Hypos 1 and 4). The length of a treaty is generally not informative about treaty exit in these models. Treaties that use more obligatory language have greater odds of treaty exit. As mentioned above, although the magnitude of a one unit increase of the permissive- obligatory score is small, those small increases can add up to approximately 11 percent in the data. However, there is large uncertainty in using the score to predict treaty exit, as illustrated in gure 3.10. The high levels of uncertainty is understandable where treaty text has to be interpreted in the context of other treaty obligations, international law generally, and the circumstances at the time. Yet, it is promising that the crude measure of obligation has a relationship to treaty exit. The results should serve as an invitation to rene the permissive-obligatory score or develop better measures to assess the treaty texts. Supervised machine learning is probably the best next step in devising a measure that can be applied to a large number of treaty texts. In short, there is support for many of the theoretical expectations and additional empirical research can help improve this understanding. 131 Figure 3.10: Marginal Eects of Permissive-Obligatory Score on Pre- dicted Treaty Exit 3.8 Conclusion This broad analysis of treaty behavior, particularly treaty exit, is the rst| based on current literature|that looks at all types of treaties and incorpo- rates both treaty actions and treaty text in its analysis. It moves beyond looking at subsets of treaties, namely only those that create international or- ganizations or multilateral treaties. It also expands the literature on treaty design and bridges it to understand how design can aect later behavior. Although some of the key measures here are noisy, they can be rened to sharpen our understanding. 132 As described in this chapter, the approach taken to assess treaty exit across decades and numerous treaties presented a number of limitations. The rst is the limitation of measuring various key variables of interest. The most rigorous analysis of treaty texts to assess obligations and exibility requires human coding, specically by highly trained or experienced international attorneys. This of course is possible, but extremely time consuming and expensive|prohibitively so with this volume of treaties. Even reducing the number of treaties to a couple hundred would remain costly. The coding process for the data used in Koremenos (2016), covering 243 agreements instead of 623 or 1,400 here, required years of coding by numerous teams of research assistants. Automation avoids the costs but sacrices signal for noise to some unmeasured extent. Additionally, certain variables are tremendously dicult to quantify on a broad scale. Determining whether a state breaches its obligations is ex- tremely fraught. A breach may be secret, eluding detection from the re- searcher and other treaty members alike. However, even when a suspected breach occurs and is made public, the breach is often not conceded by the state in breach. The analysis of determining breach is the job of the lawyers and judges who assess the facts against the treaty provisions and law, gen- erally. To do this across hundreds of treaties is not feasible in the scope of this project. However, we can assess the eect of breach on bargaining and treaty exit by focusing on a small number of cases. In doing so, not only can we assess how arguments over breach aect bargaining, but we can also investigate the other mechanisms at work in the theory. In the next 133 three chapters, I will explore three dierent treaty institutions to assess the mechanisms and unmeasured variables to expand on the broad analysis in this chapter. 134 Chapter 4 ICC: Institutional Flexibility and Relative Importance The continent has suered the consequences of unaccountable governance for too long to disown the protections oered by the I.C.C. [{] Those leaders seeking to skirt the court are eectively looking for a license to kill, maim and oppress their own people without consequence. . . . Most of all, they believe that neither the golden rule, nor the rule of law, applies to them. [{] But they know that they cannot say these things in public without repercussions. Instead, they conveniently accuse the I.C.C. of racism. . . . Far from being a so-called white man's witch hunt, the I.C.C. could not be more African if it tried. . . . We . . . need the continent's heavyweights, Nigeria and South Africa, to ex- ercise leadership and stop those who don't like the rules from attempting to rewrite them. Far from a ght between Africa and the West, this is a ght within Africa, for its soul. | Rev. Desmond Tutu 1 1 Desmond Tutu \In Africa, Seeking a License to Kill." The New York Times (Oct. 10, 2013), available at nytimes.com/2013/10/11/opinion/in-africa-seeking-a-license-to- kill.html 135 4.1 Introduction The International Criminal Court (ICC) was created by the Rome Statute, which went into force on July 1, 2002, as a permanent international court to hold individuals accountable for atrocities during con ict. Its task is to prosecute individuals for some of the most heinous crimes, specically the international crimes of genocide, crimes against humanity, war crimes, and aggression. 2 Key to the ICC's success in accomplishing its goals is main- taining its power through legitimacy. The ICC's legitimacy is signicantly in uenced by the number of states that agree to be bound by its jurisdiction, especially populous, powerful, and in uential states. Over the two decades the court has been in operation, it has had to navigate the turbulent politics of carrying out its mission while maintaining cooperation from its members. In maintaining cooperation, claims that the court is not acting legitimately are a signicant challenge. Scholars, state parties, commenters, and liti- gants have criticized the court on a number of dimensions, each a factor in the ICC's legitimacy as an institution. The ICC's legitimacy is aected by perceptions of the composition of the judicial bodies and whether they represent the individuals who the court serves and prosecutes. The ICC is progressive for an international body with a global reach in that it strives to include judges from diverse backgrounds. However, some question whether this aspiration for diversity of race, sex, and backgrounds exists in practice. The ICC is also criticized for its apparent regional bias, focusing almost all of 2 The Assembly of States Parties to the Rome Treaty recently decided to activate jurisdiction of the crime of aggression starting July 17, 2018. 136 its investigations and prosecutions on African states, without pursuing any powerful Western states. Strangely, states like the U.S. also criticize the ICC for its anti-U.S. bias. 3 Additionally, the procedures of the court have also come under criticism as being unfair, infeasible, or inecient. Each of these broad categories of criticisms has lead some to claim that the legitimacy of the ICC is eroding. When the ICC's legitimacy erodes, the weight of continued membership in the institution increases for the member states. Leaders fear prosecution, governments feel constrained, and states may receive external and internal pressure leveraging the court in their arguments. In such a case, leaders need to arbitrage some of these costs in one way or another. Where breach- ing the Rome Statute carries signicant costs, in terms of reputational and potential material costs, states often use the threat of withdrawal in their bargaining with the other member states and the institution itself. Within that context, the ICC may weigh the relative importance of the states to the institution, and the likely eect of the exit of that state on its broader legitimacy, in determining whether to concede during a bargaining process. The structure of the institution allows it selective exibility in adapting to the continuing bargain between the members. Having broken up the insti- tution into constituent parts that can make independent, strategic choices allows for it to function as a nimble institution while building legitimacy. For example, the Prosecutor and Pre-Trial Chamber have signicant discretion 3 Siobh an O'Grady. \John Bolton isn't alone in condemning the ICC." The Washington Post. (Sept. 11, 2018), available at https://www.washingtonpost.com/world/2018/09/11/john-bolton-isnt-alone- condemning-icc/?noredirect=on&utm term=.79ca8d0a7867. 137 over case selection and wield that strategically. Moreover, the Assembly of State parties can internally address member concerns or even adjust court procedures, prosecutors, or judicial appointments to increase exibility, as discussed more in this chapter. In this chapter, I explore how the ICC is an example of an institution that places signicant obligations on its members but is exible enough to ad- dress potential breakdowns in cooperation. Interestingly though, the ICC exercises its exibility strategically in assessing the importance of a potential exit from a member state. In the next section, the chapter summarizes the history of international criminal tribunals, the eventual creation of the ICC, and the early situations prosecuted at the court. Next, the chapter details the court's eorts to promote its legitimacy, increase its eectiveness, and how critiques have arisen about whether the court has met those goals. In particular, the section addresses accusations of bias at the court in two main areas: in judicial representation and regional bias. With an understanding of the structure and critiques of the court, the chapter then details recent cases where the court was faced with a member threatening to leave the organization. Here, looking at the cases of Kenya, South Africa, Burundi, and The Philippines reveals how the ICC strategically exercises its levers of exibility contingent on its concern for the threat of exit to its legitimacy. These episodes support the theory presented in this dissertation and point to an additional area of research on the selective use of exibility in the bargaining process. 138 4.2 Background: Legacy, Creation, and Early Cases 4.2.1 Brief History of International Criminal Tribunals The modern development of international criminal tribunals can trace its roots to the aftermath of the rst world war. Although international hu- manitarian law has a modern heritage coming out of discrete treaties and rules agreed to in the early nineteenth century, examples being the Treaty of Guadalupe Hidalgo (1848) between the U.S. and Mexico and the U.S. Lieber Code (1863), international criminal tribunals to prosecute war crimes did not develop until the twentieth century. After World War I, the victorious Allied Powers recommended that the leaders of the defeated powers, includ- ing the German Kaiser, should be prosecuted for violating the laws of war on the basis of command responsibility. 4 The Treaty of Versailles (1919), memorializing the Central Power defeat, included a recognition by the Ger- man government that individuals could be prosecuted for war crimes before a military tribunal, including the Kaiser (art. 228), and laid individual re- sponsibility on the Kaiser for those crimes (art. 227). However, with the Kaiser in the Netherlands, which refused to turn him over, only a few trials occurred in Leipzig with mixed results (Mullins 1921; Kramer 2006). Ad- ditionally, the Allies accused Turkish ocials of committing crimes against humanity against the Armenian population. The post-World War I Turk- ish government indicted, tried, and convicted the accused through domestic 4 \Commission on the Responsibility of the Authors of the War and on Enforcement of Penalties." The American Journal of International Law 14.1-2 (1920): 95{154. 139 military tribunals; however, the sentences were never carried out (Kramer 2006). 5 In the aftermath of World War II, the Allies created two tribunals to pros- ecute individuals in both the European and Pacic theaters. In Europe, the International Military Tribunal (IMT) or Nuremberg Trials were con- ducted from November 1945 to October 1946 to try 24 defendants, who were high ranking ocials in the Nazi military and government. Among those defendants were Rudolf Hess, Adolf Hitler's Deputy F uhrer; Joachim von Ribbentrop, Nazi Minister of Foreign Aairs; Karl D onitz, head of the Kriegsmarine (Nazi Navy) and head of state after Hitler's death; and in- famous Hermann G oring, original head of the Gestapo, Commander of the Luftwae, and Reichsmarschall|making him the highest ranking Nazi of- cial to be tried at Nuremberg. The Allies had planned for some form of accountability of crimes committed by the Nazis even before the war had ended but diered on the format (Heller 2011; Hirsch 2020). In 1945, France, the United Kingdom, the Soviet Union, and the U.S. signed the Nuremberg Charter to create the IMT (Heller 2011, 4{27, 107{138). The Charter included the innovation that individuals could be held liable for violating international law, rather than states. The defendants were charged with conspiracy to commit crimes against peace; planning, initiating, and waging wars of aggression; war crimes; and crimes against humanity. This set the precedent for later courts leading to the ICC. Each signatory had 5 See also \Armation of the United States Record on the Armenian Genocide Reso- lution" H.Res.316 (109th Congress (2005-2006)),x2(1)-(7). 140 a prosecuting team, with the overwhelming majority coming from the U.S. The U.S.'s chief prosecutor, Supreme Court Justice Robert H. Jackson, said of the proceedings during his opening statement: \That four great nations, ushed with victory and stung with injury stay the hand of vengeance and voluntarily submit their captive enemies to the judgment of the law is one of the most signicant tributes that Power has ever paid to Reason." 6 In defense, the accused raised three broad categories of defenses: jurisdic- tional, substantive, and military necessity (Heller 2011, 295{312). Regard- ing the jurisdiction of the IMT, the accused argued that the Charter of the IMT was adopted after the crimes had been committed, constituting ex post facto criminalization (Schabas 2017, 4{8). Other specic defenses included selective-prosecution of defendants, lack of liability for following orders, immunity based on ocial position, mistake of law, self-defense, and the tu quoque defense essentially accusing the prosecuting powers of the same crimes. These defenses, although not broadly successful for the defendants, would generate substantial jurisprudence to be used by later defendants and critics of international criminal tribunals. Ultimately, the judges issued a verdict on October 1, 1946 convicting 19 defendants and acquitting three. Twelve of the defendants were sentenced to death and ten were hanged. Thereafter, the U.S. would conduct additional military tribunals to prose- cute Nazi war crimes from December 1946 to April 1949. 6 Robert H. Jackson. \Opening Statement before the International Military Tri- bunal." (Nov. 21, 1945), available at https://www.roberthjackson.org/speech-and- writing/opening-statement-before-the-international-military-tribunal/ 141 In the Pacic, U.S. General Douglas MacArthur, Supreme Commander for the Allied Powers in Allied-occupied Japan, created a tribunal in Tokyo by decreeing the International Military Tribunal for the Far East Charter (IMTFE Charter) or Tokyo Charter, which was modeled on the Nuremberg Charter (Boister and Cryer 2008). The Tokyo Charter set out three cat- egories of crimes to be prosecuted: crimes against peace, war crimes, and crimes against humanity. Although the post-World War II tribunals were similar, the tribunal in Tokyo diered in a few key ways. Chief among them was the exemption of Emperor Hirohito from prosecution. The Emperor's exception from prosecution was and remains controversial. The Tokyo IMT developed arguments regarding collective responsibility, conspiracy in com- mitting crimes against peace, and how the crime of aggression removes all justication for wartime killing (Boister 2010). After the Cold War, a series of mass atrocities prompted the creation of a number of ad hoc international criminal tribunals. To respond to the the allegations of war crimes during the Yugoslav Wars, the U.N. Security Council unanimously issued Resolution 827 on May 25, 1993 establishing the International Criminal Tribunal for the former Yugoslavia (ICTY) based in the The Hague, Netherlands. 7 The ICTY had jurisdiction to prosecute grave breaches of the Geneva Conventions, war crimes, genocide, and crimes against humanity. The rst indictment was issued in 1994, the rst of 161 7 The Security Council built upon its earlier resolutions 713 (1991), noting the ght- ing in Yugoslavia \constitutes a threat to international peace and security," 780 (1992), expressing concern over \widespread violations of international humanitarian law," and 808 (1993) determining to establish a tribunal to prosecute individuals accused of those violations of international humanitarian law. 142 indictments to be made over the life of the court. The ICTY also charged the former President of Yugoslavia, Slobodan Milo sevi c, with war crimes and crimes against humanity; however, he died in detention awaiting trial. Milo sevi c did not appoint counsel for himself during the proceedings, arguing that the ICTY was illegitimate and \illegal" having not been approved by the U.N. General Assembly. 8 A year after establishing the ICTY, the U.N. Security Council issued Reso- lution 955 creating the International Criminal Tribunal for Rwanda (ICTR). The ICTR was charged with investigating and prosecuting individuals ac- cused of committing atrocities during the horric Rwandan genocide. Sitting in Arusha, Tanzania, the ICTR prosecuted individuals for genocide, crimes against humanity, war crimes, and contempt of the Tribunal. Compared to the ICTY, the ICTR was charged with addressing crimes in the context of an internal con ict. Some have criticized the tribunal for carrying out \victor's justice" where the Rwandan Patriotic Front and its leader Paul Kagame were not prosecuted. In the case against Jean-Paul Akayesu, the ICTR set the legal precedent that gendered violence of genocidal rape fell under the crime of genocide. 9 After the verdict, presiding judge Navanethem Pillay made clear the advancement of the law in this respect: \From time immemorial, rape has been regarded as spoils of war. Now it will be con- sidered a war crime. We want to send out a strong message that rape is no longer a trophy of war" Ellis (2006). 8 The Prosecutor v. Slobodan Milo sevi c, IT-02-54 (MICT-13-58), Transcript of Pro- ceedings (3 July 2001). 9 The Prosecutor v. Jean-Paul Akayesu, ICTR-96-4-A (MICT-13-30), Judgment (1 June 2001). 143 The ICTY and ICTR functioned in tandem, sharing similar founding statutes, similar mandates, and even shared institutional resources (Schabas 2017, 12). By overlapping provisions, the U.N. Security Council tried to create an ecient system where, at the start, both tribunals shared a prosecu- tor and appellate body. As the tribunals completed their work, they were both subsumed into the U.N. International Residual Mechanism for Crimi- nal Tribunals, which was created to wind down the courts and to carry on any remaining work. 10 Aside from these two tribunals, the international community created a number of other ad hoc tribunals and \hybrid courts" including the Special Court for Sierra Leone (SCSL) created in 2002; Special Tribunal for Lebanon (STL) created in 2007; and the Extraordinary Cham- bers in the Courts of Cambodia (ECCC) in 2001 to prosecute crimes of the Khmer Rogue in the Cambodian genocide. These tribunals were established in situations where the ICC either may not have temporal jurisdiction or was unable to prosecute at the time (Schabas 2017, 14). Across the ad hoc tribunals and the ICC, approximately 390 individuals have been indicted, 228 convicted (58 %), and 47 acquitted (12 %). Figure 4.1 shows the number of charges, convictions, and acquittals in each interna- tional criminal tribunal. As evident in gure 4.1, the ICC has had far fewer charges and convictions than the ICTY and ICTR, which were the most active or productive. This may be due to the discrete scope of the tribunals and the fact that they were established by the U.N. Security Council versus the broad and diverse members of the Rome Statute. The low volume of the 10 UN Doc. S/RES/955 (1994). 144 Figure 4.1: Results of Major Int'l Criminal Tribunals cases at the ICC may be attributable to its eorts to negotiate the politics of maintaining its membership, which is explored more below. The work of the ICTY and ICTR in uenced many to see the need for a permanent international criminal court to address atrocities, which unfor- tunately continue to occur. The jurisprudence of those tribunals was and is tremendously in uential in the creation of the ICC and its functioning (Schabas 2017, 13). Article 8 of the Rome Statute, dening war crimes, was largely based on the ndings of the ICTY in the Tadi c case. 11 The dictum of the ICTY Appeals Chamber stating that crimes against humanity could occur in times of peace, not only in war, was adopted in article 7 of the Rome Statute, which denes the crime. Additionally, some jurisprudence 11 The Prosecutor v. Tadi c, IT-94-1-AR72, Decision on the Defense Motion for Inter- locatory Appeal on Jurisdiction (2 Oct. 1995). 145 was explicitly rejected in the Rome Statute, for example drafters included the defense of duress (art. 31(1)(d)) that was rejected by the ICTY in Erdemovi c. 12 4.2.2 The Rome Statute Again, the ICC was created by the Rome Statute, which was agreed to in Rome on July 17, 1998 by 120 states that voted to approve the treaty. In 1994, the U.N. General Assembly agreed to pursue eorts to create an in- ternational criminal court, using the International Law Commission (ILC)'s draft statute as a foundation (Bassiouni 1999). The Rome Statute and cre- ation of a permanent international criminal court was the result of years of preparatory work and weeks of intense negotiations (Wippman 2004). Dur- ing those extensive negotiations at the two ad hoc committees established by the General Assembly, 13 the delegates addressed the complicated issues of how to come to agreement on sensitive matters including the jurisdiction of the court and the substance of the crimes to be prosecuted. As part of those negotiations, the delegates supported the idea of \complementarity" for ju- risdiction, where the court would only act if domestic courts were unable or unwilling, as well as dening not simply enumerating the covered crimes. Some of these decisions and preferences diverged from initial proposals and 12 The Prosecutor v. Erdemovi c, IT-96-22-A, Sentencing Appeal (7 Oct. 1997). 13 The two committees were the Ad Hoc Committee on the Establishment of an Inter- national Criminal Court (1995) and Ad Hoc Committee was replaced by the Preparatory Committee on the Establishment of an International Criminal Court (1996{1998). 146 the ILC draft, creating concerns for potential members, chief among them the United States. Figure 4.2: The ICC at The Hague The ICC complex in The Hague, the Netherlands. Photo credit: author. The U.S. relationship with the ICC has been fraught from the negotiations over the Rome Statute. After President George H.W. Bush expressed reser- vations to the creation of a permanent international criminal court, Pres- ident Bill Clinton declared the U.S.'s \signicant and positive interest" in the ICC (Amann and Sellers 2002, 382{383). The U.S. was instrumental in drafting the ILC statute. President Clinton sent Ambassador-at-Large for War Crimes Issues, David J. Scheer, to Rome to participate in the ne- gotiations. The congressional opposition to the US joining the court and submitting to its jurisdiction was erce. U.S. Senator Jesse Helms (R) of North Carolina, who chaired the Senate Committee on Foreign Relations 147 was opposed to the U.S. joining the court. During hearings on whether join- ing the ICC would be in the U.S.'s best interest, Senator Rod Grams (R) of Minnesota set the tone as he opened the hearings: \The United States must aggressively oppose this court each step of the way, because the treaty establishing an international criminal court is not just bad, but I believe it is also dangerous." 14 Fearing that the United States would lose control of the court, he continued: The proposed ICC is not a part of the international system. It sits alone and above the system, and that is by design. At present international law regarding peace and security is largely what- ever the Security Council says that it is. With the creation of the International Criminal Court, that will no longer be the case. . . . The fact remains, the most eective deterrent is the threat of military action; and this court is undermining the ability of the United States to do that very thing." 15 Finally, addressing U.S. sovereignty, he declared that the U.S. \will not cede its sovereignty to an institution which claims to have the power to override the United States legal system and to pass judgment on our foreign pol- icy actions." 16 Senator Diane Feinstein (D) of California shared many of the same concerns but thought the ICC was important to ll the \interna- tional criminal justice" gap. 17 At the hearing, the senators raised concerns over universal jurisdiction, proprio motu (self-initiating prosecutor), and the crime of aggression, among other concerns. These concerns would resurface 14 \Is a U.N. International Criminal Court in the U.S. National Interest?" Hearing before the Subcommittee on International Operations of the Committee on Foreign Relations United States Senate (July 23, 1998) S. HRG. 105{724 (105th Congress, 2d session), 1. 15 Id. 16 Id. at 3. 17 Id. at 4{5. 148 during later eorts by the U.S. to undermine and pressure the court under the administrations of George W. Bush and Donald Trump. Shortly before leaving oce, President Clinton signed the Rome Statute on December 31, 2000 for the express purpose of being in a \position to in uence the evolution of the court." 18 In eect, Clinton attempted to keep one foot in the door to continue to negotiate from the inside. President Bush reversed course and unsigned the treaty in May 2002. 19 The Bush Administration claimed support for \accountability for atrocities and war crimes" but through ad hoc tribunals rather than a permanent court (Amann and Sellers 2002, 384). Despite the lack of support and active opposition from the U.S., the Rome Statute was eventually ratied by the minimum 60 states (Rome Statute, art. 126). At the time, ratication was expected to be slow given the nature of the treaty and the need for countries to adopt new domestic laws to align with the Rome Statute (Schabas 2017, 23{24). Senegal was the rst state to ratify the treaty on February 2, 1999. By the second anniversary of the signing of the treaty, 14 states had ratied. By the third anniversary, 37 states had ratied. The sixtieth ratication came on April 11, 2002, celebrated at a ceremony at the U.N. Four states, Sudan, Israel, Russia, and the United States had signed the Rome Statute but later informed the 18 \Statement on the Rome Treaty on the International Criminal Court." U.S. Dept. of State (Dec. 31, 2000). 19 Letter dated May 6, 2002 from Under Secretary of State for Arms Control and In- ternational Security, John R. Bolton, to U.N. Secretary General Ko Annan stating \the United States does not intend to become a party to the treaty[, and,] [a]ccordingly, the United States has no legal obligations arising from its signature on December 31, 2000." 149 U.N. Secretary General that they no longer intended to become parties and, therefore, the treaty has no binding eect on them. The wave of ratications had surprised observers at the time and is depicted in gure 4.3, which shows the number of states ratifying or acceding to the Rome Statute each year from 1999 to 2019. Kiribati was the most recent state to join when it acceded on November 26, 2019. Figure 4.3: Number of States Ratifying or Acceding Over Time In September 2002, the Assembly of State Parties met to adopt instruments regarding elements of crimes and procedures, along with appointing judges and the prosecutor. The Assembly of State Parties is a distinct organization, with its own secretariat, responsible for managing the court and providing a forum for the member states to propose and adopt amendments to the Rome Statute or address other matters regarding the institution. Each member 150 of the Rome Statute has one representative. The Assembly established a Bureau with two working groups, one in The Hague and one in New York. The working group in The Hague is responsible for the operations of the ICC. The working group in New York is responsible for addressing issues of participation by state parties, including increasing ratication by new states. The Assembly and the Bureau's working groups meet annually but may meet more frequently if needed. When the Rome Statute went into force on July 1, 2002, it set the starting point for the court's temporal jurisdiction. The Rome Statute details the jurisdiction, functioning, and nancing of the ICC in its 128 articles. Its primary task is to prosecute individuals for the international crimes of geno- cide, crimes against humanity, war crimes, and aggression (art. 5). Articles six through eight dene the crimes of genocide, crimes against humanity, and war crimes, respectively. The crime of aggression, intentionally left for later negotiations (art. 123), was eventually dened at the Review Confer- ence of the Rome Statute, which took place from May 31 to June 11, 2010, in Kampala, Uganda (the \Kampala Review Conference"). In comparing the Rome Statute to other establishing documents, it be- comes evident that it is more detailed and encompassing than the others. Figure 4.4 shows the length of the charters or statutes that established the main international criminal tribunals. As is clear, the Rome Statute is by far the longest and most complicated establishing document, by a factor of three. This is largely due to the fact the that Rome Statute is a perma- nent court with jurisdiction over numerous states (123 members as of 2022). 151 Figure 4.4: Length of Documents Establishing Int'l Criminal Tribunals Because this wide geographical and temporal reach is costly to states, the drafters had to bargain over many other provisions to clarify the parame- ters of the institution. Their eorts were designed to create a meaningful institution with strong obligations but exible enough to continue its work despite contentious politics by the state members. 4.2.3 Early Cases After the court was founded, it began the work of prosecuting the crimes under its mandate. It was slow to start at rst. Upon taking oce as the rst prosecutor, Luis Moreno-Ocampo determined that the rst situation to be investigated would be the violence in Ituri in the Democratic Republic of 152 the Congo (DRC). This began the process of the prosecutor pursuing cases of civil war in central Africa, including the situations in Uganda and Cen- tral African Republic. 20 In these cases, the member states self-referred the situations to the prosecutor. Because the ICC relies heavily on state coop- eration for its prosecutions, Prosecutor Ocampo made the decision to focus on opposition groups in his prosecutions in hopes of obtaining cooperation from the referring states (Schabas 2017, 33). Table 4.1: Situations under Investigation at the ICC Situation Region Date Opened Referred by Dem. Republic of the Congo (DRC) Africa 2004 DRC Uganda Africa 2004 Uganda Darfur, Sudan Africa 2005 U.N. Security Council Central African Republic I Africa 2007 Central African Republic Kenya Africa 2010 Prosecutor Libya Africa 2011 U.N. Security Council Cote d'Ivoire Africa 2011 Prosecutor Mali Africa 2013 Mali Central African Republic II Africa 2014 Central African Republic Georgia Eastern Europe 2016 Prosecutor (authorized) Burundi Africa 2017 Prosecutor (authorized) Bangladesh-Myanmar Asia 2019 Prosecutor (authorized) Afghanistan Asia 2020 Prosecutor Venezuela I Latin America 2021 Multiple (6) State Parties Palestine II Middle East 2021 Prosecutor Philippines Asia 2021 Prosecutor (authorized) Ukraine Eastern Europe 2022 Multiple (39) State Parties As the U.S. position regarding the ICC warmed, especially during the Obama Administration's policy of \positive engagement" (Fairlie 2011), the U.N. Security Council made referrals to the prosecutor for the situations in Dar- fur, Sudan 21 and Libya 22 in 2005 and 2011, respectively. The U.N. Security Council had attempted to refer the situation in Syria to the ICC in 2014, but 20 The Central African Republic made two referrals to the court, one in 2004 and another in 2014. 21 UN Doc. S/RES/1593 (2005). 22 UN Doc. S/RES/1970 (2011). 153 the resolution was blocked by Russia and China who both vetoed the draft resolution. 23 The Darfur matter, pushed by the U.S. through Secretary of State Colin Powell, end up planting the seeds of controversy for the court for years to come, as discussed below (See generally Peter 2016, 16{21). Prosecutor Ocampo waited several years before using his authority to initiate cases using his proprio motu power under Article 15. He did so in cases of post-election violence in Kenya and Cote d'Ivoire. In 2010, the Pre-Trial Chamber granted the Prosecutor's request for authorization to conduct an investigation of the 2008 violence in Kenya. 24 As discussed later, while the proceedings were beginning at the ICC, two of the primary defendants were elected president and vice president. Like the situation in Sudan, where the Prosecutor pursued the state leaders, the court received signicant resistance from its members. Table 4.1 lists the situations under investigation by the ICC over its years of operation. The list includes those investigations for which the Prosecutor sought and was granted authority. Other preliminary investigations by the prosecutor exist for circumstances in Colombia, Nigeria, and Iraq. Of the seventeen investigations by the Prosecutor listed, ten are with regard to situations and individuals from an African state. The early dominance of cases from African states, whether because of self-referral or not, prompted 23 Somini Sengupta. \China and Russia Block Referral of Syria to Court." The New York Times (May 22, 2014), available at https://www.nytimes.com/2014/05/23/world/middleeast/syria-vote-in-security- council.html 24 Situation in Kenya (ICC-01/09), Decision Pursuant to Article 15 of the Rome Statute on the Authorization of an Investigation into the Situation in the Republic of Kenya (31 Mar. 2010). 154 criticism that the ICC was biased and perpetrating a form of neo-colonialism. This criticism challenged the legitimacy of the court, which is addressed in detail in the next section. 4.3 Legitimacy, Eectiveness, and Growing Cri- tiques The ICC undeniably is reliant on perceived legitimacy to be eective and rel- evant. As mentioned previously, arguments about the court's ineectiveness and illegitimacy are often used by members who are unhappy with court activity. In essence, these are often the arguments raised in the bargain- ing process between the state and the institution after a shock redistributes the costs among the parties. Understanding the critiques of the ICC based on illegitimacy and ineectiveness gives a foundation for assessing the spe- cic arguments used when members threatened to withdraw from the Rome Statute. In general, legitimacy is promoted through various avenues, including the demographic background of the judicial body, selection of cases, and pro- cedural fairness. As Shapiro (2013) posited, judges use symbols of legal authority and objective legal techniques to neutrally apply preexisting rules to create a \noble lie" of judicial neutrality. The \noble lie" allows the judge to extend her power and the power of the state. Legitimacy is not only at 155 play during proceedings at the court, but also after a decision, when com- pliance is crucial, and before any prosecution when actors make decisions in the shadow of potential prosecution (Alter 2014). The ICC is a \new-style" international court, i.e. a court with compulsory jurisdiction or that is accessible to non-state actors, that aects politics in the area of international humanitarian and criminal law (Alter 2014). The ICC has a role in enforcing humanitarian and criminal law in the interna- tional sphere by holding individuals accountable for mass atrocities. This enforcement role is both self-binding on state parties, by checking their own behavior, and other-binding, imposing the will of the powerful, where cases can be initiated by the U.N. Security Council (Rome Statute, Art. 13(b)). The ICC's power is manifest through its direct jurisdiction in conjunction with embedded international law in domestic systems, allowing the ICC to connect with compliance constituencies. The court's eectiveness arises from various institutional mechanisms in na- tional jurisdictions. Jo and Simmons (2016) argue that two broad channels of deterrence of violence exist as prosecutorial deterrence and social deter- rence. Using formal theory and examining Kenya as a case study, Chaudoin (2016) argues that institutions have the greatest ability to induce compli- ance when domestic pro- and anti-compliance groups have similar desires to win compliance contests, as well as similar costs to take action to pur- sue their goals. An institution like the ICC can leverage the competition of those domestic groups to push for compliance from within. Institutions have a weaker marginal eect when groups are imbalanced. This evidence 156 supports the theories that international courts, and international law more broadly, works in part through compliance constituencies (Keck and Sikkink 1998; Alter 2014; Simmons 2009). The ICC's eectiveness at accomplishing its central mission is also crucial in promoting and maintaining its legitimacy. A growing body of scholarship nds that the ICC can and has been eective at decreasing violence. ICC investigations can deter governments from committing human rights viola- tions by imposing a variety of costs on them that decrease their expected payos for engaging in human rights abuses (Appel 2018). Appel nds that leaders from states that have ratied the Rome Statute commit lower levels of human rights abuses than non-ratier leaders. However, the existence of the ICC and the specter of possible investigation or prosecution have unintended consequences. The prospect of international criminal prosecution may increase the length of civil con ict, which makes those con icts more costly and deadly (Nalepa and Powell 2016). In ad- dition, the grand goals of the ICC have also caused a paradox in that the ICC is by its nature \weaker" than its domestic counterparts, and yet, it is called upon to handle extremely complex and sensitive cases (Damaska 2009). This often leads to a legitimacy gap between the ends and the means, where the institution's aspirations are not proportional to its capabilities. Damaska (2009) argues that streamlining the court's procedure and scaling back the ambitions of the court is key in maintaining its legitimacy. Despite its practical implications, this prescription is unsatisfying. The appeal of the ICC is exactly its ambitious goal of holding accountable those guilty 157 of some of the most heinous crimes and shedding light on those crimes for posterity. Curtailing those ambitions may do more to decrease the court's legitimacy than otherwise. Representation on the the Bench: Composition and Diversity The composition of the judges on a court is signicant in promoting le- gitimacy. Arguments in favor of diversity of a judicial body broadly fall into two related camps: arguments of function and arguments of legitimacy (Shikhelman 2017). Generally, diversity is benecial for functional reasons by bringing a wide and representative background of experience to the judi- cial process. Each judge brings with her life experiences that will in uence the way she will view a case or argument. These varying viewpoints can contribute values or ideas to the court that would otherwise be missed. Ju- dicial diversity may also help correct for bias (Berzon 2012). This increases the perceived legitimacy for the acts and decisions of the judges. Legitimacy is essential for international courts because their eectiveness and ability to enforce their decisions is largely dependent on political and social leverage, as explained above (Grossman 2009; Alter 2014). Shikhelman (2017) investigates how the individual backgrounds of decision- makers may aect outcomes and legitimacy. In empirically investigating the U.N. Human Rights Committee, she nds that the backgrounds of Commit- tee Members matter in a subset of issue areas, like immigration cases involv- ing western states. There are also many issues where background does not 158 matter. She explains that the ICC's Article 36 diversity provision is \excep- tionally detailed" (Id. at 7). Other studies nd that factors like gender or race also in uence decision-making in domestic courts (Peresie 2005; King and Greening 2007). Using anthropological and psychological methods, Se- cunda (2010) argues that values act unconsciously on judicial decision mak- ing in what is termed \cultural cognition." The future U.S. Supreme Court Justice Sonia Sotomayor recognized the in uence of her background on her own decision-making as a judge (Sotomayor 2002). Representation from both sexes is important for a number of reasons, in- cluding the creation of role models, elevating disadvantaged subgroups, and normative and socially legitimacy (Grossman 2016b). In addition, gender diversity would be an essential factor in cases involving gender-based crimes. For the most part, international courts and tribunals are imbalanced, favor- ing men to women (Grossman 2016b). Grossman (2016a) investigates the claim made by Charlesworth, Chinkin and Wright (1991) that the struc- tures of international law \privilege men." Grossman's data shows that in mid-2015, women made up 20 percent or less of nine of twelve interna- tional courts. Grossman (2016a,b) shows that where selection procedures are closed, opaque, and there is no quota or aspirational target for a sex- balanced bench, women obtain judgeships at disproportionality low num- bers. She challenges the argument that the paucity of female judges is the result of a shallow pool of qualied female candidates. Despite these hurdles, other research shows that women generally, and women of color specically, 159 are making gains on international courts (Dawuni 2017). Research on judi- cial representation of judges identifying as non-binary or a member of the LGBTQ community, especially at the international level, is lacking. It is important to note that judicial decision-making cannot simply be re- duced to some deterministic formula based on a judge's sex, gender, race, ethnicity, or experience. As professionals, judges generally strive to make decisions as impartially as they can based on the facts and the law presented to them. This is essential in maintaining the \noble lie" so that the judge's decision is viewed with legitimacy. With that said, it would be unrealistic to view a judge as a cold machine without a worldview that, at a minimum, subconsciously in uences decision-making. In short, diversity is necessary for legitimacy but not sucient without the overarching framework of fair- ness and impartiality. Regionalism and Bias The historical context of international criminal law is critical in understand- ing \legal pluralism" (Stewart and Kiyani 2017). At its most basic level, the ICC re ects an embedded European approach to international law enforce- ment (Alter 2014, 146). The ICC has been vigorously accused of bias against some states or regions, particularly in Africa (Schmitt 2016; Seymour 2016). Clarke (2009) argues that the ICC is linked to North-South inequalities and the rise of neoliberal practices and ideas; individual liability promotes a Western elite view of justice. Beyond Africa, Asia and the Middle East are 160 generally outsiders as a result of their relative lack of international courts (Alter 2014, 137{138). Middle Eastern states, like Iran, view the ICC as both a tool to use against the aggression of other states and as problematic in that its principles con ict with cultural norms and Shari'a law (Abtahi 2005). Various crimes and punishments under Iranian criminal law, like cap- ital punishment, whipping, stoning, and removal of limbs, present signicant friction with the Western-oriented norms of the Rome Statute. Moreover, under Shari'a principles, Muslims may now be judged by non-Muslim judges. This is both problematic and benecial for states like Iran, which may not see any decisions of the court as binding on its Muslim citizens. Asia has a notoriously uneasy relationship with the Court (See generally Schuldt 2015; Sperfeldt and Palmer 2016; Chesterman 2016). Kim (2017) notes that Asia, followed by, Africa are the least represented regions of signa- tories to the Rome Statute. The uneasy relationship is attributed to the his- torical attitudes of sovereignty and non-interference; a culture of impunity; and limited nancial, technical, and human resources. Asian and African countries shared many of these attitudes and features given their similar histories. However, the African states developed a more positive orientation to the ICC based on their recent experience of atrocities, in particular in Rwanda, and the coordination of many nations under organizations like the Southern African Development Community (SADC) or the Organisation of African Unity (OAU), which served as a focal point for advocacy in the negotiations of the Rome Statute. States in Asia did not have the transna- tional cooperation and regional unity that the African states had as the ICC 161 took shape. Asian countries make up only 35 percent of the membership of the ICC, despite constituting a majority of the world's population (Kim 2017). China and India, the two most populous nations on the planet, are both critical of the ICC and have yet to sign the Rome Statute (Jianping and Zhixiang 2005; Ramanathan 2005). The legitimacy of international courts is not simply an academic exercise. The legitimacy crisis has led to a number of states threatening withdrawal from the Rome Statute, but currently, only two have in fact withdrawn, i.e. Burundi and the Philippines (discussed in more detail below). This looming crisis lead to various shifts in the personnel at the court. Ms. Fa- tou Bensouda, a Gambian lawyer, was appointed for a term beginning in 2012 and ending in 2021. 25 The Prosecutor's duties include the investiga- tion and prosecution of the crimes under the jurisdiction of the ICC, which arguably was an attempt by the court and state parties to address concerns of legitimacy over the selection procedure of cases that are investigated and brought to the court. As Hafetz (2017) explains, the ICC's selection deci- sions leave much to be desired in prompting legitimacy. He argues that the international criminal courts should focus more on the \distributive consid- erations" in choosing among the unfortunate multitude of atrocities in the world. However, prosecutorial decisions and situation selection are inher- ently political, even if guided by law, which should not itself undermine the court's legitimacy (See Zakerhossein 2017; Rashid 2021). 25 Farouk Chothia, \Africa's Fatou Bensouda is new ICC chief prosecutor," BBC (Dec. 12, 2011), available at http://www.bbc.com/news/world-africa-16029121. 162 4.4 Bargains and Exits: Reforming the Institution for the Right Member As with most international treaties, the Rome Statute explicitly allows a state to exit. Article 127 of the Rome Statute provides the details of with- drawal under the treaty, stating that withdrawal shall take eect one year after the date of receipt of notication. The Rome Statute is arguably a fairly rigid institution subjecting individuals in member states to prosecution both in domestic and international courts. The principles of admissibility and complementarity (art. 17), which gives a member state the opportunity to enforce the provisions of the treaty itself, allow exibility in whether an investigation or case is brought. However, matters may be referred to the ICC by other members and the UN Security Council, as well as the pro- prio motu authority of the prosecutor (arts. 13{15, 53), posing signicant risks of prosecution that may be outside a member state's control. Mem- bers are also required to cooperate with the ICC and align their national legal systems with its provisions (arts. 86{93). Additionally, member states are required to fund the ICC based on a U.N. scale that assesses a state's population and wealth (art. 117), which adds a direct material or nancial cost to participating. Despite the fact that the ICC provides both public and private goods, many African leaders have criticized the ICC for being imperialistic and biased against African states. Led by Kenyan President Uhuru Kenyatta and South 163 African President Jacob Zuma, the African Union (AU) held a special sum- mit on the issue of calling for withdrawal from the ICC on October 13, 2015. 26 Those working on diplomacy between the ICC and African states saw the push for the AU withdrawal as coming from non-Rome Statute members, namely Sudan and Egypt. 27 The AU failed to endorse mass with- drawal for lack of support; however, the seeds had been planted. Burundi announced that it would withdraw from the ICC on October 7, 2016, becoming eective on October 27, 2017. 28 South Africa and The Gambia both notied an intent to withdraw in 2016; however, they rescinded those notices a few months later. 29 On March 17, 2018, the Philippines also announced its intent to withdraw from the Rome Statute because of what it saw as eorts to \politicize and weaponize human rights." 30 This came less than two years after Rodrigo Duterte, the country's populist and nationalist president, took oce. President Duterte's term has been characterized by 26 Somini Sengupta, \As 3 African Nations Vow to Exit, International Court Faces Its Own Trial," The New York Times (Oct. 26, 2016), available at https://www.nytimes.com/2016/10/27/world/africa/africa-international-criminal- court.html. 27 Author interviews (Interviews E, G), October 2019. 28 Jina Moore, \Burundi Quits International Criminal Court" The New York Times (Oct. 27, 2017), available at https://www.nytimes.com/2017/10/27/world/africa/burundi-international-criminal- court.html. 29 Norimitsu Onishi, \South Africa Reverses Withdrawal From Interna- tional Criminal Court," The New York Times (Mar. 8, 2017), available at https://www.nytimes.com/2017/03/08/world/africa/south-africa-icc-withdrawal.html. 30 Patricia Lourdes Viray. \Philippines formally informs UN of ICC withdrawal" The Philippine Star (Mar. 16, 2018), available at https://www.philstar.com/headlines/2018/03/16/1797330/philippines-formally-informs- un-icc-withdrawal. 164 a bloody domestic drug war often criticized by international observers. 31 Tridgell (2018) points out that the recent withdrawals provide a challenge for the Court and Prosecutor to broaden the scope of investigations and cases in a compressed timeframe. Yet, the ICC has addressed potential exit in a number of cases. In certain situations, it strategically exercised its exibility to reform the institution to avoid the exit. This decision is largely based on the nature of the state threatening the withdrawal and whether the ICC believes that exit would be a long-term blow to the institution. 4.4.1 Kenya The bargaining and ultimate success of the ICC in avoiding Kenya's exit is instructive to understanding how the court's exibility was strategically em- ployed. In 2007, violence broke out in Kenya following mishandled national elections. The elections were characterized by tribal hostility between the Kikuyu, Embu, and Meru, on the one hand supporting Mwai Kibaki, and the Luo, Luhya, Kalenjin, and Muslim constituencies, on the other hand supporting Raila Odinga. Kibaki was declared the winner, but Odinga also claimed victory. Protests broke out across the country, including in Mom- basa and Nairobi. Police responded with deadly force, shooting demonstra- tors, which en amed tensions. The violence is estimated to have cost over 31 \Philippines 2017/2018 Report." Amnesty International, available at https://www.amnesty.org/en/countries/asia-and-the-pacic/philippines/report- philippines/. 165 1,000 lives and displaced hundreds of thousands more. 32 Kenyatta was ac- cused of instigating the violence to support Kibaki and target his opponents. Following the violence, the Kenyan government created the international Commission of Inquiry on Post-election Violence chaired by the Hon. Justice Philip Waki, Judge of the Court of Appeal of Kenya (\Waki Commission"). The Waki Commission's mandate was to investigate the matter and oer recommendations to the government (Peter 2016, 21). After considering the violence, the Commission produced a bold report that called for the end to impunity and the creation of a local court to hold those responsible for the violence accountable. 33 \There must be an end to this cowardice and pious sense of self preservation in public service if impunity will be eradicated in this country!" 34 The report did not, however, publicly disclose the names of the alleged perpetrators of the the crimes (Peter 2016, 22). Instead, the list of alleged perpetrators was given to U.N. Secretary General Ko Anan, who then passed it on to ICC Prosecutor Ocampo. 35 The Waki Commission's recommended that a special tribunal be established was considered by the Kenyan government, but it was never able to eectively act. The Kenyan Parliament in fact voted twice in favor of referring the matter to the ICC. There were competing factions in the country, some favoring prosecution 32 \Kenya election: Uhuru Kenyatta wins presidency," BBC (Mar. 9, 2013) available at http://www.bbc.com/news/world-africa-21723488. 33 Waki Commission Report, pp. 472{476. 34 Waki Commission Report, p. 460. 35 \Annan did not ambush Kenya, says Justice minister." Nation (July 13, 2009, updated July 3, 2020), available at https://nation.africa/kenya/news/1056-623306- ukx8q9/index.html 166 and others opposed. 36 The Pre-Trial Chamber II granted the Prosecutor's request to open an in- vestigation into alleged crimes against humanity in Kenya on March 31, 2010. A summons was issued for Kenyatta and others almost a year later. He was accused of ve counts of crimes against humanity in the context of the 2007-2008 post-election violence in Kenya. The conrmation of charges \rocked" Kenyan politics (Peter 2016, 23). Charges against some of the accused individuals were not conrmed by the Pre-Trial Chamber II shortly after proceedings began. In the midst of the legal proceedings, Kenyatta was elected President of Kenya. This led to the court's decision to allow the proceedings to proceed without Kenyatta or Ruto, Deputy-President, need- ing to be physically present for all proceedings. The proceedings included the participation of 725 victims. Shortly after the charges were conrmed, MP Issac Ruto (supporter of the accused Ruto but not related) rallied support in the Kenyan parliament to endorse his 2010 motion to withdraw Kenyan from the Rome Statute. 37 Although the motion passed, the Kenyan Prime Minister Raila Odinga (a competitor of Kenyatta's in the presidential election) rejected it and refused 36 James Macharia. \Kenya's Iron Lady backs ICC trials for poll violence." Reuters (April 4, 2011), available at https://www.reuters.com/article/ozatp-kenya-icc-opposition- 20110404-idAFJOE7330H520110404 37 \Parliament pulls Kenya from ICC treaty." Nation (Dec. 22, 2010, updated July 04, 2020). 167 to pull out of the ICC. 38 Article 2(6) of the 2010 Kenyan Constitution pro- vides that any treaty or convention ratied by Kenya is made a part of the law of Kenya subject to the Constitution. In addition, the Treaty Making and Ratication Act No. 45 of 2012 provides for the procedure of treaty mak- ing and ratication pursuant to article 2(6) of the Constitution. Generally, \the national executive shall be responsible for initiating the treaty making process, negotiating and ratifying treaties" (section 4(1).) Section 17 of the Act provides that when the government wishes to withdraw from a treaty, it must produce a \memorandum indicating the reasons for such an intention" but the general rules for entering into a treaty apply. In September 2013, again the Kenyan Parliament threatened to exercise the country's right to withdraw from the Rome Statute under Article 127. 39 The vote was led by majority leader Adan Duale, who argued: \Let us protect our citizens. Let us defend the sovereignty of the nation of Kenya." 40 At the time, no other country had withdrawn from the Rome Statute. With the threat of Kenya following through on its threat to leave the Rome Statute, and be the rst to do so, the Court and its members sprung into action to negotiate behind the scenes. 41 38 \Kenya `opposes' leaving ICC." Aljazeera News (Dec. 24, 2010), available at https://www.aljazeera.com/news/2010/12/24/kenya-opposes-leaving-icc; \Kenyan par- liament dismisses plea for ICC pull-out." Reuters (Dec. 21, 2010), avail- able at https://www.reuters.com/article/uk-kenya-icc/kenyan-parliament-dismisses-plea- for-icc-pull-out-idUKTRE6BK3N020101221 39 \Kenya MPs vote to withdraw from ICC." BBC News (Sept. 5, 2013), available at https://www.bbc.com/news/world-africa-23969316 40 Id. 41 Author interviews (Interviews D, E, F), October 2019. 168 The Rome Statute is silent as to the issue of reconsideration of decisions of the Appeals Chamber (Schabas 2017, 317). However, the Assembly of States may act to eectively \overrule" the Appeals Chamber. 42 In Ruto, et al. and Kenyatta, the Trial Chamber ruled that the accused could waive his right to be present at trial because of his ocial duties as president or deputy-president of Kenya. The Trial Chamber imposed a number of condi- tions, requiring the accused's presence at certain points in the trial. In the ICC, the judges in the Trial Chamber are given broad latitude in determining the procedure and ow of trial (Rome Statute, art. 64). The Appeals Cham- ber held that the Trial Chamber had exercised its discretion \too broadly" and reversed. 43 However, shortly after the Appeals Chamber decision, the Assembly of States amended the Rules of Procedure and Evidence using lan- guage from the Trial Chamber, essentially overruling the Appeals Chamber (Schabas 2017, 317{318). Thus, the Trial Chamber's broad authority was reinforced. Kenyatta eventually did appear before the court, being the rst sitting head of state to do so, even though technically appearing as a private citizen| having temporarily transferred his powers to his deputy. Days later, the new ICC prosecutor, Fatou Bensouda, dropped the charges against Kenyatta in December 2014, partly due to a lack of cooperation between Kenya and the court in the process of gathering evidence (Peter 2016, 37). This was seen as 42 See Ruto, et al. (ICC-01/09-01/11), Separate Further Opinion of Judge Eboe-Osuji to the \Reasons for the Decision on Excusal from Presence at Trial under Rule 134 quater", 18 February 2014, para. 22; and Ruto, et al. (ICC-01/09-01/11), Reasons for the Decision on Excusal from Presence at Trial under Rule 134 quater, 18 February 2014, para. 79. 43 Kenyatta (ICC-01/09-02/11), Decision on Defense Request for Conditional Excusal from Continuous Presence at Trial, 18 October 2013, para 63. 169 a major blow to the court and the Prosecutor, especially where it appeared that the previous Prosecutor|Luis Moreno-Ocampo was replaced by Fatou Bensouda in the midst of the proceedings|had aggressively pursued the case in the novel circumstance of electoral violence. Despite dismissing the case, the Prosecutor made all the evidence against Kenyatta public and reported Kenya to the Assembly of State parties. The episode with Kenya highlighted the strain and excessive burden put on states when their current leaders are the target of prosecution (Peter 2016, 21{26). The AU broadly sought to limit prosecutions of heads of state and strengthening immunity, which was largely initiated by the indictment of al-Bashir as discussed below (Sadat and Cohen 2016). The ICC and Assembly of State Parties worked behind the scenes to per- suade Kenya against leaving the Rome Statute. 44 Key actors were diplo- mats from Latin American countries who worked on compromise positions. 45 They sometimes did so using incentives of preferential treatment or security benets. 46 Kenyan's relationship with states like the United Kingdom and the United States, as well as the E.U., were likely important in its ulti- mate decision not to withdraw. 47 When the charges were announced, US President Obama urged cooperation with the ICC: I urge all of Kenya's leaders, and the people whom they serve, to cooperate fully with the ICC investigation and remain focused on implementation of the reform agenda and the future of your 44 Author interviews (Interviews D, E, F), October 2019. 45 Author interview (Interview E), October 2019. 46 Author interviews (Interviews E, F), October 2019. 47 Author interviews (Interviews D, E, F), October 2019. 170 nation. Those found responsible will be held accountable for their crimes as individuals. No community should be singled out for shame or held collectively responsible. Let the accused carry their own burdens|and let us keep in mind that under the ICC process they are innocent until proven guilty. As you move forward, Kenyans can count on the United States as a friend and partner. 48 The US continued to urged Kenyan cooperation with the court. Addition- ally, Europeans states, many of which were also members of the Rome Statute, felt the strain of the debate over the ICC case and threatened Kenyan withdrawal. 49 The E.U. had favored a \local judicial solution to the post-election violence" to avoid potential problems with a prosecution by the ICC based in Europe. 50 Also, the ICC likely understood that Kenyans were divided over the issue, allowing room to bargain. The public in Kenya did not favor withdrawal from the Rome Statute, with only 34 % supporting and 55 % opposing with- drawal in 2014. 51 In general, Kenyan's support over the years has generally favored ICC intervention to prevent impunity by perpetrators. 52 48 \Statement by President Obama on the International Criminal Court announcement." White House (Dec. 15, 2010), available at https://obamawhitehouse.archives.gov/the-press-oce/2010/12/15/statement-president- obama-international-criminal-court-announcement 49 Je Otieno. \The ICC tested the strength of the EU-Kenya rela- tionship but we survived." The East African (Mar. 14, 2015), available at https://www.theeastafrican.co.ke/tea/news/east-africa/the-icc-tested-the-strength-of- the-eu-kenya-relationship-but-we-survived{1333658 50 Id. 51 Afrobarometer Policy Paper no. 23 (August 2015), available at https://afrobarometer.org/publications/pp23-support-international-criminal-court- africa-evidence-kenya 52 Id. 171 The member states and the court worked with extra vigor to avoid Kenya leaving, seeing it as an important member to the ICC's legitimacy. 53 The Assembly of Parties had agreed to take up Kenya's proposal to prevent future prosecution of current heads of state 54 , eventually amending its procedures. Kenya was one of the \critical states" in that the institution worried that a Kenyan withdrawal would begin a cascade of withdrawals. 55 Further, Kenya was a populous and important country in Africa that had been working toward moving to improving justice. Leaving the Rome Statute would have been a substantial blow to the project of promoting justice and holding powerful perpetrators of violence to account. 4.4.2 South Africa Like Kenya, South Africa is an important and in uential state on the African continent. South Africa quickly joined the ICC on November 27, 2000. At the same time, the South African Parliament enacted the Implementation of the Rome Statute of the International Criminal Court Act 27, which incorporated the Rome Statute into domestic law in accordance with the country's constitution. After the period of apartheid, South Africa generally took pride in moving toward justice and accountability. That pride was ultimately challenged under the leadership of President Jacob Zuma. 53 Author interview (Interview E), October 2019. 54 Judie Kaberia. \Win For Africa As Kenya Agenda Enters ICC Assembly." Capital News (Nov. 20, 2013), available at https://www.capitalfm.co.ke/news/2013/11/win-for- africa-as-kenya-agenda-enters-icc-assembly/ 55 Author interview (Interview E), October 2019. 172 As mentioned above, the U.N. Security Council referred the situation in Darfur, Sudan to the prosecutor in 2005. 56 U.S. Secretary of State Colin Powell led eorts at the U.N. to pursue the referral to the ICC, invoking article VIII of the Genocide Convention. After the Sudanese government ig- nored the Pre-Trial Chamber's arrest warrants for a government ocial and militia leader, the Prosecutor obtained a warrant for the country's leader, Omar al-Bashir. Igniting a \restorm," this caused concern in the African Union (AU) because the Prosecutor's actions might undermine the AU's peace eorts in Darfur (Schabas 2017, 36{38). The AU submitted a request to the U.N. Security Council to defer the investigation and prosecution at the ICC pursuant to Article 16 of the Rome Statute. 57 The Security Coun- cil noted the AU concerns but did not take action. 58 In response, the AU adopted a resolution that because its request had not been acted upon, the AU member states would refuse to cooperate in the arrest or surrender of President al-Bashir. 59 Thereafter, al-Bashir traveled throughout the AU in deance of the ICC's arrest warrants, knowing that the AU member states would not cooperate with the court. South Africa hosted the 25th Summit of the AU in Johannesburg in early June 2015. On June 13, 2015, al-Bashir landed in South Africa to attend the summit. According to Article 86 of the Rome Statute, member states 56 UN Doc. S/RES/1593 (2005). 57 African Union Peace and Security Council Decision (PSC/MIN/Comm (CXLII)), 21 July 2008,{{ 3, 5, 9, 11(i). 58 UN Doc. S/RES/1828 (2008). 59 Decision on the Meeting of the African States Parties to the Rome Statute of the International Court (ICC), Dec. Assembly/AU/13 (XIII),{ 10. 173 are required to \cooperate fully with the Court in its investigation and pros- ecution of crimes within the jurisdiction of the Court." This provision had been incorporated into domestic law. However, Zuma's cabinet had agreed secretly with the AU to grant diplomatic immunity to all delegates attending the AU summit. 60 The Southern Africa Litigation Centre, a Johannesburg- based non-prot supporting human rights lawyers in southern Africa coun- tries, had successfully sued to force the government to arrest al-Bashir while in the country. 61 However, the Zuma government did not detain al-Bashir and had misrepresented whether he was in the country to the courts, war- ranting the High Court to call the government's behavior \disgraceful." 62 This led to an internal crisis in South Africa. The domestic and interna- tional backlash led to the Zuma government's threat to withdraw from the Rome Statute. South Africa submitted its formal notice to withdraw on October 19, 2016. 63 It notied the ICC along with Burundi and The Gambia, worrying some that the future of the institution was at stake. While the withdrawals of 60 Qaanitah Hunter, Mmanaledi Mataboge, and Phillip De Wet. \How Zuma and ministers plotted Omar al-Bashir's escape." Mail & Guardian (Jun 18, 2015), available at https://mg.co.za/article/2015-06-18-how-zuma-and-ministers-plotted-omar-al-bashirs- escape/ 61 Southern African Litigation Centre v Minister of Justice and Constitutional Devel- opment and Others, Gauteng Division of the High Court of South Africa (27740/2015) Decision (23 June 2015). 62 \South African court rules failure to detain Omar al-Bashir was `disgraceful.' " The Guardian (Mar. 15, 2016), available at https://www.theguardian.com/world/2016/mar/16/south-african-court-rules-failure- to-detain-omar-al-bashir-was-disgraceful 63 Sewell Chan and Marlise Simons. \South Africa to Withdraw From In- ternational Criminal Court," The New York Times (Oct. 21, 2016), avail- able at https://www.nytimes.com/2016/10/22/world/africa/south-africa-international- criminal-court.html. 174 smaller, less important nations might be tolerable, the withdrawal of South Africa would be a signicant blow to the court. One commentator made this explicit: \The Burundi decision was easy to dismiss as a government seeking to avoid direct scrutiny; South Africa's is much more signicant. The African Union has been a forum for anti-I.C.C. sentiment, and countries like Kenya and Uganda may now seek to capitalize on the momentum." 64 Other African nations, including Kenya, paid close attention to South Africa's withdrawal and the domestic wrangling over whether it was valid. 65 South Africa's decision was embarrassing to many in the country, which generally prides itself on valuing human rights and international coopera- tion after casting o the yoke of apartheid, as mentioned above. 66 It was particularly troubling how the Zuma government had outed the order of the judiciary over the al-Bashir controversy. To add to the embarrassment, the South African government's decision to withdraw appeared to be less of a deliberate process and designed more to avoid the issues. 67 The ICC was concerned by South Africa's decision to withdraw being an important member, especially among African nations. 68 The ICC inter- nally takes accusations of bias and prejudice seriously, especially when they 64 Id. 65 See, e.g., Linda Awuor and Faith Wanjiku. \The Decision by the National Execu- tive to Deliver the Notice of Withdrawal of South Africa from the Rome Statute of the ICC without Prior Parliamentary Approval was Unconstitutional and Invalid." Kenya Law: Blog (Mar. 22, 2017), available at http://kenyalaw.org/kenyalawblog/notice-of- withdrawal-of-south-africa-from-the-rome-statute-of-the-icc/ 66 Author interview (Interview G), October 2019. 67 Author interview (Interview G), October 2019. 68 Author interview (Interview F), October 2019. 175 come from in uential states. 69 However, the ICC was aware, like the public generally, that South African civil society and the judiciary was active in addressing the government's decision. With that backdrop, the ICC and Assembly of States, although concerned, was less anxious about the eects of a South African exit. Eventually, that strategy was vindicated when the South African High Court ruled the withdrawal unconstitutional and it was rescinded on March 7, 2017 (See generally Bradley and Helfer 2017; Woolaver 2017, 2019). 70 Less than a year later, Jacob Zuma was forced to resign the presidency under pressure from his own party and the public. 4.4.3 Burundi and The Philippines From the ICC's perspective, the withdrawal of other states was less threat- ening to the institution. Both Burundi and The Philippines exited from the Rome Statute in 2016 and 2018, respectively. 71 In each case, the leaders of the countries faced the prospect of personal prosecution based on their brutal use of force on their own populations. Burundi is a small, landlocked country in sub-saharan Africa, which has unfortunately suered numerous bloody con icts and civil wars into its 69 Author interview (Interview E), October 2019. 70 Norimitsu Onishi, \South Africa Reverses Withdrawal From Interna- tional Criminal Court," The New York Times (Mar. 8, 2017), available at https://www.nytimes.com/2017/03/08/world/africa/south-africa-icc-withdrawal.html. 71 Jina Moore, \Burundi Quits International Criminal Court" The New York Times (Oct. 27, 2017), available at https://www.nytimes.com/2017/10/27/world/africa/burundi-international-criminal- court.html; Patricia Lourdes Viray. \Philippines formally informs UN of ICC withdrawal" The Philippine Star (Mar. 16, 2018), available at https://www.philstar.com/headlines/2018/03/16/1797330/philippines-formally-informs- un-icc-withdrawal. 176 modern history. Nonetheless, Burundi had joined the Rome Statute on September 21, 2004 as a sign of progress. In 2015, after a constitutional crisis involving President Pierre Nkurunziza's desire to extend his rule for a third term, the military's coup d'etat failed. After the 2015 coup attempt, President Nkurunziza purged the government and arrested opposition g- ures, creating a humanitarian crisis in the region. State agents, police, and members of the \Imbonerakure" 72 systematically attacked the civilian pop- ulation. Evidence also surfaced of mass graves. 73 The U.N. Human Rights Council has establish a commission of inquiry to investigate the alleged hu- man rights abuses in the country. A preliminary investigation by the ICC prosecutor was announced on April 25, 2016. Six months later, Burundi announced its intention to leave the ICC. 74 However, the Pre-Trial Chamber III authorized the Prosecutor to open an investigation of alleged crimes that took place in Burundi from December 1, 2004 to October 26, 2017. 75 Given the timing of the announced investi- gation and notice of withdrawal, it appears the government is simply trying 72 A rebel group developed out of disarmed ghters from the ruling party's previous incarnation. The word \Imbonerakure" means \those who see far" in the Kurundi lan- guage. Human rights groups worry that the group with tens of thousands of members had become more radical and violent in recent years. 73 \Burundi: Satellite evidence supports witness accounts of mass graves." Amnesty International (Jan. 28, 2016), available at https://www.amnesty.org/en/latest/news/2016/01/burundi-satellite-evidence-supports- witness-accounts-of-mass-graves/ 74 Jina Moore, \Burundi Quits International Criminal Court" The New York Times (Oct. 27, 2017), available at https://www.nytimes.com/2017/10/27/world/africa/burundi-international-criminal- court.html. 75 Public Redacted Version of \Decision Pursuant to Article 15 of the Rome Statute on the Authorization of an Investigation into the Situation in the Republic of Burundi", ICC-01/17-X-9-US-Exp, 25 October 2017 (ICC-01/17-9-Red) Pre-Trial Chamber III (09 November 2017). 177 to thwart eorts at prosecuting government ocials. The situation in the The Philippines is similar. The Philippines joined the ICC on November 1, 2011. Rodrigo Duterte, who had served as Mayor of Davao City for decades, o and on, was elected president of The Philippines in 2016. While Duterte was Mayor of Davao, the \Davao Death Squad" al- legedly committed widespread extra-judicial killings of suspected drug users and petty criminals. 76 On assuming oce as president, he carried over these practices with his \war on drugs." 77 Opponents of Duterte were also subject to retaliation, including the jailing of Senator Leila de Lima on trumped up drug charges. Human Rights Watch estimates that over 12,000 individuals have been killed as the result of Duterte's drug war. 78 Duterte also lashed out at international criticism from the U.S., E.U., and U.N. 79 On February 8, 2018, the ICC Prosecutor announced that she will investi- gate the situation in The Philippines based on reports of the government sponsored extra-judicial killings. 80 Just over a month later, the Duterte gov- ernment announced that The Philippines would withdraw from the Rome 76 Felipe Villamor. \Ex-Ocer in Philippines Says He Led Death Squad at Duterte's Behest." The New York Times (Feb. 20, 2017), available at https://www.nytimes.com/2017/02/20/world/asia/rodrigo-duterte-philippines-death- squad.html 77 \Philippine death squads very much in business as Duterte set for presidency." Reuters (May 24, 2016), available at https://www.reuters.com/article/us-philippines- duterte-killings-insight-idUSKCN0YG0EB 78 \Philippines: Duterte's `Drug War' Claims 12,000+ Lives." Human Rights Watch (Jan. 18, 2018), available at https://www.hrw.org/news/2018/01/18/philippines- dutertes-drug-war-claims-12000-lives 79 Felipe Villamor. \Philippine Police Resume War on Drugs, Killing Dozens." The New York Times (Feb. 2, 2018), available at https://www.nytimes.com/2018/02/02/world/asia/philippines-drug-war.html 80 \Statement of the Prosecutor of the International Criminal Court, Fatou Ben- souda, on opening Preliminary Examinations into the situations in the Philippines 178 Statute. On 15 September 2021, Pre-Trial Chamber I authorized the Pros- ecutor to begin an investigation to alleged crimes that occurred in the con- text of the \war on drugs" in The Philippines between November 2011 and March 2019. 81 Like the case in Burundi, the timing of the events appear to show that the country left the ICC to thwart eorts to prosecute. The high risk of prosecution by the ICC based on these events likely in- creased Burundi, namely President Nkurunziza's|as well as Philippine Pres- ident Duterte's|status quo costs of remaining in the ICC's jurisdiction. It also likely raised its perceived costs of breach where both presidents valued using violent tactics of suppression to continue his hold on power. However, despite both countries withdrawing from the court, the investigations can continue since the alleged crimes occurred while each states were parties (Rome Statute, art. 127(2)). Sta at the ICC were less concerned with the withdrawals of Burundi and The Philippines, despite the troubling nature of the crimes, where they saw these decisions to be idiosyncratic based on the particular leaders. 82 They also did not see a signicant risk of contagion based on those exits because those leaders had little support from the international community. Some at the ICC took a patient approach, believing that a change in leadership in those countries would likely be sucient for them to rejoin. 83 and in Venezuela." Oce of the Prosecutor, International Criminal Court (8 Febru- ary 2018), available at https://www.icc-cpi.int/news/statement-prosecutor-international- criminal-court-fatou-bensouda-opening-preliminary-0. 81 Decision on the Prosecutor's request for authorisation of an investigation pursuant to Article 15(3) of the Statute (ICC-01/21-12) Pre-Trial Chamber I (15 September 2021). 82 Author interviews (Interviews E, F), October 2019. 83 Author interview (Interview D), October 2019. 179 4.5 Assessing the Theory As the cases described above show, the ICC, Assembly of Parties, and its members are able to bargain to avoid a potential exit. However, the ICC is strategic on how it employs its tools of exibility. This is largely based on the relative importance of the member and risk that the withdrawal will have a contagion eect. As a staer at the ICC noted \countries matter for dierent reasons," particularly in their ability to entice other states to join (or leave). 84 Broadly, this conrms one aspect of the theory developed in Chapter 2, where a bargain to account for a shock should be more feasible where the treaty member is important to the value of the treaty, all else being equal (Claim 2). This case can speak to other expectations of the theory. As described above, the Rome Statute is a detailed treaty that imposes heavy obligations on its members states. Using the metrics developed in Chapter 3, the Rome State is approximately 35,315 words long and has a permissive-obligatory score of 315 (see Appendix C). As described in table 3.1 of Chapter 3, the Rome Statute falls on the longer and more obligatory portion of the data analyzed there. Out of the 623 treaties examined in Chapter 3, the average permissive-obligatory score was 31.5. According to that measure, the Rome Statute is ten times more obligatory than the average treaty. Its length also decreases the room for interpretation, and thus it is less exible. As such, we should expect it to suer from an exit and more exits (Hypos 1 and 4), 84 Author interview (Interview E), October 2019. 180 which appears to initially be conrmed. Yet, the ICC's bargaining strategy described here seems to highlight how an institution can strategically employ its exibility to avoid an exit in any particular case. As mentioned in Chapter 3, the Rome Statute does not allow reservations (art. 120). Therefore, no reservations would be registered to lighten the obligations in the treaty (Hypo 3). The fact that reservations are banned is an explicit indication that this is a treaty with strong obligations from the start. Without the ability to reserve any obligations from the treaty, at least 16 interpretative declarations have been registered in relation to the Rome Statute. 85 That number is signicantly larger than the average for treaties on the whole, and we should understand this to be a signal of the treaty's strong obligations (Hypo 2). The Rome Statute allows has a high number of members, which should make it dicult to negotiate reform (Hypo 5). However, as mentioned, strategic bargaining by the ICC and Assembly of Parties appears to mitigate the po- tential problems of collective bargaining across a wide group of states. As mentioned above, key actors in working behind the scenes were diplomats from Latin American countries who worked on compromise positions. 86 Fur- ther, the ICC leveraged domestic actors to not only encourage compliance but avoid exit (See Alter 2014; Chaudoin 2016). Again, this allowed the institution to avoid some of the problems incumbent on collective action. 85 The following states registered declarations: Australia, Belgium, Cape Verde, Colom- bia, Egypt, France, Guatemala, Israel, Jordan, Luxembourg, Malta, New Zealand, Portu- gal, Sweden, United Kingdom, and Uruguay. 86 Author interview (Interview E), October 2019. 181 From the perspective of the ICC, each state faced with a decision to with- draw from the Rome Statute was forced into that position by exogenous circumstances. 87 Based on interviews of sta at the ICC and a withdraw- ing member state, the participants generally resist the broad narratives re- garding the rationale for withdrawal. 88 Especially in the context of South Africa, they saw the decision to withdraw to not be based on AU pressure but rather internal pressure from civil society that demanded action in light of al-Bashir's visit to the country. 89 Yet, they acknowledge that certain states use economic incentives to discourage exit or encourage accession. The states active in this type of behavior are western European countries and Canada. 90 That external in uence can sometimes hurt the ICC's bargaining position. Members of the ICC see the arguments of powerful and in uential non- members, especially the U.S., as fuel for the arguments of member states that the court is illegitimate or exceeding its mandate in some way. 91 As was the case with the AU call for mass withdrawal, African states pointed to the allegations of abuses and crimes by the U.S. in Iraq and the lack of prosecution as hypocrisy (Seymour 2016). Furthermore, the U.S. has not only opposed the ICC rhetorically but even went so far as to sanction its members. 92 87 Author interviews (Interviews E, F, G), October 2019. 88 Author interviews (Interviews E, F, G), October 2019. 89 Author interview (Interview G), October 2019. 90 Author interviews (Interviews D, E), October 2019. 91 Author interviews, October 2019. 92 \International Criminal Court ocials sanctioned by US." BBC News (Sept. 2, 2020), available at https://www.bbc.com/news/world-us-canada-54003527 182 The ICC and Assembly of States largely takes a public \hands-o" ap- proach to bargaining with members, but it will internally work hard to resolve con ict through diplomacy in the Assembly. 93 Many less serious is- sues are addressed through bargaining, including horse-trading on judicial appointments. 94 The ICC also quietly leverages its members, NGOs, and civil society in the bargaining process (See Chaudoin 2016). For example, organizations like the Coalition for the International Criminal Court work across borders to promote the work of the ICC and team up with local part- ners to change policy within states with regard to the ICC. The public of African states remain largely supportive of the mission of the ICC, diverging from the preferences of their leaders. 95 With this divergence, the ICC can sometimes aord to wait until governments or policies change. 96 Where the Prosecutor has shown a willingness to act independently to pros- ecute individuals for violent crimes, the ability to eectively cheat on obli- gations and avoid punishment diminishes. Moreover, the structure of the ICC's jurisdiction makes it dicult to untether a state from its obligations after violations have been committed. This is the case in Burundi and The Philippines. Therefore, the ICC can aord to make less concession in those cases. Thus, exit would be more likely in those circumstances. 93 Author interviews, October 2019. 94 Author interview (Interview G), October 2019. 95 Afrobarometer Policy Paper no. 23 (August 2015), available at https://afrobarometer.org/publications/pp23-support-international-criminal-court- africa-evidence-kenya 96 Author interview (Interview D), October 2019. 183 4.6 Conclusion Despite the wave of criticism that bueted the ICC over the last decade, it has weathered the storm by employing its selective exibility. Some fear that the world is entering a challenging time of great power competition and diminished resolve to address atrocities committed during armed con ict. This \crisis of multilateralism" is particularly troubling for institutions like the ICC. 97 Nonetheless, as shown in this chapter, the ICC is able to nimbly navigate many of those challenges by strategically employing its levers of exibility. Where a state is important to the long-term legitimacy and survival of the court, it may be willing to take bold action. This is not to say that the judges and prosecutor do not follow the lead of the fact and the law. The other branches of the institution, namely the Assembly of State Parties, is quietly active in continuing to bargain with its members to hold the orga- nization together. Moreover, the sta at the court work with governments and transnational groups to promote the ICC's work. As Zvobgo (2019) shows through survey experiments, attitudes toward the ICC can be im- proved by demonstrating its eectiveness and impartiality. Thus, there is continued hope that the court can leverage its exibility and external help in preventing a detrimental exit from the institution. With this in mind, we should most expect exit from the institution when the obligations to avoid mass atrocities are based on past crimes of a leader who does not have much 97 Author interview (Interview E), October 2019. 184 support internationally and the state's exit will not be seen as a blow to the court's legitimacy. 185 Chapter 5 Arms Control: Breach or Exit Cooperation and understanding are built on deeds, not words. Complying with agreements helps; violating them hurts. | Ronald Reagan, 40th President of the United States 1 No, I don't think we're violating; I think we're withdrawing from it. | John Bolton, U.S. National Security Advisor 2 1 Ronald Reagan. \Address to the Nation and Other Countries on United States-Soviet Relations." (Jan. 16, 1984), available at https://www.reaganlibrary.gov/archives/speech/address-nation-and-other-countries- united-states-soviet-relations 2 Press Brieng by National Security Advisor John Bolton on Iran (May 08, 2018) https://www.presidency.ucsb.edu/node/335839 186 5.1 Introduction Examining treaty exit in the context of arms control agreements highlights the in uence of treaty compliance and detection in decisions to withdraw from a treaty. Arms control treaties are often sharply dened in scope given that they address critical security matters between states. The specic types of arms and how those arms may be produced, deployed, used, and disposed of can be detailed in specic terms. Further, states are highly incentivized to demand compliance for security matters; however, they often result to informal eorts to allow for exibility (Koremenos 2016, 226{260). Verica- tion was and is a hallmark of modern arms control agreements, particularly in the Cold War and post-Cold War period. The United States and the So- viet Union spent signicant energy and resources in negotiating verication provisions and carrying out those terms. In this context, the \case" examined in this chapter is not dened as one treaty, but a series of treaties between two countries over time. These treaties are both bilateral and multilateral. This allows for comparisons across treaties and administrations under one issue area. Security agree- ments, like arms control treaties, are important to study given the nature of the subject matter that is governed. The weapons that are addressed in these treaties have the capability of widespread and grievous harms to broad swaths of the populations of not only the parties to the treaty, but the global population. Further the issue of compliance is also signicantly important, often hotly contested, and many times specically identied. 187 In analyzing international cooperation between the US and Russia, this chapter relies on both primary and secondary sources, accounts in the news, elite interviews conducted in 2019 and 2020, and focused quantitative data of both state's behavior with regard to treaty disengagement. Determining the motives of states is challenging, particularly in states without open access to policymaking like Russia. To overcome that challenge, we can use the diplomatic practice of a state across fora or \sites" as empirical evidence of a broader foreign policy strategy (Schmitt 2020b). What follows is a summary of the arms control agreements between the two superpowers and then a fuller description of the two treaties and their place in the broader arms control system. Then, a description of the stresses put on the agreements is presented highlighting the countries' eorts to bargain to account for those stresses along with a description of accusations of breach and rationales for eventually leaving the treaties. As described herein, the rationales for each state's perceptions and actions are often interrelated and have their genesis decades before the eventual exit. These cases are useful in highlighting that dynamic, especially in understanding the role that an- ticipated, suspected, and conrmed treaty breach plays in a state's calculus. After that description, a brief digression to each country's general behavior involving disengagement from international cooperation is presented to give a baseline understanding of the states' practices. The chapter concludes with an assessment of the theory as applied to these cases of treaty withdrawal, with a discussion of implications and further questions. 188 5.2 Summary of Arms Control between the Su- perpowers Eorts to manage the military capabilities of states has a long history, but arms control as it is known today can trace its history to the dawn of the twentieth century with the two International Peace Conferences at the Hague in 1899 and 1907 (Goldblat 2002; Williams Jr and Viotti 2012). There, the Emperor of Russia, lagging economically and militarily, urged the other participants to work toward universal peace and a reduction in \excessive" armaments. Although the eorts to limit the calibre of naval guns, thickness of armor plating, and velocity of projectiles failed, eorts to limit arms would again be revisited after World War I and in the interwar period. After World War II, as the world came to terms with a new class of weapons of mass destruction, eorts emerged to limit the spread, use, testing, and number of these weapons. \Victory in World War II brought no sense of security, therefore, to the victors" (Gaddis 2006). 3 Early proposals to sig- nicantly restrict the possession and use of nuclear weapons, for example the Baruch Plan, were rejected. 4 Nonetheless, the superpowers were able to 3 After the war, the states focused on increasing security, primarily through arming themselves. In the aftermath of World War II and spending signicantly on the Marshall plan, producing more atomic weapons was still far cheaper than bringing the US military forces back to World War II levels (Gaddis 2006). 4 The Baruch Plan, delivered to the United Nations by Bernard Baruch, called on an international organization to regulate and possess the technology for atomic weapons. The US proposed decommissioning its atomic weapons and transferring the weapons technology to the U.N. as long as no other country produced the weapon and a strict regime of verication, monitoring, and inspection was in place. Joseph Stalin of the USSR rejected the proposal, fearing it would entrench the US nuclear monopoly. (See generally Gaddis 2006). 189 cooperate on arms control despite their on-going suspicions other each oth- ers' intentions. Table 5.1 summarizes the signicant arms control treaties between the US and Russia since the end of World War II. 5 At the start of the Cold War, multilateral eorts to ban nuclear weapons from Antarctica, outer space, and the seabed produced a number of treaties. 6 The Antarctic Treaty signed in 1959 is characterized as the \rst true arms control treaty of the modern era . . . " (Williams Jr and Viotti 2012, 223). In 1968, the Non-proliferation of Nuclear Weapons Treaty (NPT) was agreed to in an eort to prevent the spread of nuclear weapons beyond those states already in possession and to aspire to eventual elimination of the weapons. The early multilateral eorts of the early part of the Cold War gave way to bilateral agreements between the countries on reducing the number and type of nuclear weapons produced and deployed in their arsenals. In the 1970s, the Strategic Arms Limitation Talks (SALT) produced the Anti- ballistic Missile (ABM) Treaty, limiting strategic missile defenses, and In- terim Agreement, limiting intercontinental ballistic missiles (ICBM) and submarine-launched ballistic missile (SLBM) forces in both countries. In 1973, the two countries signed the Agreement on the Prevention of Nuclear War where the parties agreed to \exercise restraint" and commit to con- sultations in times of crisis. In 1975, the Biological Weapons Convention (BWC) prohibiting the development, production, acquisition, or stockpiling 5 Table 5.1 is by no means exhaustive but useful for background and context. 6 Antarctic Treaty (1959); Treaty on Principles Governing the Activities of States in the Exploration and Use of Outer Space, including the Moon and Other Celestial Bodies (1967); Treaty on the Prohibition of the Emplacement of Nuclear Weapons and Other Weapons of Mass Destruction on the Seabed and the Ocean Floor and in the Subsoil Thereof (1971). 190 of biological weapons entered into force, applying to both the US and the Soviet Union. Thereafter, the US and Russia agreed to SALT II in 1979, further limit- ing deployment of ICBMs, SLBMs, and adding limits to strategic bomber- based nuclear forces. However, after the Soviet Union invaded Afghanistan, President Jimmy Carter asked the US Senate not to consider SALT II for advice and consent. In the 1980s, President Reagan proposed the Strategic Arms Reduction Treaty (START I) to reduce deployed strategic arsenals and eliminate excess delivery vehicles under verication through on-site in- spection. START I's entry into force was delayed given the collapse of the Soviet Union, and eorts to denuclearize Ukraine, Kazakhstan, and Belarus. The latter three countries gave up their nuclear arms, transferring them to Russia, and joined the NPT and START I agreements. On December 8, 1987, US President Ronald Reagan and Soviet General Secretary Mikhail Gorbachev signed the Intermediate-Range Nuclear Forces (INF) Treaty to completely eliminate land-based intermediate- and short-range missiles, en- forced by an intrusive inspection regime. After the break up of the Soviet Union, Presidents George H.W. Bush and Boris Yeltsin agreed to continue eorts begun in START I by sign- ing START II in 1993. START II also called for a reduction in strategic weapons and banned the deployment of multiple independently targetable reentry vehicle (MIRV) ICBMs. Before ratication of START II was com- plete, Presidents Bill Clinton and Boris Yeltsin came to an understanding at their Helsinki Summit that the countries should ratify START II and begin 191 Table5.1: Signicant Arms Control Agreements between US and Russia (USSR) Treaty Signed EIF End Date Status Antarctic Treaty Dec. 1, 1959 June 23, 1961 In force Partial Test Ban Treaty Aug. 5, 1963 Oct. 10, 1963 In force Outer Space Treaty Jan. 27, 1967 Oct. 10, 1967 In force NPT July 1, 1968 Mar. 5, 1970 In force Seabed Treaty Feb. 11, 1971 May 18, 1982 In force BWC April 10, 1972 Mar. 26, 1975 In force SALT I May 26, 1972 Oct. 3, 1972 Oct. 3, 1977 Expired ABM Treaty May 26, 1972 Oct. 3, 1972 June 13, 2002 Terminated Threshold Test Ban Treaty July 3, 1974 Dec. 11, 1990 In force SALT II June 18, 1979 Never EIF INF Treaty Dec. 8, 1987 June 1, 1988 Aug. 2, 2019 Terminated START I July 31, 1991 Dec. 5, 1994 Dec. 5, 2009 Expired Open Skies Treaty Mar. 24, 1992 Jan. 2, 2002 Nov. 22, 2020 Terminated START II Jan. 3, 1993 Never EIF CWC Jan, 13, 1993 April 29, 1997 In force START III Never Negotiated SORT May 24, 2002 June 1, 2003 Feb. 5, 2011 Replaced by New START New START April 8, 2010 Feb. 5, 2018 Feb. 5, 2026 In force negotiations on START III for further strategic weapons reductions. After a long delay, START II was ratied by both the US Senate and Russian Duma. Yet, the treaty did not enter into force because Russia had made continued US compliance and participation in the ABM Treaty a pre-condition. US President George W. Bush, unilaterally withdrew from the ABM Treaty in 2002, which precipitated Russia to announce it would not be bound by START II. Having never been fully negotiated, START III was replaced by the Strategic Oensive Reductions (SORT) or Moscow Treaty signed in 2002. In its brief two pages (compared to the 400 pages of START I), SORT continued the countries' commitment to reducing operationally deployed strategic weapons, without verication or accounting of the weapons elim- ination. After START I expired in 2009, the parties continued to abide by SORT, even though it lacked the strong verication provisions of START I. 192 Presidents Barack Obama and Dmitry Medvedev signed the New START Treaty in 2010, which replaced SORT and included extensive measures for verication. New START had an initial 10-year duration, with an option for a ve-year extension that Presidents Joe Biden and Vladimir Putin exercised in 2021. The US and Russia also entered into other agreements to decrease risks of con ict and improve trust between the nations. In 1992, NATO members and former members of the Warsaw Pact signed the Open Skies Treaty, allowing participating states the right to over ight of each other's territory with unarmed xed-wing aircraft for observation. Once a party obtained information from the over ight, it was obligated to share the information with other members. Citing Russia violations, the US eventually withdrew from the Open Skies Treaty in 2020, 7 followed shortly thereafter by Russia in 2021. The post-Soviet era also saw the agreement of the Chemical Weapons Convention (CWC) in 1993 and negotiations over the Comprehensive Test Ban Treaty, although the later was not ratied by the US Senate. Members of NATO and former members of the Warsaw Pact entered into the Treaty on Conventional Armed Forces in Europe (CFE Treaty) in 1990 dividing the area covered, the Atlantic to the Urals (ATTU), into two zones broadly corresponding to the two alliances' territories with limits on conventional arms in each zone. Further agreements between the parties were made to amend or modify the agreement in the CFE Treaty into the late 1990s. 7 The Biden Administration informed Russia that the US does not intend to rejoin the treaty. As of the writing of this dissertation, the Biden Administration has not stated a change in that policy. Commentators cite the concerns over rejoining at a time of poor relations and Russian aggression in Ukraine. 193 On July 14, 2007, President Vladimir Putin announced Russia's intention to suspend compliance with the CFE Treaty in response to the proposed deployment of missile defense systems in Europe by the US. Cooperation between the US and Russia under the CFE Treaty continued to deteriorate, and Russia has not come back into compliance. Over six decades of relations, the US and Russia have negotiated and coop- erated over security issues in an eort to minimize security concerns, lower arming costs, and promote peace. This is all in addition to other areas of cooperation between the countries, including coordination in outer space, eventual adoption of human rights norms, and an opening of economic re- lations after the fall of the Soviet Union. As described here, many of the agreements established during the height of the Cold War have either expired or terminated. In particular, the agreements with some of the most robust verication procedures have now ended, leaving New START as one of the last agreements between the states allowing inspections. New START is set to expire in 2026, and prospects of a new arms control agreement appear slim given the state of relations between the US and Russia in light of the 2022 invasion of Ukraine. Yet, the countries have found common ground in the past and there can be more opportunities for cooperation in the future. Turning from a general overview of the arms control landscape between the US and Russia, I will now focus on the relations between the two countries with respect to two major treaties: the ABM Treaty and the INF Treaty. Using these cases as a focus allows us to track the rationales for the eventual 194 unilateral withdrawal by the US in 2002 and 2020, respectively. These with- drawals were seen as a dramatic break in the ties between the countries and, on that ground, are worthy of exploration. Yet, in investigating the appli- cation of the theory proposed in Chapter 2, these cases of exit also provide ample material to explore mechanisms and test the theory's explanatory power. The US withdrawals from the two treaties occurred under dierent administrations. Examining the decisions of multiple administrations allows for opportunities to compare and contrast their approaches. Additionally, by examining a long time period, I can collect more evidence. While the primary sources for the more recent withdrawal under the Trump Adminis- tration are more dicult to secure, I leverage recent memoirs, new reports, and interviews. Because the withdrawal from the ABM Treaty occurred over two decades ago, I have the opportunity to gather a wider array of evidence from primary sources and rst-person accounts. 5.2.1 ABM Treaty In an eort to curtail the expensive arms race between the superpowers, the ABM Treaty was entered into by the US and the Soviet Union to limit the anti-ballistic missile systems used by each country in defending against ballistic missile-delivered nuclear weapons, namely ICBMs. As mentioned above, the ABM Treaty was a part of the SALT I agreement. The treaty was designed to end and prevent expensive and risky arms races in missile defenses. At the Saryshagan test site in the Kazakhstan desert, the Soviet Union had been rst to successfully test an interception of a ballistic missile 195 in March 4, 1961 (Gruntman 2015). SALT I, including the ABM Treaty, was a part of US President Richard Nixon's strategy of d etente (Russian: razryadka), which would \wrap the Soviet Union in a web of new inter- national agreements and understandings that would make the Cold War manageable and less threatening" (Homan 2009, 21). President Richard Nixon and General Secretary Leonid Brezhnev signed the treaty during the 1972 Moscow Summit on May 26, 1972. The US Senate later ratied the treaty on August 3, 1972. The ABM Treaty \shall be of unlimited duration" (art. XV), with reviews of the treaty by the parties every ve years (art. XIV). The ABM Treaty remained in force, even after the fall of the Soviet Union. 8 Russia, Belarus, Kazakhstan, and Ukraine (each possessing nuclear weapons or material after the collapse) became successor states to the treaty. 9 The treaty's obligations are expressed largely in terms of the parties \under- taking" to implement the treaty obligations (arts. I, III, V, VI, IX-XI). The treaty includes verication by \national technical means" (art. XII) and the creation of a Standing Consultative Commission (art. XIII); however, compliance is largely left up to the states. This pithy treaty gives a signif- icant level of exibility to the state parties to comply, even contemplating 8 Some in the US argued that the ABM Treaty ceased to have eect after the collapse of the USSR (Feith 2009, 231{232). 9 Memorandum of Understanding on Succession (MOUS) established the parties to the ABM Treaty with the four former Soviet-republics considered to be the USSR Successor States. 196 likely violations in article VIII. The treaty includes two escape clauses, one allowing amendment (art. XIV) and one allowing withdrawal (art. XV). 10 5.2.2 INF Treaty After the negotiation failure at the 1986 Reykjavik Summit between Pres- ident Reagan and General Secretary Gorbachev, the parties continued to negotiate, leading to the 1987 Washington Summit the following year. At that summit, Reagan and Gorbachev signed the INF Treaty on December 8, 1987, entering into force on June 1, 1988. This culminated negotiation eorts going back to the 1980 Geneva Summit during the Carter Administration. The INF Treaty called on the US and Soviet Union to eliminate all ground- based ballistic and cruise missiles with ranges between 500 and 5,500 kilo- meters. These weapons were not covered under the SALT II Treaty. This was the rst arms control treaty between the countries to eliminate a whole class of nuclear weapons. The INF Treaty consisted of seventeen articles and two protocols on the elimination of the systems and on-site inspections. The parties also signed an accompanying Memorandum of Understanding (MOU) that included public declarations on the numbers and locations of armaments covered by the treaty. Because the missile systems covered by the treaty were small and mobile, capable of both conventional and nuclear armaments, verication was key to the treaty's eectiveness. At the signing, Reagan said \We have listened to 10 For a detailed analysis of the text and purpose of the ABM Treaty, see Chayes and Chayes (1985). 197 the wisdom of an old Russian maxim . . . doveryai, no proveryai|trust, but verify." 11 The parties were to use both cooperative measures (e.g., on-site in- spections) and national technical means (NTM) (e.g., use of satellite sensing) for verication. Unique among arms control agreements between the states, the treaty included an intrusive enforcement mechanism of on-site inspec- tions (arts. X{XIII), which was later used as a model for START I. Article XIII establishes a Special Verication Commission to assist the parties in resolving any disputes over compliance and to improve the eectiveness of the treaty. The treaty was to continue indenitely (art. XV). The US and USSR met their goal in destroying the missiles by June 1, 1991. The US removed missiles, including its Pershing IIs, from military bases in Belgium, Italy, the Netherlands, West Germany, and the United Kingdom. The Soviet Union also removed its missiles, including its SS-20s, from Eu- ropean states like East Germany and Czechoslovakia. After the break up of the Soviet Union, the former republics of Belarus, Kazakhstan, and Ukraine, along with Russia, joined the INF Treaty. On site inspections ended in 2001 and compliance was primarily veried by satellite monitoring. 12 11 Ronald Reagan. \Remarks on Signing the Intermediate- Range Nuclear Forces Treaty." (Dec. 8, 1987) available at https://www.reaganlibrary.gov/archives/speech/remarks-signing-intermediate-range- nuclear-forces-treaty 12 John Russell, On-Site Inspections Under the INF Treaty: A Post- Mortem. VERTIC Brieng Paper 01/02 (August 2001), available at http://www.vertic.org/media/Archived Publications/Brieng Papers/Brieng Paper 01 2.pdf 198 5.3 Stresses on Arms Control Regime 5.3.1 Star Wars and the ABM Treaty Reinterpretation De- bate The ABM Treaty limited anti-ballistic missile defense to preserve the strate- gic balance between the countries through the logic of mutually assured destruction (MAD). The treaty also expressly banned the development of space-based systems for territorial defense of the country (art. V). In Octo- ber 1985, President Reagan proposed the Strategic Defense Initiative (SDI), derisively known as \Star Wars," which was essentially a space-based an- tiballistic missile system for the defense of the US. Reagan's focus on missile defense can be traced, at least in the form that would surface in his presi- dency, to his campaign. After visiting Cheyenne Mountain, Colorado in the summer of 1979, Reagan was perturbed by the fact the US had no eec- tive defense to stop an incoming Soviet missile. His policy advisor, Martin Anderson from the Hoover Institution at Stanford University, devised a set of potential national defense policy positions for Reagan. One path was to continue with SALT II, another was to engage in an arms race. Nei- ther option was appealing, so they came up with a third way focusing on a \Protective Missile System" (Homan 2009, 28{29). The position was 199 added to the 1980 Republican Party Platform 13 but not made part of Rea- gan's stump speeches. In speaking about his disapproval of Nixon's policy regarding the ABM Treaty, he said: \We bargained that away in exchange for nothing" (Homan 2009, 29, n. 5). As Gaddis (2006) later highlighted, Reagan \was the only nuclear abolitionist ever to have been president of the United States." At Reykjavik in 1986, Gorbachev and Reagan discussed the future of arms control. Gorbachev proposed a 50 percent reduction in \strategic oensive arms", eliminating certain classes of missiles in Europe, prohibition of nu- clear testing, and a promise to remain in ABM Treaty for 10 more years (Homan 2009, 261). Reagan stuck with his dream of SDI and proposed that the US share the technology once developed and do away with the ABM Treaty. Gorbachev chaed at the suggestion, nding it hard to believe the US would share such sensitive technology. The subject of SDI ultimately scuttled the talks (Homan 2009, 261{270). Later that same year, the US accused the Soviet Union of violating the terms of the ABM Treaty by constructing a radar station at Krasnoyarsk in Siberia, which could be used for a nationwide anti-missile defense system (Gorbachev 1996, 454). The radar was a permitted type of radar used for early warning 13 Republican Party Platform of 1980. The Republican Na- tional Convention, Detroit, Michigan (July 15, 1980) available at https://www.presidency.ucsb.edu/documents/republican-party-platform-1980. The Platform called for \Vigorous research and development of an eective anti-ballistic missile system, such as is already at hand in the Soviet Union, as well as more modern ABM technologies." It went on: \Republicans pledge to revitalize America's military research and development eorts, . . . to enable us to deal with possible breakthroughs in anti-missile defense, anti-satellite killers, directed energy systems, and the military and civilian exploitation of space." 200 against missile attacks but it was placed in an impermissible location and facing inward rather than outward. Internally, the USSR acknowledged the violation of Art. 6b of the Treaty, but the USSR would not publicly acknowledge the violation for at least two more years (Homan 2009, 275{ 276). Accusations of other missile defense developments by proponents of SDI were greatly exaggerated where the \Soviet technology was way behind" (Homan 2009, 276{277). Soviet leadership tried to use concessions on the Krasnoyarsk radar station as a bargaining chip with the US (Homan 2009, 293). Yet, the US refused to agree to the Soviet proposal of a period of non-withdrawal while the USSR was violating the treaty (Gorbachev 1996, 453{455). To avoid violating the treaty, the Reagan Administration sought to \rein- terpret" the treaty to permit SDI without rst consulting the Senate or the Soviet Union. 14 The unilateral decision started years of internal debate, especially with key Senators, including Sam Nunn (D), Chairman of the Senate Armed Services Committee. The Senate held hearings and debates on the issue. Congress withheld appropriations for SDI tests that violated the treaty. The Senate reported the ABM Treaty Interpretation Resolution, rearming the original understanding of the treaty. In 1988, the Senate at- tached a condition to the Intermediate-Range Missile Treaty specifying that the US would interpret the treaty in line with the understanding shared by the President and Senate at the time of the Senate's advice and consent. Eventually, the Reagan Administration announced it would comply with the 14 The history of the debate is reviewed in Koh et al. (1988). 201 original interpretation as a matter of policy, without conceding that their interpretation was legally incorrect. In 1993, President Clinton rescinded the Reagan interpretation and conrmed that the US would abide by the original interpretation. Despite President Reagan's desire to pursue missile defense and original objection to the terms of the ABM Treaty, he did not decide to exit. In fact, his dialogue with Gorbachev lead to the creation of a new arms control treaty, the INF Treaty in 1987, which increased the cooperation between the super powers to eliminate a class of weapons under rigorous verication procedures. The conclusion of the treaty largely was possible when it was unlinked from the other thorny issues between the states (Homan 2009, 279). \By signing the INF treaty we had literally removed a pistol held to our head" (Gorbachev 1996, 444). Further, when a new president, George H.W. Bush, and Gorbachev signed the START I Treaty in 1991, there was little mention of SDI or other contentious issues of missile defense (Homan 2009, 367{368). By this time the Soviet economy was imploding and \[t]he powerful riptide of the economy pulled everything [including the military- industrial complex] down with it" (Homan 2009, 368). Yet, the cooperation between the US and a reforming USSR would be challenged in the coming years as the Soviet Union itself was about to fall apart. 202 5.3.2 Extending Arms Control after the USSR Collapse On December 25, 1991, Mikhail Gorbachev appearing on television, seated at a desk before the red Soviet ag and a plain blank wall, resigned as president of the Soviet Union. \I am leaving my post with apprehension, but also with hope, with faith in you, your wisdom and force of spirit. We are the heirs of a great civilization and its rebirth into a new, modern and dignied life now depends on one and all." 15 The Soviet hammer and sickle ag was lowered that evening from where it ew for generations above the Kremlin, to be replaced by the tricolored ag of an independent Russian state lead by Boris Yeltsin. Over 4,800 miles away, in the Oval Oce of the White House, President George H.W. Bush addressed the American people and the world regarding the dissolution of the USSR. \We stand tonight before a new world of hope and possibilities and hope for our children, a world we could not have con- templated a few years ago. The challenge for us now is to engage these new states in sustaining the peace and building a more prosperous future." 16 The US quickly recognized the newly independent states and set on a policy of reducing the risk of nuclear proliferation and threat reduction. \Our ene- mies have become our partners, committed to building democratic and civil societies. They ask for our support, and we will give it to them." 17 15 \Text of Gorbachev's Resignation Speech." AP News (Dec. 25, 1991) https://apnews.com/article/bcdbf49be92683caa5b4723886baafcd 16 \END OF THE SOVIET UNION; Text of Bush's Address to Nation on Gorbachev's Resignation." The New York Times (Dec. 26, 1991), available at https://www.nytimes.com/1991/12/26/world/end-soviet-union-text-bush-s-address- nation-gorbachev-s-resignation.html 17 Id. 203 In the Bush I Administration, 18 top-level advisors wanted to take a dierent, strategic, and more calculated approach to addressing arms control (Bush and Scowcroft 1998, 11{14). US Secretary of State James Baker and Soviet Minister of Foreign Aairs Eduard Shevardnadze were able to make progress on arms reduction when the USSR unlinked the issue of SDI and missile defense from other issues (Baker and Glasser 2021). As the Soviet Union was imploding, the US rushed to secure reductions in nuclear weapon stockpiles and deployment of other weapons. The US pro- posed unilateral cuts in the nal months before the USSR's collapse and began eorts that eventually lead to START II in 1993. Bush announced the US policy, which included a request that the USSR \agree to permit only limited deployment of defenses for protection against limited ballistic missile strikes, whatever the source" (Bush and Scowcroft 1998, 546). According to Bush, Gorbachev's only sticking point was testing but the Soviet's were un- sure about where they were going with arms control, despite an orientation of cooperation between the countries. The break up of the Soviet Union posed various challenges to the countries' arms control eorts. Within the administration, Secretary Baker proposed \ve principles" on the independence of the new states to attempt to in- uence a peaceful transition. The principles included self-determination through democratic means; respect for existing borders; respect for democ- racy and rule of law; human rights; and adherence to international law and 18 To distinguish the two Bush administrations and for eciency, I use \Bush I" to refer to the senior Bush and \Bush II" to represent his son. 204 the USSR's existing treaty obligations. The national security team added a \sixth" principle of central control over nuclear weapons and safeguards against internal and external proliferation (Bush and Scowcroft 1998, 544{ 545). Brent Scowcroft, Bush's National Security Advisor, considered the break up of unied control as a possible benet because limiting the size of a potential attack would dilute the risks to the US (Bush and Scowcroft 1998, 544). However, he acknowledged the dangers of weakened capacity of the new states and risk of proliferation. Eventually, the nuclear-armed, former Soviet republics relinquished their nuclear armaments and acceded to START and the NPT, with security guarantees from the US, United Kingdom, and Russia (Reed and Stillman 2010, 204{219). 19 President Bill Clinton continued the work of his predecessor strengthening the Nunn-Lugar Cooperation Threat Reduction Program 20 and completing the denuclearization of Ukraine, Kazakhstan, and Belarus, obtaining signa- tures on the Tripartite Accord with Yeltsin and Ukrainian President Leonid Kravchuk (Stent 2014, 27{29). The Clinton Administration also proposed pursuing a missile defense system, but even if less ambitious than SDI it would still violate the ABM Treaty. The administration faced a tension where Russia was weak and internal pressure from military, industry, and congressional republicans favored missile defense. 21 Some commenters ar- gued that a system designed to intercept North Korean or Iranian missiles 19 See also Budjeryn (2016). 20 Named for the founding US Senators Sam Nunn (D) of Georgia and Richard Lugar (R) of Indiana, who shepherded the passage of the Soviet Threat Reduction Act in 1991. Presidents Bush and Yeltsin signed the \Umbrella Agreement" in 1992 to allow for the programs between the US and Russia. 21 Author interview (Interview M), March 2020. 205 would not violate the \spirit" of the treaty. 22 Russians saw missile defense and its limitation under the ABM Treaty as a \fundamental cornerstone of the US-Russian nuclear relationship" (Stent 2014, 30). At the 1997 Helsinski Summit, Clinton and Yeltsin discussed arms con- trol, in particular START II reductions and missile defense. Clinton saw a national missile defense system as a violation of the ABM Treaty, but a the- ater missile defense system as permitted (Clinton 2004, 752). Yeltsin and the Russians were concerned about the dividing line between permissible and prohibited programs, leading to an agreement on technical denitions. A package of ABM-related agreements were entered into in 1997, cover- ing matters of treaty succession, establishing a \demarcation line" between permitted and restricted systems, and condence-building measures. In 1998, as part of the Standing Consultative Commission (SCC) estab- lished by the treaty, the state parties to the ABM Treaty met in Geneva to review the ABM Treaty and its continued eectiveness. At the confer- ence, the parties \agreed that the Treaty continues to operate eectively and rearmed the fundamental importance of the Treaty, as a cornerstone of strategic stability, for strengthening international security and for pro- moting the process of further reductions in strategic oensive arms." At the 22 In 2000, Ashton Carter, future Secretary of Defense, explained: \A limited ABM system of the kind envisioned by the Clinton administration would violate the letter of the ABM Treaty, but it is arguable whether it would violate the spirit of the treaty. . . . Thus logic would indicate the Russians should not object to it. Russia fears that a limited system will be replaced over time by a much larger system." Ash Carter. \Star Wars Redux? Ashton Carter on Missile Defense." Press Release. (May 08, 2000), available at https://www.belfercenter.org/publication/star-wars-redux-ashton-carter-missile-defense 206 time, the state parties stressed the importance of the 1997 ABM-related agreements. Congressional Republicans were strongly in favor of a robust missile defense system targeting long-range missiles (Stent 2014, 30). These diering views would arise in legislation. Clinton vetoed the National Defense Authoriza- tion Act for Fiscal Year 1996 23 because it called for the deployment of a national missile defense system, which would violate commitments under the ABM Treaty and jeopardize Russia's implementation of START I and its ratication of START II (Clinton 2004, 690). 24 During the second Clin- ton Administration, Congress introduced the Defend of America Act of 1996 authorizing the deployment of an anti-ballistic missile system. 25 Democratic Senators libustered the bill and it did not receive a oor vote. Competing bills were introduced supporting the president's position. 26 Some in Congress and the government held the view, extreme as it was, that the ABM Treaty was non-binding after the fall of the Soviet Union, arguing that when the Soviet Union ceased to exist in 1991 the treaty lapsed (Feith 2009, 231{232). This included US Senators Trent Lott (R) of Missouri and Jesse Helms (R) of North Carolina (Costa et al. 1998), Senators who worked hard to thwart eorts to preserve the ABM Treaty{even holding up ratication of the Comprehensive Test Ban Treaty. 23 H.R.1530 (104th Congress (1995-1996)). 24 See also Presidential Message, H.Doc. 104-155 (104th Congress, 2d Session). 25 S.1635 (104th Congress (1995-1996)). See also H.R.3144 (104th Congress (1995- 1996)). 26 See Ballistic Missile Defense Act of 1996, H.R.3489, 104th Congress (1995-1996). 207 Clinton later tried to renegotiate the ABM Treaty to amend it to allow the US to deploy a national missile defense system (Clinton 2004, 908). Clinton saw his inability to obtain concessions from Vladimir Putin as a result of Putin being aware of the US domestic political reality and the up-coming election. \The Republicans had been enamored of missile defense since the Reagan era, and many of them wouldn't hesitate to abrogate the ABM Treaty in order to deploy it. . . . Putin didn't want to have to deal with this twice" (Clinton 2004, 908). President Clinton pursued the negotiations to amend the treaty, despite his belief that the costs of missile defense were too high; that attacks from terrorists possessing small weapons of mass destruction (WMDs) was more likely; and the potential security spiral that would result from Russia, China, Pakistan, and India building more missiles to overcome defenses. As an incentive, Clinton thought a missile defense system could be shared|similar to Reagan's intentions: \If we ever did develop a workable system, I thought that we should oer the technology to other nations and that, in so doing, we could probably persuade the Russians to amend the ABM Treaty" (Clinton 2004, 908). In 1999, President Clinton signed the National Missile Defense Act, 27 which put the US on a course to deploy a national missile defense system. The Ad- ministration argued that the development phase was treaty compliant and set about reforming the treaty through negotiations with Russia. Republi- cans, still contending that the ABM Treaty was too limiting, also introduced 27 H.R.4 (106th Congress (1999-2000)). 208 the Freedom From Mutually Assured Destruction Act of 1999 28 to prohibit the administration from complying with the treaty. After visiting Moscow in the summer of 2000, Presidents Clinton and Putin issued a Joint Statement on Principles of Strategic Stability acknowledging their discussions regarding amending the treaty and noting \the importance of the consultative process and rearm their determination to continue con- sultations in the future . . . ." 29 However the prospects of an amendment in the US's favor were dim with opposition from Russia's Putin and French President Jacques Chirac. 30 5.3.3 Changing Security Concerns in the 21st Century Under the Bush Administration, the US sought to \remake" US-Russia re- lations. In the summer of 2001, the US Defense Department considered how to approach the Russian government in these eorts. The US desired to exit from the ABM Treaty, preferably by agreement with the Russians (Feith 2009, 45). Within the administration, the desire was for a \mutual abroga- tion" to allow small-scale defenses by both countries (Rice 2011, 59{60). In reshaping the US approach to Russia, policymakers at the Defense Depart- ment, chief among them Secretary of Defense Donald Rumsfeld, sought to 28 H.R.2022 (106th Congress (1999-2000)). 29 Joint Statement on Principles of Strategic Stability. The White House. (June 4, 2000) https://clintonwhitehouse5.archives.gov/WH/New/Europe-0005/factsheets/js{ strategic-stability.html 30 Russia, EU Oppose Any Changes to ABM Treaty." RFE/RL Newsline, vol. 4, no. 211, part I, 31 October 2000, available at https://nuke.fas.org/control/abmt/news/treaty- abm-001031.htm and \Putin oers arsenal cut." CNN (July 2, 2001), available at https://www.cnn.com/2001/WORLD/europe/07/02/chirac.putin/index.html 209 tie security issues to the political and economic issues that were wrapped up in Russia's desire for respect, standing, and foreign investment. (Feith 2009, 45). The US saw \rogue states" like North Korea as leveraging the limitations imposed by the ABM Treaty, which prohibited the US from building de- fenses to ICBMs (Feith 2009, 231). At a May 2001 speech at the National Defense University in Washington, President Bush stated: \We need a new framework that allows us to build missile defenses to counter the dierent threats of today's world. To do so, we must move beyond the constraints of the 30-year-old ABM Treaty. This treaty does not recognize the present, or point us to the future. It enshrines the past." 31 Where the Clinton Administration tried to reconcile missile defense with the ABM Treaty, the Bush II Administration saw it as an unacceptable im- pediment (Stent 2014, 72). Both Vice President Dick Cheney and National Security Advisor Condoleezza Rice shared President Bush's view that the treaty was outdated. They saw eorts to preserve those institutions as part of an eort to preserve the status quo of when Russia was a superpower on par with the US. \Arms control and the ABM Treaty were integral to that reality and thus talisman against decline" (Rice 2011, 60). This appears to have been Putin's estimation as well (Stent 2014, 52{53). Cheney (2011, 325) put it in blunt terms: \The treaty was of advantage to the Soviets." 31 \President Bush Speech on Missile Defence." White House (May 1, 2001), available at https://nuke.fas.org/control/abmt/news/010501bush.html 210 Additionally, the fact that the ABM Treaty memorialized the MAD doc- trine of the Cold War was seen as part of the contemporary problem of the agreement. Chiding critics and praising the Reagan policy, Rice (2011, 58) said of the opposition: \But among the high priests of arms control, SDI was a threat to strategic stability." Within the administration, there was debate about how to handle the desire of missile defense. Cheney, who had for a long time advocated for abrogating the ABM Treaty, found resistance from the State Department. Colin Powell was concerned withdrawing would strain relations with Russia, so he argued that \we should stop short of abro- gation and negotiate loopholes in the treaty for developing missile defenses" (Cheney 2011, 325). Cheney disagreed, seeing action on exit coming before bargaining through diplomatic eorts. Exit rst, bargain later. Carrying over the views of others like Senators Trent Lott and Jesse Helms, Secretary of Defense Donald Rumsfeld held the view that the ABM Treaty was no longer binding after the Soviet Union collapsed (Rumsfeld 2011, 308). When returning from China, Secretary of State Colin Powell stated: \we've got to get beyond the constraints of the ABM Treaty because it keeps us from doing things that we need to pursue missile defenses." 32 Secretary Powell reiterated the US was attempting to negotiate with Russia, but his comments suggested that despite discussions on exibility the US was headed toward an exit. 32 Secretary Colin L. Powell. \En route to Washington, DC, from Shanghai, Peo- ple's Republic of China." U.S. Department of State. (Oct. 21, 2001), available at https://nuke.fas.org/control/abmt/news/102101powell.html 211 On December 13, 2001, two months after the September 11, 2001 terrorist attacks on the US, President George W. Bush gave Russia, the other remain- ing nuclear power and treaty member at the time, 33 the required six months notice of the United States' withdrawal from the ABM Treaty under Article XV(2) of the Treaty. The US had determined that it needed to develop a national missile defense system against \rogue states," which would not be permissible under the ABM Treaty (Feith 2009, 231). 34 John Bolton, Under Secretary of State for Arms Control and International Security, stated with joy: \With President Bush's announcement of the US withdrawal from the ABM Treaty on December 13, a dangerous relic of the Cold War was o- cially put to rest" (Bolton 2008, 74). Bush characterized his \cooperation" with Putin on arms control and leaving the ABM Treaty as \[o]ne of the biggest achievements" of his work with the Russian leader (Bush 2010, 432). Although Russia's initial response to the unilateral decision by the US to withdraw was tepid, it caused deep consternation and worry in Russia. Putin pushed Bush for a new arms control treaty in its wake, eventually leading to SORT the following year (Stent 2014, 73). The US moved forward with its missile defense activities, with agencies like the Missile Defense Agency (MDA) continuing its work. 35 Secretary Rumsfeld believed that \[n]o well- informed Russian ocial seriously worried that the United States's missile 33 Belarus, Kazakhstan, and Ukraine surrendered their nuclear arsenals by 1996. 34 \Announcement of Withdrawal from the ABM Treaty" (December 13, 2001), avail- able at https://georgewbush-whitehouse.archives.gov/news/releases/2001/12/20011213- 2.html 35 The MDA, formerly named the Ballistic Missile Defense Organization (BMDO), was a continuation of prior agencies and eorts stretching back to Reagan's SDI. The MDA is a section of the US Department of Defense. 212 defense program would protect America against a massive nuclear strike from Russia. I suspected that their real concern might have been that US missile defenses could damage Russia's image as a world power" (Rumsfeld 2011, 307). Rumsfeld claims Putin admitted as much to him at a later meeting, but that Putin could not allow a US missile defense or a withdraw from the ABM Treaty without protest (Rumsfeld 2011, 309). At the Munich Security Conference in February 2007, President Vladimir Putin argued that the INF Treaty should be reconsidered in light of the changing security environment where non-parties to the agreement were not restrained like Russia and the US by its provisions. 36 At the same time, Russia's Chief of the Russian General Sta, General of the Army, Yuri Baluyevsky, stated: \The INF Treaty between Moscow and Washington is of an indenite nature, but there is a possibility of withdrawing from it if one of the parties provides convincing evidence that it is necessary to withdraw." 37 Chief among Russia's complaints were the development of medium-range missiles and the US establishment of missile defense systems in Poland and the Czech Republic. Understanding Vladimir Putin was crucial for US policymakers at this time. Hill and Gaddy (2015, 349, 351, 352) explain that Putin is a strategic thinker but has a poor understanding of the West. \Putin thinks, plans, and acts 36 Vladimir Putin. \Speech and the Following Discussion at the Mu- nich Conference on Security Policy." (Feb. 10, 2007), available at http://en.kremlin.ru/events/president/transcripts/24034 37 \Russia `may withdraw' from US missile treaty." (translated from Russian) BBC Russian.com (Feb. 15, 2007), available at http://news.bbc.co.uk/hi/russian/russia/newsid 6365000/6365413.stm 213 strategically. . . . [F]or Putin, strategic planning is contingency planning. There is no step-by-step blueprint" (Hill and Gaddy 2015, 352). As such, Putin likely understood that protesting too loudly to the US withdrawal with little chance of changing the US position would make Russia look impotent. Moreover, since Putin was suspicious of US motives and assumed that the US position could change in the future, planned for the worst and worked toward building up Russia's own defenses. At the last formal summit between Bush and Putin, missile defense was a central issue (Stent 2014, 157). Although the US did not intend its missile defense goals to be aimed at Russia, the Russians were concerned that once deployed the systems could be used against Russia (Stent 2014, 153{158). The US tried to calm Russian anxieties by oering to allow Russia mili- tary observers at missile defense sites in Poland and the Czech Republic, despite those countries' reservations (Stent 2014, 157). No agreement was ever reached on missile defense, and Russia announced a number of vague countermeasures to the US deployments. When coming into oce, Obama, like Bush, sought to \reset" relations be- tween the US and Russia. In working toward a reset, Obama and Medvedev worked on cooperating on various security and non-security issues (Obama 2020, 339{343, 461{465). Included in these eorts were reductions to each country's respective nuclear stockpile, eventually negotiated into the Strate- gic Arms Reduction Treaty (New START). The Obama Administration saw the conclusion and ratication of New START as crucial in its eorts to address the threat of nuclear proliferation in Iran (Obama 2020, 608). To 214 secure passage of ratication through the US Senate, the Obama Adminis- tration pledged to undertake a multi-year, multi-million dollar moderniza- tion of the US's nuclear arsenal (Obama 2020, 608). During the Obama Administration, the US was already reconsidering its land-based missile defense systems in Europe (Obama 2020, 342). At the 2010 NATO Summit in Lisbon, the leaders issued a statement making missile defense in Europe a priority of the alliance. 38 The US sought more eective defenses against countries like Iran, a point made to the Russian leaders (Obama 2020, 342, 465, 471). As of August 2021, Iran's Sejjil ballistic missile had a maximum range of 2,000 km, 39 which could potentially reach Turkey, Greece, North Macedonia, Bulgaria, and Romania but not the rest of the NATO alliance. The US decision to withdrawal from the ABM Treaty under the Bush Ad- ministration and plan to house missile defense systems near Russia's borders continued to be a source of strategic concern to Russia and Vladimir Putin during the Obama Administration (Obama 2020, 464). The US's placement of anti-ballistic missile systems|Aegis Ashore|in Romania 40 and Poland 41 38 See paragraphs 2, 23, 30, 36{38, 45 of the \Lisbon Summit Declaration" (Nov. 20, 2010), available at https://www.nato.int/cps/en/natolive/ocial texts 68828.htm 39 Missile Defense Project, \Missiles of Iran," Missile Threat, Center for Strategic and International Studies, June 14, 2018, last modied August 10, 2021, available at https://missilethreat.csis.org/country/iran/ 40 \United States Ballistic Missile Defense Site at Deveselu Air Base in Ro- mania." US Embassy in Romania, available at https://ro.usembassy.gov/our- relationship/implementing-missile-defense-europe/united-states-ballistic-missile-defense- site-at-deveselu-air-base-in-romania/ 41 \As Ukraine war rages, Poles divided over US missile base." France24 (Mar. 18, 2022), available at https://www.france24.com/en/live-news/20220318-as-ukraine-war- rages-poles-divided-over-us-missile-base 215 likely fueled Russian concerns. Russia sought to counter balance US missile defense eorts by building up its nuclear weapon capacity. 42 In 2014, the US publicly accused Russia of violating the INF Treaty by de- veloping and testing ground-based cruise missile (9M729 missile) that had a range in excess of the limits agreed to in the treaty. 43 Russia denied the accusations. In fact, the Russians blurred the line between nuclear and conventional weapons such that it could make arguments against accusa- tions of breach. 44 Yet, for years, the Obama Administration was \stuck in [a] zombie state" of how to approach Russian breaches, eventually deciding on a plan to slowly escalate pressure through negotiations, declarations of breach, expand US missile defense, and continued negotiations. 45 Russia accused the US of its own violations by placing missile defense launch sys- tems in Europe, which could be used to re cruise missiles (using targets in missile defense tests) and using armed drones that were equivalent to ground-launched cruise missiles. 46 Some US ocials, with a hint of disdain, believed US non-compliance to be unlikely: \[a]s with all US treaties, the Defense and State Departments were overgrown with lawyers; we couldn't 42 Dave Majumdar. \Russia's Nuclear Weapons Buildup Is Aimed at Beat- ing U.S. Missile Defenses." The National Interest (Mar. 1, 2018), available at https://nationalinterest.org/blog/the-buzz/russias-nuclear-weapons-buildup-aimed- beating-us-missile-24716 43 Unclassied Adherence to and Compliance with Arms Control, Nonproliferation, and Disarmament Agreements and Commitments. US Department of State (July 2014) pp. 8{10. 44 Author interview (Interviews J, K, L, M), March 2020. 45 Author interviews (Interviews J, K), March 2020. 46 Unclassied 2015 Report on Adherence to and Compliance with Arms Control, Non- proliferation, and Disarmament Agreements and Commitments. US Department of State (June 5, 2015). 216 violate a treaty if we wanted to" (Bolton 2020, 160). 47 H.R. McMaster, who served as National Security Advisor before John Bolton attributed Russia's violations as part of its eort to develop the capacity to carry out their doc- trine of \escalation control"|threatening a nuclear attack on Europe and, thus, creating a dilemma for the US in how to respond (McMaster 2020). Arms control had not been high on the Trump Administration's agenda in the early years (Bolton 2020, 130). Yet, it may not come as a surprise that some in the Trump White House desired to end the US participation in the INF Treaty (Bolton 2020, 159). 48 Internally, there was little consideration of the eects of withdrawal or the impact on the other members of the INF Treaty, namely Belarus, Kazakhstan, and Ukraine. 49 Before ultimately an- nouncing the US's intention to withdraw from the treaty, the Trump Admin- istration made mild eorts to coax Russia back into compliance or bargain for exibility. In discussing his 2018 meeting in Geneva with Nikolai Patrushev soon after becoming US National Security Advisor, H.R. McMaster explained the un- derlying rationale for dialogue with Russia: \Russia is, of course, a nuclear power, and a strained relationship is better than no relationship, if for no other purpose than to prevent misunderstandings that might increase the chance of war" (McMaster 2020, 24). This came after Secretary of State 47 John Bolton appears to have a penchant for disparaging lawyers, of which he is one: \Putin recalled that both of us were lawyers, saying, `We could go on talking like this until dawn,' and we then exchanged jokes about lawyers" (Bolton 2020, 169). 48 Others in the administration, including Tim Morrison, Senior Director for Europe and Russia on the National Security Council, also subscribed to the same views as Bolton. Author interviews (Interviews H, M), March 2020. 49 Author interviews (Interviews J, K), March 2020. 217 Rex Tillerson had tried to use his background as an oil executive to improve US-Russia relations (McMaster 2020, 26). McMaster also noted that the US sought to compel a change in Russian behavior by threatening withdrawal from the INF Treaty in 2017 through the State Department (McMaster 2020, 308). McMaster explained that the US tried to impose costs on Russia as part of their bargaining and to push it back into compliance (McMaster 2020, 25). Those costs were clearly unsuccessful at persuading the Russians. In August 2018, US National Security Advisor John Bolton and his Russian counterpart Nikolai Patrushev met in Geneva to negotiate over arms control after Trump's infamous July meeting with Putin at Helsinki. Bolton argued that Russia was consistently breaching the treaty, that the provisions put the US at a disadvantage compared to China and Iran, and it was techno- logically outdated in being limited to land-launched missiles (Bolton 2020, 159{160). During the negotiations, both sides accused the other of breach- ing their obligations. They also discussed amending or \universalizing" the INF Treaty to include China, Iran, and others; however, the prospect of getting those countries to agree was low. Bolton intended to impress onto Patrushev that the US threat of withdrawal was credible \even though there was no ocial US position" (Bolton 2020, 160). NATO allies and some in the government tried to delay a decision to with- draw to buy more time for the bargaining process (Bolton 2020, 170). Ger- many in particular sought a 60-day delay in the decision, which US Secretary of Defense Jim Mattis supported. Bolton urged the president to \pull the trigger on December 4" to avoid delay and further eorts at negotiation. 218 Bolton also convinced the president to \to announce simultaneously the suspension of our treaty obligations because of Russia's material breach, a concept separate from withdrawal, which would allow us to begin `violating' the treaty" before the end of the withdrawal notice period (Bolton 2020, 171). After agreeing with Bolton, Trump was persuaded by Angela Merkel to delay the formal withdrawal for 60 days to allow for more bargaining. Trump did not seem to comprehend or acknowledge that a 180-day waiting period existed for the withdrawal to take eect. President Trump announced the US intention to terminate the INF Treaty because of Russian breach and worries about China's growing missile ca- pabilities. Secretary of State Mike Pompeo declared Russia in \material breach" of the treaty. The ocial withdrawal notice was given to the other members on February 2, 2019, that the US will withdrawal unless Russia come into compliance by eliminating its ground-based missiles that violated the treaty. Russia did not comply and the US withdrawal was eective on August 2, 2019. 5.4 Cooperation and Withdrawal of Superpowers 5.4.1 US Approach to Disengagement During the four years of the Trump Presidency, many wondered if the US had entered into a new era of disengagement with the international community. During his campaign, Donald Trump made no secret of his intentions to 219 pull the US out of a number of international commitments, especially those made by President Obama. He railed against the Trans-Pacic Partnership (TPP) and the Paris Climate Accord. Once in oce, he made good on his campaign promises and pulled the US out of a number of international organizations and agreements, threatening to do so to even more. Was Donald Trump unique in his antagonism, or at least skepticism, of international cooperation? Were the US exits under his presidency more than other presidents? Opposition to international cooperation in the US is not new. Going back to the founding, some in the US opposed high levels of engagement in international aairs. Famously, George Washington warned against foreign entanglements and endorsed a policy of renversement des alliances (\reversal of alliances") by advising to \steer clear of permanent alliances with any portion of the foreign world." 50 In his recommendations for how the US should approach foreign policy, Washington emphasized his prescription for \temporary" agreements that may need to be abandoned: Harmony, liberal intercourse with all nations are recommended by policy, humanity, and interest. But even our commercial pol- icy should hold an equal and impartial hand, neither seeking nor granting exclusive favors or preferences; consulting the nat- ural course of things; diusing and diversifying by gentle means the streams of commerce, but forcing nothing; establishing with powers so disposed, in order to give trade a stable course, to de- ne the rights of our merchants, and to enable the Government to support them, conventional rules of intercourse, the best that present circumstances and mutual opinion will permit, but tem- porary and liable to be from time to time abandoned or varied 50 George Washington. \Farewell Address" (1796). 220 as experience and circumstances shall dictate; constantly keep- ing in view that it is folly in one nation to look for disinterested favors from another; that it must pay with a portion of its inde- pendence for whatever it may accept under that character; that by such acceptance it may place itself in the condition of having given equivalents for nominal favors, and yet of being reproached with ingratitude for not giving more. There can be no greater error than to expect or calculate upon real favors from nation to nation. It is an illusion which experience must cure, which a just pride ought to discard. 51 The US later ended its alliance with France set forth in the 1778 Treaty of Alliance at the Convention of 1800 where the countries signed the Treaty of Mortefontaine. 52 Upon taking oce as president, Thomas Jeerson adopted this position as described in his inaugural address by detailing his adminis- tration's approach to foreign policy as \peace, commerce, and honest friend- ship with all nations, entangling alliances with none." 53 In 1823, the Monroe Doctrine established the US foreign policy of avoiding involvement in Eu- ropean aairs in exchange for European nations avoiding involvement in the Western Hemisphere. President Abraham Lincoln secured congressional ratication of his termination of a treaty with the United Kingdom in 1865 (Bradley 2013). Isolationism was prevalent in the early twentieth century. President Woodrow Wilson was forced to modify his campaign for his second term, winning in 1916 on the slogan \He kept us out of war" (Beschloss 2019). President Franklin Roosevelt was constrained by the Neutrality Acts of the 1930s, 51 (Id., emphasis added.) 52 The treaty ended the existing alliance, ended a quasi-war, and established most favored nation trade relations between the US and Napoleonic France. 53 Thomas Jeerson, Inaugural Address (1801). 221 which largely lead the administration to devise the lend-lease program with the United Kingdom to avoid running afoul of the law (Beschloss 2019). Senator Arthur Vandenberg decried Lend-Lease stating: \We have torn up 150 years of traditional foreign policy. We have tossed Washington's farewell address into the discard." 54 After World War II, voices of non-intervention continued, most prominently by members of Congress like Ohio Senator Robert A. Taft who opposed the creation of NATO and criticized the Tru- man Administration's handling of the Korean War. Taft's stance stands in contrast to other Republicans like President Dwight D. Eisenhower, who saw alliances and cooperation crucial during the Cold War, and of Senator and presidential candidate Barry Goldwater who favored aggressive anti- communist internationalist policies. In the modern era, approaches to engagement and international cooperation have largely fallen along ideological lines. Conservatives in the US largely take a \realist" approach generally favoring policies, at least since the 1980s, that support the themes of conservatism, neoconservatism, and isolationism (Rathbun 2008). Long-serving Senator Jesse Helms of North Carolina (R) supported engage- ment in the context of security arrangements, especially between Russia, 55 54 H.R. 1776, 1941 (77th Congress). The bill was defeated largely along party lines (28-54-13). 55 Helms supported the Partnership for Peace and partnered with fellow Senator Joe Biden in demanding that the Bush administration obtain nuclear arms reductions with Russia by formal treaty. 222 but was key in blocking any ratication of the Rome Statute of the Interna- tional Criminal Court|going so far as to introduce the American Service- Members' Protection Act or the \Hague Invasion Act" (passed in 2002), authorizing, among other things, the president to use \all means necessary and appropriate" to seek the release of US personnel detained at or on be- half of the ICC. In 2007, Congressman Ron Paul, Republican from Texas, introduced the American Sovereignty Restoration Act 56 calling for the US to withdraw from the United Nations and expel the organization's headquar- ters. 57 That legislation has been reintroduced multiple times, most recently in 2019 by Congressman Mike D. Rogers, Republican from Alabama. 58 The Trump Administration's exit from a number of treaties led to action by Congress and a owering of legal scholarship on the ability of a US President to unilaterally withdraw from a treaty (Galbraith 2017; Bradley and Goldsmith 2017; Koh 2018; Amirfar and Singh 2018; Lapointe 2020). In the US, the federal government has exclusive authority over international relations, as opposed to the states, including the ability to enter into a treaty. 59 The Constitution is explicit about the procedure for entering into a treaty but leaves open who has authority to leave a treaty. The Treaty Clause, as it is called, is found in article II, section 2, clause 2 of the US Constitution. It reads: \[The President] shall have Power, by and with the 56 H.R.1146. 110th Congress (2007{2008). 57 The Republican Party of Texas has called for withdrawal from the U.N. since at least 1998. 58 H.R. 204. 116th Congress (2019{2021). 59 US Constitution, art. I, sec. 10, clause 1 (1789) (\The Contracts Clause"). 223 Advice and Consent of the Senate, to make Treaties, provided two thirds of the Senators present concur . . . ." Once a treaty is ratied, it becomes binding federal law. As such, Congress has the power to modify or repeal treaties through legislative action, even if this breaches the treaty under international law. In the Head Money Cases, 60 the Supreme Court held that treaties hold the same|not a privileged| position as other domestic, federal law. For most the nation's history up until the twentieth century, Congress as- serted a more active role in determining whether to end a treaty (Bradley 2013). The rst unilateral Presidential withdrawal occurred in 1899{over a century after the adoption of the Constitution{when the McKinley ad- ministration terminated certain portions of an 1850 commercial treaty with Switzerland. Unilateral withdrawal by the President became more common thereafter, picking up considerably during the administration of Franklin D. Roosevelt. Unilateral presidential treaty termination continued for the following decades, leading to a signicant shift toward the president's power in the 1970s after the Carter Administration successfully terminated the 1954 mutual defense treaty with Taiwan over the objections and lawsuit of members of Congress. 61 Since then, Presidents have continued to act unilat- erally, without signicant resistance from Congress, with notable exceptions during the Bush II and Trump administrations. Bradley (2013) attributes 60 112 U.S. 580 (1884), also referred to as Edye v. Robertson. 61 See Goldwater v. Carter 444 U.S. 996 (1979). 224 the shift as part of the broad accretion of presidential power in the twentieth century. 62 However, was the Trump Administration truly an anomaly? Figure 5.1 shows all the treaty exits made by the US from 1945 to the beginning of 2021 by president and president's party. The treaty exits are not uniformly distributed across presidents, with signicant peaks in treaty exit during the Carter, Clinton, and to a much lesser extent, Trump administrations. In looking at the volume of exits by democratic presidents, what appears to be happening is a signicant restructuring of the US international relation- ships. In the case of President Carter, the president terminated the Bancroft treaties in consultation with the Senate Committee on Foreign Relations in 1980. 63 The Bancroft treaties were nineteenth and early twentieth century bilateral and multilateral agreements 64 between the US and other countries recognizing the right of each party's nationals to change their nationality through naturalization under dened parameters. After a series of supreme court rulings called into question the constitutionality of Bancroft treaties, 65 the administration determined to terminate the treaties and establish new immigration policies and agreements. 66 62 Also, imperial presidency. 63 Department of State Bulletin, v. 80, no. 2045, December 1980, pp. 79{80. 64 From 1868 to 1937, the US entered into 25 Bancroft treaties covering 34 foreign states. 65 See Reid v. Covert, 354 U.S. 1 (1957); Schneider v. Rusk, 377 U.S. 163 (1964); and Afroyim v. Rusk, 387 U.S. 253 (1967). 66 Although, issues regarding these treaties have risen in more recent litigation. See Kuper v. Mulrean, 209 F.Supp.2d 1079 (S.D. Cal. 2002). 225 Trump's withdrawals dier from those of the Carter and Clinton administra- tions in that, although he often framed the withdrawal as part of a broader bargaining process, he also tapped into the anti-internationalist rhetoric for domestic political purposes. Trump was not against entering into new agree- ments that he saw as transactionally benecial to the country|or himself politically. Thus, his treaty withdrawals, although worrisome to many, did not signal the end of all cooperation. Rather, it was cooperation on his terms. Other than the presidents who exited frequently, some presidents have very few if not no registered treaty exits. That by no means that their administrations did not threaten to withdraw. Now that withdrawal has become such a partisan issue, it is possible we will see US behavior vacillate as control of the White House changes. Figure 5.1: Treaty Exits by the US On April 26, 2019 during a speech at an annual NRA meeting, President Trump announced that the US would withdraw its signature from the U.N. 226 Arms Trade Treaty (ATT) and signed an order to the eect on stage. 67 Trump unwound another Obama era policy by the stroke of his pen, which the 45th President threw into an ecstatic audience. President Obama had entered into the multilateral agreement in 2013. Trump's high prole ac- tions have motivated Congress to reassert restraints on unilateral presiden- tial withdrawal. At the 2018 Helsinki summit, President Trump, standing next to President Vladimir Putin, sided with the Russian leader and against U.S. intelligence agencies in stating \I have President Putin, he just said it's not Russia. I will say this, I don't see any reason why it would be." 68 Trump's statements were widely denounced by members of both parties in Washington. Sena- tor John S. McCain called it a \disgraceful performance," adding \No prior president has ever abased himself more abjectly before a tyrant." 69 Fearing Trump may withdrawal the US from NATO, a bipartisan group of Senators reintroduced a joint resolution requiring a president to seek the advice and consent of the Senate before leaving NATO. 70 Without Senate approval, the bill prohibits any funding to be used to leave NATO and authorizes the 67 Bill Chappel. \Trump Moves To Withdraw U.S. From U.N. Arms Trade Treaty." NPR News (April 26, 2019), available at https://www.npr.org/2019/04/26/717547741/trump-moves-to-withdraw-u-s-from-u- n-arms-trade-treaty 68 \Trump sides with Russia against FBI at Helsinki summit." BBC News (July 16, 2018), available at https://www.bbc.com/news/world-europe-44852812 69 Id. 70 The resolution had been introduced in the prior Congress without a vote in committee or oor vote. See S.J.Res.62 115th Congress (2017-2018). 227 Congressional Legal Counsel to challenge the president in court. 71 A sim- ilar resolution was introduced in the House of Representatives. 72 Senator Tim Kaine of Virginia (D), a co-sponsor of the resolution, cited the Admin- istrations withdrawals from the \Iran nuclear deal" and the Paris Climate Accord, among other reasons, for why Congress should take action to re- strain the president. The Senate Resolution passed a committee vote in December 2019, but it did not receive a vote on the Senate oor before the end of the 116th Congress. 73 Figure 5.2 shows the activity of the US Congress with respect to treaty exits. The gure shows the number of bills or resolutions introduced by each chamber from 1975 to 2020, color coding the bars by the party of the bill or resolutions sponsor. To assess congressional activity in comparison to the president, the name and party of the president from 1975 to 2020 is also included on the right. As gure 5.2 shows, activity in congress has uctuated greatly over the last four decades. Both chambers, the House and Senate, are particularly active in the early 1990s, mid-1990s, and in the Trump administration. Moreover, in some of those peak years, it appears that the opposition party is most active in introducing legislation on the topic| ostensibly seeking to constrain or compel presidential action. Congress is also a locus of bargaining between congress members, committee staers, advocacy groups, and foreign governments. 74 In the US system, this system 71 See S.J.Res.4 116th Congress (2019-2020). 72 H.J.Res.41 116th Congress (2019-2020). 73 See S.J.Res.4 116th Congress (2019-2020). 74 Author interviews (Interviews H, I), February 2020. 228 Figure 5.2: Congressional Actions re Treaty Exit of checks and balances likely has a signicant impact on the US's decisions to withdraw from treaties. 229 5.4.2 Russian Approach to Disengagement In assessing the Russian approach to international cooperation and dis- engagement, we have to start with an understanding of how Russian ap- proaches international law generally. The Soviet legacy carries over to the the modern Russian approach in a number of ways. From the start, the USSR developed an \anti-Western," regional view of in- ternational law (M alksoo 2015, 4). Based in part on its Marxist and Leninist foundations, the Soviet Union took a view of competition with the capitalist West and often acted in ways that to the West would be labeled \revision- ist." These tendencies carried over into Russia's modern engagement with the international community and have become more explicit in recent years (M alksoo 2015, 158). \In fact, there are reasons to think that Russia's ide- ological energy to shape international law and challenge Western normative projects continues to be higher than China's" (M alksoo 2015, 11). Key ideas underlying the Russian view of international relations and law are the idea of a sphere of in uence (groraum), the \Russian World" (russkyi mir), and derzhava, the idea that Russia is a great power. This last concept, derzhava, is important in how Russia sees itself in the community of states and whether its actions can violate international law. For Russia, in a general sense, international law is the post-World War II power structure, and defending international law is defending Russia's 230 privileged place in that order (M alksoo 2015, 175). 75 In that way, Rus- sia is the guardian of international law, the katechon or restrainer against Western unilateralism (M alksoo 2015, 148{153). Based in this view of the global power structure, symmetry among the Great Powers is essential. As Vladimir Putin explained as he rst took oce as president: \A strong state for Russia is not an anomaly, but, on the contrary, a source of guarantee and order" (Stent 2014, 52). Russia's shock and dismay regarding NATO intervention in 1999 was used as justication for military responses in South Ossetia and Abkhazia. Scholars have explicitly characterized Russia as res- urrecting imperialist views of international politics (Plokhy 2015; Snyder 2018). Russia is a \derzhava with imperialist re exes that is inclined to use international law in a primarily instrumental way" (M alksoo 2015, 182). US Diplomat George F. Kennan characterized the Soviet Union's foreign policy in stark realist terms: \[T]he eye of the Kremlin can distinguish, in the end, only vassals and enemies." 76 Tellingly, separatists in eastern Ukraine have voiced these same views. An- drei Purgin, a politician from the Republic of Donetsk in Eastern Ukraine, shared a particularly Russian view of international relations and its unique heritage: \We are rooted in the Byzantine tradition. Western laws are for us secondary and a sense of justice primary" (M alksoo 2015, 183). This pulls 75 Some in the US Congress voiced similar views when considering US membership in the 1998 Rome Statute. \Is a U.N. International Criminal Court in the U.S. National Interest?" Hearing before the Subcommittee on International Operations of the Commit- tee on Foreign Relations United States Senate (July 23, 1998) S. HRG. 105{724 (105th Congress, 2d session), 1. 76 Quoted in: Fareed Zakaria. \ ` A Guest of My Time' " The New York Times (Feb. 21, 2014), available at https://www.nytimes.com/2014/02/23/books/review/the-kennan- diaries-by-george-f-kennan.html 231 from Mikhail Nikiforovich Katkov, a popular journalist during the reign of tsar Alexander III, who wrote: \the West relies on contract, the East on justice." Domestically, it can be challenging to assess Russian practices under law. Throughout the Soviet Union and into modern Russia, legal texts often do not comport with state practice. For example, the 1977 Soviet Constitution established the fteen Soviet Republics as \sovereign" but only in the sense that sovereignty was dominated by decisions from Moscow, especially in the context of foreign relations (M alksoo 2015, 6). In addition, Russian scholarship tends to be theoretically and dogmatically oriented, rather than focused on the practice of national or international courts (M alksoo 2015, 77{97). \Law, in other words, has a dierent meaning and function in society" (M alksoo 2015, 6). In a 2013 poll, 60 percent of Russian respondents thought the government did not respect the constitution partly or at all (M alksoo 2015, 7). For Russian scholars, international law is rst and foremost what the exec- utive branch does, not what the courts do. The Russian political system is dominated by a \hegemonic" executive, in an oce shaped by Russia's rst two, strong presidents: Boris Yeltsin (1991{1999) and Vladimir Putin (2000{2008) (Willerton 2013). Over that time, \Russia managed no demo- cratic changes of executive power" (Snyder 2018, 79). Under the Russian Constitution, the president represents the country in international relations and directs foreign policy, including the negotiation 232 and signing of treaties (Constitution of the Russian Federation (1993), arts. 80, 85, 86). The president is also the Supreme Commander-in-Chief of the Russian Armed Forces (art. 87.) Article 106 of the 1993 Constitution requires that laws of the Duma, lower house, about treaty ratication and denunciation be reviewed by the Council of Federation, upper house. Russian treaty law and practice is distinct from that in the US and many other countries. Russia, continuing the practice from Soviet years, has en- acted special legislation regarding the procedures for concluding, ratifying, and terminating treaties (Butler 1995). Under the 1936 Constitution of the USSR, the Presidium of the Supreme Soviet of the USSR held the power to ratify and denounce international treaties signed by the USSR (art. 49(l)). This power was carried over to the 1977 Constitution of the Soviet Union where the Presidium and Council of Ministers of the USSR held that au- thority (arts. 121(6), 131(6)). In Soviet practice, denunciation of treaties was governed by the special law regarding ratication and denunciation of treaties. 77 Additionally, the USSR considered a treaty annulled when a party to it violated its treaty obligations. 78 In 1995, Russia enacted the Russian Federal Law on International Treaties, which essentially incorporated the 1969 Vienna Convention on the Law of Treaties into Russian law (Butler 1995). Articles 35 through 40 concern termination and suspension of operation of treaties to which Russia is a 77 Memorandum on the Soviet Doctrine and Practice with Respect to the Law of Treaties." U.N. General Assembly. A/CN.4/37 (21 Nov. 1950), Arts. VII, XXII, pp. 13{16, 39, available at https://legal.un.org/ilc/documentation/english/a cn4 37.pdf [hereafter \UN Memo of Soviet Treaty Doctrine and Practice"]. 78 UN Memo of Soviet Treaty Doctrine and Practice, Art. XVIII, p. 34. 233 party. In general, the president can receive or make a proposal to termi- nate a treaty, which is then submitted to the Federal Assembly as proposed legislation. Once approved by the legislature, the president signs the law formalizing the treaty termination. In 2021, President Putin followed this procedure when withdrawing from the 1992 Open Skies Treaty. 79 Since 1945, Russia (Soviet Union) registered a withdrawal or denunciation with the UNTS less than twenty times. Figure 5.3 shows Russian treaty exits over time from both the Soviet Union and Russia. On the whole, General Secretary Brezhnev and President Yelstin were the leaders who exited the most during their term of leadership. Figure 5.3: Treaty Exits by the Russia (Soviet Union) The signicant dierence in exit behavior between Russia and the US is 79 \Putin submits legislation on Russian withdrawal from Open Skies treaty." Reuters (May 11, 2021), available at https://www.reuters.com/world/putin-submits-legislation- russian-withdrawal-open-skies-treaty-2021-05-11/ 234 fascinating and likely telling about how each country engages with the in- ternational community. Generally, the US was and is more active in making and ending treaties, being a central participant in the global legal order since 1945. Moreover, the practice of unilateral withdrawals by US presi- dents might lead to more exits based on ideological or partisan shifts with regular elections. Russia's low volume of exits could be the result of diering domestic legal structure, but this seems unlikely where the president holds \hegemonic" authority over the government. Other actors in the Russian system, like the Federal Assembly and the judiciary do not serve as sig- nicant checks on presidential predilections. Another possible explanation could be the more stable ideological and partisan make up of the Soviet and Russian governments, where policies remained largely consistent for long pe- riods of time. In that case, the leadership would have less desire to exit from treaties it likely signed and agreed to in the past. Further, Russia may have chosen other means of addressing changed circumstances, like bargaining or breach, with dierent frequency than the US. If the costs of breaching are lower for Russia, then that may be a more acceptable outcome in its strategic choice. 5.5 Analysis In reviewing the relationship between the US and Russia on arms control, we can particularly assess how breaching in an environment of extreme vig- ilance aects a state's strategic preferences across bargaining, breaching, 235 and exit. This is particularly relevant to Claim 1 derived from the theory in Chapter 2. In arms control, the states come to the table with suspicion. At the start, US President Reagan distrusted treaties with the Soviet Union, in uenced by the book The Treaty Trap by the founder of the Screen Ac- tors Guild, Laurence W. Beilenson (Homan 2009, 30). Especially in the context of security, the states are consistently pressured to defend their se- curity and appease domestic constituencies regarding eorts to protect their populations. A thread throughout the security relationship between the US and Russia was missile defense. Out of all the issues the US and Russia have faced in recent decades, missile defense remains particularly challeng- ing (Stent 2014). Missile defense was important to both parties, along with their allies, and potentially shifted the strategic balance in their nuclear armaments. The Soviet Union's Moscow defenses and the US defenses in Europe would routinely play a role in the states' bargaining for decades. Here, we will assess the theory's application to the case of bargaining of arms control from the Reagan Presidency to Trump's, focusing rst on the ABM Treaty and then on the INF Treaty. In the analysis, I will assess the obli- gations imposed by the treaties, the eorts to amend or bend the treaties, accusations of breach, 80 and eventually the rationales for exit. In reviewing the history of arms control between the two countries since the 1980s, it is intriguing to see the preference in form of the treaties change with adminis- trations. Interestingly, there was a distinct dierence in preference between 80 The US State Department's view and arguments on treaty compliance and breach are often based on a common law understanding of contract, especially where most of the State Department lawyers were trained in American law schools. Author interview (Interview L), March 2020. 236 Bush II and John Bolton (Bolton 2020, 161) for simpler treaties (e.g., 2002 Treaty of Moscow at less than 500 words) and the more detailed treaties like START (approximately 13,500 words) and new START (approximately 3,800 words not counting the 165-page protocol or 174-pages of annexes). Shorter agreements are less precise, likely have less stringent obligations, and generally leave more room for interpretation. In short, shorter treaties left more exibility to the parties in setting their behavior. In assessing these cases, it is useful to consider counterfactuals in assessing the explanatory power of the theory within the historical context (Fearon 1991; Levy 2015). Here, we might imagine counterfactual worlds where the US or Russia make dierent decisions either leading to the US not with- drawing or neither country violating one of the arms control treaties. Coun- terfactuals to some aspects of the Cold War are hard to imagine. Could the uncomfortable allies in World War II have come out of that con ict more trusting each other? What if they jumped straight into a shooting war in Europe rather than a cold war? Those counterfactuals would completely change the arms control environment examined here. As such, we will take the Cold War as a necessary backdrop to this analysis. \For all its dangers, atrocities, costs, distractions, and moral compromises, the Cold War|like the American Civil War|was a necessary contest that settled fundamental issues once and for all" (Gaddis 2006). Useful in assessing accusations of non-compliance, the US State Department 237 produces a report, pursuant to section 403 of the Arms Control Disarma- ment Act, 81 which analyzes the US and other nations' compliance with arms control, nonproliferation, and disarmament agreements in force. Of course, the report is generated by the government of one of the treaty participants being assessed, raising doubts that the US would declare itself in violation of its treaty commitments. Moreover, the reports are generated across ad- ministrations and with varying standards. 82 Nonetheless, it is a useful tool in having a multi-year look at the US view of compliance with detailed rationale. 5.5.1 ABM Treaty In short, the US withdrew from the ABM Treaty because no bargain was acceptable to either side to address the US desire for greater defense capa- bilities and the US was ultimately not willing to openly breach the treaty. The shock that prompted the US's withdrawal is generally explained as the collapse of the Soviet Union and the end of the Cold War, coupled with the new threat of terrorism made visible on September 11, 2001. Fitzmaurice and Elias (2005, 185{198) argue that this was a case where the US's with- drawal was, although explicitly based on the withdrawal provision in the ABM Treaty, it also implicated the principle of a fundamental change of circumstances set out in Article 62 of the VCLT allowing withdrawal. Thus, the US had multiple justications for its withdrawal. 81 22 U.S.C.x 2593a. 82 Author interview (Interview L), March 2020. 238 The ABM Treaty most directly provided a private good to its members that allowed both the US and the Soviet Union (as well as its successors) to avoid the costs of an expensive arms race. However, preventing a nuclear arms race may also be a public good, especially for European states, in that it decreased the likelihood of a nuclear war|something that is in everyone's favor. In the lead up to the withdrawal, the private benets of the ABM Treaty likely decreased for the US because any savings in decreasing the arms race were less valuable when Russia not competing as stridently. The public benets of the treaty likely remained neutral; however, the costs of remaining in the treaty likely increased as it restricted the US's policy choices in other areas, including counterterrorism. The repercussions and costs of the decision to withdraw may include President Vladimir Putin's later announcement that Russia was developing new missile systems. 83 The US explicitly noted the ABM Treaty as being an impediment to its pref- erence to increase its missile defense capabilities. Rumsfeld counted with- drawing from the ABM Treaty as one of three main obstacles to achieving success in missile defense, where exiting the treaty was the legal obstacle, as- suring allies that the systems could protect them a diplomatic obstacle, and successful tests the technical obstacle (Rumsfeld 2011, 613). However, when US policy shifted to favor missile defense, it did not immediately decide to end the ABM Treaty. It both tried to reform the treaty and reinterpret it to allow for some level of breach for decades. 83 Vladimir Putin, Presidential Address to the Federal Assembly (March 1, 2018, Moscow) http://en.kremlin.ru/events/president/news/56957 239 Applying expectations from the theory and ndings of the previous chapters, we start by assessing the treaty text. The ABM Treaty has a 5,220 words and a permissive-obligatory score of20 (see Appendix C). The treaty is slightly longer than the average treaty analyzed in Chapter 3 but only by 755 words. Additionally, its permissive-obligatory score is permissive generally and much more permissive compared to the average treaty (31.5). Based on our theoretical expectations, we should expect exit to be relatively unlikely because of its average length and permissive score (Hypos 1 and 4). As explained herein, the treaty was in force for approximately three decades. That longevity may be attributed to the treaties use of permissive language. Additionally, because the ABM Treaty was a bilateral agreement between the US and Russia (USSR), the number of participants is as low as can be, at two members. With only one other member to negotiate with for half of the treaty's lifespan, bargaining should be relatively easier decreasing the likelihood of treaty exit (Hypo 5). Even after the USSR successor states acceded to the treaty, the number of participants was still relatively low at ve members (Hypo 5). Nonetheless, we might consider whether the multilateralization of the treaty in the 1990s as increasing the risk of treaty exit. Neither state submitted a reservation or declaration, precluding any conclusions as to whether the treaty was more or less obligatory or exible (Hypo 2 and 3). Regarding its reinterpretation, the Reagan Administration rst tried to nd a way for the US to pursue SDI as the treaty existed, i.e. without com- ing to an understanding with the Soviet Union. At that point, the US had 240 been unsuccessful in enticing the Soviets with a bargain on SDI. Armed with clever treaty reading, the US sought to bend the treaty to its limits. Koh (1996) argues that the interpretation debates and confrontations within the US were an example of the transnational legal process in action. Actors like Senator Nunn, \norm entrepreneur" Gerard Smith, and NGOs like the Arms Control Association and the National Committee to Save the ABM Treaty worked as an \epistemic community" to internalize a narrow inter- pretation of the treaty. Koh (1996, 2649, fn. 249) points to the failure of the Reagan Administration's interpretation as leading to the new strategy of calling for withdrawal from the treaty. In taking the exit rst, bargain later approach, the Bush II administration forfeited leverage to continue the bargaining process, which was later repeated by the Trump Adminis- tration. 84 Both administrations expressed a desire to update or reform the security relationship with Russia; however, once the U.S. exited the treaties, it was much harder to bargain. The exits damaged trust between the states in context of security arrangements. Where bargaining fails and breaching is too costly, the US was left with the option to exit. Although Russia's immediate response to the withdrawal of the US from the ABM Treaty was muted, it had a deep eect on Russian policy going for- ward (Roberts 2018). Putin withdrew Russia from START II in retaliation. Gorbachev (2016, 314, 355) attributes much of the blame for souring rela- tions between the US and Russia on growing US \unilateralism," which was manifest in the US's withdrawal from the ABM Treaty and Kyoto Protocol, 84 Author interviews conrm this dynamic. Author Interview (Interview H), February 2020. 241 along with resistance to new international institutions like the International Criminal Court (ICC). The US withdrawal from the ABM Treaty precipi- tated a decoupling of the countries in the area of security and arms control. Of course, the exit did not prevent further cooperation, as evidenced by the conclusion of New START between Obama and Medvedev in 2010. In examining the breakdown of the ABM Treaty, we can highlight at least two important insights from this episode: the relative importance of treaties and exits' aect on future cooperation. First, certain treaties may be more important or foundational than others, even within a particular issue area. As explained in this chapter, the US and Russia have and had concluded a series of arms control and disarmament treaties for decades. These treaties created a layered regime of rules and expectations as between the two nations as to how they can can expect the other to arm and conduct itself in the context of security. Within that multi-layered relationship, certain treaties likely were more important to the parties than others. A number of the arms control treaties had nite durations, and while important, would have a discrete impact on the parties' relationship. Other treaties of unlimited duration would lock in expectations into the future. Disputes over the less crucial treaties or even exits from them may have less impact. Disputes and exits over more important treaties can have signicant spill over eects to other treaty relationships and destabilize broader cooperation. Here, the ABM Treaty locked-in the strategic logic of mutual vulnerability agreed to by the states in the adolescence of the Cold War. That logic, and thus a treaty like the ABM Treaty, would be central to the states' broader 242 understanding of their relationship on the issue of security. A change in preferences within the ABM Treaty relationship would shift the calculation of the parties in other existing or potential treaty relationships. Reagan's insistence in pursuing missile defense and distaste for the logic of MAD shifted ocial US policy for decades to come, especially given his place in the pantheon of modern conservative US politics. Missile defense, whether it was SDI, theater missile defense, or defenses against \rogue states," resurfaced again and again in negotiations between the states. One might wonder whether this was a case of one side, Russia, bringing up an issue as part of a bargaining strategy of obstructing or sabotaging the bargaining process, knowing the strong US preference for missile defense. Of course, in the context of international bargaining, states will use whatever leverage they have available. In this case, the Russians' fear of the strate- gic destabilizing eect of US missile defense and loss of reputation from a recognition of is relative weakness, likely prompted Russians to favor less exibility and avoid signicant concessions. Even in the 1990s when Russia was more open to cooperation with the West generally, Russia was con- cerned that even the theater missile defense system proposed by the Clinton Administration could quickly lead to a larger more capable defense system. This commitment problem was never eectively resolved, especially when the US took action to push for broad missile defense in the lead up to Bush II's exit. Bush's exit, although not immediately damaging to the security relationship| the countries agreed to SORT shortly thereafter|was used as a justication 243 for the exit from a number of other related treaties. Having largely ended the war in Chechnya and consolidated his power after his rst term, it ap- pears Putin took a dimmer view of cooperation with the US and Europe on issues of security. During the War on Terror and with the prospect of a rising China, the US's focus was less concerned with the security relation- ship with Russia. For Russia, especially for leaders like Putin, that lack of focus by the US was part of the problem. To keep its status, Russia had to remain an equal in terms of bargaining over security issues. Unshackling the parties from the Cold War doctrine of mutual vulnerability provided Moscow with fodder for future arguments to justify arming, meddling, and retaliatory actions in other treaties. Going forward, Russia could consis- tently point to the US exit as an unambiguous rationale that the US does not want to cooperate on security. Had the US not withdrawn from the treaty, Russia could not use the bright- line exit as a justication. It may have argued that the US was in breach, as it did with other treaties like the INF Treaty. But those arguments would have been harder to make convincingly. Also, the US and Russia may have agreed to mutual concessions or amendments to the ABM Treaty to allow for defenses directed at non-Russian threats. The parties did cooperate on reducing strategic arms, thus cooperation on missile defense was theoreti- cally possible. In the case of the INF Treaty, US allies had asked the US to do more to address Russian concerns over the US's increase use of drones and radar testing; however, the US waved o those concerns. 85 Further, 85 Author interviews (Interviews J, K), March 2020. 244 maintaining not only the balance of power but the balance of respect be- tween the states likely would have been welcomed by the Russians. The US argument that missile defense was limited and would not be eective against a Russian, or Chinese, nuclear assault suered from a simple commitment problem. How could the US make these assurances into the future? Possible solutions, coming from our understanding of rational treaty design, could have been to increase the exibility of the ABM Treaty by creating nite durations for certain proposals. Those set durations would allow the parties to adjust with a set timeline, something proposed in talks between Gor- bachev and Reagan. Moreover, it could have more explicitly linked missile defense to strategic arms reduction, being that the two issues are related. With more defense, each side has an incentive to increase stockpiles or speed to break-though the defenses. By setting caps on both, the parties could have managed the strategic vulnerability in such a way to reform the treaty relationship rather than shed it aside completely. 5.5.2 INF Treaty In looking at the treaty text, the INF Treaty was longer than the ABM Treaty by approximately 1,800 words and had a much higher permissive- obligatory score of 103 (see Appendix C). This score is almost three times greater than the average score for the treaties considered in Chapter 3. Therefore, we should consider the INF Treaty as more obligatory and less exible than the ABM Treaty. Thus, the risk of treaty exit would be greater for this treaty than the ABM Treaty (Hypos 1 and 4). Like the ABM 245 Treaty, neither state registered any reservations or declarations (Hypos 2 and 3). The INF Treaty was a bilateral agreement with only two participants, making negotiations over treaty terms easier and more exible (Hypo 5). As described above, the later US exit from the INF Treaty in 2019 was linked to the earlier exit from the ABM Treaty in 2002. As described by senior ocials in the Trump Administration, they drew a direct line from wanting to pursue missile defense and Bush II's exit to the desire to leave other Cold War era treaties. Chief among them was National Security Advisor John Bolton who campaigned internally to in uence the President's decision. However, even Bolton's predecessor, McMaster, thought the exit from the ABM Treaty was correct and allowed for the US to properly pursue its security interests. Yet, it is surprising that those in the administration discounted the negative eect of Bush II's exit on Russia's later compliance behavior. Based on what little we know about the internal debates, if any occurred at all, the issue of Russian INF Treaty compliance was viewed narrowly in terms of one treaty rather than a broader view of the relationship between the states. This could partly be due to the lack of any real deliberations of arms con- trol policy within the Trump Administration. 86 In the White House, the Na- tional Security Council (NSC) often favors security or diplomatic approaches depending on the president. 87 Based on interviews with individuals who had worked in government on these policies, as well as staers on Capitol Hill, 86 Author interviews (Interviews H, I), February and March 2020. 87 Under Republican presidents, the NSC favor defense. Under Obama, the NSC sided more with the State Department. Author interviews (Interviews J, K), March 2020. 246 there was no or very little interagency process in the decision leading up to the US exit from the INF Treaty. 88 In stark contrast to the numerous and iterative memoranda of the Rumsfeld Defense Department and National Se- curity Advisor Condoleezza Rice's eorts at promoting interagency harmony under the Bush Administration, the Trump Administration functioned from the top down. Decisions would be made by the White House ahead of in- teragency debate on policy. This left many in the government scrambling to catch up to iron out the details of the new US policy. Thus, there was rarely a clear bargaining position or ask from the US to avoid exit or facil- itate breach. Despite eorts by Secretary of State Rex Tillerson and NSA McMaster, it was unclear how the US could convince Russia to come into compliance after leaving the ABM Treaty and pursuing missile defense de- ployments in Europe. There simply was no oer that both sides were willing to accept and the INF Treaty was not exible enough to allow for changed strategy or violation. It is important to note Russia's more antagonistic behavior as it pertains to the prospects of avoiding a US exit from the INF Treaty. Analysts and researchers note two main motives for Russia's revisionist behavior in recent decades: status-seeking and security. In seeking status, Russia seeks to re- claim its role as a global superpower, which was lost after the collapse of the Soviet Union. The strategic arms control and the ABM Treaty were areas where Russia was still considered a superpower and dealt with as an equal to the US (Stent 2014, 30). Russia's view of western actions in the years after 88 Author interviews (Interviews H, I, J, K, L, M), February through March 2020. 247 the end of the Cold War conrmed their suspicions that the West did not respect Russia. Particularly in light of the 1999 Kosovo intervention, the West did not even consult with Russia in preparations for the intervention. To be left out of the decision-making for a matter so close to Russia in many respects was an in ection point in Russia-Western relations. Moreover, con- cerns over security have been fueled by the expansion of NATO and the EU, the US withdrawal from arms control agreements like the ABM Treaty, and western military engagement in Kosovo, Iraq, and Libya. In part, Russia saw western actions as disingenuous or hypocritical. To be sure, the status-seeking and security motivations are intertwined and overlap. Russia's security concerns of NATO expansion also exacerbate its view that the West does not properly respect Russia's status as a great power entitled to a sphere of in uence. \The basic line of Russian foreign policy through 2011 was not that the European Union and the United States were threats. It was that they should cooperate with Russia as an equal" (Snyder 2018, 79). Yet, \Putin chose empire over integration" (Snyder 2018, 81). Where Russia held this view, how could a bargain with the US ever have been reached to save the INF Treaty? It is conceivable that bargains that acknowledged Russia as an equal, as vaguely dened by the Kremlin, were plausible. Except after Russia's eorts to dominate and later dismantle Ukraine (not to mention election interference in Europe and the US) it is hard to imagine concessions from the West being palatable. Thus, losing the forum that the ABM Treaty provided likely degraded the prospects of future security negotiations, narrowing the bargaining range in the context 248 of treaties like the INF Treaty. 5.6 Conclusion In examining these two episodes of withdrawal by the US, we see how bar- gaining, breaching, and exit can be linked across treaties even bilateral ones, especially within the same issue area. Here, the US and Russia were often bargaining with respect to one treaty in the context of a web of treaties, as detailed in table 5.1. Missile defense, the primary driver of contention between the countries in the context of arms control, was often linked to negotiations over reductions in strategic nuclear arms or other restrictions on intermediate and short range weapons. While the model presented in the theory chapter focuses on one treaty relationship, the theory could be expanded to include a layered series of treaty relationships by which states bargain across linked issues and dierent fora. Treaty benets or costs may be substituted or shifted across treaties, in uencing decisions to exit (See, e.g., Dalton 2021). As described in this chapter, the desire to develop and deploy missile defense systems was a contributing factor in the decision of the US to leave at least the ABM Treaty. The US has deployed missile defense systems in Alaska, California, Guam, and South Korea to counter threats from North Korea and 249 China. 89 With the rise in competition between the US, Russia, and China to develop hypersonic missiles and corresponding missile defense systems, it seems the lack of institutions like the ABM Treaty or INF Treaty has led to a new arms race. However, this is not likely the end of cooperation between the US and Russia on the issue of security. Despite the fact that the relationship between the states is in shambles in the wake of the 2022 Russia-Ukraine War, there may be room for future cooperation. At some point, leadership will change, wars end, and priorities change. It is dicult to make predictions, but we can see how in the past enemies became allies and adversaries agreed to work together for the common good. The international community is facing severe challenges and is likely negotiating a transition to a new framework for international relations. As military budgets soar 90 and other problems divert attention, states may nd it in their best interest to minimize those costs or divert resources to other challenges. After a period of decoupling, the US and Russia will have an opportunity to build new relationships based on lessons from the past, including those learned from the ABM and INF treaties. 89 The US is deploying and testing multiple systems, including its Terminal High Al- titude Area Defense (THAAD) and \Iron Dome" systems, to create a multi-layered missile defense. Alastair Gale. \U.S. Tests Israel's Iron Dome in Guam as Defense Against Chinese Cruise Missiles." The Wall Street Journal (Nov. 9, 2021), available at https://www.wsj.com/articles/u-s-tests-israels-iron-dome-in-guam-as-defense-against- chinese-cruise-missiles-11636455224 90 \World military expenditure passes $2 trillion for rst time." Stockholm International Peace Research Institute (SIPRI) (April 25, 2022), available at https://www.sipri.org/media/press-release/2022/world-military-expenditure-passes- 2-trillion-rst-time 250 Chapter 6 Brexit: Bargaining Failure 6.1 Introduction On a brisk evening in 2019 next to the black fences surrounding the Palace of Westminster, a group of people stood on the street with signs decrying Brexit and stating that leaving the European Union will "hurt working class Britons." Standing before the blue EU ag, a handful of protestors describe their concern that leaving the EU with allow the U.K. government to pull back worker rights that were secured by EU law. 1 In the process of express- ing concern for his nephew's future, one of the \Remainers" was shouted at by a passerby. \Traitor!" yelled the man, as he continued with expletives. 1 Author interview (Interview C), October 2019. 251 A London police ocer approached and reduced the tension, but the inter- action highlighted the lingering tensions in the country three years after the historic Brexit referendum. The 2016 referendum vote in the U.K. to leave the E.U. shocked many across the globe. The E.U., the decades-long European integration project, appeared to be unraveling. The U.K. became a member of the European Community, a forerunner of the E.U., by entering the Treaty of Accession of 1972. For over four decades the U.K. participated in the European project. The 51 to 48 percent vote in favor of leaving the E.U. led to the resignation of then Prime Minister David Cameron and the new Prime Minister Theresa May triggering the Article 50 two-year clock to negotiate its exit from the Union. Brexit was one of the most prominent and important treaty withdrawals of the modern era. It is also unique in that it involved a direct vote of the coun- try's population. 2 For the most part, decisions to withdraw are made by the executive, or sometimes legislative, branches of a state. Here, the people of the United Kingdom were given a direct voice in making that decision|even if Parliament and Theresa May ultimately exercised the authority to with- draw. In the years since, explanations and accounts of the causes and ram- ications of the decision abound. Numerous books and articles have been published cataloging the events leading up to the referendum and oering explanations for the ultimate outcome. Most of the explanations of Brexit 2 In 1982, Greenland voted to withdraw from the European Community after being granted home rule by Denmark. It left the E.C. in 1985. 252 see deep roots of the exit in the history of the U.K.'s uneasy relationship with Europe (See, e.g, Menon and Salter 2016). Many of those books focus on the domestic politics, whether the internal political battles between lead- ers in the U.K. (Shipman 2016) 3 or changes in attitudes of the U.K. public because of domestic politics, especially rise of UKIP, and weighing of eco- nomic and immigration factors (Clarke, Goodwin and Whiteley 2017b,a). Others saw the long-term transformation of the country demographically as creating con icts that could be exploited in British politics (Sobolewska and Ford 2020). Tooze (2018) traces the causes of Brexit, at least in the near term, to the fallout from the Great Recession. Still others, criticized the political and social rationales for Brexit, including English nationalism (O'Toole 2018). Those who participated in the events also contributed their perspectives, including former Prime Minister David Cameron, whose de- cisions largely put the country on the path to the referendum (Cameron 2019). As mentioned, numerous books and articles have been written on Brexit over the last few years. Much of the work shifted to focusing on the long drawn out divorce between the U.K. and the E.U., as well as looking to the future of the relationship between the British Isles and the continent. This chapter adds to that substantial literature by focusing on the bargaining process and the features of the E.U. that in uenced the ultimate decision to leave the union. As posited in Chapter 2, the U.K. and the E.U. members could have come to a bargain under the right parameters to avoid exit. However, Brexit 3 Those in government found Shipman (2016) to be an accurate description of the domestic politicking of the time. Author interview (Interview B), October 2019. 253 raises interesting and important questions as to why that failed. How did a national referendum on the decision to withdraw disrupt the usual calculus? In a counterfactual world, the U.K. leadership|namely Cameron and his allies|would have happily agreed to a bargain with Europe that granted concessions while preserving the U.K.'s membership. Yet, Cameron was not only negotiating with the E.U. and its 27 other members, but also very publicly negotiating with the U.K. citizens as to why the concessions he had obtained warranted remaining a member of the E.U. Cameron tied his hands by committing to the referendum and, as explained below, underestimated the challenges of controlling the referendum process. This all limited the bargaining space to such an extent that it was highly unlikely there was a bargain that would placate those calling for exit on \sovereignty" grounds while allowing the E.U. to preserve its fundamental tenets. As described in Chapter 4 in the context of the ICC's approach to Kenya's threatened withdrawal, the U.K. was an important member within the E.U., in terms of its size and economy, particularly the nancial sector based in London. Yet, that importance was not unlimited. The E.U. was willing to make concessions to keep the more than 60 million U.K. citizens as European citizens. The U.K. was not a founding member of the E.U. and joined later, if not wholeheartedly. Thus, the E.U. had functioned without the U.K. and could imagine functioning without it again. Further, as explained below, the U.K. did not appear to seriously consider an outright breach of its obligations to the E.U., especially with regard to the key issue of immigration. Thus, this case poses challenges and insights to the theory put forward in this 254 dissertation. This chapter is divided into two parts. In the rst part, it summarizes the U.K.'s relationship with the E.U., its decision to belatedly join, its un- easy relationship, prior referendum on membership, and ultimate challenges leading up to the June 2016 referendum. From there, the chapter focuses on the bargaining process between the U.K. and E.U. leaders, the campaign to sell the proposed bargain, the vote, and the immediate aftermath. In the second portion, I use the bargaining over Brexit to test the expectations of the theory from Chapter 2, in particular, whether the obligations of the union became too heavy for the U.K. to bear with no exibility to escape them. The chapter concludes with insights for extending the theory and implications for further research. 6.2 Long Road to Brexit: A History 6.2.1 UK-EU Relationship Before the 1940s, the U.K.'s focus was to protect the English Channel and preserve its independence from Europe (Anderson 2021). In a speech to the House of Commons on March 1, 1848, Lord Palmerston explained: \We have no eternal allies, and we have no perpetual enemies. Our interests are eternal and perpetual, and those interests it is our duty to follow." 4 A sentiment echoed by world leaders for generations, including U.S. President's 4 Lord Palmerston, speech, House of Commons, 1 March 1848 in Oxford Essential Quotations (4 ed. Susan Radclie 2016). 255 George Washington and Thomas Jeerson (See discussion in Chapter 5), this encapsulated the U.K.'s reluctance to be tied to any permanent relationship with the other European powers. In defeating the fascist governments of Europe, the U.K. saw itself as victo- rious along with its war-time partner, the United States. In the eyes of the U.K., they were superior to the nations of Europe that had been destroyed by war (Anderson 2021). This superiority also could be found in the lingering legacy of early industrialization and vastness of its empire (O'Rourke 2019). Thus, the post-war eorts at building a project of integration intent on pre- venting another war was not immediately appealing to the U.K. The U.K. was not a part of the eort in creating the Treaty of Rome, establishing the European Economic Community (EEC), the predecessor to the European Union. However, the U.K. was willing to accede to joint defense agreements with European states like the Treaty of Dunkirk (1947), Treaty of Brussels (1948), and North Atlantic Treaty (1949). The U.K. was also an original signatory to the 1949 Statute of the Council of Europe, signed in London, after the public urging of Winston Churchill in the preceding years. \Europe [as created through numerous treaties and international institutions] was a soft landing after empire" (Snyder 2018, 72). 5 In the early 1960s, the U.K.'s view of Europe began to change. During the Suez Crisis, the U.S. reaction prompted a rethinking in the U.K., whereby 5 (Snyder 2018, 68) explains: \The major European powers had never been nation- states: before the Second World War they had been empires, where citizens and subjects were unequal; afterwards, as they lost their empires, they had joined a process of European integration in which sovereignty was shared." 256 it needed more like-minded partners in the world (Anderson 2021). Further, the U.K.'s economy was struggling with rationing and seemed to be falter- ing when the economies of Europe, especially Germany, were ascendent. In the 1960s and 1970s, the U.K. was sometimes characterized as the \sick man of Europe" given its economic woes and poor performance. 6 The Con- servative government under Harold McMillian was encouraged by the U.S. Kennedy Administration to join the EEC. The U.K. would be a bridgehead for the English-speaking countries into Europe. French President Charles DeGaulle viewed the U.K. with suspicion, simply a Trojan horse for the United States. In January 1963, France vetoed the U.K.'s membership with le grand \non". 7 Under a new Labour government led by Prime Minister Harold Wilson in the U.K., Parliament agreed to join the European Com- munities (on a motion carried in the House of Commons by 487 to 26 votes), but again DeGaulle vetoed membership. France had lifted the draw bridge to economic union. In the 1970s, DeGaulle was replaced by Georges Pompidou and Conserva- tive, pro-Europe Prime Minister Heath took over for Wilson in the U.K. Heath negotiated terms for the U.K. to join within eighteen months and then secured bipartisan consensus in Parliament to join. Parliament passed the European Communities Act of 1972 (301 to 284 votes), granting legal 6 Buttonwood. \Britain: back to being the sick man of Europe?" The Economist (July 19, 2017), available at https://www.economist.com/buttonwoods- notebook/2017/07/19/britain-back-to-being-the-sick-man-of-europe 7 \Modern History Sourcebook: President Charles de Gaulle: Le Grand `Non": Britain's Proposed Entry Into The Common Market, May 16, 1967" Internet History Sourcebooks Project, Fordham University, available at https://sourcebooks.fordham.edu/mod/1967-degaulle-non-uk.asp 257 authority for the U.K. to accede to the European Communities. Yet, while in opposition, Heath had promised that joining the EEC would only happen with the \full-hearted consent" of Parliament and the public. No referendum on entry occurred and that promise remained unfullled. When Labour retook control of government, Heath renegotiated terms with Pompidou and then planned a referendum on the revised terms. Opposition arose from both the left and the right, including from Peter Shore arguing against the loss of sovereignty and Enoch Powel|infamous for his \Rivers of Blood" speech|but the opposition was unsuccessful (Norton 1978). In June 1975, the U.K. then had a referendum on joining the E.C. with 64% turnout and 67% of the public approving of joining Europe. Margaret Thatcher and other \mainstream" politicians supported joining. When she took over as Prime Minister, Thatcher worked to strengthen the ties with Europe and opportunities for London's nancial sector. This changed in the mid- to late-1980s when Thatcher rejected Europe's desire to work toward a monetary union. John Major, Thatcher's successor, ne- gotiated an opt-out from the single currency and agreed to the Treaty of Maastricht in 1992. This was seen as a victory for the U.K. until the Danes rejected the Maastricht Treaty by referendum. This prompted calls for a referendum in the U.K. After substantial division, the treaty was eventually ratied in May 1993. The Conservatives lost power by wide margins in 1997, leaving Tony Blair a 258 commanding majority with which to govern. Gordon Brown, Blair's Chan- cellor of the Exchequer, opposed the single currency and, aside from signing on to the Social Chapter, the U.K. government proceeded under Maastricht with a Eurozone opt-out. Britian's economic strength into the 2000s incen- tivized Brown to oppose joining the Economic and Monetary Union (EMU), contrary to Blair's position. Brown's position prevailed, and he eventually became Prime Minister in 2007. Under Brown, the U.K. signed the Treaty of Lisbon in 2007, only to be faced with the nancial crisis in 2008. The Lisbon Treaty amended the Treaty on European Union to specify the procedure for a member to withdraw from the E.U. New Labour was trounced in 2010 and David Cameron and the Tories led the country through austerity measures in the wake of the nanical crisis. Before the election, Cameron had promised a referendum on the Lisbon Treaty, but retreated after all other EU members ratied it. In 2011, Euroskeptic Conservatives forced the European Union Act through Parliament obligating the government to hold a referendum on any further treaty revisions. 6.2.2 2008 Recession and Eurozone Crisis In the 21st century, the City of London was the driving force of the United Kingdom's economy. Yet, the success of the nancial sector was outwardly focused while much of the rest of the economy suered during the Great Recession. The 2008 nancial crisis hit U.K. banks hard, prompting more regulation and oversight. The banks then refocused their eorts to becoming 259 the gateway to China and the renminbi|to the consternation of the United States (Tooze 2018). The U.K.'s latest desire to renegotiate the terms of its membership in the E.U. was driven, in part, by the Eurozone crisis and the widely held belief that greater scal and nancial integration was needed as cure Tooze (2018, 545). As Tooze (2018, 543) puts it: \The 2008 crisis marked a break in Britain's relations with the EU." The seeds of the eort to renegotiate can be found in the years shortly after the nancial crisis. In 2010, the Daily Express began campaigning for Brexit in November 2010. (Tooze 2018, 544). In 2011, at the height of the crisis, 80 Tory back benchers called for a referendum on the Lisbon Treaty (Tooze 2018, 544). In 2013, Prime Minister Cameron gave a speech 8 about the \European Union and how it must change," laying out his vision for robust negotiations with E.U. leaders on addressing numerous concerns of the U.K. public and trou- blesome backbenchers. Echoing the sentiment of Lord Palmerston from the 1800s, Cameron explained: And it's true that our geography has shaped our psychology. [{] We have the character of an island nation: independent, forthright, passionate in defence of our sovereignty. [{] We can no more change this British sensibility than we can drain the English Channel. [{] And because of this sensibility, we come to the European Union with a frame of mind that is more practical than emotional. [{] For us, the European Union is a means to an end|prosperity, stability, the anchor of freedom and democracy 8 \David Cameron's EU speech - full text." The Guardian (Jan. 23, 2013), https://www.theguardian.com/politics/2013/jan/23/david-cameron-eu-speech- referendum 260 both within Europe and beyond her shores|not an end in itself. [{] We insistently ask: how, why, to what end? Acknowledging the various crises facing the E.U., Cameron explained that the E.U. needed to address three major challenges: the weaknesses of the Eurozone, the lack of competitiveness of the European economy, and the democracy gap within the institution. He warned that failing to address those core issues posed the risk \that Europe will fail and the British peo- ple will drift towards the exit." He set out his goal of bargaining for more exibility within the E.U., stating: A real choice between leaving or being part of a new settlement in which Britain shapes and respects the rules of the single market but is protected by fair safeguards, and free of the spurious regu- lation which damages Europe's competitiveness. . . . [{] In other words, a settlement which would be entirely in keeping with the mission for an updated European Union I have described today. More exible, more adaptable, more open|t for the challenges of the modern age. [{] And to those who say a new settlement can't be negotiated, I would say listen to the views of other par- ties in other European countries arguing for powers to ow back to European states. In addition to the nancial crises facing the EU, the problems raised by Ukraine and the in ux of refugees from Syria also strained the relation- ship of the bloc. The obligations of each member were being increased and their commitment to those obligations tested. The warning signs could be seen in the European election results in 2014. This was the rst European Parliament election since the then on-going Eurozone crisis. In response to the crisis, European leaders implemented harsh austerity measures, which 261 caused a backlash throughout Europe. Approval of EU leadership dropped signicantly in Southern European countries including Portugal, Spain, and Greece. 9 Spain saw the largest drop going from an approval rate of 59% in 2008 to 27% in 2013. Ireland also experienced a large drop in approval, moving from 70% in 2008 to 47% in 2013. Approval of EU leadership topped fty percent in only four of the 27 member states, namely Denmark, Bel- gium, Germany, and Luxembourg. Approval in the U.K. dropped from a 36% approval in 2008 to 29% in 2013. The elections resulted in signicant gains for EU-skeptic and anti-establishment parties in the European Parliament. Voter turnout was the lowest for any European Parliament election since 1979, at 42.6%. 10 The elections saw the parliament as a whole move more to the right and Euroskeptism become an increasing frame for members of the European Parliament. This further signaled the discontent of the population and constituent state members to how the institution was functioning. In the U.K., the U.K. Independence Party (UKIP) outperformed both Labour and the Conservative parties, with Prime Minister Cameron's conservative party coming in third overall. 11 European leaders viewed the election re- sults as a warning for the E.U. institution and calls for reform were made 9 Anna Manchin. \EU Leadership Approval at Record Low in Spain, Greece" Gallup (Jan. 8, 2014) https://news.gallup.com/poll/166757/leadership-approval-record- low-spain-greece.aspx 10 Drew Desilver. \Euroskeptics are a bigger presence in the European Parliament than in past." Pew Research Center (May 22, 2019) https://www.pewresearch.org/fact- tank/2019/05/22/euroskeptics-are-a-bigger-presence-in-the-european-parliament-than- in-past/; \2019 European election results" European Parliament. (Oct. 22, 2019) https://www.europarl.europa.eu/election-results-2019/en/turnout/ 11 Andrew Grice and Nigel Morris. \European elections results 2014: Farage and Ukip top poll as Europe swings to the right." The Independent. (May 26, 262 by various leaders, including Cameron. 12 Cameron was forced to take the results into account in opposing the nomination of Jean-Claude Juncker as European Commission President given the U.K.'s view that he was an \old- style European federalist." 13 Cameron noted: \The European Union cannot just shrug o these results and carry on as before. We need change. We need an approach that recognises that Europe should concentrate on what matters, on growth and jobs, and not try to do so much." 14 6.2.3 Promised Negotiations on Reform The U.K. had to negotiate with the E.U., which is composed of twenty seven other members and the E.U. organization itself. Any reforms would have to be agreed to unanimously. Cameron noted frustration in not only nego- tiating with leaders from the member countries and the E.U., but also the \ocials deeply imbued with Brussels orthodoxies" (Cameron 2019, 640). As Cameron (2019, 622) described it: \It came in two parts: a renegotiation and a referendum. Both were necessary, and both were long overdue. I didn't see them as separate endeavours; they were two sides of the same coin. I could never have undertaken a proper renegotiation without a referendum. I needed that endpoint to focus the minds of the leaders from whom I was 2014) https://www.independent.co.uk/news/uk/politics/european-elections-results- 2014-sweeping-gains-for-nigel-farage-s-ukip-deal-further-body-blow-to-main-parties- 9434042.html 12 Paul Taylor and Luke Baker. \After seismic elections, EU leaders assess damage." Reuters (May 27, 2014) https://news.gallup.com/poll/166757/leadership-approval-record- low-spain-greece.aspx 13 Id. 14 Id. 263 seeking change." Thus, as to Cameron's understanding, the status quo was not acceptable to the U.K. To remain entailed reform. \The polling was clear: there was only a majority for remaining in Europe if longstanding problems were addressed. Indeed, I was acutely aware of the danger of not getting enough from an intransigent EU to satisfy an impatient public. `I do worry that what is negotiable is not sellable and what is sellable is not negotiable,' I confessed at the time on tape" (Cameron 2019, 622{623). Cameron sought to commence negotiations with the European Council in June 2015, with the goal of securing an agreement by December 2015 or February 2016. On that timetable, a referendum could occur in June 2016 (Cameron 2019, 627). He chose the timeline in part based on the scheduled elections in Germany and France, discussed in more detail below. Although hoping for broader treaty change, the U.K.'s demands were more limited than what some desired. However, broader European reform was likely unrealistic. Cameron viewed the British people's goal as: \What they wanted to know was simply that we could live in our own version of Europe with an annual bill we could stomach, immigration levels we could handle, and rules and regulations we could see the sense in" (Cameron 2019, 627). Particularly, the focus of the negotiations was to address popular enmity of some aspects of membership like deeper political union, regulations, and lack of control of immigration, and national interests like addressing how decisions in the Eurozone aected the U.K. Cameron sought reform in four categories, or \baskets" as he called them: sovereignty, competitiveness, 264 fairness, and immigration (Shipman 2016, 120{122, 129{130, 146). Again, to remain was to reform. Immigration and the movement of people within Europe was one of the key issues for the negotiation between the U.K. and the E.U. Cameron char- acterized some of the European court decisions on the issue of freedom of movement as resulting from \mission-creep," expanding the right without limit (Cameron 2019, 632). Cameron's government was concerned about the use of welfare by migrants or \benet tourism" as he put it. Yet, leaders of other E.U. members did not view the U.K.'s concerns as an immigration problem because they saw internal movement as movement by E.U. citizens, not migration from outside the union (Cameron 2019, 640). The core elements of the E.U. were redlines in the negotiation process. \The EU couldn't allow the UK to dictate the future course of integration. Free- dom of movement and the equal treatment of EU citizens were nonnego- tiable" (Tooze 2018, 547). \[T]here was no deal on earth that would have enabled us to stay in the EU and not sign up to the `four freedoms'|the free movement of goods, services, capital and people. The EU is a quid pro quo: we get full access to the single market only if we accept its conditions" (Cameron 2019, 634) Thus, a state could not remain in the E.U. and reject free movement. If free movement was problematic for the U.K., nding a compromise would be dicult. The U.K. government was also concerned that without broad treaty changes, 265 any concessions by the E.U. could not be guaranteed. They sought a \Dan- ish Model" whereby reforms would be implemented and backed by legally binding instruments lodged with the U.N. and incorporated into the treaties the next time they opened up. Denmark had rejected the Maastricht Treaty by referendum in 1992 and reached a similar arrangement. The issue of se- curing a guarantee would dog Cameron in the domestic campaign to sell his proposed deal (Cameron 2019, 660). Nonetheless, Cameron saw the U.K.'s concerns as arising out of broader problems in the E.U. \With the impact of the nancial crash, EU enlargement, mass migration and European courts' judgements, it was failing too. Fraying support in the UK was just the beginning of the unravelling" (Cameron 2019, 635). Cameron's demand for reform was couched as reform or exit to E.U. mem- bers (Cameron 2019, 638). However, skepticism about whether the U.K. would follow through undermined the eectiveness of the negotiation strat- egy. \[T]he thing I was picking up from most leaders was that they simply did not think Britain would leave. For them, the referendum was a ruse to get more out of the renegotiation" (Cameron 2019, 640{641). Essentially, Cameron's threat was not credible. This may have been partly because he and his government also saw exit as unlikely. 15 Even though fundamental remaking of the U.K.'s position in the E.U. was not possible, Cameron settled for a proposal so that he could proceed with the promised referendum (Tooze 2018, 547). The U.K. would receive an \emergency brake" allowing the U.K. to limit benet payments to migrants 15 Author Interview (Interview B), October 2019. 266 for a \one o" seven-year period and the E.U. agreed to recognize that fur- ther political integration of the E.U. did not apply to the U.K. This proposal was viewed as weak in the U.K. and not addressing the main criticisms in the U.K. In uential members of Cameron's own party, Mayor of London Boris Johnson and education minister Michael Gove, split with him to join the Brexit campaign. 16 The resignations should have been an early indication of trouble for Cameron's eorts. 17 Outside the U.K., Europeans generally saw the U.K. leaving the E.U. in negative terms (a median of 70% of the public in nine EU states), although 32% of the public in France believed Brexit would be a positive for the E.U. 18 Although, in 2016, views of the E.U. in France and Greece were more unfavorable than in the U.K. Roughly two thirds of the public in the U.K. (65%) and Greece (68%) believed some powers should be returned to national governments from the E.U. In that climate, the stage was set for a potentially monumental referendum. 6.2.4 Campaign Deciding on the mechanics of the referendum was crucial. Those decisions also likely aected the ultimate outcome. Of the main decisions to be made, the wording of the referendum question, the threshold vote needed to pass, and timing of the vote were paramount. 16 Although, Johnson's position may have been more driven by political ambition among Euroskeptic Tories than a true belief that Brexit would succeed (Applebaum 2020, 70). 17 Author Interview (Interview B), October 2019. 18 Bruce Stokes. \Euroskepticism Beyond Brexit." Pew Research Center (June 7, 2016) https://www.pewresearch.org/global/2016/06/07/euroskepticism-beyond-brexit/ 267 As part of that decision making process, phrasing the question to the public was the subject of substantial internal debate. Cameron had wanted the question to be \Should the United Kingdom remain a member of the Eu- ropean Union|Yes/No?" with \Yes" being \stay in the EU on the basis of the new deal," mirroring the formulation of the 1975 referendum (Cameron 2019, 624). However, the Electoral Commission recommended a dierent formulation, framing the choice as whether to \remain" or \leave" the E.U.: \the question be \Should the United Kingdom remain a member of the European Union or leave the European Union?" (Cameron 2019, 624{625). The issue of whether a particular threshold of the vote would need to be obtained by one side or the other was raised but not fully considered. For his part, Cameron believed that anything other than a simple majority of the entire voting population would have been problematic (Cameron 2019, 625). Requiring more than fty percent of the vote would appear like the establishment was putting a nger on the scales|something it was already accused of doing by some and something Cameron, in hindsight, claimed to work to minimize. Another option was to require that a majority of each of the four nations of the United Kingdom would need to vote for leave. Cameron saw that option as encouraging separatism. The U.K. had survived a Scottish referendum on independence only a couple years prior. As for timing, Cameron sought to leverage an o-cycle election in France and Germany to maximize his bargaining position (Cameron 2019, 626). Pursuant to U.K. law, the referendum must be held by December 2017; however, Germany and France were set to hold elections in 2017. Thus, a 268 referendum in 2016 would allow the U.K. to negotiate with those countries in a year where the two biggest countries were not engaged in electoral domestic politics. Cameron posits that E.U. leaders were reluctant to address changes that would require input from their own voters because of this dynamic (Cameron 2019, 627). Moreover, Cameron felt the U.K. was strong and wanted to capitalize on that strength, which may wane in the future (Cameron 2019, 626). Yet, as Tooze (2018, 545) notes: \London operated from the hubristic assumption that Britain could change the EU's course." Further, the U.K. should have foreseen its limited in uence in the organization when it campaigned against the 2014 nomination of Jean-Claude Juncker as Commission President at the European Council. Only David Cameron and Viktor Orban voted against Juncker, signaling the U.K.'s outlier position among the members. The U.K.'s relationships with other states, outside the E.U., was also a signicant factor in the campaign. During a visit to the U.K., President Obama remarked that should the U.K. exit from the E.U., it might nd itself in a worse position trying to secure a trade deal with the U.S. because the U.S. would be incentivized to secure a deal with a big trading bloc, like the E.U. On the other end, outside in uence on the campaign also came from actors hoping to weaken Europe. In particular, the in uence of bots and media with Russian backing was focused on separating the U.K. from Europe (Snyder 2018, 104{106). \Brexit was a major triumph for Russian foreign policy, and a sign that a cyber campaign directed from Moscow could change reality" (Snyder 2018, 106). 269 However, the campaign was fraught from the start. Some of Cameron's own ministers desired to campaign against the government's position (Ship- man 2016). In particular, Ministers Chris Grayling and Theresa Villers threatened to resign if they could not campaign for leave. With the threat of defections from his government, Cameron caved and agreed in January 2016 to allow ministers to campaign for Leave if they wished. However, this decision did not stop high-prole defections from Cameron's own party, including Michael Gove, Ian Duncan Smith, and Boris Johnson. Cameron's campaign to sell his proposed reforms to the E.U.-U.K. relation- ship was problematic in that it focused on the pain of a potential Brexit rather than the benets of remaining in the E.U. Dubbed \Project Fear," the government's campaign highlighted the nancial costs to U.K. families should Brexit pass, as rst put forward by the Treasury Department in April 2016 (Shipman 2016). 19 George Osborne, the Chancellor of the Exchequer, warned that Brexit would cause house prices in the U.K. to substantially drop. On June 9, 2016, former prime ministers Tony Blair and John Major traveled to Northern Ireland and warned that Brexit would break up the U.K. Cameron even went so far as to speculate that Brexit could lead to war in Europe (Shipman 2016). The bombastic rhetoric, and at times out right lies, was matched on the Leave side. Boris Johnson argued that the E.U. was pursuing the same superstate as Adolf Hitler and Nazi Germany using \dierent methods" (Shipman 2016). Much of the strategizing was or- chestrated by Dominic Cummings, who was well aware of the government's 19 In the U.K., the Treasury Department assesses and must approve any policies that aect taxes or public expenditures. 270 weak points having worked in government for most of his career (Shipman 2016, 93). Cummings is credited with devising the \Take back control" slo- gan for the Leave Campaign, along with the infamous claim that the U.K.'s National Health Service (NHS) would receive $350 million after Brexit|a claim plastered on the side of a bus that toured the country (Shipman 2016). On June 16, 2016, approximately a week before the vote, Jo Cox, a pro-E.U. member of parliament was shot and killed in West Yorkshire by a man who later told a court his name was \death to traitors, freedom for Britain." 20 The tension in the country had reached a boiling point. Cameron's eorts at combatting the false claims were also unsuccessful. In retrospect, Cameron conceded that he believed a weakness of the Remain campaign was that it focused on \technical" arguments whereas the Leave Campaign focused on \emotional arguments" (Cameron 2019, 667). George Osbourne echoed this sentiment in the years that followed. 21 The stakes for the U.K. were tremendously high. At the time, half of the U.K.'s trade, i.e. $200 billion, was with the E.U. (Tooze 2018, 548). Half of the U.K.'s $1.2 trillion in FDI was from the E.U. It was estimated that trade between the U.K. and the E.U. was 55% greater being a member of the trading bloc than otherwise. Also, the dispersion of citizens in the U.K. 20 Robert Booth, et al. \Jo Cox murder suspect tells court his name is `death to traitors, freedom for Britain'." The Guardian (June 18, 2016) avail- able at https://www.theguardian.com/uk-news/2016/jun/18/thomas-mair-charged-with- of-mp-jo-cox 21 \George Osborne `regrets' mistakes that led to Brexit vote." BBC News (October 31, 2018), available at https://www.bbc.com/news/uk-politics-46039623 271 and across Europe was signicant, with 3.2 million EU citizens living in the U.K. and 1.2 million U.K. citizens living in the other members of the E.U. 6.2.5 The Vote On June 23, 2016, the voters took to the polls to vote on the contentious and highly anticipated referendum. The ocial results were not announced until the following morning, shocking many. That day, June 24, Cameron conceded defeat and announced his resignation in the coming months. He was succeeded by Theresa May on July 13, 2016 as Prime Minister, charged with carrying out the will of the electorate and withdrawing the U.K. from the E.U. A Brexit vote was not truly a vote against globalism, as many commentators described, but a vote \to restore Britain's greatness and freedom. . . . Brexit was a vote for autonomy" (Tooze 2018, 554). This was a vote for the U.K. to play a larger role in the world outside of and separate from the E.U. \But there was a nostalgia for something else: a world in which England made the rules" (Applebaum 2020, 63). Supporters of Brexit generally held lower levels of education and income, and they were particularly hard hit by austerity measures since 2010. Those measures aected areas that had been declining and made the decline worse. Brexit voters also were generally located in areas where UKIP did well in 272 2014. A preference for Brexit aligned with preferences for authoritarianism, security, death penalty, and public ogging. 22 Figure 6.1: Uncertain Future for the U.K. Palace of Westminster and Big Ben behind scaolding, awaiting Brexit to take eect in 2019. Photo credit: author. 6.2.6 Aftermath Brexit was not as a moment but a process (Tooze 2018). To eect the withdrawal, the U.K. was required to invoke Article 50 of the Treaty on the European Union. The U.K. government was largely unprepared for a successful Brexit vote and scrambled to devise a plan to negotiate the exit. 23 22 E. Kaufmann, \It's NOT the Economy, Stupid: Brexit as a Story of Personal Values," LSE European Politics and Policy (blog) (July 9, 2016). 23 Author interview (Interview B), October 2019. 273 Prime Minister May delayed invoking Article 50 until March 2017, after seeking approval from Parliament as ruled by the U.K. Supreme Court. 24 After the vote, under the leadership of Germany, the E.U. agreed on a tough bargaining position (Tooze 2018, 559). Trade talks would not begin until the terms of Brexit were set. The UK must meet its nancial obligations to the E.U. There could be no access to the common market without freedom of movement. The Court of Justice of the European Union (CJEU) would retain jurisdiction over EU citizens. The UK government's response was to make threats of reconguring itself as the \low-tax Singapore of the West" 25 and, as Tooze (2018, 560) puts it \the muscular state would turn the City of London and Britain's oshore posi- tion into a battering ram." Notwithstanding the thorny negotiations with Europe, the U.K. expanded its treaty-making powers and embarked on ne- gotiating a number of new agreements to address its status outside the E.U. Larik (2020). In one way, Brexit ended one, collective treaty relationship to be replaced with a number of new treaties to account for the change in circumstances. In recent years, views of the E.U. have improved, with a median of 67% of the public in 14 E.U. member states holding favorable views of the institu- tion. 26 Favorability of the E.U. in Greece has jumped 26 percentage points 24 R (Miller) v. Secretary of State for Exiting the European Union, [2017] UKSC 5 (24 January 2017). 25 G. Parker, J. Ford and A. Barker. \Is Theresa May's Brexit Plan B an Elaborate Blu?" Financial Times (Jan. 19, 2017). 26 Richard Wike, et al. \European Public Opinion Three Decades After the Fall of Communism: 4. The European Union." Pew Research Center (Oct. 14, 2019) https://www.pewresearch.org/global/2019/10/14/the-european-union/ 274 from 2016 to 2019. Twelve of the 14 EU states surveyed have majorities of their populations that believe EU membership has beneted their country. In the Czech Republic, only 40% of the public believes that EU member- ship has benetted the country and 20% believe membership has hurt the country. Nonetheless, the overall trend in Europe is greater satisfaction with membership than during the Eurozone crisis. Views of the E.U. in the U.K.'s neighbor, Ireland, are mostly positive|as has historically been the case. 27 These views of the E.U. were likely in uenced by the painful post-Brexit ne- gotiations. However, with more distance between the Eurozone and refugee crises, a later Brexit vote may have had a dierent outcome. 6.3 Application of Theory In Chapter 2, the bargaining model of treaty exit produced various expecta- tions about when treaty members would fail to devise a bargain to address an external shock and increase in treaty obligations. The treaty's baseline level of obligations and exibility in allowing a state to breach with minimal costs aects the prospects of a successful bargain to avoid exit (Claims 2 and 3). In the case of Brexit, the 2008 nancial crisis, European sovereign debt crisis, con ict in Ukraine, and Syrian refugee crisis were all shocks that caused the E.U. to implement policies that increased the perceived burden of remaining within the organization (Tooze 2018). 27 Standard Eurobarometer 94, Public Opinion in the European Union, National Re- port: Ireland, Winter 2020-2021, p. 5. 275 Joining the E.U. comes with signicant obligations and has limited exibil- ity to negotiate new bargains on treaty obligations or allow breach. These obligations are so extensive that E.U. law has developed into its own prac- tice area and subject of instruction in law schools around the globe. Most large law rms with a presence in Europe have oces in Brussels, the E.U. headquarters. The European Court of Justice has made clear that this new E.U. law \constitutes a new legal order of international law for the benet of which the states have limited their sovereign rights . . . ." 28 E.U. law is supreme within the union, as conrmed by the court in a number of cases, superseding national law in most respects. 29 Further, not only can the or- ganization enforce E.U. law, primarily through the European Commission, but each member may enforce E.U. obligations as to each other through legal action (see Treaty on the Functioning of the European Union (TFEU), art. 259). At the core of E.U. law are two treaties: the Treaty on Euro- pean Union and the TFEU (referred to collectively as the consolidated E.U. treaties). Both the Lisbon Treaty and the consolidated E.U. treaties are long, de- tailed, and contain high levels of obligatory language. As the governing instruments for the E.U., the consolidated E.U. treaties have 72,503 words 28 Case C-26/62, Van Gend en Loos v Nederlandse Administratie der Belastingen, 1963 E.C.R. 00001. 29 See, e.g., Case 6-64, Flaminio Costa v E.N.E.L., 1964 E.C.R. 00585 (conrming member states transferred sovereign rights to the Community and Community law could not be overridden by national law); Case 106/77, Amministrazione delle Finanze dello Stato v Simmenthal SpA., 1978 E.C.R. 00629 (ruling domestic courts are obligated to strike down any domestic legislation contrary to Community law); Case C-106/89, Marleasing SA v La Comercial Internacional de Alimentaci on SA, 1990 E.C.R. I-04135 (ruling national law must be interpreted to avoid con ict with Community law). 276 and a permissive-obligatory score of 952 (see Appendix C). This exceeds anything in the sample of treaties considered in Chapter 2. This is sensible where the consolidated E.U. treaties are designed to bind Europe together in deep ways that are unique among states. As such, we should consider the consolidated E.U. treaties as highly obligatory and prone to treaty exit (Hypo 1 and 4). No state registered a declaration or reservation to the Lis- bon Treaty, which precludes an assessment of the treaties' obligations and exibility using those treaty actions (Hypos 2 and 3). 30 The E.U.'s membership has grown over its lifespan, reaching 28 member states before Brexit. The rules governing how each state is represented in the decision-making of the E.U. make negotiations complicated and costly, on the whole (see Lechner and Ohr 2011)|as Cameron noted when negotiating on his proposed reforms (Cameron 2019, 640). Thus, the institution would be less exible and more prone to exit (Hypo 4). Further, as discussed in this chapter, the integration of laws and regulations within the E.U. makes breaches of E.U. law more costly than other international agreements. As will be discussed in detail below, the E.U. monitors compliance and uses investigations, lawsuits, and sanctions to compel compliance. One of the arguments made in favor of Brexit was that the U.K. was con- tributing too much nancially without receiving a commensurate benet. This high obligation would make any external shock that increased those costs that much worse. Thus, the shocks that the E.U. experienced may 30 Interestingly, the E.U., as an organization, has used reservations in its own treaty practice (Durand 2018). 277 have pushed the perceived costs to levels that precluded any possible bar- gain. To assess the U.K. relative contribution to the E.U., we can determine the U.K.'s net nancial benet from the E.U. overtime. Figure 6.2 presents the net contribution of all E.U. members from 2000 to 2020 using data made available by the European Commission. Figure 6.2: Net EU Contributions In the gure, the U.K.'s net contribution is highlighted to show its rela- tive contribution. For the 20-year period covered in the data, the U.K. is a net contributor to the E.U., often one of the biggest net contributors to the union. Germany is the largest net contributor to the E.U., with France and Italy also contributing large shares. Examining the trend of the U.K.'s contribution, it largely increased peaking in 2015, just before the 2016 ref- erendum. In the years after the referendum, the U.K. remains one of the 278 top contributors to the E.U. As mentioned previously, the U.K. beneted from E.U. membership in nu- merous ways. In quantiable nancial terms, at the time of the referendum, half of the U.K.'s trade, i.e. $200 billion, was with the E.U. (Tooze 2018, 548). Also, half of the U.K.'s $1.2 trillion in FDI was from the E.U. Further, the U.K.'s trade with the E.U. was estimated to be 55% greater from being a member of the trading bloc than otherwise. Finally, the U.K. benetted from the labor of individuals from other E.U. members who lived and worked in the U.K., with 3.2 million E.U. citizens living in the U.K. and 1.2 million U.K. citizens living in the other members of the E.U. Although the U.K. was a net contributor to the E.U. system, it also received substantial economic benets that could be weighed against those costs. In the end, being one of the biggest net contributors made for a ripe argument for leaving the E.U. The other metrics were much harder to communicate to the average voter. Aside from the nancial burden, the E.U. imposes signicant obligations on its members, primarily through the implementation of E.U. law and policy within each member state. Walter Hallstein, the rst President of the European Commission referred to the E.U. as a \community of law" or Rechtsgemeinschaft. 31 The harmonization of policy and law across Europe is one of the dening features of the E.U., and one that constrains the 31 For a summary of the complexities of translation and contestation over mean- ing in German, see Thorben Kl under, \What is Rechtsgemeinschaft?" Perspectives, Max Planck Law (Sept. 7, 2020) https://law.mpg.de/perspectives/2020/09/07/what- is-rechtsgemeinschaft/, adapted from German version, Thorben Kl under, \Was ist die europ aische Rechtsgemeinschaft? Nichts. Alles!" VerfBlog, (June 25, 2020) https://verfassungsblog.de/was-ist-die-europaeische-rechtsgemeinschaft-nichts-alles/ 279 options to adjust to shocks through breach. Snyder (2018, 80) described the integration project as largely being centered on the spread and reinforcement of the rule of law: The politics of integration were fundamentally dierent from the politics of empire. The EU was like an empire in that it was a large economic space. It was unlike an empire in that its organiz- ing principle was equality rather than inequality. . . . A century later, the boring EU had solved this problem. Its tedious process of accession involved the export of the spirit of the law. Euro- pean integration was a means of transporting the idea of the rule of law from places where it functioned better to places where it functioned worse. Because the rule of law and compliance with the law is so engrained in the E.U. structure, deviations from the law are costly for various reasons. These potentially high costs disincentive cheating but also constrain the ability of member states to breach to address their concerns after a shift in the distribution of perceived costs and benets. As Cameron explained during negotiations, it was either reform or leave, there was no consideration of breaching (Cameron 2019, 638). One dilemma in developing a systematic assessment of the compliance of E.U. members states with their obligations is the broadness of the obliga- tions across the many dierent issue areas governed by E.U. law. Yet, the E.U. does have a system of accountability and monitoring for each mem- ber's eorts to apply E.U. law within their borders. According to Article 4(3) of the TFEU, the member states and the European Commission (EC or the Commission) share a responsibility to ensure compliance with E.U. 280 law. Pursuant to a request made by the European Parliament in 1984, the European Commission prepares and presents an annual report on the monitoring of the application of E.U. law in the member states. The an- nual Monitoring the Application of E.U. Law Report, \provides an account of the Commission's work to ensure the correct application, implementation and enforcement of EU law . . . ." 32 This goes beyond the disputes that make their way into a courtroom, although some of those actions are also included in the report. Generally, the Court of Justice (CJ), one of two bodies of the Court of Justice of the European Union (CJEU), 33 interprets E.U. law and ensures the uniform application across all E.U. member states as outlined by TFEU. As discussed above, the rulings of the CJ have made clear the E.U. law is supreme and binding on member states. Under article 258 of the TFEU, the CJ may determine whether a member state has fullled its obligations under E.U. law. These matters are primarily brought before the CJ by the Commission, through what is called an infringement procedure. The infringement procedure is divided into two phases: a pre-litigation phase and litigation phase (TFEU, art. 258). In the rst phase, the Commission sends the member state a letter of formal notice requesting an explanation for the apparent breach within a specied time limit. If the member state's response is unsatisfactory or the state fails to respond, the Commission then send a reasoned opinion to the state asking for the state to comply within 32 \Monitoring the Application of the European Union law, 2019 Annual Report." Eu- ropean Commission, 2. 33 The other being the General Court dealing principally with competition law, State aid, trade, agriculture, trademarks. 281 a set time limit. Should the member state fail to comply after receiving the reasoned opinion, the process moves to the second phase of litigation. The Commission may also seek nes if the case involves the failure of the member state to transpose a EC directive (TFEU, art. 260(3)). Figure 6.3: Complaints Against the U.K. re EU Law The format, completeness, and coverage of the annual reports has varied. However, in recent years, the EC's report has been largely standardized and comprehensive. Figures 6.3 and 6.4 summarize two metrics of com- pliance among the member states from approximately 2007 to 2019. Fig- ure 6.3 shows the number of new complaints against member states that the Commission received for each year, highlighting the complaints against the United Kingdom. Although the Brexit referendum occurred in 2016, as highlighted in gure 6.3, the U.K. did not leave the E.U. until January 31, 282 2020 after years of negotiations. As such, the U.K. was still subject to E.U. law through 2019, giving a chance to see trends both before and after Brexit. Figure 6.4: Infringement Cases Against the U.K. re EU Law The Commission receives complaints against member states on a wide range of issues, including consumer, environment, employment, taxation, internal market, mobility, and migration issues, among others. The Commission can decide to pursue a complaint by launching the E.U. Pilot mechanism or a formal infringement procedure. From 2015 to 2019, the Commission processed the complaints in 40 weeks on average. 34 The complaints can serve as noisy a barometer of how the public views a member state's compliance with E.U. law. 34 \Monitoring the Application of the European Union law, 2019 Annual Report." Eu- ropean Commission, 14. 283 In gure 6.3, it is apparent that the U.K. had a high number of new com- plaints in 2009, but the number of complaints ultimately were not the most in the E.U. Spain, Italy, Germany, and France (not surprisingly the four most populous countries in the E.U.) consistently have more complaints than the U.K. Italy hits a peak of new complaints in 2016, the same year as the Brexit referendum. After 2010, the U.K. has a fairly consistent level of new complaints at an annual average of 178, compared with France with an annual average of 285. In recent years, less populous states like Romania and the Netherlands have been the targets of more complaints than the U.K. Thus, on the surface, it does not appear that the U.K.'s implementation or misimplementation of E.U. law has generated outsized reaction from the public. Moving from public complaints to Commission actions, gure 6.4 shows the number of new infringement cases initiated against each member states from 2008 to 2019, again highlighting the U.K. Given that infringement cases are generally brought by the Commission itself, the variation on the number of new cases brought against the member states is smaller. In 2019, the states with the highest number of new cases brought against them were Spain, Greece, Italy, Poland, and Bulgaria. As with the case of public complaints, the number of cases initiated against the U.K. sits rmly within the midst of the number of cases against other member states. Other than tracking a community uptick in cases in 2011, the number of annual new cases against the U.K. has remained relatively stable with an annual average of 59 new cases a year. Spain, Italy, and Greece all have an annual average of over 80 284 new cases. Thus, in examining metrics of the U.K.'s compliance with E.U. law and regulations, it appears that the U.K. did not try to breach E.U. law more before or after Brexit. It is possible the Commission has had a light hand in pursuing the U.K., but the public complaints have also largely been stable before and after Brexit. Thus, the U.K. likely concluded that breaching was not an option. It did not unilaterally stop the ow of migration or stop making contributions to the E.U.|even after the referendum. Thus, the stakes for a bargain were heightened. It is also possible that the U.K.'s con- tinued compliance with its obligations undermined its threats of withdrawal during the negotiations with E.U. leaders. In other cases of treaty exit, the exit is preceded by attempts to cheat and prot without substantial costs. Where there is little to no exibility to breach, it might be hard to signal an intention to exit or shift expectations during the bargaining process. Additionally, adding the referendum to the bargaining process constrained Cameron's ability to negotiate. This is not a direct feature of the E.U. per se, but it does correspond to the principles of rule of law and consent. In this case, it is not only the preferences of the U.K.'s elected representatives, but also the preferences of the public at large that will determine the success or failure of a bargain to avoid exit. Although the model presented in Chapter 2 focuses on states making decisions as a unied actor|or at least one with a unied preference of outcomes|it is useful to consider the calculations made by the voters, even where Parliament ultimately had to ratify any decision regarding U.K. membership in the E.U. 285 As reported by Clarke, Goodwin and Whiteley (2017b), U.K. voters did weigh the costs and benets of supporting Brexit, in the contexts of the economy and immigration. Other emotional reactions to the EU and heuris- tics about leaders also played a role in voters' decision making. Clarke, Goodwin and Whiteley (2017b) build on the logic of Hooghe and Marks (2005), which posits that attitudes to European integration are shaped by three broad factors. The rst factor is the distributional costs and benets of integration, namely who wins and who loses. The second factor is the community consideration of social identities. Individuals who acknowledge multiple identities favor integration while those with an exclusive national identity are more Eurosceptic. Finally, the third factor is cues or heuristics that voters receive from elites, including politicians and party leaders (See also Zaller et al. 1992). Elite opinion on European integration is most in u- ential when it is unied (Ray 2003). Clarke, Goodwin and Whiteley (2017b) nd support for each of those factors and surmise that a change to any one could have changed the outcome for such a slim vote margin. Interestingly, attitudes about the E.U. in the U.K. have varied signicantly over the last two decades but only resulted in an exit when the shock of economic crisis and immigration posed an insurmountable obstacle to successfully bargain- ing for accommodations. Although the voters could and did weigh costs and benets of exit as ra- tional actors, they also employed heuristics that allowed a wider range of opinions that could misalign with the preferences of the government. In a counterfactual world, the Cameron government may have done a better job 286 at managing the remain campaign or even reading public sentiment during the bargaining process. The government could have taken unilateral action on a key issue, like immigration, to appease some members of the public at the cost of being punished for the breach. However, the government did not do so because it likely estimated the costs of breaching as being prohibitive costly compared to exiting. As Cameron (2019, 630, 632{634) explained in hindsight, the U.K. sought deregulation within the E.U. and limitations on the free movement of EU citizens{not an outright ban. Arguably, the U.K. had the ability to reduce movement of people from the E.U., as it did with non-E.U. citizens (Cameron 2019, 634). However, violating its obligations at the border would have had tremendous costs, especially at the border of Northern Ireland and the Republic of Ireland. 35 The U.K. eventually did enact law in 2020 36 that ended the free movement of people into the country, drawing accusations of breach from the EC and the beginning of infringement proceedings in the E.U. 37 35 The border issue on Ireland played a signicant role before, during, and after Brexit. The U.K. government is bound by the Good Friday Agreement (1998), which does not explicitly prohibit border controls but seeks to \return as possible to normal security arrangements in Northern Ireland . . . " (Agreement, \Security", para. 2). Irish citizens, who are also E.U. citizens, have special status in the U.K. and can travel unhindered within the Common Travel Area (CTA), an arrangement dating back to 1922 and codied in multiple pieces of legislation in both Ireland and the U.K. These complicating factors and the eorts the U.K. government has taken to avoid violating these provisions highlight its posture of avoiding breach on many key issues related to the E.U. 36 Immigration and Social Security Co-ordination (EU Withdrawal) Act 2020, receiving Royal Assent on November 11, 2020. 37 The EC contended that the U.K. was bound by the obligation of the free move- ment of people until the end of the post-Brexit transition period. Press release: \May infringements package: key decisions." European Commission (May 14, 2020), available at https://ec.europa.eu/commission/presscorner/detail/en/inf 20 859 287 Brexit focused both policy and research on the issues of international cooper- ation and exit. As explained in this dissertation, to understand treaty exit is to understand international cooperation broadly. In recent years, Gastinger (2021) has developed a \EU Exit Index" to estimate each member state's propensity to leave the E.U. Using the index, which factors together a num- ber of variables he associates with a risk to exit, he nds that the U.K. was and remains an outlier among E.U. member states. He also estimates that Austria, Sweden, Cyprus, Czechia, and Denmark have the next highest propensity to exit the E.U., but that the risk is low based on current data. The measure is useful in assessing relative risks of exit among E.U. mem- bers, yet it has limitations|seemingly acknowledged|that perceptions of the U.K.'s post-Brexit \success" could change and encourage further exits. Figure 6.5: Continued Opposition to Brexit E.U. supporters or \Remainers" protest Brexit in 2019. Photo credit: author. 288 A vote for Brexit was not really a vote against globalism (as commentators described) but a vote \to restore Britain's greatness and freedom" Tooze (2018). \Brexit was a vote for autonomy." (Tooze 2018, 554). Similar senti- ments were expressed by remain protesters outside the Palace of Westmin- ster more than three years later. 38 With that autonomy, the U.K. can adjust to shocks without the burdens put on it by the E.U. Yet, the U.K. also loses out on all the benets that came with membership to the E.U. Even years after the referendum, it remains to be seen whether the U.K. can thread the needle of benetting from trade and commerce with its neighbors while avoiding the burden of complying with the internal law and regulations of the union. 6.4 Conclusion In this chapter, I used the monumental decision of the U.K. to leave the E.U. as a vehicle to understanding why bargaining without an eective option to breach or sucient room to reform leads to treaty exit. Key to the Brexit narrative is the direct vote on the topic by the people. By pairing renegotia- tion and referendum, Cameron had tied his hands and limited his bargaining options. Further, he opened up the bargaining process to the messiness of domestic politicking rather than the formal and staid negotiations among 38 Author interview (Interview C), October 2019. 289 state leaders and diplomats. Although Gastinger (2021) concludes that an- other exit is not on the horizon, that is contingent on the nature of future shocks like the COVID-19 pandemic. How will the pandemic aect each E.U. member state's calculation of the benet of remaining within the union? The response to the pandemic put the organization under stress, again, showing cracks between some of the members. Early in the pandemic, some states like Germany and France had an abundance of personal protective equipment (PPE) but did not share their surplus with other states in dire need like Italy. 39 Further, the E.U.'s bargaining regarding the vaccines and roll-out was bungled and mishandled at multiple steps. The E.U. failed to assert priority from the vaccine man- ufacturers, falling behind the U.S. and U.K. when supplies were short. 40 Further, a few member states have had frustrating handling of vaccination, where in Germany residents could only obtain appointments by telephone and needed to make numerous calls. 41 All of this, on top of seeing the suc- cesses in places like Israel, U.S., and U.K., have caused frustration among the European public. This would seem particularly troubling given the per- ceived early ineptitude of the U.S. and U.K. governments during the pan- demic, as well as the U.K.'s recent exit. Is this proof that the U.K. really 39 Ben Hall, Miles Johnson, and Martin Arnold. \Italy wonders where Europe's soli- darity is as coronavirus strains show." The Financial Times (March 13, 2020), available at https://www.ft.com/content/d3bc25ea-652c-11ea-b3f3-fe4680ea68b5 40 Elian Peltier and Josh Holder. \How Europe, After a Fumbling Start, Overtook the U.S. in Vaccination." The New York Times (July 29, 2021), available at https://www.nytimes.com/interactive/2021/07/29/world/europe/europe-us- vaccination.html 41 Caroline Copley. \Tip-os and cold calls: Germans hustle for a COVID-19 shot." Reuters (May 19, 2021), available at https://www.reuters.com/business/healthcare- pharmaceuticals/tip-os-cold-calls-germans-hustle-covid-19-shot-2021-05-19/ 290 is better o going it alone? We might imagine the consternation of ocials in Brussels over the disparity and publics in other states contemplating how their own fortunes might change outside the union. On the other side of the scale is the benet many of the smaller states re- ceived from bargaining for vaccines as a block. These smaller states beneted from the group bargaining power with vaccine manufacturers. Additionally, the integration of the continent and free ow of people necessitates that the group needs to come out of the pandemic together to regain economic well- being and normalcy. Brexit and the long process of divorce informed the E.U. on how to adjust its bargaining strategy to avoid another signicant defection. 291 Chapter 7 Conclusion 7.1 Contribution, Implications, and Remaining Ques- tions As shown in the data and in looking at both historical and modern cases, treaty exit is not uncommon while being relatively dwarfed in comparison to the creation of treaties. Over the last seven decades, states have created far more treaty relationships than they have ended by leaving. However, those exits can be disruptive to international cooperation and may cause more harm than not entering into an agreement in the rst place. Ending a relationship may harm trust more than never having started the relationship. To save cooperation, states bargain to reform their relationship in the shadow 292 of possible breach and exit. This dissertation provides a framework to un- derstand treaty withdrawal as a part of states' strategic calculations in the repeated bargaining within a treaty relationship. Building on existing and parallel ideas of this relationship, this project works to formalize assump- tions and mechanisms to add rigor to our theoretical explanation for state behavior. As explained in Chapter 2, the exibility and level of obligations of a treaty dene a states' preferences over whether it abides with the status quo, attempts to unilaterally adjust to a shock, or withdraw from the rela- tionship. In considering how extensively it (or they) will make concessions, the states on the other side of the bargain will assess the importance, as broadly dened, of the potential exiting state to the institution. If there is a risk to the survival of the institution through a loss of contributions or contagion, the remaining states will be more willing to reform the institution and make more substantial adjustments. Aside from making assumptions explicit, formalizing a theory of treaty exit highlights the distribution and commitment problems that may make secur- ing a bargain for reform unattainable. These problems were key in some signicant exit scenarios, including in the case of Brexit. Further, the sim- plicity of a formal model allows for a focus on particular aspects of the bar- gain that may underlie many dierent cases. These foundational dynamics can explain a wide variety of behavior enriching pre-existing theories. Exten- sions and modications, of course, are possible and encouraged. Engaging in that further work and charting a path is made clearer when our theories are formalized. 293 The theory was tested and conrmed in part using a mix of quantitative and qualitative methods. First, the aspects of the theory regarding exibil- ity and obligation were tested on a new dataset of treaties collected from the U.N. Treaty Series. This new data expands the corpus of information used to study treaty behavior. Particularly in the context of treaty exit, this data moves beyond membership in international organizations and multi- lateral treaties to include bilateral treaties and treaties over a wide range of subjects. Further, the new data includes treaty actions, like signatures, ratications, reservations, and declarations, as well as treaty texts. This approach and the data it created expands the range of data used to test our understanding of international cooperation. The data here admittedly could be improved by collecting more treaty texts, more cleaning and pre- processing of treaty texts, and employing additional methods of automated text analysis. This data, and the data built on this work, can be used to study a wide variety of state behavior by pairing it with other events and actions. Beyond the quantitative analysis, Chapters 4, 5, and 6 examine three cases of treaty withdrawal to uncover the mechanisms at work in the theory and to discover additional wrinkles in how the theory applies. Each case high- lights how a treaty's level of obligation and exibility interact to create the conditions for a successful bargain to avoid treaty exit. As expected by the theory and empirical observations discussed above, treaties that include more obligatory language, are more detailed, and have numerous members are more likely to see treaty exit. However, as the cases highlight, those 294 variables are only part of what in uences treaty exit. The exibility, or lack thereof to cheat, also can in uence whether a treaty suers an exit. Further, the cases illustrate how institutions can be strategically exible depending on the importance of the member state at issue. Each case study provided more context to each episode of exit and allowed an exploration of variables that are dicult to quantify. In Chapter 4, the diering postures of the International Criminal Court (ICC) to the threat of withdrawal from its members highlighted the percep- tion of the importance of the exiting state in the bargaining process. There, the ICC|through the Assembly of Parties|was willing to essentially over- rule the Appeals Chamber by rewriting the rules of procedure. It did this because Kenya was an important actor because its exit could risk conta- gion. South Africa's exit was also a threat to the institution; however, in that case, the domestic opposition and strength of the state's institutions allowed the ICC to wait for a domestic change before conceding too much. Finally, the idiosyncratic cases of Burundi and The Philippines show that the ICC will not concede in cases where the leader of the exiting state is under investigation and there is little risk of contagion. In Chapter 5, treaty breach and exit are examined in the context of the arms control relationship between the U.S. and Russia over many decades. In looking at that relationship, the exibility of the treaties to allow cheat- ing weighed on the leaders decisions of whether to continue cooperation. Especially in the context of security, the level of obligation and ability to exibly respond to threats are often at the forefront of leaders' calculations. 295 During the Cold War, the leaders of the U.S. and Russia were constantly concerned that the other would obtain some advantage that would threaten their security. President Reagan's dislike of the doctrine of mutually assured destruction led his administration to seek ways to pursue missile defense unilaterally, leading to extensive domestic debate over the interpretation of the treaty's obligations. Later U.S. presidents sought to adjust to the changing security environment within the bounds of the treaty obligations, like the Clinton Administration's theater missile defense proposal. The ten- sion between the constraints on missile defense and the changing security environment exerted prolonged pressure on the U.S. for multiple admin- istrations. Moreover, the bilateral nature of the agreements allowed for a signicant amount of bargaining and potential exibility between competing superpowers. Ultimately, the ABM Treaty prevented the U.S. from eec- tively cheating without signicant costs and the two sides could not reach a bargain to reform the agreement to accommodate limited missile defense. This led to the ultimate cascade of exits that unraveled much of the Cold War-era arms control regime between the U.S. and Russia. Finally, in Chapter 6, the fascinating interplay between bargaining between states and between leaders and their public is revealed as substantially signif- icant to whether bargaining can be successful. The E.U. system is premised on the fundamental respect for the rule of law through the harmonization of E.U. law across the member states. The baseline obligations of the union are signicantly higher than the average treaty, making signicant shocks to a country more pronounced. For the U.K., the uneasy relationship with the 296 E.U.'s scal, monetary, and free movement of people policies was strained after a series of crises in the early twenty-rst century. With no real option to breach to adjust, the U.K. was faced with a choice between reform or exit. In coming to a tentative deal with the E.U., the U.K. government misunderstood the public's perceived level of obligation within the E.U., despite numerous metrics showing dwindling appetite for modest changes to the U.K.-E.U. relationship. Because the U.K.'s primary demands went to the heart of the integration project and there was little to no room for possible breach, exit became the only viable outcome. As the quantitative and qualitative analyses reveal, there are additional questions to be answered about why states decide to leave a treaty rela- tionship. As highlighted by the most frequent exiting states over the last 80 years and the cases of the U.S. and U.K.'s exit from high-prole treaties, there appears to be some relationship between a state's respect for the rule of law and reputation for following commitments and its preference for us- ing treaty exit as a bargaining tool. In those cases, breaching is almost completely excluded from the viable options to addressing a shock in the level of obligations. Those states attached so much weight to compliance that breaching would be prohibitively costly. In those scenarios of reform or exit, threatened exit is a useful and legal bargaining tool that the states are comfortable employing. Further examination of the state-level preferences for legal actions might shed more light on the variation of behavior between states. Other questions of interest include the dynamics of future negotiations after 297 a treaty exit and the linkage of bargaining across dierent treaty relation- ships. As was the case in Chapter 5, the bargaining between states often happens in consideration of multiple treaty relationships. Finally, further exploring the rhetoric and reactions of treaty exit would be valuable for our understanding of international cooperation. States often go to great lengths to explain the reasons for their withdrawal, even when unconditionally per- mitted by the treaty itself. As described in the introduction, the Trump Administration publicly emphasized the behavior of other states, the un- fairness of the treaty obligations, and the impact to domestic constituencies when announcing the U.S.'s intention to leave the Paris Climate Accord. Like most withdrawal provisions, the Paris Climate Accord does not require any explanation for the withdrawal to have eect (see art. 28). 1 When and why do they feel the need to overexplain? Relatedly, even when a with- drawal is permissible in the terms of an agreement, other states may use statements of disapproval as part of their eort to shame or coerce the with- drawing state into dierent behavior. Each of these questions and behaviors is worthwhile to assess because answers will increase our understanding of how international cooperation falls apart and how it can be improved. 7.2 Where do we go from here? Because of the consequential nature of Brexit and the numerous withdrawals of the Trump Administration in the U.S., there are fears that international 1 3156 UNTS 1. 298 cooperation is unraveling. It is true that the international legal order is under pressure on numerous fronts. The challenge of addressing pandemics, climate change, and economic turbulence is testing the ability of the interna- tional community to address monumental issues facing humanity. Further, revisionist actors like Russian President Vladimir Putin and his supporters seek to test the resiliency of the international legal order in the face of un- bridled aggression and brutality. Russia's unprovoked war in Ukraine poses a test of the ability of international law and institutions to respond to injus- tice and human suering. Some may point to the invasion as an example of international law's impotence to restrain pure power. Yet, although missiles and jets move with breath-taking speed, the law reacts slowly, absorbing the trauma of war to eventually spring back to restore order. Moreover, \Putin's war" has, for the time-being, fortied cooperation be- tween certain states. NATO is more unied than it has been in decades. The shock of the war has even caused the fractious U.S. Senate to reexamine and reinforce the U.S.'s international commitments. After generations of ceding authority over treaty withdrawal to the executive branch, the U.S. Senate has shown signs that it is more interested in protecting some of America's most important treaty obligations. On April 15, 2021, U.S. Senators Tim Kaine of Virginia (D) and Marco Rubio of Florida (R) reintroduced Senate Joint Resolution 17, \A joint resolution requiring the advice and consent of the Senate or an Act of Congress to suspend, terminate, or withdraw the United States from the North Atlantic Treaty and authorizing related 299 litigation, and for other purposes." 2 The resolution requires a president to seek the advice and consent of the Senate before leaving NATO. Without Senate approval, the bill prohibits any funding to be used to leave NATO and authorizes the Congressional Legal Counsel to challenge the president in court. In part a reaction to the Trump presidency, the resolution has gained traction during the Russia-Ukraine con ict, attracting Senators Mitt Romney of Utah (R) and Ted Cruz of Texas (R) in the weeks after the Rus- sian invasion. On March 29, 2022, the Senate Foreign Relations Committee passed the resolution, which may proceed to the full Senate for a vote. However, we know that when communities feel abandoned or forgotten by their government and society as a whole, they can develop resentment that may manifest itself in electoral politics (See, e.g., Hill 2021). France re- cently reelected Emmanuel Macron as president, after a strong challenge from Marine Le Pen, who questioned Europe's place in the world, France's role in Europe, and how the international community will rise to address various challenges. 3 Ms. Le Pen, a repeat contender for the presidency, has an illiberal vision for France and seeks to disengage from key international institutions. 4 2 A previous version of the resolution passed a committee vote during the nal weeks of the Trump presidency, but did not receive a vote on the Senate oor before the end of the 116th Congress. See S.J.Res.4 116th Congress (2019-2020). A similar resolution was introduced in the House of Representatives. H.J.Res.41 116th Congress (2019-2020). 3 Paul Kirby. \French election result: Macron defeats Le Pen and vows to unite divided France." BBC News (April 24, 2022), available at https://www.bbc.com/news/world- europe-61209058 4 \Le Pen wants France out of NATO integrated command, backs NATO-Russia links." France24 (April 13, 2022), available at https://www.france24.com/en/france/20220413-le- pen-wants-france-out-of-nato-integrated-command-backs-nato-russia-links 300 Moreover, not only will the Russian invasion of Ukraine have repercussions to international law and order, but so will the valiant Ukrainian resistance and consistent calls for action to the international community. In addressing the U.N. Security Council, President Zelensky put forward his challenge: If this continues, the nale will be that each state will rely only on the power of arms to ensure its security, not on international law, not on international institutions. [{] Then, the UN can simply be dissolved. . . . The power of the UN Charter must be restored immediately. [{] The UN system must be reformed immediately so that the right of veto is not a right to kill. . . . There can be no more exceptions, privileges. Everyone must be equal. All participants in international relations. Regardless of economic strength, geographical area and individual ambitions. [{] The power of peace must become dominant. The power of justice and the power of security. As humanity has always dreamed of. 5 Further, the European-ness of the con ict has and will likely continue to highlight divisions across the world, especially in terms of arguments of hypocrisy to the West's forceful reaction to the invasion. 6 Long-settled norms of state sovereignty, the bedrock of the international legal system, are being challenged. 7 The world is not entering a new Cold War but a more fractured, suspicious, and guarded era. States will be more strategic 5 Speech by the President of Ukraine at a meeting of the UN Security Council on April 5, 2022, available at https://www.president.gov.ua/en/news/vistup-prezidenta-ukrayini- na-zasidanni-radi-bezpeki-oon-74121 6 See, e.g., Amrit Dhillon. \Indians reluctant to denounce Russian `brothers' over Ukraine" The Guardian (Mar. 27, 2022), available at https://www.theguardian.com/world/2022/mar/27/indians-reluctant-to-denounce- russian-brothers-over-ukraine 7 Tanisha M. Fazal. \The Return of Conquest? Why the Future of Global Order Hinges on Ukraine." Foreign Aairs (May/June 2022), avail- able at https://www.foreignaairs.com/articles/ukraine/2022-04-06/ukraine-russia-war- return-conquest 301 in their relationships just at a time when global challenges require collec- tive solutions. Encouraging international cooperation will be key to those eorts and understanding how cooperation breaks down is essential. This dissertation adds to that understanding. All is not lost. Throughout history, cooperation has ebbed and owed, like a tide often coming after a time of turbulence. We should not despair at the decoupling of states and increased competition. Interests will realign and new generations will have dierent priorities. Like the elements formed in the stars, things fall apart and come together again producing new building blocks for cooperation. Bibliography 302 Bibliography Abbott, Kenneth W. 1999. \International relations theory, international law, and the regime governing atrocities in internal con icts." American Journal of International Law 93(2):361{379. Abbott, Kenneth W and Duncan Snidal. 2000. \Hard and soft law in inter- national governance." International organization 54(3):421{456. Abbott, Kenneth W, Robert O Keohane, Andrew Moravcsik, Anne-Marie Slaughter and Duncan Snidal. 2000. \The concept of legalization." Inter- national organization pp. 401{419. Abtahi, Hirad. 2005. \The Islamic Republic of Iran and the ICC." Journal of International Criminal Justice 3(3):635{648. Adejumobi, Saheed A. 2007. The history of Ethiopia. Greenwood Publishing Group. Almeida, Paula Wojcikiewicz and Ma ra Fajardo Linhares Pereira. 2013. \The obligations of signatory states of an international treaty: from good faith to international obligation." Revista Direito GV 9(1):171{197. Bibliography 303 Alter, Karen J. 2014. The new terrain of international law: Courts, politics, rights. Princeton University Press. Amann, Diane Marie and Mortimer NS Sellers. 2002. \The United States of America and the International Criminal Court." The American Journal of Comparative Law 50(suppl 1):381{404. Amirfar, Catherine and Ashika Singh. 2018. \The Trump Administration and the Unmaking of International Agreements." Harv. Int'l LJ 59:443. Anderson, Perry. 2021. \The Breakaway." The London Review of Books 43(2). Appel, Benjamin J. 2018. \In the Shadow of the International Criminal Court: Does the ICC Deter Human Rights Violations?" Journal of con- ict resolution 62(1):3{28. Applebaum, Anne. 2020. Twilight of democracy: The seductive lure of au- thoritarianism. Doubleday. Arend, Richard J and Darryl A Seale. 2005. \Modeling alliance activity: an iterated prisoners' dilemma with exit option." Strategic Management Journal 26(11):1057{1074. Axelrod, Robert. 1984. The Evolution of Cooperation. New York: Basic Books. Azaria, Danae. 2017. \Secret Treaties in International Law and the Faith of States in Decentralized Enforcement." AJIL Unbound 111:469{473. Bibliography 304 Baker, Peter and Susan Glasser. 2021. The Man Who Ran Washington: The Life and Times of James A. Baker III. Anchor. Barber a, Pablo, Amber E Boydstun, Suzanna Linn, Ryan McMahon and Jonathan Nagler. 2021. \Automated text classication of news articles: A practical guide." Political Analysis 29(1):19{42. Barnett, Michael and Martha Finnemore. 2012. Rules for the World. Cornell University Press. Barnett, Michael N and Martha Finnemore. 1999. \The politics, power, and pathologies of international organizations." International organiza- tion 53(4):699{732. Baron, Joshua. 2014. Great Power Peace and American Primacy: The Ori- gins and Future of a New International Order. Springer. Bas, Muhammet A and Andrew J Coe. 2018. \Give peace a (second) chance: A theory of nonproliferation deals." International Studies Quar- terly 62(3):606{617. Bassiouni, M Cherif. 1999. \Negotiating the Treaty of Rome on the estab- lishment of an International Criminal Court." Cornell Int'l LJ 32:443. Baylis, Elena. 2009. \Outsourcing investigations." UCLA J. Int'l L. Foreign A. 14:121. Bearce, David H and Stacy Bondanella. 2007. \Intergovernmental organi- zations, socialization, and member-state interest convergence." Interna- tional Organization 61(4):703{733. Bibliography 305 Bennett, Andrew. 2014. \Disciplining our conjectures: Systematizing pro- cess tracing with Bayesian analysis." Process tracing: From metaphor to analytic tool pp. 276{98. Bennett, Andrew and Jerey T Checkel. 2014. Process tracing: from philo- sophical roots to best practices. In Process tracing: From metaphor to analytic tool. Cambridge University Press chapter 1, pp. 3{38. Bentham, Jeremy. 1843. Principles of Judicial Procedure, with the Outlines of a Procedure Code. In The Works of Jeremy Bentham, Published under the Superintendence of his Executor, John Bowring (1838{1843), ed. J. Bowring. Vol. II chapter 2, pp. 11{15. Berendsen, Bernard, ed. 2011. Common goods in a divided world. Amster- dam: KIT Publishers. Bernhagen, Patrick. 2008. \Business and international environmental agree- ments: domestic sources of participation and compliance by advanced industrialized democracies." Global Environmental Politics 8(1):78{110. Berzon, Marsha S. 2012. \Dissent," Dissentals," and Decision Making." California Law Review pp. 1479{1492. Beschloss, Michael. 2019. Presidents of War: The Epic Story, from 1807 to Modern Times. Broadway Books. Black, Henry Campbell, Bryan A Garner, Becky R McDaniel, David W Schultz and West Publishing Company. 2004. Black's Law Dictionary. 8th ed. West Group St. Paul, MN. Bibliography 306 Boehmer, Charles, Erik Gartzke and Timothy Nordstrom. 2004. \Do inter- governmental organizations promote peace?" World Politics 57(1):1{38. Boister, Neil. 2010. \The Application of Collective and Comprehensive Criminal Responsibility for Aggression at the Tokyo International Mil- itary Tribunal: The Measure of the Crime of Aggression?" Journal of International Criminal Justice 8(2):425{447. Boister, Neil and Robert Cryer. 2008. The Tokyo International Military Tribunal: A Reappraisal. Oxford University Press. Boix, Carles, Michael Miller and Sebastian Rosato. 2013. \A complete data set of political regimes, 1800{2007." Comparative Political Studies 46(12):1523{1554. Bolton, John. 2008. Surrender is not an option: Defending America at the United Nations. Simon and Schuster. Bolton, John. 2020. The room where it happened: A White House memoir. Simon and Schuster. Boyd, Christina L. 2016. \Representation on the courts? The eects of trial judges' sex and race." Political Research Quarterly 69(4):788{799. Bradley, Curtis A. 2013. \Treaty Termination and Historical Gloss." Tex. L. Rev. 92:773. Bradley, Curtis A and Jack L Goldsmith. 2017. \Presidential Control over International Law." Harv. L. Rev. 131:1201. Bibliography 307 Bradley, Curtis A and Laurence R Helfer. 2017. \Treaty Exit in the United States: Insights from the United Kingdom or South Africa?" AJIL Un- bound 111:428{433. Budjeryn, Mariana. 2016. The Power of the NPT: International Norms and Nuclear Disarmament of Belarus, Kazakhstan and Ukraine, 1990-1994 PhD thesis Central European University. Bull, Hedley. 1977. The Anarchical Soceity: A Study of Order in World Politics. New York: Columbia University Press. Bush, George H.W. and Brent Scowcroft. 1998. A World Transformed. New York: Alfred A. Knopf. Bush, George W. 2010. Decision points. Broadway Books. Butler, William E. 1995. \Russian federation: Federal law on international treaties." International Legal Materials 34(5):1370{1392. Cameron, David. 2019. For the Record. William Collins. Cassese, Antonio. 2005. International Law. 2nd ed. Oxford university press. Chapman, Terrence L, Johannes Urpelainen and Scott Wolford. 2013. \In- ternational bargaining, endogenous domestic constraints, and democratic accountability." Journal of Theoretical Politics 25(2):260{283. Chapman, Terrence L and Scott Wolford. 2010. \International organizations, strategy, and crisis bargaining." The Journal of Politics 72(1):227{242. Bibliography 308 Charlesworth, Hilary, Christine Chinkin and Shelley Wright. 1991. \Femi- nist approaches to international law." American Journal of International Law 85(4):613{645. Chaudoin, Stephen. 2016. \How Contestation Moderates the Eects of In- ternational Institutions: The International Criminal Court and Kenya." The Journal of Politics 78(2):557{571. Chayes, Abram and Antonia Handler Chayes. 1985. \Testing and Develop- ment of Exotic Systems under the ABM Treaty: The Great Reinterpre- tation Caper." Harv. L. Rev. 99:1956. Chayes, Abram and Antonia Handler Chayes. 1993. \On compliance." In- ternational organization 47(2):175{205. Chayes, Abram and Antonia Handler Chayes. 1998. The new sovereignty: compliance with international regulatory agreements. Harvard University Press. Chayes, Antonia Handler, Abram Chayes and Ronald B Mitchell. 1995. Active compliance management in environmental treaties. In Sustainable Development and International Law, ed. Winifred Lang. London: Graham & Trottman Ltd. Checkel, Jerey T. 2001. \Why comply? Social learning and European identity change." International organization 55(3):553{588. Cheney, Richard B. 2011. In my time: A personal and political memoir. Simon and Schuster. Bibliography 309 Chesterman, Simon. 2016. \Asia's ambivalence about international law and institutions: Past, present and futures." European Journal of Interna- tional Law 27(4):945{978. Choi, Seung-Whan. 2021. \Nationalism and withdrawals from intergov- ernmental organizations: Connecting theory and data." The Review of International Organizations pp. 1{11. Clark, Richard. 2020. \Bargain Down or Shop Around? Outside Options and IMF Conditionality." Working Paper . Clarke, Harold D, Matthew Goodwin and Paul Whiteley. 2017a. Brexit Why Britain Voted to Leave the European Union. University of Essex Publication. Clarke, Harold D, Matthew Goodwin and Paul Whiteley. 2017b. \Why Britain voted for Brexit: An individual-level analysis of the 2016 referen- dum vote." Parliamentary Aairs 70(3):439{464. Clarke, Kamari, Abel S. Knottnerus and Eefje de Volder. 2016. Africa and the ICC: An Introduction. In Africa and the ICC: Perceptions of Justice, ed. Kamari Clarke, Abel S. Knottnerus and Eefje de Volder. Cambridge University Press chapter 6, pp. 107{126. Clarke, Kamari Maxine. 2009. Fictions of justice: The International Crim- inal Court and the challenge of legal pluralism in Sub-Saharan Africa. Cambridge University Press. Clinton, Bill. 2004. My life. Vintage. Bibliography 310 Coe, Andrew J and Scott Wolford. 2020. East Asian History and Interna- tional Relations. In East Asia in the World: Twelve Events That Shaped the Modern International Order, ed. Stephan Haggard and David C. Kang. Cambridge University Press chapter 15, pp. 263{281. Collier, David. 2011. \Understanding process tracing." PS: Political Science & Politics 44(4):823{830. Collier, Paul and Nicholas Sambanis. 2002. \Understanding civil war: a new agenda." Journal of Con ict Resolution 46(1):3{12. Combs, Nancy Amoury. 2009. \Testimonial Deciencies and Evidentiary Uncertainties in International Criminal Trials." UCLA J. Int'l L. Foreign A. 14:235. Conrad, Courtenay R and Emily Hencken Ritter. 2013. \Treaties, tenure, and torture: The con icting domestic eects of international law." The Journal of Politics 75(2):397{409. Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staan II Lind- berg, Svend-Erik Skaaning, Jan Teorell, Joshua Krusell, Moa Olin, Josene Pernes, Johannes Roemer et al. 2018. \V-Dem [Country- Year/Country-Date] Dataset v8." Varieties of Democracy (V-Dem) Project . Costa, Keith J, Trent Lott, Don Nickles, Larry E Craig, Jon Kyl, Jesse Helms, Connie Mack, Paul Coverdell and Bob Smith. 1998. \GOP raps Clinton for'contrivances': SENATE DECLARES ABM TREATY NULL Bibliography 311 AND VOID; BLASTS WHITE HOUSE POSITION." Inside Missile De- fense 4(21):4{6. Crampin, Joseph. 2020. \Treaty withdrawal and recalcitrant States." Cam- bridge International Law Journal 9(2):225{240. Dai, Xinyuan. 2005. \Why comply? The domestic constituency mecha- nism." International Organization 59(2):363{398. Dai, Xinyuan. 2007. International institutions and national policies. Cam- bridge University Press. Dalton, Taylor R. 2020. \Asking for the Moon: A Bargaining Model of Treaty Exit." Working Paper . Dalton, Taylor R. 2021. \The Ties that (Un)bind: A Network Analysis of Treaty Exit." Working Paper . Damaska, Mirjan. 2009. \The International Criminal Court between aspi- ration and achievement." UCLA J. Int'l L. Foreign A. 14:19. D'Amato, Anthony. 1975. \International Law and Rawls' Theory of Justice." Denv. J. Int'l L. & Pol'y 5:525. Dawuni, Josephine J. 2017. \African Women Judges on International Courts: Symbolic or Substantive Gains." U. Balt. L. Rev. 47:199. De Mesquita, Bruce Bueno, Alastair Smith, Randolph M. Siverson and James D. Morrow. 2005. The Logic of Political Survival. MIT press. Bibliography 312 Donaldson, Megan. 2017. \The survival of the secret treaty: publicity, se- crecy, and legality in the international order." American Journal of In- ternational Law 111(3):575{627. Donaldson, Megan. 2020. \The League of Nations, Ethiopia, and the Mak- ing of States." Humanity: An International Journal of Human Rights, Humanitarianism, and Development 11(1):6{31. Dorussen, Han and Hugh Ward. 2008. \Intergovernmental organizations and the Kantian peace: A network perspective." Journal of Con ict Res- olution 52(2):189{212. Downs, George W and David M Rocke. 1990. Tacit bargaining, arms races, and arms control. University of Michigan Press. Downs, George W, David M Rocke and Peter N Barsoom. 1996. \Is the good news about compliance good news about cooperation?" International Organization 50(3):379{406. Durand, Pierre. 2018. \How and why the European Union makes reserva- tions to international agreements." Common Market Law Review 55(5). Eichensehr, Kristen E. 2012. \Treaty Termination and the Separation of Powers." Va. J. Int'l L. 53:247. Eilstrup-Sangiovanni, Mette. 2020. \Death of international organizations. The organizational ecology of intergovernmental organizations, 1815{ 2015." The Review of International Organizations 15(2):339{370. Bibliography 313 Ellis, Mark. 2006. \Breaking the silence: Rape as an international crime." Case W. Res. J. Int'l L. 38:225. Fairlie, Megan A. 2011. \The United States and the International Crimi- nal Court post-Bush: A beautiful courtship but an unlikely marriage." Berkeley J. Int'l L. 29:528. Fassbender, Bardo, Anne Peters, Simone Peter and Daniel H ogger. 2012. The Oxford handbook of the history of international law. OUP Oxford. Fearon, James D. 1991. \Counterfactuals and hypothesis testing in political science." World politics 43(2):169{195. Fearon, James D. 1995. \Rationalist explanations for war." International organization 49(3):379{414. Fearon, James D. 1997. \Signaling foreign policy interests: Tying hands versus sinking costs." Journal of Con ict Resolution 41(1):68{90. Fearon, James D. 1998. \Bargaining, enforcement, and international coop- eration." International organization 52(02):269{305. Feith, Douglas J. 2009. War and Decision: Inside the Pentagon at the Dawn of the War on Terrorism. Harper Collins. Finnemore, Martha and Kathryn Sikkink. 1998. \International norm dy- namics and political change." International organization 52(4):887{917. Fitzmaurice, Malgosia and M Fitzmaurice. 2015. Whaling and international law. Cambridge University Press. Bibliography 314 Fitzmaurice, Malgosia and Olufemi Elias. 2005. Contemporary issues in the law of treaties. Eleven International Publishing. Fontanelli, Filippo and Paolo Busco. 2016. \The Function of Procedural Jus- tice in International Adjudication." The Law & Practice of International Courts and Tribunals 15(1):1{23. Franck, Thomas M. 1990. The power of legitimacy among nations. Oxford University Press on Demand. Franck, Thomas M. 1998. \Fairness in international law and institutions.". Franck, Thomas M and Thomas M Franck. 1995. Fairness in international law and institutions. Vol. 51 Clarendon Press Oxford. Gaddis, John Lewis. 2006. The Cold War: a new history. Penguin. Galbraith, Jean. 2012. \Treaty options: Towards a behavioral understanding of treaty design." Va. J. Int'l L. 53:309. Galbraith, Jean. 2017. \The President's Power to Withdraw the United States from International Agreements at Present and in the Future." AJIL Unbound 111:445{449. Gartzke, Erik. 1999. \War is in the Error Term." International Organization 53(3):567{587. Gastinger, Markus. 2021. \Introducing the EU exit index measuring each member state's propensity to leave the European Union." European union politics 22(3):566{585. Bibliography 315 Ge, Zoe X. 2022. \Self-Defeating Treaties: Global Value Chains and the Termination of Bilateral Investment Treaties." Presented at the Grad- uate Students in International Political Economy (GSIPE) Spring 2022 Workshop Series on April 18, 2022. George, Alexander L and Andrew Bennett. 2005. Case studies and theory development in the social sciences. mit Press. Gilpin, Robert. 2001. Global political economy: understanding the interna- tional economic order. Princeton university press. Goertz, Gary and James Mahoney. 2012. A tale of two cultures: Qualita- tive and quantitative research in the social sciences. Princeton University Press. Goldblat, Jozef. 2002. Arms Control: The New Guide to Negotiations and Agreements with New CD-ROM Supplement. Sage. Goldsmith, Jack L and Eric A Posner. 2005. The limits of international law. Oxford University Press. Goldstein, Judith L, Douglas Rivers and Michael Tomz. 2007. \Institutions in International Relations: Understanding the Eects of the GATT and the WTO on World Trade." International organization 61(1):37{67. Gorbachev, Mikhail. 1996. Memoirs. Doubleday. Gorbachev, Mikhail. 2016. The New Russia. John Wiley & Sons. Granovetter, Mark S. 1973. \The strength of weak ties." American Journal of Sociology 78(6):1360{1380. Bibliography 316 Gray, Julia. 2018. \Life, death, or zombie? The vitality of international organizations." International Studies Quarterly 62(1):1{13. Gray, Julia. 2020. \Life, death, inertia, change: The hidden lives of inter- national organizations." Ethics & International Aairs 34(1):33{42. Greenhill, Brian. 2015. Transmitting rights: International organizations and the diusion of human rights practices. Oxford university press. Greif, Avner. 1993. \Contract enforceability and economic institutions in early trade: The Maghribi traders' coalition." The American economic review pp. 525{548. Grimmer, Justin and Brandon M Stewart. 2013. \Text as data: The promise and pitfalls of automatic content analysis methods for political texts." Political analysis 21(3):267{297. Grossman, Nienke. 2009. \Legitimacy and International Adjudicative Bod- ies." Geo. Wash. Int'l L. Rev. 41:107. Grossman, Nienke. 2016a. \Achieving sex-representative international court benches." American Journal of International Law 110(1):82{95. Grossman, Nienke. 2016b. \Shattering the Glass Ceiling in International Adjudication." Va. J. Int'l L. 56:339. Gruntman, Mike. 2015. Intercept 1961: The Birth of Soviet Missile Defense. American Institute of Aeronautics and Astronautics, Inc. Grzybowski, Kazimierz. 1970. Soviet public international law: doctrines and diplomatic practice. Sijtho. Bibliography 317 Gulati, Mitu and Robert E Scott. 2013. The three and a half minute transac- tion: Boilerplate and the limits of contract design. University of Chicago Press. Guzman, Andrew T. 2002. \A compliance-based theory of international law." Calif. L. Rev. 90:1823. Guzman, Andrew T. 2008. How international law works: a rational choice theory. Oxford University Press. Hafetz, Jonathan. 2017. \Fairness, Legitimacy, and Selection Decisions in International Criminal Law." Vand. J. Transnat'l L. 50:1133. Hafner-Burton, Emilie M. 2008. \Sticks and stones: Naming and sham- ing the human rights enforcement problem." International organization 62(4):689{716. Hafner-Burton, Emilie M and Alexander H Montgomery. 2006. \Power posi- tions: International organizations, social networks, and con ict." Journal of Con ict Resolution 50(1):3{27. Hafner-Burton, Emilie M and Kiyoteru Tsutsui. 2005. \Human rights in a globalizing world: The paradox of empty promises." American journal of sociology 110(5):1373{1411. Hagan, John and Sanja Kutnjak Ivkovic. 2009. \Structural Pre- Conditionality, Smoking Gun Evidence and Collective Command Respon- sibility for War Crimes in the Former Yugoslavia." UCLA J. Int'l L. For- eign A. 14:149. Bibliography 318 Hamilton, Rebecca J. 2014. \The ICC's Exit Problem." NYUJ Int'l L. & Pol. 47:1. Hanqin, Xue. 2012. \Chinese contemporary perspectives on international law: history, culture and international law.". Harari, Yuval Noah. 2014. Sapiens: A brief history of humankind. Random House. Hart, H.L.A. 1961. The Concept of Law. oxford university press. Hathaway, Oona. 2002. \Do treaties make a dierence? Human rights treaties and the problem of compliance." Yale Law Journal 111(8):1932{ 2042. Hathaway, Oona A and Scott J Shapiro. 2017. The internationalists: How a radical plan to outlaw war remade the world. Simon and Schuster. Hauk, Esther and Rosemarie Nagel. 2001. \Choice of partners in multiple two-person prisoner's dilemma games: An experimental study." Journal of con ict resolution 45(6):770{793. Helfer, Laurence R. 2005. \Exiting Treaties." Virginia Law Review pp. 1579{ 1648. Helfer, Laurence R. 2010. \Exiting Custom: Analogies to Treaty With- drawals." Duke J. Comp. & Int'l L. 21:65. Helfer, Laurence R. 2012. Terminating Treaties. In The Oxford Guide to Treaties, ed. Duncan B Hollis. Oxford University Press chapter 25, pp. 634{650. Bibliography 319 Helfer, Laurence R. 2017. \Introduction to Symposium on Treaty Exit at the Interface of Domestic and International Law." AJIL Unbound 111:425{ 427. Helfer, Laurence R. 2019. \Taking Stock of Three Generations of Re- search on Treaty Exit: Masterclass European Society of International Law (ESIL) Research Forum Hebrew University of Jerusalem Faculty of Law, Jerusalem, Israel 28 February 2018." Israel Law Review 52(1):103{117. Heller, Kevin Jon. 2011. The Nuremberg Military Tribunals and the origins of international criminal law. Oxford University Press on Demand. Henisz, Witold J. 2002. \The institutional environment for infrastructure investment." Industrial and corporate change 11(2):355{389. Henkin, Louis. 1979. How nations behave: law and foreign policy. Columbia University Press. Henkin, Louis. 1995. \International law: politics and values." Developments in International Law 18. Hertz, Allen Z. 1990-1991. \Medieval Treaty Obligation The Theory of the States in the Third World and the Problem of Constitutionalism Sympo- sium: Comparative Legal Theory: A Meeting of East and West: Panel 2: Legal Theory and Constitutional Law." Connecticut Journal of Interna- tional Law 6:411. Hill, Fiona. 2021. There Is Nothing For You Here: Finding Opportunity in the Twenty-First Century. Mariner Books. Bibliography 320 Hill, Fiona and Cliord G Gaddy. 2015. Mr. Putin: operative in the Kremlin. Brookings Institution Press. Hirsch, Francine. 2020. Soviet judgment at Nuremberg: a new history of the international military tribunal after World War II. Oxford University Press. Hirschman, Albert O. 1970. Exit, voice, and loyalty: Responses to decline in rms, organizations, and states. Vol. 25 Harvard university press. Homan, David. 2009. The dead hand: the untold story of the cold war arms race and its dangerous legacy. Anchor. Holm, Fynn. 2019. \Japan's Whaling Policy: The Reasons for Leaving the International Whaling Commission." Japan pp. 126{151. Hooghe, Liesbet and Gary Marks. 2005. \Calculation, community and cues: Public opinion on European integration." European Union Politics 6(4):419{443. Hudson, Manley O. 1925. \The registration and publication of treaties." American Journal of International Law 19(2):273{292. Huneeus, Alexandra and Ren e Urue~ na. 2017. \Treaty Exit and Latin Amer- ica's Constitutional Courts." AJIL Unbound 111:456{460. Huntley, Wade L. 2004. \Ostrich Engagement: The bush administration and the North Korea nuclear crisis." The Nonproliferation Review 11(2):81{ 115. Bibliography 321 Jensen, Nathan M and Edmund J Malesky. 2018. \Nonstate actors and compliance with international agreements: An empirical analysis of the OECD anti-bribery convention." International Organization 72(1):33{69. Jianping, Lu and Wang Zhixiang. 2005. \China's Attitude towards the ICC." Journal of International Criminal Justice 3(3):608{620. Jo, Hyeran and Beth A Simmons. 2016. \Can the International Criminal Court Deter Atrocity?" International Organization 70(3):443{475. Johns, Leslie. 2007. \A servant of two masters: communication and the selec- tion of international bureaucrats." International Organization 61(2):245{ 275. Johns, Leslie. 2015. Strengthening international courts: The hidden costs of legalization. University of Michigan Press. Johns, Leslie and B Peter Rosendor. 2009. Dispute settlement, compliance and domestic politics. In Trade disputes and the dispute settlement under- standing of the WTO: An interdisciplinary assessment. Emerald Group Publishing Limited. Johnston, Alastair Iain. 2001. \Treating international institutions as social environments." International studies quarterly 45(4):487{515. Kang, David C. 2010. East Asia before the West: Five centuries of trade and tribute. Columbia University Press. Kasner, Stephen D. 1995. Power politics, institutions and transnational relations. In Bringing Transnational Relations Back In: Non-State Actors, Bibliography 322 Domestic Structures and International Institutions, ed. Thomas Risse- Kappen. Cambridge University Press pp. 257{279. Kaul, Inge, Isabelle Grungberg and Marc A Stern, eds. 1999. Global public goods: International cooperation in the 21st century. New York: Oxford University Press. Kaul, Inge, Pedro Concei c~ ao, Katell Le Goulven and Ronald U Mendoza. 2003. Providing global public goods: managing globalization. New York: Oxford University Press. Keck, Margaret E and Kathryn Sikkink. 1998. Activists beyond borders: advocacy networks in international politics. Cornell University Press. Keohane, Robert O. 1984. After hegemony: Cooperation and discord in the world political economy. Princeton University Press. Kim, Heejin. 2017. \Missed opportunities in the judicialisation of inter- national criminal law? Asian states in the emergence and spread of the Rome Statute system to punish atrocity crimes." Netherlands Quarterly of Human Rights 35(4):246{271. King, Gary and Langche Zeng. 2001. \Explaining rare events in international relations." International Organization 55(3):693{715. King, Gary, Robert O Keohane and Sidney Verba. 1994. Designing social inquiry: Scientic inference in qualitative research. Princeton university press. Bibliography 323 King, Kimi Lynn and Megan Greening. 2007. \Gender justice or just gender? The role of gender in sexual assault decisions at the International Criminal Tribunal for the Former Yugoslavia." Social Science Quarterly 88(5):1049{ 1071. Kinne, Brandon J. 2012. \Multilateral trade and militarized con ict: Cen- trality, openness, and asymmetry in the global trade network." The Jour- nal of Politics 74(1):308{322. Kinne, Brandon J. 2013a. \IGO membership, network convergence, and credible signaling in militarized disputes." Journal of Peace Research 50(6):659{676. Kinne, Brandon J. 2013b. \Network dynamics and the evolution of interna- tional cooperation." American Political Science Review pp. 766{785. Kinne, Brandon J. 2018. \Defense cooperation agreements and the emer- gence of a global security network." International Organization 72(4):799{ 837. Klotz, Audie. 1995. \Norms reconstituting interests: global racial equal- ity and US sanctions against South Africa." International Organization 49(3):451{478. Kobayashi, Lisa. 2005. \Lifting the International Whaling Commission's moratorium on commercial whaling as the most eective global regulation of whaling." Environs: Envtl. L. & Pol'y J. 29:177. Bibliography 324 Koh, Harold Hongju. 1996. \Why do nations obey international law." Yale lj 106:2599. Koh, Harold Hongju. 2018. \Presidential Power to Terminate International Agreements." Yale LJF 128:432. Koh, Harold Hongju, John Nowak, Grover Joseph Rees and Abraham D Sofaer. 1988. \The Treaty Power." U. Miami L. Rev. 43:101. Kolma s, Michal. 2021. \International pressure and Japanese withdrawal from the International Whaling Commission: when shaming fails." Aus- tralian Journal of International Aairs 75(2):197{216. Koremenos, Barbara. 2001. \Loosening the Ties that Bind: A Learning Model of Agreement Flexibility." International organization 55(02):289{ 325. Koremenos, Barbara. 2005. \Contracting around international uncertainty." American Political Science Review 99(04):549{565. Koremenos, Barbara. 2013. \The Continent of International Law." Journal of Con ict Resolution 57(4):653{681. Koremenos, Barbara. 2016. The Continent of International Law. Cambridge Univ Press. Koremenos, Barbara and Allison Nau. 2010. \Exit, No Exit." Duke J. Comp. & Int'l L. 21:81. Bibliography 325 Koremenos, Barbara, Charles Lipson and Duncan Snidal. 2001. \The Ra- tional Design of International Institutions." International Organization 55(04):761{799. Koskenniemi, Martti. 2012. A history of international law histories. In The Oxford handbook of the history of international law. Oxford University Press. Kramer, Alan. 2006. \The rst wave of international war crimes trials: Istanbul and Leipzig." European Review 14(4):441{455. Lahno, Bernd. 1995. \Trust, reputation, and exit in exchange relationships." Journal of con ict resolution 39(3):495{510. Langer, Maximo. 2009. \Trends and Tensions in International Criminal Procedure: A Symposium." UCLA J. Int'l L. Foreign A. 14:1. Lapointe, Joseph M. 2020. \Treaty Termination and the Presidency: Us- ing Custom to Solve Separation of Powers Disputes." Charleston L. Rev. 14:133. Larik, Joris. 2020. \Brexit, the EU-UK Withdrawal Agreement, and global treaty (re-) negotiations." American Journal of International Law 114(3):443{462. Lawal, Olawale. 2012. \Pariah State System and Enforcement Mechanism of International Law." Journal of Alternative Perspectives in the Social Sciences 4(1). Bibliography 326 Lechner, Susanne and Renate Ohr. 2011. \The Right of Withdrawal in the Treaty of Lisbon: A game theoretic re ection on dierent decision processes in the EU." European Journal of Law and Economics 32(3):357{ 375. Leeds, Brett Ashley and Burcu Savun. 2007. \Terminating alliances: Why do states abrogate agreements?" The Journal of Politics 69(4):1118{1132. Levy, Jack S. 2015. \Counterfactuals, causal inference, and historical anal- ysis." Security Studies 24(3):378{402. Lim, Claire SH, Bernardo S Silveira and James M Snyder. 2016. \Do judges' characteristics matter? ethnicity, gender, and partisanship in texas state trial courts." American Law and Economics Review 18(2):302{357. Lipson, Charles. 1991. \Why are some international agreements informal?" International Organization 45(4):495{538. Lupu, Yonatan. 2016. \Why do states join some universal treaties but not others? An analysis of treaty commitment preferences." Journal of Con- ict Resolution 60(7):1219{1250. Lupu, Yonatan and Brian Greenhill. 2017. \The networked peace: Inter- governmental organizations and international con ict." Journal of Peace Research 54(6):833{848. Lupu, Yonatan and Vincent A Traag. 2013. \Trading communities, the networked structure of international relations, and the Kantian peace." Journal of Con ict Resolution 57(6):1011{1042. Bibliography 327 M alksoo, Lauri. 2015. Russian approaches to international law. Oxford University Press, USA. Manseld, Edward D and Helen V Milner. 2012. Votes, vetoes, and the political economy of international trade agreements. Princeton University Press. Manseld, Edward D, Helen V Milner and B Peter Rosendor. 2000. \Free to trade: Democracies, autocracies, and international trade." American political science review 94(2):305{321. Manseld, Edward D, Helen V Milner and B Peter Rosendor. 2002. \Why democracies cooperate more: Electoral control and international trade agreements." International Organization 56(3):477{513. Mantilla, Giovanni. 2020. Lawmaking under Pressure: International Hu- manitarian Law and Internal Armed Con ict. Cornell University Press. Marcus, Harold G. 2002. A history of Ethiopia. Univ of California Press. Marshall, Monty G, Ted R Gurr and Keith Jaggers. 2016. \Polity IV Project: Political Regime Characteristics and Transitions, 1800-2015.". Marshall, Peter D. 2011. \A comparative analysis of the right to appeal." Duke J. Comp. & Int'l L. 22:1. Martin, Lisa. 2013. Against compliance. In Interdisciplinary Perspectives on International Law and International Relations: The State of the Art,, ed. Jerey L. Duno and Mark A. Pollack. Cambridge University Press pp. pp. 591{610. Bibliography 328 Mattes, Michaela, Brett Ashley Leeds and Naoko Matsumura. 2016. \Mea- suring change in source of leader support: The CHISOLS dataset." Jour- nal of Peace Research 53(2):259{267. McLachlan, Campbell. 2019. \The assault on international adjudication and the limits of withdrawal." International & Comparative Law Quarterly 68(3):499{537. McMaster, H.R. 2020. Battlegrounds: The Fight to Defend the Free World. Harper. M egret, Fr ed eric. 2009. \Beyond fairness: understanding the determinants of international criminal procedure." UCLA J. Int'l L. Foreign A. 14:37. Menon, Anand and John-Paul Salter. 2016. \Brexit: initial re ections." International Aairs 92(6):1297{1318. Merryman, John Henry and Rogelio P erez-Perdomo. 2007. The civil law tra- dition: an introduction to the legal systems of Europe and Latin America. Stanford University Press. Milkias, Paulos and Getachew Metaferia. 2005. The Battle of Adwa: Re ec- tions on Ethiopia's historic victory against European colonialism. Algora Publishing. Morelli, Antonio. 2021. The Legal Framework of Withdrawal from Multilat- eral Treaties. Brill Nijho. Bibliography 329 Morin, Jason L. 2014. \The Voting Behavior of Minority Judges in the US Courts of Appeals: Does the Race of the Claimant Matter?" American Politics Research 42(1):34{64. Morrow, James D. 1994. \Modeling the forms of international cooperation: distribution versus information." International Organization 48(3):387{ 423. Morrow, James D. 2007. \When do states follow the laws of war?" American Political Science Review 101(3):559{572. Mulesky, Suzie and Wayne Sandholtz. 2018. \Human Rights Treaty Obli- gations and State Commitment." Working paper. Mulesky, Suzie, Wayne Sandholtz and Kelebogile Zvobgo. 2020. \Do Human Rights Treaty Obligations Matter for Ratication?" Available at SSRN 3383165 . Mullins, Claud. 1921. The Leipzig trials: an account of the war criminals' trials and a study of German mentality. HF & G. Witherby. Nalepa, Monika and Emilia Justyna Powell. 2016. \The Role of Domestic Opposition and International Justice Regimes in Peaceful Transitions of Power." Journal of Con ict Resolution 60(7):1191{1218. Norton, Philip. 1978. Conservative Dissidents: Dissent Within the Parlia- mentary Conservative Party, 1970{74. London: Temple Smith. Oatley, Thomas. 2015. International political economy. Routledge. Bibliography 330 Obama, Barack. 2020. A promised land. Crown. Ohlin, Jens David. 2009. \A meta-theory of international criminal proce- dure: vindicating the rule of law." UCLA J. Int'l L. Foreign A. 14:77. Ohlin, Jens David. 2015. The assault on international law. Oxford University Press, USA. Olson, Mancur. 1965. The logic of collective action. Harvard University Press. O'Rourke, Kevin. 2019. A short history of Brexit: From Brentry to backstop. Penguin UK. O'Toole, Fintan. 2018. Heroic failure: Brexit and the politics of pain. Head of Zeus Ltd. Pauwelyn, Joost and Rebecca J Hamilton. 2018. \Exit from International Tribunals.". Pemstein, Daniel, Kyle L Marquardt, Eitan Tzelgov, Yi-ting Wang, Joshua Krusell and Farhad Miri. 2018. \The V-Dem measurement model: la- tent variable analysis for cross-national and cross-temporal expert-coded data.". Peresie, Jennifer L. 2005. \Female judges matter: Gender and collegial deci- sionmaking in the federal appellate courts." Yale Law Journal pp. 1759{ 1790. Bibliography 331 Peter, Chris Marina. 2016. Fighting impunity: african states and the inter- national criminal court. In The International Criminal Court and Africa: One Decade On, ed. Evelyn A. Ankumah. Intersentia pp. 1{62. Pevehouse, Jon and Bruce Russett. 2006. \Democratic international governmental organizations promote peace." International organization 60(4):969{1000. Pevehouse, Jon C. 2002. \Democracy from the outside-in? International or- ganizations and democratization." International organization 56(3):515{ 549. Pevehouse, Jon CW, Timothy Nordstrom, Roseanne W McManus and Anne Spencer Jamison. 2020. \Tracking organizations in the world: The Correlates of War IGO Version 3.0 datasets." Journal of Peace Research 57(3):492{503. Phelan, Steven E, Richard J Arend and Darryl A Seale. 2005. \Using an iterated Prisoner's Dilemma with exit option to study alliance behavior: Results of a tournament and simulation." Computational & Mathematical Organization Theory 11(4):339{356. Plokhy, Serhii. 2015. The last empire: the nal days of the Soviet Union. Hachette UK. Pollack, Mark A and Gregory C Shaer. 2009. When cooperation fails: The international law and politics of genetically modied foods. Oxford University Press on Demand. Bibliography 332 Preiser, Wolfgang and R Bernhardt. 2008. \History of international law, Ancient Times to 1648." The Max Planck Encyclopedia of Public Inter- national Law 4:884. Pretorius, Joelien and Tom Sauer. 2022. \When is it legitimate to abandon the NPT? Withdrawal as a political tool to move nuclear disarmament forward." Contemporary Security Policy 43(1):161{185. Putnam, Robert D. 1988. \Diplomacy and domestic politics: the logic of two-level games." International organization 42(3):427{460. Ramanathan, Usha. 2005. \India and the ICC." Journal of International Criminal Justice 3(3):627{634. Rapp, Kyle. 2020. \Law and contestation in international negotiations." Review of International Studies 46(5):672{690. Rashid, Farid Mohammed. 2021. Prosecutorial Discretion in the Interna- tional Criminal Court: Legitimacy and the Politics of Justice. Routledge. Rathbun, Brian C. 2008. \Does one right make a realist? Conservatism, neo- conservatism, and isolationism in the foreign policy ideology of American elites." Political Science Quarterly 123(2):271{299. Rathje, Steve, Jay J Van Bavel and Sander van der Linden. 2021. \Out- group animosity drives engagement on social media." Proceedings of the National Academy of Sciences 118(26). Raustiala, Kal. 2005. \Form and substance in international agreements." American journal of international law 99(3):581{614. Bibliography 333 Raustiala, Kal and Anne-Marie Slaughter. 2002. International law, interna- tional relations, and compliance. In The Handbook of International Rela- tions, ed. Thomas Risse Walter Carlnaes and Beth Simmons. Sage Pub- lications, Ltd. Rawls, John. 1971. A Theory of Justice. Oxford University Press. Rawls, John. 1999. A theory of justice. Harvard university press. Ray, Leonard. 2003. \When parties matter: The conditional in uence of party positions on voter opinions about European integration." The Jour- nal of Politics 65(4):978{994. Reed, Thomas and Danny Stillman. 2010. The nuclear express: A political history of the bomb and its proliferation. Zenith press. Reid-Henry, Simon. 2020. Empire of democracy: The remaking of the west since the Cold War. Simon & Schuster. Reus-Smit, Christian. 2003. \Politics and international legal obligation." European Journal of International Relations 9(4):591{625. Rice, Condoleezza. 2011. No higher honor: A memoir of my years in Wash- ington. Crown. Richard, Henry. 1877-1878. \On the Obligation of Treaties." Law Magazine and Review; a Quarterly Review of Jurisprudence, and Quarterly Digest of All Reported Cases 3:91. Roberts, Kari. 2018. The United States. In Routledge Handbook of Russian Foreign Policy. New York: Routledge pp. 237{253. Bibliography 334 Rosendor, B Peter. 2005. \Stability and rigidity: politics and design of the WTO's dispute settlement procedure." American Political Science Review 99(3):389{400. Rosendor, B Peter and Helen V Milner. 2001. \The optimal design of international trade institutions: Uncertainty and escape." International Organization 55(04):829{857. Rubenson, Sven. 1964. \The Protectorate Paragraph of the Wichal e Treaty." The Journal of African History 5(2):243{283. Rumsfeld, Donald. 2011. Known and unknown: a memoir. Penguin. Russett, Bruce, John R Oneal and David R Davis. 1998. \The third leg of the Kantian tripod for peace: International organizations and militarized disputes, 1950{85." International Organization 52(3):441{467. Sadat, Leila Nadya and Benjamin Cohen. 2016. Impunity through immu- nity: the kenya situation and the international criminal court. In The International Criminal Court and Africa: One Decade On, ed. Evelyn A. Ankumah. Intersentia pp. 101{124. Sandholtz, Wayne. 1999. \Globalization and the Evolution of Rules." Glob- alization and governance pp. 77{102. Sandholtz, Wayne. 2008. \Dynamics of international norm change: Rules against wartime plunder." European Journal of International Relations 14(1):101{131. Bibliography 335 Sandholtz, Wayne and Kendall W Stiles. 2009. International norms and cycles of change. Oxford University Press. Sandholtz, Wayne and Mark M Gray. 2003. \International integration and national corruption." International Organization 57(4):761{800. Sarvarian, Arman, F Fontanelli, A Zidar, V Tzevelekos and R Baker. 2015. Procedural Fairness in International Courts and Tribunals. British Insti- tute of International and Comparative Law. Schabas, William A. 2017. An introduction to the international criminal court. 5th ed. Cambridge University Press. Scharf, Michael P. 1998. \Case Analysis: The Prosecutor v. Slavko Dok- manovi c. Irregular Rendition and the ICTY." Leiden Journal of Interna- tional Law 11(2):369{382. Schimmelfennig, Frank. 2005. \Strategic calculation and international so- cialization: Membership incentives, party constellations, and sustained compliance in Central and Eastern Europe." International Organization 59(4):827{860. Schmitt, Clayton J. 2020a. \The Future Is Today: Preparing the Legal Ground for the United States Space Force." U. Miami L. Rev. 74:563. Schmitt, Olivier. 2020b. \How to challenge an international order: Rus- sian diplomatic practices in multilateral security organisations." European Journal of International Relations 26(3):922{946. Bibliography 336 Schmitt, Paul D. 2016. France, Africa, and the ICC: The Neocolonialist Critique and the Crisis of institutional Legitimacy. In Africa and the ICC: Perceptions of Justice, ed. Kamari Clarke, Abel S. Knottnerus and Eefje de Volder. Cambridge University Press chapter 7, pp. 127{151. Schnitzer, David A. 2012. \Into Justice Jackson's Twilight: A Constitutional and Historical Analysis of Treaty Termination." Geo. LJ 101:243. Schuldt, Lasse. 2015. \Southeast Asian Hesitation: ASEAN Countries and the International Criminal Court." German Law Journal 16(1):75{104. Seawright, Jason. 2016. Multi-method social science: Combining qualitative and quantitative tools. Cambridge University Press. Seawright, Jason and John Gerring. 2008. \Case selection techniques in case study research: A menu of qualitative and quantitative options." Political Research Quarterly 61(2):294{308. Secunda, Paul M. 2010. \Cultural Cognition at Work." Fla. St. UL Rev. 38:107. Seymour, Lee J M. 2016. The ICC and Africa: Rhetoric, Hypocrisy Man- agement and Legitimacy. In Africa and the ICC: Perceptions of Justice, ed. Kamari Clarke, Abel S. Knottnerus and Eefje de Volder. Cambridge University Press chapter 6, pp. 107{126. Shannon, Megan, Daniel Morey and Frederick J Boehmke. 2010. \The in- uence of international organizations on militarized dispute initiation and duration." International Studies Quarterly 54(4):1123{1141. Bibliography 337 Shapiro, Martin. 2013. Courts: a comparative and political analysis. Uni- versity of Chicago Press. Shelton, Dinah. 1997. \Compliance with international human rights soft law." Stud. Transnat'l Legal Pol'y 29:119. Shelton, Dinah. 2000. Commitment and compliance: The role of non-binding norms in the international legal system. Oxford University Press on De- mand. Shikhelman, Vera. 2017. \Diversity and Decision-Making in International Judicial Institutions: The United Nations Human Rights Committee as a Case Study." Berkeley Journal of International Law 36. Shipman, Tim. 2016. All out war: the full story of how Brexit sank Britain's political class. William Collins London. Simmons, Beth. 2010. \Treaty compliance and violation." Annual Review of Political Science 13:273{296. Simmons, Beth A. 1998. \Compliance with international agreements." An- nual review of political science 1(1):75{93. Simmons, Beth A. 2000. \International law and state behavior: Commit- ment and compliance in international monetary aairs." American Polit- ical Science Review 94(4):819{835. Simmons, Beth A. 2009. Mobilizing for human rights: international law in domestic politics. Cambridge University Press. Bibliography 338 Simmons, Beth A. 2014. \Bargaining over BITs, arbitrating awards: The regime for protection and promotion of international investment." World Politics 66(1):12{46. Simpson, Gerry. 2004. Great Powers and Outlaw States: Unequal Sovereigns in the International Legal Order. Vol. 32 Cambridge University Press. Singer, J. D. 2011. \National Material Capabilities (v4. 0)." Correlates of War Project . Singer, J David. 1988. \Reconstructing the correlates of war dataset on material capabilities of states, 1816{1985." International Interactions 14(2):115{132. Slaughter, Anne-Marie, Andrew S Tulumello and Stepan Wood. 1998. \In- ternational law and international relations theory: A new generation of interdisciplinary scholarship." American Journal of International Law 92(3):367{397. Snyder, Timothy. 2018. The Road to Unfreedom: Russia, Europe, America. Tim Duggan Books. Sobolewska, Maria and Robert Ford. 2020. Brexitland: Identity, diversity and the reshaping of British politics. Cambridge University Press. Sohn, Louis B. 1964. \Expulsion or forced withdrawal from an international organization." Harvard Law Review pp. 1381{1425. Sotomayor, Sonia. 2002. \A Latina judge's Voice." Berkeley La Raza LJ 13:87. Bibliography 339 Sperfeldt, Christoph and Emma Palmer. 2016. \(Not) All Roads Lead to Rome: Ending Impunity for International Crimes in Southeast Asia." FICHL Policy Brief Series 66. Starr, Sonja B. 2009. \Ensuring Defense Counsel Competence at Interna- tional Criminal Tribunals." UCLA J. Int'l L. Foreign A. 14:169. Stein, Arthur A. 1982. \Coordination and collaboration: regimes in an anarchic world." International organization 36(2):299{324. Stent, Angela E. 2014. The Limits of Partnership: U.S.-Russian Relations in the Twenty-First Century. Princeton University Press. Stewart, James G and Asad Kiyani. 2017. \The Ahistoricism of Legal Plu- ralism in International Criminal Law." The American Journal of Com- parative Law 65(2):393{449. Stone, Randall W. 2008. \The scope of IMF conditionality." International organization 62(4):589{620. Tooze, Adam. 2018. Crashed: How a decade of nancial crises changed the world. Penguin. Trachtman, Joel P. 2014. \The Economic Structure of the Law of Interna- tional Organizations." Chi. J. Int'l L. 15:162. Treier, Shawn and Simon Jackman. 2008. \Democracy as a latent variable." American Journal of Political Science 52(1):201{217. Tridgell, Jennifer. 2018. \The Departed: Implications of the Philippines' Withdrawal from the ICC." Opinio Juris . Bibliography 340 URL: http://opiniojuris.org/2018/04/12/the-departed-implications-of- the-philippines-withdrawal-from-the-icc/ Vabulas, Felicity. July 1, 2016. \Brexit isn't all that special. Here's why nations leave international organizations." The Washington Post: Monkey Cage . URL: https://www.washingtonpost.com/news/monkey- cage/wp/2016/07/01/brexit-isnt-all-that-special-heres-why-nations- leave-international-organizations/ van Aaken, Anne and Bet ul Simsek. 2021. \Rewarding in International Law." American Journal of International Law 115(2):195{241. Verosta, Stephan. 2007. \History of International Law, 1648 to 1815." Max Planck Encyclopaedia of Public International Law . Voeten, Erik. 2001. \Outside options and the logic of Security Council action." American Political Science Review 95(4):845{858. von Borzyskowski, Inken and Felicity Vabulas. 2018. \The Costs of Membership Withdrawal from Intergovernmental Organizations." Polit- ical Economy of International Organizations Conference, University of Wisconsin, Madison. URL: http://wp.peio.me/wp-content/uploads/2018/02/Borzyskowski- and-Vabulas Consequences-of-Withdrawal-paper-53.pdf von Borzyskowski, Inken and Felicity Vabulas. 2019. \Hello, goodbye: When do states withdraw from international organizations?" The Review of International Organizations 14(2):335{366. Bibliography 341 von Borzyskowski, Inken and Felicity Vabulas. 2022. \On IGO withdrawal by states vs leaders, and exogenous measures for inference." The Review of International Organizations 17(1):217{222. Voon, Tania and Andrew D Mitchell. 2017. \Ending International Invest- ment Agreements: Russia's Withdrawal from Participation in the Energy Charter Treaty." AJIL Unbound 111:461{466. Wagner, R Harrison. 2000. \Bargaining and war." American Journal of Political Science pp. 469{484. Waibel, Michael. 2017. \Brexit and Acquired Rights." AJIL Unbound 111:440{444. Weeks, Jessica L. 2008. \Autocratic audience costs: Regime type and sig- naling resolve." International Organization 62(1):35{64. Weiler, Joseph HH. 1985. \Alternatives to Withdrawl from an International Organization: The Case of the European Economic Community." Isr. L. Rev. 20:282. Whiting, Alex. 2009. \Lead evidence and discovery before the International Criminal Court: the Lubanga case." UCLA J. Int'l L. Foreign A. 14:207. Willerton, John P. 2013. Presidency. In Routledge Handbook of Russian Politics and Society. Routledge pp. 91{101. Williams Jr, Robert E and Paul R Viotti. 2012. Arms Control: History, Theory, and Policy [2 volumes]: History, Theory, and Policy. ABC-CLIO. Bibliography 342 Wippman, David. 2004. \The International Criminal Court." CAMBRIDGE STUDIES IN INTERNATIONAL RELATIONS 96:151{188. Woolaver, Hannah. 2017. \Domestic and International Limitations on Treaty Withdrawal: Lessons from South Africa's Attempted Departure from the International Criminal Court." AJIL Unbound 111:450{455. Woolaver, Hannah. 2019. \From joining to leaving: Domestic law's role in the international legal validity of treaty withdrawal." European Journal of International Law 30(1):73{104. Young, Alison L. 2017. \Brexit, Miller, and the Regulation of Treaty With- drawal: One Step Forward, Two Steps Back?" AJIL Unbound 111:434{ 439. Zakerhossein, Mohammad H. 2017. Situation selection regime at the inter- national criminal court: law, policy, practice. Cambridge: Intersentia. Zaller, John R et al. 1992. The nature and origins of mass opinion. Cam- bridge university press. Zvobgo, Kelebogile. 2019. \Human rights versus national interests: Shifting US public attitudes on the international criminal court." International Studies Quarterly 63(4):1065{1078. Zvobgo, Kelebogile, Wayne Sandholtz and Suzie Mulesky. 2020. \Reserv- ing rights: Explaining human rights treaty reservations." International Studies Quarterly 64(4):785{797. Appendices 343 Appendix A Formal Model Analysis and Proofs A.1 Analyzing the Bargaining Model of Coopera- tion, Breach, and Withdrawal The game is solved using backward induction and subgame perfection. To do so, we can analyze each subgame in turn, working back to the start of the game where nature assigns a random value to the shock to the distribution of value to the treaty (e). To start, Player 2 punishes if: Appendices 344 P >:P v 2 +ex(1q)s +(v 2 +ex)>v 2 +ex +(v 2 +e) x(1q)s>x x +xqs> 0 xqs> 0 xq>s x> s q Player 2 punishes if its costs divided by the probability of success, s q , is less the what Player 1 proposes or takes, x. On the extremes, if the probability of success, q, is zero the function is undened. When the probability of success is certain, Player 2 punishes so long as the cost of punishing, s, is less than what it would lose,x. In short, because the probability of successful punishment covers the unit space, [0; 1), the lower the probability of success the more costly punishing becomes. Uncertainty compounds the costs of punishment. If Player 2 does not punish, then Player 1's preferences are as follows: Appendices 345 C =B v 1 e +(v 1 e) =v 1 e +x +(v 1 e) 0 =x W =B v 1 e =v 1 ex +(v 1 e) 0 =x +(v 1 e) (v 1 e) =x W =C (v 1 e) = 0 v 1 +e = 0 e =v 1 e =v 1 All else being equal, Player 1 would choose to withdraw rather than remain and cooperate under the status quo if the shock is greater than the value of the treaty to Player 1,e>v 1 . When the shock does not overcome the value of the treaty, Player 1 is either indierent or prefers remaining in the treaty. Appendices 346 Player 1 breaches where what it can obtain from cheating,x, is greater than 0, all else being equal. However, if the shock is greater than the value of the treaty to Player 1,e>v 1 , the value of cheating must exceed the discounted balance of the treaty value to Player 1 less than the shock,(v 1 e). Put another way, as the magnitude of the shock increases above the value of the treaty, the greater the value from cheating needs to be for breaching to be protable. Therefore, a large enough shock makes breaching unattractive and a state would simply favor leaving the treaty. If Player 2 does punish, then Player 1's preferences are as follows: Appendices 347 W >B v 1 e>v 1 e +x(1q)k(q) +(v 1 e) 0>xxqkq +(v 1 e) kq(v 1 e)>xxq kq(v 1 e)>x(1q) kq(v 1 e) (1q) >x C >B v 1 e +(v 1 e)>v 1 e +x(1q)k(q) +(v 1 e) 0>x(1q)k(q) kq>x(1q) kq (1q) >x W =C kq(v 1 e) (1q) = kq (1q) kq(v 1 e) =kq (v 1 e) = 0 v 1 +e = 0 e =v 1 e =v 1 Appendices 348 When Player 2 punishes, Player 1 preferences as to whether to breach or not are dependent on the likelihood of suering a sanction from Player 2. Like before, if the shock is greater that the value of the treaty to Player 1, e>v 1 , Player 1 would prefer to withdraw compared to continue to cooper- ate. Player 1 breaches if the amount it can take is greater than the likely cost of punishment, kq (1q) , or the likely cost of punishment and cost of future participation in the treaty kq(v 1 e) (1q) . As the probability of successful pun- ishment increases, q, the greater the costs to Player 1. In such a situation, Player 1 has less exibility to unilaterally adjust by taking what it needs to adjust (x) to the shock (e). Where the punishment is severe, largek, there is a high probability of punishment, largeq, or the shock is signicantly large, large e, Player 1 would prefer to withdraw than to breach. Player 2 accepts the proposal under the following conditions: Appendices 349 A>RjB(:P ) v 2 +ex +(v 2 +ex)>v 2 +ex +(v 2 +e) v 2 +ex>v 2 +e x> 0 x> 0 A>RjB(P ) v 2 +ex +(v 2 +ex)>v 2 +ex(1q)s +(v 2 +e) x +v 2 +ex>x +xqs +v 2 +e x>xqs s>xq +x s>x(q +) s q + >x A>RjW v 2 +ex +(v 2 +ex)>v 2 +e +(v 2 +eI) x +(v 2 +ex)>(v 2 +eI) x +v 2 +ex>v 2 +e +I xx>I x(1 +)>I x> I (1 +) xRjC v 2 +ex +(v 2 +ex)>v 2 +ex +(v 2 +e) x +v 2 +ex>v 2 +e xx> 0 x(1 +)> 0 x> 0 x< 0 In light of Player 1 withdrawing, Player 2 accepts Player 1's proposal if the proposal is greater than Player 1's importance to the treaty relationship, x > I (1+) . That importance diminishes into the future, (1+) , which cor- responds to our intuition that a player's importance may be high now but is less so in the future where adjustments can be made to account for that importance. This dynamic highlights the states' immediate preferences. Ad- ditionally, as Player 1's importance, I, increases, the greater the proposal, x, must be for Player 2 to accept it. If Player 1 will cooperate, it is unlikely that Player 2 will reject a proposal but stay with the status quo because x will not be negative. It would only be plausible when x is 0. In other words, if Player 1 oers something to Player 2, it is better to accept than reject and live with the status quo. Similarly, when Player 1 breaches without Player 2 punishing, Player 2 will accept any proposal greater than zero. Appendices 351 A.1.1 Making a Deal or Exit In this game, Player 2 accepts Player 1's proposal, x, as long as Player 1 proposes something, x> 0, and the proposal is not greater than Player 1's discounted importance, x < I (1+) , and the costs of uncertain punishment to Player 2, s q+ >x. As punishment becomes more certain, q! 1, the less room there is for a bargain because s q+ decreases as q increases. Player 1 makes a proposal, x, that is greater than the discounted value of the treaty with the shock,x> (ev 2 ) 1+ , and greater than the cost of being punished, x> kq (+q) . Because, as dened, any positive oer will be greater than the cost of being punished, the second condition will be satised. Thus, the parties will reach a bargain as long as Player 1 proposes a shift in costs that accounts for the future costs of the shock and the importance of Player 1 to the treaty and costs of punishing a breach. When the shock is greater than the value of the treaty to Player 1, e>v 1 , Player 1 will need to make a proposal to account for the shock. However, for Player 1 to be satised with a new deal, the proposal will exceed Player 1's importance to the treaty and Player 2's costs to successfully punish a breach. Appendices 352 W 0. In this case, Player 1 cooperates (C) if q> d (d+k) , breaches (B) if q< d (d+k) and is indierent otherwise. If punishment is certain to be eective, (q = 1), Player 2 will always punish (P ) and Player 1 will cooperate (C), as long as k> 0. Player 1 will be indierent if k = 0 because punishment will have no sting. Whenq = 0, Player 2 is indierent between punishing (P ) and not punishing (:P ). Player 1 will breach (B) as long as d> 0. If d = 0, then Player 1 is Appendices 358 indierent. Thus, to keep the games interesting, it is assumed that d > 0 and k> 0. Here, the model illustrates how the certainty that the punishment is eective is crucial in Player 2 being able to deter Player 1 from breaching the treaty. Player 2 desires that the probability that punishment is eective (q) is suf- ciently high, q > d (d+k) , that Player 1 always chooses to cooperate. This could be accomplished by increasing detection or targeting the punishment. Relatedly, Player 2 can achieve the same deterrence eect if k is suciently high becauseq is decreasing ink, i.e., ask increases,q decreases. This makes sense intuitively where the more costly the punishment, the less certain its eectiveness needs to be to prevent Player 1 from breaching. Conversely, the more valuable to the extra benet of breaching d, the harder it will be for Player 2 to deter Player 1 from breaching. This model shows the relationship between certainty and the benets and costs of breach and punishment. However, Player 2 almost always punishes as long a punishment has some probability of success. This is likely because punishment in this model is costless. Thus, in the next model, a cost to punishment is added to investigate how it aects the equilibria. Appendices 359 A.4 Model Version 2: Uncertainty with Costly Punishment This version of the game introduces a new parameter to represent the cost of punishment (s). If Player 2 chooses to punish, it incurs this cost (s). This cost can be thought of as the simply administrative or transaction cost of instituting some punishment. A state that punishes must allocate some of its nite resources to implementing a punishment. This parameter may also capture collateral costs in implementing the punishment. For example, it might be the case that a punishment is some economic sanction that also hurts the punishing state's economic activity. Figure A.3 is the game tree for the variation of the game focusing on uncertainty with costly punishment (s). Figure A.3: Uncertainty with Costly Punishment P1 (b 1 c 1 ;b 2 c 2 ) C P2 (b 1 c 1 +d(1q)k(q);b 2 c 2 d(1q)s) P (b 1 c 1 +d;b 2 c 2 d) :P B Player 2 punishes if q > s d , does not punish if q < s d , and is indierent otherwise. Assuming q > s d and Player 2 punishes, Player 1 will breach if q < d (d+k) and cooperate if q > d (d+k) . If Player 2 does not punish, q < s d , then Player 1 breaches as long as d> 0. Appendices 360 Adding the cost of punishment adds new equilibria to the game. Particularly, Player 1 breaches and Player 2 does not punish as long as the probability of eective punishment is less than the cost of punishment (s) divided by the defection benet (d). Thus, the probability of eectiveness is increasing in the cost of punishment and decreasing in the defection benet. If pun- ishment is particularly costly, then the probability of success must be high enough to warrant the punishment. Otherwise, Player 2 is better o letting Player 1 breach and not incurring the cost of punishment. Additionally, if the defection benet, or the value Player 2 stands to lose (sucker's payo), is high, Player 2 would be willing to punish and incur the cost with a lower probability of success. A.5 Model Version 3: Removing Uncertainty from Costly Punishment The third variation looks at the game without any uncertainty, removing q, and with costly punishment. Here, if Player 2 chooses to punish (P ) it is always eective. Figure A.4 is the game tree for the variation of the game removing uncertainty from costly punishment. Without the uncertainty, the game boils down to whether punishment is costly (s) and then a simple balancing between d and k. As long as pun- ishment is costly, Player 2 will not punish and Player 1 will breach if d 0 and cooperate if d = 0. As long as breaching as some benet, d > 0, then Appendices 361 Figure A.4: Removing Uncertainty from Costly Punishment P1 (b 1 c 1 ;b 2 c 2 ) C P2 (b 1 c 1 +dk;b 2 c 2 ds) P (b 1 c 1 +d;b 2 c 2 d) :P B Player 1 will breach and Player 2 will not punish because it is too costly to do so. A.6 Conclusions The second variation game has the most interesting equilibria, including equilibrium where breach occurs without punishment. This would not hap- pen unless punishment had zero chance of success. This also conrms that when some of the parameters are zero, then the games are less interesting. Breach without punishment intuitively corresponds with understanding of how states work in the international system. If the probability that a costly punishment will be successful is low enough, or the cost of punishment high enough, then punishment would not be rational. Without punish- ment, breach is the best action. Breach remains the best action until the probability that punishment is successful increases or the strength of the punishment increases. These dynamics comport with intuitions and may be useful in formulating the broader theory regarding institution exibility. Appendices 362 Proofs Variation 1 If q> 0, Player 2 will always punish (P ). Thus, Player 1 chooses to breach (B) over cooperate (C) based on the following: u 1B;P =b 1 c 1 +d(1q)k(q)>b 1 c 1 =u 1C d(1q)k(q)> 0 ddqkq> 0 dqkq>d q(d +k)>d q> d (d +k) q< d (d +k) Player 1 will choose to cooperate when q > d (d+k) and is indierent when q = d (d+k) , following similar simplication. Therefore, out of the four possible strategy sets, two (C;P ) and (B;P ) are equilibria given whetherq is greater than, less than, or equal to d (d+k) . Appendices 363 C;P if q> d (d+k) C;:P Not SPNE B;P if q< d (d+k) B;:P Not SPNE If q = 1, Player 2 always punishes and Player 1 always cooperates as long as k > 0, as shown below. If k = 0, then Player 1 is indierent between its choices. u 1B;P =b 1 c 1 +d(1q)k(q) 0 Ifq = 0, Player 2 is indierent between punish (P ) and not punish (:P ). In that case, Player 1 will breach as long as either potential outcome is greater than cooperate (C). Appendices 364 u 1B;P =b 1 c 1 +d(1q)k(q)>b 1 c 1 =u 1C b 1 c 1 +d(1)k(0)>b 1 c 1 b 1 c 1 +d>b 1 c 1 d> 0 u 1B;:P =b 1 c 1 +d>b 1 c 1 =u 1C d> 0 If d = 0, Player 1 is indierent between cooperate (C) and breach (B). Variation 2 Player 2 will punish (P ) based on the following: Appendices 365 u 2P =b 2 c 2 +d(1q)s>b 2 c 2 d =u 2:P d(1q)s>d d +dqs>d dqs> 0 dq>s q> s d Player 2 will not punish if q < s d . Thus, the cut point for Player 2 is q 2 s d >q. Assuming q< s d ,Player 2 punishes and Player 1 will breach if: u 1B;P =b 1 c 1 +d(1q)k(q)>b 1 c 1 =u 1C d(1q)k(q)> 0 ddqkq> 0 q(d +k)>d q> d (d +k) q< d (d +k) Appendices 366 Therefore, the cut point for Player 1 given q < s d is q 1 d (d+k) > q. This corresponds to the same cut point in the rst version of the model. Assuming q> s d , Player 2 does not punish and Player 1 will breach if: u 1B;:P =b 1 c 1 +d>b 1 c 1 =u 1C d> 0 With these cut points dened, the equilibria are based on the ordering of these cut points, i.e., q 1 > q 2 or q 2 > q 1 . Figure A.5 shows the equilibria graphically with the two cut points. Figure A.5: Equilibria Diagram q2 q1 ¬P ¬P P P C B C q1 C B q2 B P2 P1 (P) P1 (¬P) P2 P1 (P) P1 (¬P) Appendices 367 Being that the cost of punishment is crucial here, we solve for s given the two possible orderings. q 1 > q 2 d (d +k) > s d d 2 (d +k) >s q 2 > q 1 s d > d (d +k) s> d 2 (d +k) Whens = d 2 (d+k) , the cut points overlap. Figure A.5 shows that it is possible that Player 1 would cooperate when Player 2 failed to punish. However, this would only be true if d 0, which by denition d could only be as small as 0. Variation 3 Player 2 will punish (P ) based on the following: Appendices 368 u 2P =b 2 c 2 +ds>b 2 c 2 d =u 2:P ds>d s> 0 s< 0 However, by denition, s> 0. Therefore, Player 2 will prefer to not punish as long as s > 0. If that is the case, s > 0, Player 1 choses breach as long as: u 1B;P =b 1 c 1 +d>b 1 c 1 =u 1C d> 0 The equilibrium here would be (B;:P ) if d 0 and (C;:P ) if d = 0. C;P s 0 C;:P SPNE, d = 0 B;P s 0 B;:P SPNE, d 0 Appendices 369 If s = 0, then Player 2 is indierent. If Player 2 punishes, Player breaches as long as: u 1B;P =b 1 c 1 +dk>b 1 c 1 =u 1C dk> 0 k>d k<d In this scenario, the equilibria include the following: C;P k>d C;:P d = 0;k 0 B;P k<d B;:P d 0 Appendices 370 Appendix B Case Vignettes B.1 Treaty of Wuchale, Ethiopia (1893) After bargaining over an interpretation of the Treaty of Wuchale failed, King Menelik II of Shewa (and later Emperor of Ethiopia) denounced the treaty in 1893 (Milkias and Metaferia 2005, 25). This lead the Kingdom of Italy to attempt to forcibly impose a protectorate over Ethiopia in the First Italo-Ethiopian War. Menelik was a consequential ruler of Ethiopia (1889{1914), who expanded the country's territory, built up the country's military, and engaged with European powers with imperial ambitions in the region (Adejumobi 2007, 28{32). Menelik equipped his military using arms purchased through the international arms trade. The British eventually opposed the arms trade Appendices 371 between Europe and Ethiopia because it had imperial ambitions in north- east Africa and it sought to prevent regional powers from increasing their strength. Italy entered into a number of \friendship and commerce" treaties with Menelik between 1883 and 1887. However, disagreement between the par- ties arose regarding the interpretation of the agreements. Italy saw the agreements as useful in the politics of the region where Menelik saw the agreement as a means of obtaining arms and ammunition. In disputes be- tween the countries, Italy often resorted to accusations that Ethiopia had breached rst, despite their accuracy or lack of consideration for Italian provocations Marcus (2002). 1 In 1889, Menelik concluded the Treaty of Wichale with Count Pietro An- tonelli of Italy, which provide: formal recognition of Menelik as emperor of Abyssinia, duty-free privilege for goods passing through the port of Mas- sawa, a loan, and a promise of future arms and military supplies. Por- tions of Ethiopia were ceded for Italian uses in exchange for nancial and development assistance. Overtime, Italy's activities began to expand into Abyssianian territories. Sometime later, Menelik discovered that the treaty he signed in both Italian and Amharic contained dierent clauses regarding 1 Italians accused Ethiopia of aggression regarding the defeat of sieging forces of a small post designated as within Ethiopian territory by the Treaty of Adwa or the Hewett Treaty (1884). Decades later at the League of Nations, Italy accused Ethiopia of numerous treaty violations in the Welwel (Walwal) incident in 1935 Donaldson (2020). See also LNOJ 16 (1935): 1355fr. The tepid response from the League lead to the Italian invasion and the Second Italo-Ethiopian War, which saw Italy commit war crimes in Ethiopia. After the League's condemnation of the war and vote for weak sanctions against Italy, Italy withdrew from the League in 1937. Appendices 372 the disposition of diplomatic relations between the two states. He learned this upon receiving surprising and insulting responses from European leaders like Queen Victoria and Kaiser Wilhem revealing Italy's claim that Ethiopia was now a protectorate (Marcus 2002, 92). Article 17 of the Amharic ver- sion stated that Ethiopia could use Italy as an agent for foreign relations; the Italian version mandated that Ethiopia go through Italy. Antonelli pur- posefully created the discrepancy to secure concessions for Italy without resistance from Ethiopia Rubenson (1964). This further the Italian ambi- tion for a \second Roman empire," as espoused by Prime Minister Francesco Crispi. In August 1890, Menelik appealed to King Umberto I to correct the treaty error (Marcus 2002, 92{93). Menelik also sent letters to other powers ex- plaining the Ethiopian position. Italy also lobbied in Europe, especially with France which was an impediment to Italian plans. France repeatedly refused to recognize the Treaty of Wichale. The British sided with the Ital- ians, largely as part of its competition with France in the region. Menelik and Count Augusto Salimbeni, the Italian Resident, engaged in numerous conversations in December of that year. Menelik insisted that Italy renounce its position that Ethiopia was under an Italian protectorate. Italy argued that its national honor would be harmed if is conceded, angering Ethiopian leaders who saw the treaty as a trick to have them relinquish sovereignty. On February 27, 1893, Menelik rejected the Italian declaration, denounced Appendices 373 the treaty, and repaid the Italian loan (Milkias and Metaferia 2005, 25). 2 Italy, angered by the acts of a \Black African barbarian chieftain," pre- pared for war to save face. This breakdown in relations lead to the Italo- Ethiopian War. Ethiopia was able to muster a sizable force against the Italians, with the rst skirmish occurring in late 1895 and culminating with the Italian defeat at the Battle of Adwa on March 1, 1896. In signing the Peace Treaty of Addis Ababa, Italy renounced the Treaty of Wichale and recognized Ethiopia's absolute independence. The Ethiopian victory opened the door to an alliance with France. This set the stage for Ethiopia's accep- tance into the \family of nations" Donaldson (2020). B.2 Unraveling of the League of Nations (1925{ 1942) After the defeat of the Central Powers, 3 the League of Nations was created at the 1919 Paris Peace Conference by agreement in the Treaty of Versailles, which entered into force January 10, 1920. The League's mandate was to promote peace and settle disputes among the nations without violence. How- ever, for numerous reasons, the organization proved ineective in preventing con ict leading to another world war. It reached the apex of its membership (by number of countries) in 1934 when Ecuador joined until early 1935 when Paraguay withdrew its membership. 2 Russia also denounced the Treaty of Wuchale in 1894, supporting Ethiopia with weapons and ammunition (Milkias and Metaferia 2005, 26). 3 German Empire, Austria-Hungary, Ottoman Empire, and Kingdom of Bulgaria Appendices 374 Importantly, the US never joined the League, despite the organization being championed by US President Woodrow Wilson. Also, the Soviet Union was a member but later expelled for invading Finland in 1939. Argentina (founding member) withdrew in 1921 based on a rejection of an Argentine resolution that all sovereign states be given membership. It re- joined in 1933. Costa Rica was the rst member to withdraw permanently from the League in 1925. Brazil, a founding member, withdrew the follow- ing year (1926). A number of Latin American countries withdrew in the 1930s: Paraguay in 1935; Honduras, Guatemala, and Nicaragua in 1936; El Salvador in 1937; Chile and Venezuela in 1938; and Peru in 1939. Japan infamously withdrew from the League in 1933 over the League's criti- cism of is activities in Manchuria. Nazi Germany withdrew shortly thereafter (1933). Italy withdrew in 1937 after in light of its invasion of Ethiopia pre- cipitating the Second Italo-Ethiopian War. Spain and Hungary withdrew in 1939, and Romania followed in 1940. Vichy France withdrew in 1941; however, it was not recognized. In 1942, Haiti was the last member to withdraw before the organization was dissolved. The League was dissolved on April 18, 1946 and its assets and responsibilities transferred to the United Nations. Appendices 375 B.3 American Convention on Human Rights (ACHR), Trinidad & Tobago (1998) and Venezuela (2012) The Inter-American Conference on Human Rights was held in San Jose, Costa Rica in November 1969. 4 At the conference, the member states of the Organization of American States (OAS) drafted the American Convention on Human Rights (ACHR), which entered into force on July 18, 1978. The ACHR created two bodies to address alleged violations of human rights: the Inter-American Commission on Human Rights and the Inter-American Court of Human Rights. Article 78 of the ACHR allows for a state to withdraw from the agreement, eective one year after notice of withdrawal to the Secretary General of the OAS. Withdrawal under article 78 does not release a state party from its obligations resulting from acts that occurred prior to the eective date of denunication. The treaty contains 82 articles that seek to bring the mem- ber nation's domestic laws into framework of democratic institutions and a system of personal liberty and social justice. (Preamble.) Member states are obligated to \undertake" to respect the rights and freedoms of all under their jurisdiction and to adopt domestic law necessary to aect those rights and freedoms. (Arts. 1{2.) The language of the individual rights uses the stronger term \shall" in staking out the rights the members must respect. The treaty also details the competence, duties, and power of the Inter- American Commission on Human Rights and the Inter-American Court of 4 \Informaci on sobre La Corte Interamericana de Derechos Humanos" http://www.minrel.gov.cl/prontus minrel/site/artic/20090422/pags/20090422085224.php Appendices 376 Human Rights. (Arts. 34{69.) State parties must submit to the jurisdic- tion of the court, either generally or in a specic case (art. 62), signicantly raising the cost of remaining a member. On May 26, 1998, Trinidad and Tobago formally denounced the ACHR under article 78, starting the one-year period for it to withdraw from the institution. 5 Trinidad and Tobago's notice of denunciation and withdrawal was over the issue of the death penalty. The issue was prompted by the eect of the ruling in Pratt and Morgan v. Attorney General for Jamaica (2.A.C.1, 1994) setting strict guidelines related to appeals in capital cases. The primary issue was that appeals in capital cases must be expedited so as not to create undue delay resulting in \inhuman or degrading punishment or other treatment." Trinidad and Tobago argued that this principle was en- shrined in section 592 of its domestic constitution. The principle addressed in the case would include appeals to bodies like the Inter-American Com- mission on Human Rights, which could not guarantee that the cases could be completed in the timeframe sought. Trinidad and Tobago stated that it had tried to work with the Inter-American Commission on Human Rights to resolve the matter but was unable to receive the necessary assurances from the Commission. Thus, rather than contravene its own constitution, it chose to withdraw. At the time, Trinidad and Tobago held a Polity score of nine, was a member of the WTO, and had a high standard of living. The World Bank estimated 5 American Convention on Human Rights \Pact of San Jose, Costa Rica": Trinidad and Tobago, available at https://www.oas.org/dil/treaties B- 32 American Convention on Human Rights sign.htm Appendices 377 that in 1998, trade made up 102.33 percent of its GDP, making it highly dependent on other states. Trinidad and Tobago is a small island nation, neither a great power or pariah. Its Polity score had only slightly become more democratic from the time it entered the institution. The shock for Trinidad and Tobago was arguably the decision in Pratt and Morgan v. At- torney General for Jamaica (2.A.C.1, 1994). Trinidad and Tobago's private benets likely decreased, as the institution became less useful and its status quo costs signicantly increased because of the con ict with its constitu- tion. For Trinidad and Tobago, the institution was not suciently exible and imposed too high a cost. Thus, hypotheses one and two seem plausible in this case. The case does not quite speak to the other hypotheses. On September 10, 2012, Venezuela communicated its denunciation of the ACHR to the Secretary General of the OAS under article 78. 6 Venezuela's argument for withdrawal, albeit not completely clear, was its accusation that the Inter-American Court and Commission were undermining in its domestic aairs to destabilize the government. In 2012, Venezuela's Polity score was negative three and trade made up 50.4 percent of its GDP according to the World Bank{much of that coming from its export of oil and petroleum. It was not a great power, but given Hugo Chavez's aggressive stance agains the U.S. and support of states like Cuba, it sat close to the fringe of the international community. 6 American Convention on Human Rights \Pact of San Jose, Costa Rica": Venezuela, available at https://www.oas.org/dil/treaties B- 32 American Convention on Human Rights sign.htm The 42-page English text of the denunciation is also available on the website as a PDF document. Appendices 378 It is dicult to narrow in on what shock prompted Venezuela to withdrawal from the ACHR. Shortly after the withdrawal, President Hugo Chavez was reelected; however, he died months later. At the time, the country was also suering from severe economic problems, leading to political instabil- ity. Thus, although somewhat unclear, Venezuela's costs in the risk of being sanctioned under the institution likely increased as well as its nancial inabil- ity to continue to support the institution. Additionally, Venezuela's Polity score moved signicantly from the time of entry into the institution from a nine or democratic to negative three or slightly autocratic. This change may lead support to the change in leadership or preference constraints hy- potheses. It is perhaps relevant that these two countries are neighbors and may share similar views of the ACHR. There could be a spill over eect, but it is unclear whether there is any evidence to support this position. B.4 Collective Security Treaty Organization (CSTO), Uzbekistan (1999, 2012) The Collective Security Treaty Organization (CSTO), also referred to as the \Tashkent Pact," was created in 1992 after the the collapse of the Soviet Union in 1991. The CSTO provided for collective security of the former- Soviet states in central Asia. Uzbekistan had previously suspended its mem- bership to the group from 1999 to 2006. Azerbaijan and Georgia withdrew their membership from the CSTO in 1999. This was a result of the initial ve-year life of the treaty, which these three countries declined to extend. Appendices 379 Uzbekistan's Polity score is substantially low, providing for an opportunity to investigate how a state with an arguably small \winning coalition" ap- proaches the club good provided by a collective security agreement. Uzbek- istan is neither a great power or a pariah per se; however, its form of gov- ernment and alleged human rights abuses push it toward to periphery of the international community. Uzbekistan is not a current member of the WTO, but is in the process of becoming a member{again signaling its shift toward moving closer to the international community. The WTO estimates the trade made up 27.9 percent of Uzbekistan's GDP from 2014{2016. The CSTO Charter consists of 29 articles, with a goal toward collective security, harmonizing law enforcement, and coordinating foreign policy (arts. 7-10). The Charter of the CSTO allows states to withdraw with six months notice \[a]fter settling its obligations within the Organization." (Art. 19.) The English version 7 of the Charter uses \shall" 77 times, often referring to duties and obligations of the Organization. The Charter also spells out the funding obligations of the member states to support the Organization. (Art. 25.). Member states are also required to steps to harmonize their national legislation with the treaty obligations. (Art. 10.) The treaty appears to be relatively exible, with signicant costs of membership. Although ocially ambiguous, it was reported that Uzbekistan's 2012 deci- sion to withdrawal from the CSTO was a move away from Russia into the 7 The ocial and working language of the Organization is Russian. (Art. 28.) Thus, there may be a dierence in the obligatory language in Russian. Appendices 380 orbit of the U.S. 8 . Additionally, Russia obtained the right to veto the estab- lishment of new foreign military bases in the member states on December 21, 2011. 9 Given Uzbekistan's desire to shift its orientation to the West signals that its vale of the private security benets of the CSTO decreased, while the public benets and costs remained largely the same. Uzbekistan's costs of remaining in the treaty organization may have gone up based on action of the other members, namely Russia. However, evidence of these shifts has yet to be uncovered. 8 James Kilner. \Uzbekistan withdraws from Russia-lead military alliance" The Tele- graph (July 2, 2002). 9 Vladimir Radyuhin. \CSTO tightens foreign base norms" The Hindu (Dec. 22, 2011). Appendices 381 Appendix C Data Appendix C.1 Data Collection C.1.1 Collecting Data from UN Treaty Series The quantitative data used in this project is built on data collected from the United Nations Treaty Series (UNTS). 1 This data was generated in two phases. In the rst phase, all treaties that include a registered action of \withdrawal," \denunciation," or \termination" were collected. To do this, an \action" search 2 was run for actions containing the English words \withdrawal," \denunciation," or \termination." This was an iterative and lengthy process given the UNTS's unique search function and website design. 1 The UNTS is a publication produced by the Secretariat of the United Nations con- taining all treaties and international agreements registered or led and recorded by the Secretariat since 1945, pursuant to Article 102 of the Charter. 2 Using the \Advanced Search" webpage on the UNTS website. Appendices 382 Because the raw data included actions that were not likely relevant to the project, ltering was necessary. C.1.2 Filtering Actions The raw data included 466 variations of actions collected based on a search of registered actions \withdrawal," \denunciation," or \termination." Some of the various action titles are simply misspellings and incomplete phrases. Some of the actions are full descriptions of the actions. For example, \with- drawal from agreement," \termination of agreement between the United States and Cyprus." However, some of the actions are not pro-active actions by participants and should not be included in the analysis. An example being \withdrawal of declaration" or \withdrawal of objection." To lter actions, a list of all unique actions was generated. That list was then hand- coded as to whether the action should be included in the data using \Y" for include, \N" for do not include, and \?" if ambiguous. Using this sim- ple coding scheme, actions could be ltered in multiple ways. For the most restrictive ltering, only the actions coded as \Y" would be included. For a more expansive or exible ltering, both \Y" and \?" are included, or put another way, only the actions coded \N" are excluded. The coding of the actions was based on an assessment as to whether the action captured a decision by the participant to end its participation in the treaty. This sometimes required looking up the specic action to deter- mine whether it was a pro-active decision by the state to exit the treaty. Appendices 383 This proved more dicult for \termination" actions where treaties may be terminated for reasons other than a state's contemporary decision to exit. Terminations of only portions of an agreement, e.g. a single article, were coded as \N" because that is more akin to a renegotiation of treaty terms than an exit of the treaty regime. Partial terminations were coded as \?" because it was sometimes uncertain if partial termination referred to partial termination of some part of the agreement or partial termination such that some participants were no longer parties. On the whole, it appears many terminations were specic to relationships between two states, which were coded as \Y". Termination as to territorial application was coded as \?" because, again, it was unclear if the state was completely exiting a treaty or simply limiting applicability|potentially through a bargain with other members. The ltering process produced a dataset of 398 treaties that had at least one exit. C.2 Sampling Following King and Zeng (2001), I sampled concluded treaties with no exits. As shorthand, these are called the \no-exit" treaties. In line with King and Zeng (2001), double the number of treaties with exits (ones) is 796. To avoid the eects of likely attrition in the number of treaties to be used in the analysis (below), I sample 1; 102 \no-exit" treaties, giving a buer of 306 treaties of what is needed. Appendices 384 To determine whether the sample was representative, I examined a num- ber of dierent metrics. First, the relationship of bilateral to multi-lateral treaties appears to be largely in line with the broad list of all treaties con- cluded and registered with the UNTS, as shown in Table C.1. Table C.1: Types of Treaties in No Exit Sample Treaty Type Percent of Total Bilateral 0:95 Closed Multilateral 0:03 Open Multilateral 0:03 Figure C.1: Distribution of Treaty Conclusion Dates in Sample of No- Exit Treaties Across years, the sample has a similar distribution of the number of treaties concluded each year, as shown in gure C.1. The year with the most treaties concluded is 1997 in the sample, as compared to 1994 in the wider data. It Appendices 385 is heartening to see that the greatest number of treaties concluded is in the mid- to late-1990s. C.3 Text Analysis Table C.2 shows the dictionary used in coding the treaties as obligatory or permissive. Figure C.2 plots a truncated histogram of the permissive-obligatory scores across the 621 treaties that included a score. The scores are right skewed with most scores being positive between 0 and 100. The plot is truncated because some extreme values made the plot uninformative. Figure C.2: Distribution of Permissive-Obligatory Word Scores Appendices 386 Table C.2: Dictionary Word Category 1 shall obligatory 2 may permissive 3 oblige obligatory 4 must obligatory 5 can permissive 6 undertake permissive 7 require obligatory 8 prompt permissive 9 endeavor permissive 10 consensus permissive 11 report obligatory 12 verify obligatory 13 provide permissive 14 furnish permissive 15 collect permissive 16 conform permissive 17 establish permissive 18 maintain permissive 19 regulate obligatory 20 observe permissive 21 objective permissive 22 aspiration permissive 23 aspire permissive 24 best permissive 25 compulsory obligatory 26 principles permissive 27 rules obligatory 28 norms permissive 29 guidelines permissive 30 regulations obligatory 31 permit permissive 32 duty obligatory 33 penalty obligatory 34 impose obligatory 35 sanction obligatory 36 administer permissive 37 punish obligatory 38 regulate obligatory 39 institute permissive 40 continue permissive Appendices 387 Using the dataset made available from Koremenos (2013), 3 we can investi- gation whether the permissive-obligatory word score is correlated with any cooperation problems. Figure C.3 plots the pairwise correlations between the permissive-obligatory word score, length (words), and eight cooperation problems as coded in the COIL dataset. Analyzing texts from 96 treaties in the dataset reveals the permissive-obligatory word score is negatively cor- related with uncertainty about the world, which is sensible where states would want more permissive language to address that uncertainty at a later time. The permissive-obligatory score is positively correlated with uncer- tainty about behavior, uncertainty about preferences, distribution problems, commitment problems, and, to a lesser extent, enforcement problems. Figure C.3: Correlation Plot of Cooperation Problems and Permissive- Obligatory Word Scores 3 The more expansive data in Koremenos (2016) was not available at the time of this analysis. Appendices 388 C.4 Analysis As explained in Chapter 3, a number of statistical models were applied in assessing the relationship between the key variables of interest. Below tables C.3 and C.4 present additional results for the other models. C.5 Case Study Treaties Table C.5 summarizes select variables from the text of the treaties described in the case studies in Chapters 4, 5, and 6. Overall, the arms control treaties were shorter and much more permissive than the other treaties. Except for the ABM Treaty, all the other treaties have higher permissive-obligatory scores than the average (see Chapter 3). Appendices 389 Table C.3: Treaty Attributes Aect on Number of Treaty Exits Dependent variable: Number of Exits Poisson glm: quasipoisson negative zero-in ated Poisson glm: quasipoisson negative zero-in ated link = log binomial count data link = log binomial count data (1) (2) (3) (4) (5) (6) (7) (8) Obliagtion Score 0.002 0.002 0.0001 0.002 (0.001) (0.002) (0.002) Length (Words) 0.00002 0.00002 0.00002 0.00001 (0.00001) (0.00002) (0.00002) Multilateral 3.440 3.440 3.133 2.497 3.673 3.673 3.428 2.878 (0.153) (0.315) (0.206) (0.121) (0.228) (0.159) (0.362) No. of Members 0.007 0.007 0.010 0.007 0.007 0.007 0.013 0.006 (0.001) (0.002) (0.002) (0.001) (0.001) (0.002) (0.001) No. Reservations 0.035 0.035 0.035 0.034 0.036 0.036 0.038 0.036 (0.015) (0.031) (0.017) (0.015) (0.029) (0.018) (0.016) No. of Declarations 0.014 0.014 0.040 0.008 0.002 0.002 0.002 0.002 (0.007) (0.014) (0.023) (0.002) (0.005) (0.008) (0.003) Lifespan (Years) 0.009 0.009 0.019 0.004 0.0001 0.0001 0.004 0.008 (0.002) (0.004) (0.004) (0.001) (0.003) (0.003) (0.002) Constant 1.793 1.793 1.473 1.112 2.583 2.583 2.482 1.801 (0.164) (0.337) (0.231) (0.124) (0.233) (0.159) (0.363) Observations 621 621 621 621 1,301 1,301 1,301 1,301 Log Likelihood 1,061.810 618.284 983.972 1,670.446 987.521 1,563.750 0.750 (0.087) 0.693 (0.066) Akaike Inf. Crit. 2,139.620 1,252.568 3,352.891 1,987.043 Note: p<0.1; p<0.05; p<0.01 Data from UN Treaty Series. Appendices 390 Table C.4: Eect of Treaty Attributes on Number of Treaty Exits Dependent variable: Treaty Exit (1) (2) (3) Obliagtion Score 0.001 0.0001 (0.0004) (0.001) Length (Words) 0.00001 0.00000 (0.00000) (0.00001) Multilateral 2.788 2.453 (0.136) (0.162) No. of Members 0.002 0.002 (0.002) (0.002) No. Reservations 0.095 0.043 (0.387) (0.379) No. of Declarations 0.610 0.472 (0.117) (0.139) Year 0.004 0.004 0.006 (0.002) (0.002) (0.002) Constant 11.628 14.130 17.322 (4.157) (3.458) (4.193) Treaty Fixed Eects? Yes Yes Yes Observations 22,892 41,006 22,892 Log Likelihood 2,644.260 3,393.244 2,432.978 Akaike Inf. Crit. 5,298.521 6,800.487 4,883.955 Note: p<0.1; p<0.05; p<0.01 Data from UN Treaty Series. Appendices 391 Table C.5: Case Study Treaties Treaty Score Permissive Obligatory Words ABM Treaty -20 53 33 5; 220 EU Treaties (consol.) 952 504 1; 456 72; 503 INF Treaty 103 20 123 7; 049 Lisbon Treaty 1; 692 325 2; 017 55; 377 Rome Statute 315 345 660 35; 315 Appendices 392 Appendix D RA Instructions for Coding Treaty Obligation D.1 Summary of Approach To build new measures of treaty obligation using automated text analysis, I take a multi-pronged approach and follow the guidance of Grimmer and Stewart (2013) and Barber a et al. (2021). To start, I obtain a corpus of treaty text by collecting the English treaty text of each treaty to be analyzed from the United Nations Treaty Series (UNTS). In the UNTS, treaty text is stored in PDF format, which requires conversion to a text le and some preliminary cleaning to address issues created with the conversion. The English text of each treaty is used to avoid translation problems across texts Appendices 393 and because the English version is commonly one of the ocial versions of the treaty. At the most basic level, I use a dictionary method to categorize certain words as obligatory or permissive. This is a somewhat crude measure, based on counting and categorizing words. These coding instructions focus on the human coding needed for supervised machine learning classication. How- ever, the research assistants were also asked to validate the dictionary used in Chapter 3 (Appendix C, table C.2) based on their review of numerous treaty articles. A supervised method of classication requires human coding of treaty text to train, test, and validate a classication model to be used on out-of-sample treaty text data. To do this, I created a survey for research assistants to code treaty text (shown below). To rene the analysis, the treaty text is divided into subparts, primarily by articles. Generally, treaties are broken up into articles that each cover a specic provision of the agreement between the parties. For example, one article might outline obligations with regard to implementing treaty goals, another to create a treaty organization, an- other for the organization's governance, and another regarding funding the organization. Therefore, the articles serve as an outline of a treaty, the meta-structure of the topics within a treaty. Breaking the treaties up by articles will allow for cross-treaty comparisons of article types. The articles, along with the coding, will be recombined in later analysis. With each treaty broken up by article, the substantive articles are Appendices 394 coded as obligatory or not. What follows are the coding instructions and questions for the coders, which includes questions regarding the articles type and various parameters of the obligation. D.2 Coding Instrument and Instructions Treaty Exit: Obligation - Instructions What follows is a series of articles from a treaty text. For each article, a list of ve questions is presented to determine whether the article is best described as obligatory, permissive, or neutral. Sometimes it is dicult to categorize an article. It may also be the case that some parts of the article may fall in one category and other parts fall in a dierent category. Nonetheless, please choose the best answer for the article as a whole. Apart from coding each article as obligatory, permissive, or neutral, you will also give each article a category of article type. Treaties generally follow sim- ilar formats and include similar clauses to lay out the purpose and substance of the agreement between states. In the course of reviewing treaties, you will likely see the following types of articles: (1) an introductory section called a chapeau, which broadly denes the members' objectives, purpose of the agreement, and background to the treaty; (2) articles on conditions for participation of states and non-state actors; (3) articles on conditions for joining the treaty, including provisions regarding reservations, declarations, and notications; (4) articles on application of the treaty, in time, space, Appendices 395 and across jurisdictions, as well as relationship to other treaties; (5) articles on creation, governance, and management of a secretariat, council, commit- tee, or organization; (6) articles on voting rules for decisions; (7) articles on nancing institutions or organizations; (8) articles describing duties, re- sponsibilities, rights, or obligations of treaty members; and (9) articles on dispute settlement. Additionally, some treaty articles might contain boilerplate, miscellaneous, or general terms. Some examples include articles regarding treaty amend- ment, ratication, savings clauses, ocial language, location of depository, or withdrawal. These general articles appear in most treaties in some form or another, usually at the end of a treaty before the signatures. Their pur- pose is to lay out technical or practical aspects of the treaty that most likely don't dier greatly between treaties. As a guide, below are some denitions and synonyms: Obligatory: requiring a treaty member to take or refrain from some action without discretion and with consequences for failing to do so (compulsory, mandatory, necessary) Permissive: allowing or hoping a treaty member take or refrain from some action with discretion and without consequences for failing to do so; (vol- untary, optional, aspirational) Neutral: neither obligatory or permissive; possibly descriptive * * * Appendices 396 [Text of one article from relevant treaty] Q1: How would you categorize this article? Introductory section: broadly denes the members' objectives, pur- pose of the agreement, and background to the treaty Article on conditions for participation of states and non-state actors Article on conditions for joining the treaty, including provisions re- garding reservations, declarations, and notications Article on application of the treaty (duration, location of application, jurisdictions covered), as well as relationship to other treaties Article on creation, governance, and management of a secretariat, council, committee, or organization Article on voting rules for decisions Article on nancing institutions or organizations Article describing duties, responsibilities, rights, or obligations of treaty members Article on dispute settlement Boilerplate, miscellaneous, or general terms Other Unclear Q2: Is the language obligatory or permissive? Obligatory Permissive Neutral Appendices 397 Q3: Does the article use precise or ambiguous terms? Very precise Somewhat precise Somewhat ambiguous Very ambiguous Q4: Doesthearticleimposeanyobligationsontheagreeingparty? Yes No Unsure Q5: Are the obligations described strong or weak? Very strong Somewhat strong Somewhat weak Very weak

Abstract (if available)

Abstract Britain's decision to withdraw from the E.U., or "Brexit;" members of the African Union calling for mass withdrawal from the International Criminal Court (ICC); and the United States' announced withdrawal from the Paris Climate Accord, are all examples of an increasing number of international institutions seemingly under threat by members exercising their power to formally withdraw. Why do states choose to withdraw rather than bargain for accommodation or simply breach to meet their needs? This dissertation develops and presents a bargaining model of treaty exit that incorporates treaty features of obligation and flexibility to understand how the distributional shift in treaty costs and the effect of treaty precision and enforcement shape a member state's preference to withdraw overreaching a bargain, maintaining the status quo, or breaching. Since withdrawing is costly to international cooperation---foreclosing a forum within which to address interstate problems---it is puzzling why a state would choose to withdraw over other options. The theory is empirically tested using quantitative, large-N statistical analyses and qualitative, small-N case studies in a multi-method research design. Quantitatively, the texts and meta-data of hundreds of treaties was collected and used test aspects of the theory regarding flexibility and obligation using automated text analysis. In addition, the high-profile withdraws from the International Criminal Court (ICC); arms control agreements between the United States and Russia; and Brexit are examined in detail to uncover the mechanisms working in the states' decisions to exit. The dissertation concludes with a summary of the theory and results, implications of the findings; and reflections on the prospects for future cooperation in an increasingly competitive world.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

By word, not deed: advancing a theory of rhetorical compliance in international law

PDF

The IO contestation project: a multi-method approach towards understanding the ways states contest international organizations.

PDF

Taking orders from Brussels? external actors and regional organizations in the developing world

PDF

Refugee protection and empowerment joint venture: understanding how and why public and private actors cooperate to tackle the global refugee crisis

PDF

The origins and evolution of the U.S. alliance network: how military allies transform and transact

PDF

Code 'war' theorizing: information and communication technology's impact on international relations theorizing, negotiation, and cyber relations

PDF

Legitimizing self-determination: advancing the sovereignty of separatist movements

PDF

Challenges from below: the origins of status competitions in world politics

PDF

Explaining the variation in the extent of China's hegemony, 1279-1840 — multiple sources of ideational power beyond Confucianism

PDF

Politics behind appointments: policy, patronage, and party systems in the Andes

PDF

When bullets are not enough: determinants of chemical weapons usage and abstention

PDF

International politics and domestic institutional change: the rise of executive war-making autonomy in the United States

PDF

Forced march to modernity: State-imposed cultural change and regime stability in 20th-century east Asia

PDF

Habits and policy: the socio-cognitive foundations of foreign policy stability

PDF

Selling virtue: how human rights NGOs and their donors work together to create a better world ... for themselves

PDF

Neither purse nor sword: strategic minority opinions and judicial policy-making

PDF

Pathways to nuclear weapon reversal: exploring mechanisms and understanding non-proliferation policy

PDF

Are you as old as you feel? A quantitative and qualitative analysis of leader age and foreign policy decision-making

PDF

Organized intraethnic conflict and cooperation in divided stateless nations: the competitive dynamics of rival ethnonationalist factions

PDF

When passions run high: emotions and the communication of intentions in face-to-face diplomacy

Asset Metadata

Creator Dalton, Taylor Reeves (author)

Core Title Things fall apart: the unraveling of international institutions through withdrawal

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Political Science and International Relations

Degree Conferral Date 2022-08

Publication Date 07/25/2022

Defense Date 06/06/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag formal theory,international cooperation,international law,International relations,OAI-PMH Harvest,text analysis,Treaties

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Sandholtz, Wayne (committee chair), Coe, Andrew (committee member), Graham, Benjamin (committee member), James, Patrick (committee member), Wood, Abby (committee member)

Creator Email taylor.r.dalton@gmail.com,tdalton@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111375282

Unique identifier UC111375282

Legacy Identifier etd-DaltonTayl-10968

Document Type Dissertation

Format application/pdf (imt)

Rights Dalton, Taylor Reeves

Type texts

Source 20220728-usctheses-batch-962 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu