University of Southern California Dissertations and Theses

Legitimizing self-determination: advancing the sovereignty of separatist movements

(USC Thesis Other)

Legitimizing self-determination: advancing the sovereignty of separatist movements

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content LEGITIMIZING SELF-DETERMINATION: ADVANCING THE SOVEREIGNTY OF SEPARATIST MOVEMENTS R. Joseph Huddleston Doctor of Philosophy (POLITICAL SCIENCE AND INTERNATIONAL RELATIONS) FACULTY OF THE USC GRADUATE SCHOOL University of Southern California August 2018 2 TABLE OF CONTENTS Acknowledgements............................................................................................................................. 4 Introduction: War Still Makes the State ......................................................................... 7 1.1 Introduction ................................................................................................................................... 7 1.2 Sovereignty and Self-Determination ........................................................................................ 9 Silent Sponsors................................................................................................................. 16 2.1 Redefining Recognition ............................................................................................................. 16 2.2 The Continuum of International Sovereignty ..................................................................... 18 2.3 Developing a New Measure .................................................................................................... 24 2.4 Trends in recognition ............................................................................................................... 30 2.5 Revisiting hypotheses about recognition ............................................................................. 33 2.6 Discussion and conclusion ...................................................................................................... 37 Foulweather Friends ........................................................................................................ 39 3.1 Half Measures ............................................................................................................................. 39 3.2 Current Understanding ............................................................................................................ 40 3.3 Stability-Seekers ......................................................................................................................... 46 3.4 Analysis ........................................................................................................................................ 54 3.5 Discussion ................................................................................................................................... 61 Booed Off Stage ............................................................................................................... 63 4.1 Syria Sets the Stage ................................................................................................................... 63 4.2 Audience Cost Theory and Its Critics .................................................................................. 66 4.3 A Crucial Case: Obama’s Public Changes Course ............................................................. 68 4.4 Survey experiments are surveys ............................................................................................ 76 4.5 Research Design ........................................................................................................................ 81 4.6 Experimental Results ................................................................................................................ 83 4.7 Priming Reputation.................................................................................................................... 87 4.8 Discussion ................................................................................................................................... 89 Conclusion: Separation Anxiety ..................................................................................... 92 5.1 Separation Anxiety .................................................................................................................... 92 5.2 Limitations and Future Directions ......................................................................................... 93 5.3 Policy Implications ..................................................................................................................... 95 Bibliography ..................................................................................................................... 98 Appendices ..................................................................................................................... 113 Appendix A. Supplementary Material for Chapter 3 ............................................................. 113 Appendix B. Poll Data for Chapter 4 ........................................................................................ 114 Appendix C. Supplementary Material for Chapter 4 ............................................................. 119 3 Appendix D. Supplementary Analysis for Chapter 4 ............................................................. 125 Appendix E. Experimental Prompt for Chapter 4 .................................................................. 127 4 ACKNOWLEDGEMENTS In my case, it is possible to know who exactly to credit first in the long list of those I am indebted to for assisting me in this accomplishment. That credit goes to Nick Weller, committee member, coauthor, and friend. Nick has been a role model in every aspect of this career. The way I think about research, how I ask questions and read others’ research, and even my choice of words on paper have all been permanently affected by the excellent foundation I received under Nick’s guidance. As I move forward with the practice and puzzle of research as a central theme in my career, I will always consider myself indebted to Nick for his countless readings of my drafts, outlines, and polished papers, his Friday-afternoon-coffee-personal-research-workshops, and his generosity with his time and attention across the board. In my cohort’s foundational methods course with Nick, he exhorted us to support each other: “No one gets out of here alone, alive.” I can confidently say that the support Nick has given me—as a colleague—is unmatched by any of the other valued peers who have helped get me out of here alive. In no small part do I owe him my success in grad school, and his advice has always served me well. Next is Pat James, my dissertation chair. Aside from being an unbelievably accomplished scholar and a walking politics encyclopedia, he has been an invaluable mentor and counselor as I have learned to navigate this career. He has repeatedly dealt essential advice on what order to take my projects, what to say in applications for grants and fellowships, and how to establish a network and promote my work. Most importantly, he has answered my thousand and one little questions that make up that bigger question, how can I do this for the rest of my life? I have no doubt that Pat’s incredible understanding of the academic career and fierce advocacy for his students is at the core of their impressive record of success, and I knew he was the right person to shepherd me through this deceptively formidable stage in my professional growth. This is all aside from the repeated reads and edits he has done in my work and the brilliant guidance he has given me on getting through the practical work that goes into a dissertation. If I can do for my students what he does for his, I would consider that my utmost service to this profession. I owe the genesis of this project to Wayne Sandholtz’s excellent seminar on international law and norms. I was thrilled when he agreed to serve on my committee, and his input has been critical at every stage. Professor Sandholtz’s thinking about the influence of international norms on state behavior will continue to influence this project well beyond this point. I am grateful to Ben Graham for training me as a researcher and teacher, and for serving as an excellent model of this profession. I am grateful to my external committee member, Nick Cull, who has interjected with excellent ideas at important junctures. I am grateful to Bridget Coggins, not only for her ground-breaking work on the same topic, but for her selfless generosity with her time and attention to a grad student at another institution. I am grateful to Laurie Brand and Gerry Munck, both of whom have played important roles in influencing both my choices of topic and my approaches to empirical problems. Christian Grose, Jane Junn, David Kang, Morris Levy, Andrew Coe, Jonathan Markowitz, and Brian Rathbun have all been generous with their time and advice as well. In short, there are many faculty at USC who have contributed to my work, and all I can hope to do to pay them back is to pay it forward with my own students. I will do so. 5 I have also been blessed with truly excellent colleagues, and it pains me to bid them farewell this year. Youssef Chouhoud will always my Brooklyn brother, Therese Anders my coffee shop partner, Tom Jamieson my conference coconspirator, Ming-min Yang my high proof pal, Ronan Fu my IR theory encyclopedia, Victoria Chonn Ching my bread buddy, Suzie Mulesky my valued sparring partner, Kelly Zvobgo my critical cheerleader (if that’s a thing), and Güez Salinas my perpetual peer-encourager. Sara Sadhwani is like a sister. Josh Tuynman and Rod Albuyeh have been great friends and confidants throughout this period. I am particularly indebted to Youssef, Tom, and Therese, who have read and commented on my work numerous times, investing in the quality of my product with no expectation of anything in return. These brilliant, beautiful people have been cherished comrades as I have scaled this mountain. The foundations I brought to graduate school were solid, and I have many to thank for that as well. My parents Andy and Susie taught me how to work hard, and how to get up at 6am to study. It was endless readings of the Bible back then, but it still instilled in me the satisfaction of deep study. My dad occupied my every Saturday with massive home renovation projects, slowly building in me the ability to visualize something huge and work towards it. More than a decade later, this is one of my most valuable skills, and I am grateful to him for it. My four sisters Grace, Corrie, Melissa, and Jessica, and my two brothers Ben and Nick have all lent endless encouragement as I have pursued my graduate education, even though I was “still in school,” or as my niece put it, “in 22 nd grade.” So have Uncle Mike, Aunt Katie, and cousins Colin, Kevin, Bryan, and Chris Crow. John and Kristen Cook, Josefine Borrmann, Tiffany Seeley, you too. I have leaned on you all heavily. It would be remiss not to pay tribute to my undergraduate inspirations: Nubar Hovsepian, who first made me think critically about international politics in the Middle East; Angeliki Kanavou, who laid the foundations of my Peace Studies education in her vibrant classroom; Ann Gordon, who gave me free reign of her lab and hosted my first dabblings in the practice of social research; Stephanie Takaragawa, who inspires me to pursue my goals fiercely and be good to the people around me; and of course, the late Don Will, who was an incredible guide to me as a first gen student and academic. He is my cherished mentor and the biggest role model I have. I can scarcely enumerate the contributions my wife Jill has made to this accomplishment. As I labored away for a single degree throughout these six years, she has earned three, including two advanced degrees, and a post-master’s in nursing. She will have been hooded with a doctorate (her fourth degree) a mere week after I am, and she has worked full time the whole time. In short, Dr. Jillian Wade, DNP, AGACNP-BC, CPNP-AC, RN, CEN is the baddest lady I know, and she inspires me to be better every day. She has shared the heavy burden of the chaos, uncertainty, and frustration of academic training. She always knows the right way to get me into a better frame of mind when I stray from reason or decency. I love you, Jill. I was born in a trailer park on a chicken farm turned evangelical missions base in northwest Arkansas. I flittered to and fro between conservative homeschool groups and resource- poor public schools. We relied on the Lord, the government, and the Food Bank to ensure we were fed. I had no one to show me the ropes on getting started in an academic career, and a few even condemned university education outright. To become the first person in my family in either 6 direction to be called “Doctor” is a deeply personal accomplishment. So many have helped me get here, and I hope I have made them proud. 7 INTRODUCTION: WAR STILL MAKES THE STATE BUT OTHER STATES HELP 1.1 INTRODUCTION When representatives from 51 states signed the Charter of the United Nations in San Francisco on October 24, 1945, they unwittingly put the world’s governments into a perpetual predicament that none have yet solved. In its lofty efforts to usher in a new era of global peace, the Charter’s first article enshrined “respect for the principle of equal rights and self-determination of peoples.” 1 Self-determination, a deeply rooted democratic ideal evolving and manifesting in the works of Thomas Jefferson, John Locke, Woodrow Wilson, German philosophers, and London socialists, 2 was planted in the very first article of this new, all-encompassing global institution, meant to tame the excesses of anarchy and stabilize international relations. It satisfied longstanding American and Russian opposition towards European colonialism. 3 It encouraged initial buy-in from smaller powers and immediate buy-in from former colonies who, in the years to follow, would peel themselves away from British and French control and join the UN as 1 UN 1945, 3. 2 Unterberger 1996. 3 Fabry 2010, 154. 8 member states, including 17 in a single day in 1960. 4 To this day, self-determination is a chief principle that encapsulates the intended purpose of the United Nations. In many ways, it implies a key provision in the second article of the Charter, that “Members shall refrain…from the threat or use of force against the territorial integrity or political independence of any state.” 5 In cases in which “peoples” meant recognizable nation- states with already established borders, 6 self-determination implied the territorial integrity of states, and generally is still understood as such. Self-determination means no state can attempt to coerce any other state’s domestic political functions, the way colonial powers did. Because nation-states are understood as “peoples,” the norm against interstate territorial conquest has become well established and generally well respected. Following 1945, states continued fighting a few skirmishes as they settled their borders, but indeed the era of states swallowing other states seems to have come to a permanent close. 7 On the surface, the clear line from self-determination to territorial integrity in the Charter, body, and conduct of the UN is straight and uncompromised. However, a contradiction lies within. From a modern standpoint, the tension between these two articles is rather evident. On one hand, states have a right to territorial integrity. There can be no revision of a state’s borders without either its consent or its initiation. This provision prevents aggression from other states. On the other, it also puts states in an advantageous position over other “peoples” attempting to manifest their own self-determination by declaring themselves sovereign states. In the modern order, this necessarily entails a revision of borders, and so threatens the territorial integrity of some member. That state has all the privileges of the UN membership, 8 including the sympathies of other states driven by principles of precedent 9 and reciprocity, 10 so it places any new movement for self-determination in a necessarily difficult spot. This contradiction in the foundations of the international order has proven to be a wellspring of conflict and strife, and a gray area in which anarchy still shapes state actions. This dissertation explores that space, focusing on a broad range of states’ foreign policy activities towards self-determination movements in other states. I take these other activities to reflect international sovereignty in a way other works on international security have not. I reconceptualize recognition as an underlying trait that determines a range of foreign policy decisions, a trait that can be estimated and explored. Using this measure of “latent recognition,” I show, among other findings, that third parties support self-determination groups when violence reaches high levels, when they seize territory through war, and when major powers extend diplomatic recognition. Exploring the strange behavior of the American public towards Syria in 2012-2013 and designing and running an opinion experiment, I also demonstrate the presence of 4 UN 1960. 5 UN 1945, 3. 6 Notwithstanding a few minor border disputes 7 Gibler 2007; Gibler and Tir 2010. 8 Fazal and Griffiths 2014. 9 Hechter and Borland 2001; Coggins 2014. 10 Keohane 1986. 9 a common yet unacknowledged problem scholars may encounter when studying public support of intervention. Conflicts of self-determination are particularly difficult to prevent and resolve. They are rooted in a sort of ambiguity that exists in international law, in state practice, and in the goals of self-determination activists themselves. However, this ambiguity has a clear source, and examining that source is the best first step to understanding why these conflicts have proven to be so intractable. 1.2 SOVEREIGNTY AND SELF-DETERMINATION Concerning the breakup of Yugoslavia in 1991, the European Community proclaimed that its member states “will not recognize entities which are the result of aggression.” 11 The expectation embedded in the rules and norms of the UN and other international bodies is that nonviolent means are the only ones acceptable for revising the international system. Thus, any aspiring state, any movement for self-determination in this modern order, is told that war with a member state will not be rewarded with admission into the UN. An overview of the literature on conflict and separatism shows this not to be entirely true. Chenoweth and Stephan’s enlightening comparison of violent and nonviolent means of achieving massive political revisions finds that, in most case, nonviolent mass movements are more likely than their violent counterparts to achieve progress towards their aims. 12 However, their conclusion includes a notable caveat and exception to the rule: secessionist and separatist movements do not generally succeed nonviolently. Cunningham 13 points out that many self- determination groups, while starting out nonviolently, will reach a limit to their effectiveness. Evolving leadership and strategy very often lead to the embrace of violent means. One of Coggins’ key findings is that the successful capture and holding of the territory in question is one of the factors most likely to influence major powers’ initial decisions to recognize officially. 14 Paquin finds that those secession movements that more effectively destabilize the system through protracted conflict are those most likely to earn the official diplomatic recognition of the United States. 15 Even the enduring international legal debate surrounding self-determination and the recognition of new states has reflected this ambiguity of concept. There has been a long running debate over whether a “declaratory” or “constitutive” theory of recognition better conforms to the spirit of international law on self-determination. Declaratory theories observe an aspiring state’s assertion of sovereignty within its territory as the crucial factor determining its level of access to the international political sphere. Does the aspiring state have a monopoly on violence? Can the state establish contact with other governments? These factual questions determine 11 European Community 1992, 1487. 12 Chenoweth and Stephan 2012. 13 Cunningham 2014. 14 Coggins 2014. 15 Paquin 2010. 10 statehood, not the politically determined choices other states make to recognize an equal. This theory is applied in case law, 16 and domestic sovereignty is thought to be a precondition to recognition. Constitutive views of recognition take third-party policy as an integral part of sovereignty. An aspiring state is not a state until its peers recognize it as such. Recognition is part of acquiring statehood, a means of admission into a club. Recognition is something that can be withheld, something vulnerable to the whims of domestic and international politics. Certain international legal cases reveal the prevalence of this theory, asserting that recognition “confers certain rights and obligations under international law” and supporting such a theory across many cases. 17 Scholarly assessment of the realities of state emergence have tended towards pushing a constitutive view as well. Coggins says, “external legitimacy is the fundamental distinguishing feature between states and nonstates,” and “aspiring states need a quorum of the world’s states to consecrate their legitimacy.” 18 A tension, a seeming incentive to embrace violence, has thus also arisen in the way both international courts and state actors treat self-determination movements. Aspiring state entities are told they must assert their sovereignty, must look and act like a state, before they will be recognized. To act like a state is to gain the monopoly on violence. 19 It is counterintuitive to expect aspiring states to gain this monopoly nonviolently. Some of the most successful have managed to attain legal and executive control over some or all of the territory they claim—groups like the Kurds, the Saharawis, and Somaliland— only to stall for years without being recognized. In these cases too does the constitutive view of recognition reign. Why else would stable, democratic, economically viable Somaliland continue unrecognized, while pockmarked, perpetually faltering Somalia continues with full recognition and UN membership? For a third party to advocate a revision to the international status quo is to commit a constitutive act, not a declaratory act. Third parties are more likely to diplomatically recognize aspiring states who have captured the territory and have constructed some of the modern trappings of the state. Aspiring states have generally managed this feat in a manner similar to that described by Tilly with reference to the emergence of the modern state in Europe. 20 Organizing a people to wage months or years of war requires establishing an infrastructure for the training, movement, and governance of people. It requires buy-in. It may require an effective channeling of the yearning passions of nationalism. The collection of these institutions we call “the state” forms in these conflicts—as in those of the past—to serve as a foundation for better and better efforts to wage, 16 Brierly 1963; Chen 1951; Brownlie 2008; Harris 1979. 17 Badinter Commission in Worster 2009; Oda et al. 1996; Rolin-Jaequemyns et al. 1934 & McNair et al. 1952 in Worster 2009. 18 Coggins 2011, 435, 461. 19 Weber 1946. 20 Tilly 1990. 11 weather, and win war. In the post-WWII order, purportedly characterized by peace, nonaggression, and stable borders between UN members, war still makes the state. Third parties are reluctant to recognize new states, but they do involve themselves in other ways. Intervention in wars of secession and separation is common enough to always be considered a possibility. 21 To borrow Krasner’s terminology, 22 although third parties may not grant aspiring states the international legal sovereignty manifest in universal recognition, they will act to encourage or impede development of separatist governments’ domestic sovereignty. They will help build the de facto reality of the state while refusing to usher in de jure statehood with their declarations of diplomatic recognition. In this way are legitimacy and sovereignty bound together. When third parties involve themselves to augment or erode the power of the aspiring state, they affect its legitimacy. Perceived legitimacy is a precondition for both declaratory and constitutive acts of recognition for self-determination groups utilizing both violent and nonviolent means. Thus, third parties, even while steadfast in their policies of nonrecognition, can move these states towards their goals, both building and demanding legitimacy of potential new states. It is perhaps for this reason why so many self-determination groups—e.g. the Kurds, Saharawis, and Somalilanders— perceive an advantage to maintaining democratic structures of governance, as it is taken to appeal to international actors who may eventually confer recognition upon them. In the International Relations literature on self-determination movements, sovereignty— as it concerns third parties—has been limited to the international legal form. That other states should involve themselves in building the domestic sovereignty of separatists and self- determination groups (and therefore eroding that of the central government) has been a point absent from discussions of international sovereignty. If states do not recognize, there is no international sovereignty. If states help or hurt in other ways, that is not about sovereignty, but about international politics. The conceptual wall between domestic and international sovereignty has yielded a gap in understanding of the evolution of self-determination groups, their progress or ultimate failure. This dissertation is founded on an assertion that for third parties to support the development of breakaway states and build their domestic sovereignty is also to make their case stronger for international sovereignty. Regardless of the ascribed legal theory of recognition, regardless of the supposed relationship between legitimacy and sovereignty, and regardless of whether a group’s progress ends as a recognized or unrecognized state, international involvement in these conflicts says something about international sovereignty. Third-party states invest their resources and their international political capital in these conflicts, they involve their publics, and they stake out positions that they may maintain for years. None of this variation in policy towards self-determination groups is captured by the current conversation on international sovereignty as reflected in diplomatic recognition. The conversation on unrecognized states and that on the recognition of states are held in isolation of each other, and the contribution of third 21 Heraclides 1990; Heraclides 1991. 22 Krasner 1999. 12 parties to the development of stable yet unrecognized states has been treated as less than important to international sovereignty and the stability of the current order. In the chapters ahead, I hope to complement the field’s robust understanding of how states confer diplomatic recognition upon new states, by beginning to explore how they advance these groups’ eligibility beforehand. The research questions that drive this project are manifold. How might we reconceptualize international sovereignty to better capture the variety of stances third parties take towards these conflicts? Is there an effective measure of sovereignty or recognition that detects shifts in policy short of the momentous decision to recognize? Why do states refuse to recognize while still supporting self-determination groups in other ways? What drives these decisions? What role do third-party states’ publics play in these decisions? How do they affect leaders’ decisions? I begin to answer these questions in the following three chapters. In Chapter 2, “Silent Sponsors,” I make the case for a new concept of and method of measuring recognition. With the field’s current focus on official diplomatic recognition as the sole reflection of international sovereignty, understanding and explaining the successful progression of separatist and self- determination movements from status as nascent movement to eligibility for international recognition is needlessly difficult. Two disjoint literatures, one on partially recognized states and one on diplomatic recognition, speak to the topic, but neither has been focused much on how these states accrue international sovereignty en route to recognition. Official recognition, the dominant measure, reflects only the final stages of a long process. I argue that aside from officially recognizing aspiring states, third parties tacitly recognize them by other important means: foreign aid, military partnerships, and other forms. To understand the subtler shifts in policy third parties make towards self-determination movements, I create a latent variable model of third-party recognition, using data on military and economic aid, diplomatic exchange, IGO voting, sanctions, opposition to and support of governments facing separatists, and official recognition. With this new measure, I retest several important theoretical predictions about recognition coming from prior literature. I find: (1) that successful seizure and control of territory by separatists has a strong effect on movement towards international recognition, (2) that third parties tend to move towards recognition of their rivals’ self-determination challengers, and (3) states move away from recognition when they share security interests with the party standing to lose territory. Chapter 3, “Foulweather Friends,” advocates a theory of movement towards recognition based on international concern for regional and institutional stability. I argue that between the instability caused by bloody or protracted separatist conflict and the domestic, regional, and international destabilization involved in third-party recognition, unrecognized statehood represents a relatively stable point in the progression of self-determination movements towards statehood, at least from the standpoint of international observers. As a result, countries that prioritize status quo stability in their foreign policy platforms will at a certain point, assume an orientation that supports the separatist cause and encourages their capture and control of the territory in question, even while they remain opposed to official recognition. In other words, there is a stability “sweet spot” between conflict and recognition that can persist for years or 13 even decades. It makes sense for stability-seekers to try to move self-determination conflicts towards this point and help them stay there. As evidence, I show that third parties move towards recognition when these conflicts experience high levels of violence, measured in battle deaths, or when the self-determination group succeeds in battle. On the other hand, lower levels of violence entail movement away from recognition, because the easiest way to return to stability is to renew the silence of the minority group, be it through repression or appeasement. To put it in mathematical terms, there is a local maximum of perceived instability in these conflicts, at which point third parties switch to support the self-determination group in attaining control of the territory. In this chapter, I also show some evidence that third parties respond to state violence against civilians in these conflicts by moving towards recognition of the self-determination group. This is presumably because violence against civilians is more permanently destabilizing and difficult to recover from, aside from the fact that states may punish this violence on normative grounds. Additionally, I show that the states’ well-known concern for setting counterproductive precedent if they recognize these movements is not a driving factor for other foreign policy decisions related to these conflicts. This is a notable difference between the measure developed here and the traditional measure for international sovereignty, official diplomatic recognition. Another important feature of the measure innovated and applied in these two chapters is the ability to assess the foreign policy orientation of nonmajor powers. Most research on third party legitimization, recognition, or support of self-determination movements has focused exclusively on major powers, especially the USA, Russia, and China. This has been a necessary effect of taking diplomatic recognition as the primary reflection of international sovereignty; gathering data on the official foreign policy stances of all the world’s countries towards all its self-determination movements would constitute a very large data project. It has also been argued that as far as official recognition goes, only the decisions of major powers have any effect on other countries’ decisions or on the likelihood that one of these movements will “succeed”—that is, acquire its own state with universal recognition. 23 Even so, there is a clear dearth of understanding of how and to what level smaller powers interact with these groups, and how their policies are affected by the decision major powers make. My measure bypasses the aforementioned concerns because it does not focus on recognition, but on other policies related to recognition, for which the data is available. Every state, large and small, has the same voice in the UN General Assembly, the same ability to sign onto international sanctions, and a surprising level of engagement with the transfer of arms to parties in conflict. As an example of the utility of this comparative ability, I show in Chapter 3 that medium sized powers respond to major power diplomatic recognition of self-determination movements by themselves moving towards recognition. All in all, the measure and tests developed in Chapters 2 and 3 provide an unprecedented comparative perspective on foreign policy towards self-determination conflicts, buttressing, complementing, and adding a degree of nuance to prior work on international sovereignty. 23 Coggins 2011; Coggins 2014; Paquin 2010. 14 If Chapters 2 and 3 can be characterized as analyzing intervention policy based solely on the outputs of the “black box” of the state, Chapter 4, “Booed Off Stage,” focuses on the inputs. It proceeds in two distinct parts. In the first part, I follow the Obama administration’s statements and policies towards Syria and its president Bashar al-Assad during the Syrian civil war (which endures at the time of this writing.) In 2012 and 2013, Obama made several strong, visible foreign policy statements that can be characterized as a “threat,” but they nevertheless made clear that the US intended to intervene in Syria if the right conditions were met. I show that polls throughout the year indicated that the America public supported Obama’s stance, backing intervention should the situation come up. It did come up, but the public flipped in its evaluation of the situation in Syria, and through a relatively prominent public outcry, pressured Obama to find solutions other than American military intervention. It pressured him to back down from his threat, even in the midst of military escalation, and to turn to its Russian rival to resolve the crisis. I take the Syrian crisis as a strong candidate for a “crucial case” evaluation 24 of Audience Cost Theory, 25 which makes clear predictions on precisely this kind of scenario: a democratic leader makes a threat; the public, concerned with international reputation, pressures that leader to honor their word, and punishes them if they do not. This theory has risen to prominence in the international security literature because it: explains how democratic leaders can “marshal public opinion” to signal resolve in international crisis; 26 reveals why democratic threats should, as a rule, be taken more seriously than autocratic threats (because their audiences have more leverage); and ultimately, constitutes one of the theoretical cornerstones of the democratic peace. What makes the Syrian case interesting is that it is an abject failure of Audience Cost Theory. It has every important characteristic of the narrative of audience costs in the real world—a clear threat, and attentive and supportive audience, a condition fulfilled in a relatively short time, and an attempt to follow through—but the lynchpin character in the story, the public, does not behave as predicted. It betrays the theory and spoils the President’s stated plans. Audience cost theory has been heavily tested through survey experiments, and scholars have found the public to act consistently with theoretical predictions in those tests. To disentangle this seeming contradiction between real and simulated audience behavior, I evaluate the case through the lens of the psychology (and economic) literature on survey methodology. Might common survey response tendencies be to blame for the discrepancy in reported opinion within the Syrian case? More importantly, is it possible that survey experiments on audience costs are susceptible to the tendency to think so abstractly about hypothetical scenarios as to neglect to consider the costs of an action in the same way as one would in an imminent scenario? Construal Level Theory instructs us that people considering hypothetical scenarios tend to engage in “high-level construal,” thinking about actions in terms of desirability, whereas those pondering immediate scenarios will engage in “low-level construal,” considering costs and feasibility. To see the effect this might have on these survey tests of audience costs, I report the results of an experiment in which survey respondents were asked a question about their 24 Gerring 2007. 25 Fearon 1994. 26 Schelling 1960; Downes and Sechser 2012. 15 expectations of casualties in a foreign intervention before they were asked about presidential approval, the dependent variable. I find that priming low-level construal this way led to a marked decrease in support for intervention and a significant muting of the standard audience cost finding. This is, of course, insufficient evidence to explain the failure of Audience Cost Theory in the Syrian case, since all I have is a general overview of public opinion on intervention over the course of a year, and many other relevant pieces changed during that year. However, the change in experimental findings from a simple casualties prime is stark enough to look for the effects of this bias elsewhere in survey experimental tests of opinion on intervention policy. Chapter 5 concludes this study with a short discussion of the methodological innovations and theoretical contribution of this work, and I explore a few key policy implications and useful outlets for future research. I also outline the current limitations on this project and lay out some approaches to allay them. Overall, I emphasize the ability of this conceptualization and model of recognition to complement the conversation on international sovereignty, and I present a broad foundation for new outlets of research. The conversation on self-determination has been full of contradictions for over 100 years. Boundaries maintain peace between countries, but contribute to conflict within them. Outside forces contribute to the domestic authority of some aspiring for independence, but will not recognize what they help build. Since the dawn of this concept as an important principle in international law and foreign policy decision-making, it has always been understood as a concept that applies to some groups, but not all. However, that boundary between groups has never been delineated, and a degree of anarchy has persisted in ways unmirrored in many other foreign policy areas during the Long Peace. This study explores the effects of this uncertainty on what might be called the practice of sovereignty, and hones in on the effects of violence and instability therein. As they did when Wilson laid out his fourteen points 100 years ago, third parties make important, even deterministic, contributions to the trajectories and outcomes of these conflicts, and those actions remain a topic deserving of efforts for further understanding. 16 SILENT SPONSORS A NEW APPROACH TO MEASURING RECOGNITION OF SELF-DETERMINATION MOVEMENTS 2.1 REDEFINING RECOGNITION The United States government trains Kurdish soldiers. 27 It provides aid to the Kurdish military, 28 and it consults the Kurdish Regional Government on its engagements in the region. 29 The US funds nonmilitary aid projects in Kurdish territory, 30 and hosts the Kurdish president when he visits. 31 Many other states have similar engagements with Kurdish authorities, and similar channels of reception of Kurdish interests, even establishing consulates in the Kurdish capital of Erbil. 32 Still, despite a relatively thorough and long running relationship with Kurdish authorities, the US’ official position, and that of every other state, is that there is no sovereign Kurdistan. This remains true even after Iraqi Kurdistan’s 2017 referendum 33 and subsequent advancements by the Iraqi military into Kirkuk. 34 27 Tomson 2017. 28 McCleary 2017. 29 Cook 2016. 30 KRG 2012. 31 White House Archives 2005. 32 Caryl 2015. 33 Goldman 2017. 34 Zucchino 2017. 17 The Kurdish case demonstrates how exceedingly difficult and rare is the successful emergence of a new state in the modern era. It exemplifies the numerous, diverse challenges awaiting would-be movements for self-determination. Scholars have striven to understand these problems at nearly every step of the process: the origins of and formation of nationalist movements; 35 the justification of separatist claims; 36 the difficulties of weakening the existing governments’ claims on a territory and establishing support for separation or secession; 37 and the complications of building a case for international sovereignty. 38 This last branch of research, on the recognition phase of self-determination movements, has been somewhat disjointed. Some have examined the circumstances under which groups aspiring to statehood have managed to secure the final passage into the club. 39 Others have focused on other kinds of states, de facto and de jure states, failed and pseudo-states, in which international status does not reflect how politics operates on the ground. 40 It has been common to distinguish degrees of sovereignty by focusing on official acts of recognition as the primary indicator of international sovereignty. The many other relations between these groups and third- party states have not been treated as part of the sovereignty puzzle, and there have been few attempts to connect the different stages a movement goes through as it progresses towards recognized statehood. Diplomatic recognition of a new state has been treated as a separate and distinct phenomenon from human rights concerns over, trade with, and aid to self-identified autonomous non-state groups. Diplomatic recognition, it would seem, is the only meaningful measure of an aspiring state’s international sovereignty. In this chapter, I argue that this is only a partial view of sovereignty. I propose a unifying framework, arguing that a government’s decision to fight a separatist group in another country, take a neutral stance, engage in trade or military cooperation, or officially recognize a new state might all be conceived as actions correlated with different levels of underlying recognition of statehood. International legal sovereignty, as Krasner conceives it, 41 is reflected to an extent in how third parties support aspiring states’ domestic sovereignty. Conceptualizing international sovereignty this way solves several problems. The first advantage is that it makes for a natural connection between the two literatures mentioned above. Third parties to these conflicts make important diplomatic, military, and economic decisions that can indicate favorable (or unfavorable) stances towards each aspiring state, contribute to the process of increasing (or decreasing) aspiring states’ “eligibility” for diplomatic recognition, or both. Second, international sovereignty has generally been conceived as a characteristic that moves in only one direction, accruing but not diminishing. This principle is even enshrined in international law. However, in practice, it moves both directions, and while official recognition is rarely revoked, there is little reason to believe third parties’ views of separatist groups only 35 Gellner 1983; Anderson 1991. 36 Wellman 2010; Wellman 2005; Philpott 2001; Philpott 1998; Buchanan 1998; Buchanan 1991; Moore 2010. 37 Toft 2003; Kolstø 2006; Jenne, Saideman, and Lowe 2007; Roeder 2007; James and Lusztig 2001; Hale 2008; Siroky 2009; Carment and James 1995; Carter and Goemans 2011; Cunningham 2011. 38 Krasner 1999; Fabry 2010; Coggins 2011; Coggins 2014; Sterio 2013; Closson 2011; Geldenhuys 2009; Biswas and Nair 2010; Buzard, Graham, and Horne 2017. 39 Krasner 1999; Coggins 2014; Griffiths 2016; Paquin 2010. 40 Ramos 2013; Closson 2011; Caspersen 2011; Harvey and Stansfield 2011. 41 Krasner 1999. 18 increase in favorability. Finally, it complements official diplomatic decisions to recognize, which only capture movement on the far end of what will be conceived in this chapter as a continuum. I proceed as follows. First, I lay out the case for a new conceptualization of international recognition (or alternatively, international sovereignty), outlining the theoretical advantages of thinking about it as a continuous process, rather than a discrete foreign policy decision. Then, I describe the data and walk through the process used to estimate the latent variable model for implicit support for recognition. I choose a few self-determination groups and third-party cases to demonstrate the performance of the model through the period in question. Finally, I revisit some key questions that have thus far been explored only through examining diplomatic recognition. I compare the performance of this model with that of tests based on official recognition in prior literature. 2.2 THE CONTINUUM OF INTERNATIONAL SOVEREIGNTY The triadic nature of the context is confusing, so it is useful to clearly define the three kinds of actors in question. I define them as: (1) the host state—that is, the state that controls the territory in question; (2) the aspiring state—the separatist, secessionist, or self-determination movement that seeks to wrest control of the territory and establish a new, independent state; and (3) the third-party state—every other state that is capable of holding a position vis-à-vis either or both of these states. To name one example, in the China-Tibet dispute, China is the host state, Tibet is the aspiring state, and every other country is a third party. For the purposes of this project, the term third party does not refer to other nonstate actors, such as IGOs or NGOs. There are three reasons for this choice. The first is that the role of prominent international nonstate actors has been mixed in these conflicts. IGOs have tended to stay out of conflicts over self-determination, while some NGOs have been involved to monitor and advocate for human rights and provide relief, 42 both of which might be characterized as entailing a similar concept of recognition to mine here. The second is that the UN, the largest IGO, makes more sense to conceive of as a stage for state action, rather than an actor all to itself. Indeed, UN voting decisions are a prominent informer of the model I develop herein. The third is that most unilateral actions by international actors that have large effects on these conflicts— such as arms transfers or international sanctions—can only be meaningfully taken by states. International recognition of states has emerged as an important issue in recent work in IR, in part due to the increasing prevalence of substate groups asserting claims of self-determination. On top of Krasner’s classic conceptualization of international sovereignty as a pattern of organized hypocrisy, 43 the last few years have seen some excellent works demonstrating the role large and influential states play in the process a new state goes through to become recognized and join the system officially. 44 Coggins 45 takes official recognition and final entry into the state system as her dependent variables of interest, studying the conditions under which the major powers extend official recognition to new states. She tests the effect of official great power recognition on final entry and finds that the most influential states have de facto control over 42 Heraclides 1990, 346–347. 43 Krasner 1999. 44 Sterio 2013; Coggins 2014; Griffiths 2016. 45 Coggins 2011; Coggins 2014. 19 who are the official members of the state system. Additionally, these works share Krasner’s argument that states officially recognize such groups when it benefits them materially or diplomatically. They also share Krasner’s assertion that diplomatic recognition is the only meaningful conceptualization and reflection of international legal sovereignty. These theoretical and empirical decisions restrict the universe of considerable cases substantially. Official recognition is only one costly policy position that third-party states take towards self-determination groups; it necessarily entails escalating opposition to the claim of the host state, which may be ally or enemy, and is already a universally recognized member of the system. Furthermore, only some movements become eligible for recognition, and there are other foreign policy decisions states may take that advance the cause of an aspiring state. For example, because it is easier for third parties to recognize an aspiring state when it has already been recognized by the host state, a third party may pressure the host state into holding a binding independence referendum. US sanctions against Sudan served such a purpose in 2010. 46 Official recognition might be best characterized as the last domino to fall in the foreign policy process precipitating the emergence of new states. A broad range of other activities may come beforehand. The discussion has also been restricted in terms of direction; changes of state position in relation to these groups have generally been characterized as moving in one direction, towards recognition. A classic example proves helpful. In 1933, officials from 19 western hemisphere states, including the US, Mexico, and Brazil, convened in Montevideo, drafted, and signed a document outlining the “rights and duties of states.” At one point, it asserted, “The recognition of a state merely signifies that the state which recognizes it accepts the personality of the other with all the rights and duties determined by international law. Recognition is unconditional and irrevocable.” 47 This characterization of “declaratory” recognition is practically consensus in international law. 48 This idea of movement towards recognition as happening in only one direction is seldom pointed out, but often implied. In fact, even official recognition is sometimes revoked, as are decisions to legitimate the positions of groups seeking independence, autonomy, or self-government. Recognition might therefore be better defined as something both continuous and bi-directional. Third-party states make minor changes in their positions short of recognition, and they may also reverse those changes. Finally, the literature on international sovereignty has tended to focus only on the foreign policy decisions of large actors, the major powers. Coggins 49 and Ker-Lindsey 50 focus on how certain major players, especially the US and Russia, make decisions on international recognition. Paquin’s thorough case study 51 analyzes US policy towards six secessionist groups, with some attention paid to other international actors influencing the process. They rightly argue that major powers have the most external influence on the success of these movements in gaining statehood. This is the tendency throughout much of the foreign policy analysis literature on a variety of policy arenas. Scholars tend to focus on the major powers, especially Western powers, and build 46 Al Jazeera 2010. 47 Montevideo Convention 1933. 48 Lauterpacht 2012. 49 Coggins 2014. 50 Ker-Lindsay 2014. 51 Paquin 2010. 20 theory on a handful of cases. This project bucks that trend, taking smaller and medium sized powers’ stances towards these movement as relevant to the question of international sovereignty. One of the reasons unrecognized states—like Somaliland—are able to exist in perpetuity is that neighboring small states support and maintain a new status quo, and states of all sizes support them in IGO, economic, or military actions. To fully consider the whole spectrum of foreign policy stances toward self-determination movements requires consideration of the universe of potential recognizers. The measure expounded later makes this possible in a manner that, to my knowledge, has not been attempted before. To augment the concept formation regarding recognition of movements as new states, I conceptualize third-party states as moving along a continuum of favorability towards self- determination movements. At the far left of the continuum, a third party considers a movement for self-determination to be subversive to the ruling government and illegitimate internationally, perhaps granting them the label “terrorist” or “rebellion.” At the far right of the spectrum lies recognition of the movement as a legitimate member of the international system, a member of the club of about 200 countries with near universal recognition as self-governing. Between the two ends exist multiple positions a third-party state may take regarding a self-determination movement, as well as those towards the ruling government against which separatists stake their claims. Figure 2-1 illustrates a rough order of the positions third-party states may take towards aspiring states. Subversive or illegitimate Concern for aspiring state Criticism of host state Neutrality on group claim Legitimate claim to statehood All but name; trade or cooperation Recognition of state The positions of third-parties fall along this continuum, and changing positions can be characterized as movements back and forth. For the purposes of this chapter, only the two ends of the spectrum and the central point might be considered settled, though I have tried to organize these positions with an intuitive order through examples of degrees. Still, it is possible to describe cases with a different order. ISIS Kosovo 1991 Western Sahara Kurdistan Taiwan South Sudan East Timor Examples from US perspective Figure 2-1. The sovereignty continuum for aspiring states 21 This continuum of international sovereignty makes it easier to position the scope of the rest of this chapter in the context of prior work on recognition. For example, I start by placing the ideas of Krasner and Coggins 52 clearly on the spectrum. This is seen in Figure 2-2. In equating international legal sovereignty with diplomatic recognition, both works necessarily focus on the activity on the far right of the spectrum, asking what conditions are necessary to enable final entry into the state system. By illustrating their positions in this way, it becomes clear that the scope of their answers is somewhat limited. A stark implication of conceiving of international sovereignty in this way is that it places currently recognized states and separatist groups on equal footing (if at different points along the spectrum), at least in the eyes of third-party recognizers. Since an established state is simply the right end of continuum of possible statuses a group can take on, it differs only in its relative position. That position has some very real privileges at the international level—Fazal and Griffiths 53 point to IGO membership, improved international security, and financial benefits— and recognizing states know successful movements will be qualified for those. However, positions on this spectrum short of fully recognized states can still foster close ties with third parties and accrue many of the benefits of statehood without that title, so reluctantly given. One would be hard pressed to make a strong case that Taiwan would fundamentally change in its domestic structure and international relations if it garnered universal recognition. For state positions on this continuum to make sense, they must also be conceived of as bi- directional. With changing political circumstances, material needs, and opportunities, states move back and forth along this spectrum. The idea of recognition as something “irrevocable,” while intuitive, does not seem to work in the real international environment. 52 Krasner 1999; Coggins 2014. 53 Fazal and Griffiths 2014. Krasner Coggins Illegitimate Recognized All but name Actions run this direction Figure 2-2. Location of current research Illegitimate Recognized Instead, actions run both directions Figure 2-3. A bidirectional relationship 22 A useful example of this bi-directional nature, rooted in official recognition, is in the Western Saharan case. Figure 2-4 illustrates the position of the Western Saharan case on this continuum. The action all takes place on the right half; a 1975 International Court of Justice (ICJ) decision effectively made neutrality the de facto minimum position third parties were willing to hold in this conflict. The arrows on the right illustrate the theory, that the occurrence or threat of mass violence persuades states to forego interest or disinterest and to involve themselves in the situation by acts of recognition of self-determination movements. Acts of recognition towards Western Saharan self-determination have taken two forms: the first is by recognizing the ICJ decision in 1975 and supporting UN efforts to implement its ruling, such as founding and empowering the Mission for the Referendum in Western Sahara (MINURSO) in 1991; the second is by plainly and officially recognizing the Saharawi Arab Democratic Republic (SADR) as the rightful government of the territory, thereby assuming the results of a properly held referendum would favor independence. The years leading up to the 1991 ceasefire saw 75 sovereign nations extend recognition to an independent Western Sahara governed by the SADR. 54 This included prominent governments like those of Mexico, India, Iran, and South Africa. Since the UN brokered the ceasefire in 1991, and since violence has waned and the conflict has lost significant international attention, the SADR has witnessed a bizarre and unfortunate trend internationally. While nine states have newly recognized the SADR, including the new states of East Timor and South Sudan, 55 33 states have withdrawn, frozen, or suspended recognition of (henceforth “derecognized”) the SADR. 56 This is illustrated in Figure 2-5. 54 ARSO 2014. 55 ARSO 2002; SPSRASD 2011. 56 ARSO 2014. Recognized Illegitimate Neutral Before Ceasefire During Ceasefire Figure 2-4. Western Sahara and recognition 23 Figure 2-5. Third-party acts of recognition and derecognition. Source: ARSO 57 The Moroccan government has shrewdly applied pressure on states to derecognize the SADR. A US diplomat notes, “Getting other countries to recognize its sovereignty over Western Sahara, or to abstain from taking a position on Western Sahara, has been a dominant focus of Morocco's diplomatic efforts since 1975…Probably owing mostly to Moroccan diplomatic efforts, [33] countries which previously recognized the SADR withdrew their recognition and froze their relations with the SADR.” 58 The Moroccan government persuaded the Indian government to derecognize the SADR by leveraging India’s dependence on it for phosphates. 59 It did the same thing in Grenada. 60 In 2007, the Kenyan government found it to be “imperative” to suspend relations with the SADR in favor of those with Morocco for access to its “sizeable market,” 61 which was not open to Kenya while it recognized the SADR. The government of Dominique agreed to derecognize the SADR as part of a deal that has Morocco funding a luxury hotel there. 62 Morocco has also employed less economically based methods. It refused to deploy Moroccan troops as UN peace-keepers in Haiti in 2010 because the Haitian government recognizes the SADR, 63 and in 2008 it sent arms to the government of Comoros to resist putschists, thus pressuring the government of Seychelles to derecognize the SADR. 64 Its efforts have yielded fruit. Since the ceasefire in 1991 and during the thus far unproductive UN mission to hold a referendum, ten countries withdrew recognition in the 1990s, 57 Ibid. 58 Jackson 2009. 59 Millard 2009. 60 Grenadan Government 2010. 61 Ranneberger 2007. 62 Guguen 2013. 63 Kaplan 2010. 64 Jackson 2009. 0 10 20 30 40 50 60 70 80 0 2 4 6 8 10 12 14 16 18 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Countries per Year Recognizing Derecognizing Net • 1991 Ceasefire 24 14 in the 2000s, and at least nine more since 2010. States’ waffling on diplomatic recognition demonstrates the bi-directional nature of international sovereignty. A model of sovereignty more sensitive to this kind of fluctuation across this continuum would both connect important parts of the literature and allow new questions to be explored. For example, do the modes of operation identified by prominent scholars on international sovereignty 65 apply to cases that are further to the left on the continuum? Second, do states make decisions in the same way if they are moving towards the left vis-à-vis current and aspiring states, delegitimizing their bases for claims of sovereignty? 2.3 DEVELOPING A NEW MEASURE If international sovereignty is a continuous dimension, it requires the development of a new, continuous measure. A measure meant to complement official recognition should include other elements of the complex relationship international actors often have with separatist movements in other countries. I conceptualize support for recognition as a latent trait, assuming there is an underlying characteristic of support for recognition that cannot be measured directly, but which influences other measurable traits. Those observed outcomes are used to estimate and compare levels of this characteristic across countries and time periods, as well as develop and test ideas about changes in the theoretically important variable. Figure 2-6. Political decisions imply recognition and build legitimacy of claims. Figure 2-6 illustrates my conceptualization of recognition. Official, diplomatic recognition is one of many third-party actions that stem from an underlying perception that can also be fairly characterized as a form of recognition. Note that the order of these actions is not meant to reflect an empirically founded characterization of the order of decisions a state makes. This only reflects the idea that these actions require some kind of underlying recognition of the self-determination group’s de facto sovereignty or right to autonomy. 65 Krasner 1999; Coggins 2011; Sterio 2013. Underlying Recognition Military Intervention Voting Sanctions Economic Relations Official Recognition 25 Figure 2-7. What official recognition misses. As a brief preview of the utility of this model, Figure 2-7 shows the kind of variation one misses by focusing on international legal sovereignty as only reflected in diplomatic recognition. The US scores for four self-determination movements are displayed throughout the 1990s. Not one of these movements is diplomatically recognized by the United States, but there is substantial variation in the level of support each receives in terms of diplomatic, military, and economic aid. Moreover, one of them sees a large change in score during this period; specifically, after the Kosovo conflict escalates substantially in 1996, the US moves towards recognition (eventually extending official recognition in 2008). Detecting these subtler shifts in policy, short of diplomatic recognition, both justifies and allows for addressing the question of international sovereignty with greater emphasis on nuance in foreign policy than has been possible in prior work. I estimate this “continuous recognition” latent variable using five international diplomatic indicators and two military indicators. Following the examples of Schnakenberg and Fariss 66 and Reuning et al., 67 recognition is modeled through a dynamic item response theory (D-IRT) framework. IRT models were first developed to estimate student ability, an unobservable characteristic, based on answers to questions of various difficulty, observable responses correlated with student ability. In the model developed here, each manifest variable is analogous to a correct answer to a test question. In the testing environment, each question has a certain difficulty, with the assumption that a student’s ability to answer questions of higher difficulty implies an ability to answer less difficult questions. In a model of recognition, the analogue would be that some of the actions third-party states take are more difficult than others, and so fewer states take them. 66 Schnakenberg and Fariss 2014. 67 Reuning, Kenwick, and Fariss 2016. 26 This type of model does not require that one specify how important each variable is for estimating the latent trait. Neither does it assume equal weights for all variables, as some indices do. Rather, it uses the provided data—responses to test questions in the classical context, third- party decisions in this context—to estimate two qualities, or parameters. The first is the “difficulty” of an item; some questions are harder than others, and some foreign policy decisions are rarer than others. Questions that are answered wrong more often, and rare favorable (to separatists) decisions, are considered more difficult. In other words, there are certain foreign policy decisions that happen more often; those would be classified as less difficult, and the estimated level of difficulty is informed by the data. The second is how well a variable “discriminates” among cases—harder questions should classify students of higher ability with greater accuracy, and tougher foreign policy decisions towards separatists are more informative of how much a state supports recognition. The general idea behind the IRT model in this context is that it allows one to look at the behavior states exhibit in a set of foreign policy actions, taken to be correlated with a latent property, support for recognition. The model lets the provided data sort the third-party positions along the continuum based on how favorable their foreign policy activities are for the aspiring states or how unfavorable they are for host states, in that year and in prior years. Figure 2-8 illustrates the intuition behind this model. Latent variable models also assume local independence among indicator variables; any relationships among variables within the model are due to variation of the latent variable, not to causal relationships between the indicator variables themselves. An example of this assumption in this project would be that a country’s intervention in another’s secessionist conflict cannot cause favorable votes in the UN; instead, the decision to intervene and the decision to vote for favorable measures both stem from an underlying trait of latent recognition of the separatist group. In essence, the variables are only correlated based on their shared relationship with the underlying measure and not because one variable causes another. A classic example would be that an answer to a question on a standardized test does not cause another answer. Rather, the scores on two questions are correlated because both capture ability, the unobservable characteristic. Underlying Recognition Military Intervention UN Voting Sanctions Figure 2-8. Latent variable model of recognition illustrated. What we can see Assumed Causal Relationship Latent Recognition We can’t observe What we estimate 27 Table 2-1. Indicators for continuous recognition Item Explanation Diplomatic Recognition 68 Official acts of recognition UN Voting 69 UN votes favoring separatist territory in any way Diplomatic exchange 70 3 rd party visits with leaders in separatist territory Sanctions 71 Participation in sanctions against host state or separatists Foreign Aid 72 Levels of foreign aid to host state or separatists Military Intervention 73 Level and direction of intervention in conflict Arms transfers 74 Aid to host state or separatists in form of arms Table 2-1 displays the indicators used to estimate this model. These variables are chosen because: (1) they each constitute a unilateral decision on the part of the third party towards the self-determination conflict’s actors, (2) they require the third party to locate—to “recognize,” in the plain English sense of the term—a non-state actor with whom they can interact and/or support, and (3) these actions advance or degrade the aspiring state’s “eligibility” for official recognition. The data were derived from 10 sources. Six datasets were used for international diplomatic and non-military variables:  Coggins 75 and Griffiths and Butcher 76 secession and separatist movement data. This data was used to establish the set of cases from 1945-2015 included in my dataset. If a case was included in either of these datasets, it was included as a conflict-year.  Coggins’ recognition/conflict data. The Coggins data also includes: a country-year observation for each major power; a variable indicating whether that country officially recognizes the separatist group; a variable indicating whether any other major power recognizes the separatist group.  United Nations General Assembly Voting Data. 77 The raw data gives a description of each roll call vote before the general assembly and how each member state voted. All observations with any mention of one of the self-determination movements in either the Coggins (2014) or the Griffiths (2014) data were extracted. Depending on how the host state voted, each of these roll call votes was coded as either favorable or unfavorable to the separatist movement. In most cases a ‘yes’ vote was counted as favorable. Then, all years with multiple roll call votes for the same movement were collapsed so that each voting state had a statistic for the percentage of their votes that were favorable and 68 Coggins 2011; Fazal and Griffiths 2014. 69 Voeten 2012. 70 Bayer 2006. 71 Morgan, Bapat, and Kobayashi 2014. 72 USAID 2017. 73 Pearson and Baumann 1993; Kisangani and Pickering 2015. 74 SIPRI 2016; Harbom, Melander, and Wallensteen 2008. 75 Coggins 2014. 76 Griffiths and Butcher 2013. 77 Voeten 2012. 28 unfavorable to the separatist group. This data was considered exhaustive, meaning a lack of observation in the data was taken to mean there was no activity concerning that conflict in that year’s UN sessions.  Correlates of War (COW) diplomatic exchange data. 78 This data indicates whether and how each state is diplomatically represented in each other state. There is one observation every 5 years, and 3 variables: diplomatic representation level of side 2 at side 1; level of side 1 at side 2; and a binary variable indicating any exchange between sides 1 and 2. This data was considered exhaustive.  Threat and Imposition of Sanctions (TIES) data. 79 Each observation in this data counts: a case of international economic sanctions; start and end year; offending state country code; “sending” state country codes; the “primary sender” country code; final outcome; and a series of variables describing the type, scope, estimated costs, and domestic parties involved with each sanction-case.  US AID. 80 This data covers: the amount and type of military and non-military aid to each country in each year; which government party, office, or agency dispersed the aid; and the official purpose of the aid. This data was considered exhaustive. Four datasets were used to assess military relations with host and aspiring state actors:  International Military Intervention. 81 This data operates at the case level and runs from 1946-2005. This data includes: intervening and receiving state codes; start and end dates; level and direction of intervention, whether supporting or opposing state or non-state actor; and a series of variables describing the characteristics of the intervention. The data was expanded to a dyad-year structure.  Uppsala Conflict Data Program (UCDP) data on external support. 82 This country-year data includes: the sending state; the location of conflict; type of intervention or support; and the recipient of the assistance, whether state or non-state.  Military Intervention of Powerful States. 83 This data operates at the case level, detailing the intervention of the American, British, French, Chinese, and Russian governments into the conflicts of other states. It includes: the start and end date of intervention; location of conflict; the recipient of assistance, whether state or non-state; the target of assistance, whether state or non-state; the type and level of intervention; and several variables describing the nature of the conflict.  SIPRI arms transfer data. 84 The raw data takes the form of a “trade register” document. From this document, this was collected for each case: sending party; receiving party; weapon description; number ordered and delivered; year ordered and delivered; and description of the transaction. An indicator variable was generated to specify whether the transfer took the form of “aid,” which means the weapons were delivered with no monetary exchange. 78 Bayer 2006. 79 Morgan, Bapat, and Kobayashi 2014. 80 USAID 2017. 81 Kisangani and Pickering 2015. 82 UCDP 2017. 83 Sullivan and Koch 2009. 84 SIPRI 2016. 29 It is important to note that, although I use Coggins’ data for its sample and for a few other conflict-level and recognizer-level variables, I do not incorporate official recognition itself into my model. The theoretical justification for this omission is that the foreign policy activity I explore falls to the left of official recognition on the continuum. Recognition is a special decision, only given reluctantly even by very sympathetic third parties because of its high costs. Though related, it is distinct from underlying recognition as I characterize it. In addition, my measure and tests are meant to complement prior work on official recognition, and to include recognition in my estimates of this measure would undercut this process. Assembling these 10 data sources into a single dataset required converting it to take the unit of analysis to be third-party–conflict–year. This is because some host states face multiple separatist movements in the same years and some separatist movements have conflicts with multiple states. For example, Georgia–South-Ossetia and Georgia–Abkhazia are considered different conflicts so that US–Georgia–South-Ossetia–1995 and US–Georgia–Abkhazia–1995 have two separate observations. At the same time, Turkey-Kurds, Iran-Kurds, and Iraq-Kurds are all considered separate conflicts, and there are separate respective US–Turkey–Kurds–1995, US– Iran–Kurds–1995, and US–Iraq–Kurds–1995 observations. In this manner, the data used to estimate the latent variable model is dyadic in structure, even if not resembling the traditional dyad-year form IR scholars may be accustomed to. From these aggregated data, the following “manifest” variables (“test questions”) were used for each third-party–conflict–year: logged total foreign aid from 3 rd party state to host or aspiring state in constant 1995 USD; percentage of relevant UN roll call votes favorable to aspiring state; percentage of relevant UN roll call votes unfavorable to aspiring state; binary indicator for military aid sent to aspiring state; binary indicator for sanctions against host state for reasons related to separatist movement; and binary indicator for any diplomatic representation between third party and aspiring state. Note that the first three variables are continuous, and the latter four are binary. Recognition is modeled as a latent trait, θit, which exists for each third-party–conflict dyad, or unit, i = 1, … , N across each time period t = 1, … , T. θ is assumed to determine the values taken by a series of manifest variables, y, with k = 1, … , K representing the number of items observed. yitk is the value observed for the manifest variable k for unit i in time t. The difficulty and discrimination parameters αk and βk are estimated for each manifest variable. In a logistic regression, these parameters would be analogous to the slope and intercept. In the testing environment, β estimates the difficulty of each of the questions, and α estimates the degree that item discriminates between students in different regions on the latent ability continuum. I use a dynamic model to address temporal non-independence in the data. That is, I assumed each unit’s latent trait is autocorrelated over time. By contrast, a traditional (static) IRT model bases estimates for θ only on the items in that observation, not those of prior years. I base the priors for the latent variable θ on the estimated value for the previous year of the same unit. For the first observation period for each unit, the prior on θ is a standard normal distribution. In each subsequent time period, the prior is normally distributed with mean θi(t − 1) and an innovation variance σ, which is estimated from the data and informed by a gamma distribution. It is assumed to be the same for all units in the model. Because this model assumes the latent trait is correlated over time, some sudden jumps in the latent trait will be smoothed out in the estimation, rather than estimated per observation-year, as a static model would. 30 The priors are summarized as follows: θi1 ∼ N(0, 1) ∀i ∈ [1,N] θit ∼ N(θi (t − 1), σ) ∀i ∈ [1, N] and ∀t ∈ [2, T] σ ∼ Γ(1, 1)I(σ > 0) α ~ N(0, 10) β ~ Γ(1, 1) for dichotomous variables β ~ N(0, 10) for continuous variables Here, I is an indicator function taking the value of 1 when the relationship of the function is true, and 0 otherwise. Putting it all together, the likelihood function is in this form: ℒ = Λ ( 𝛼 − 𝛽 𝜃 ) 1 − Λ ( 𝛼 − 𝛽 𝜃 ) , , Λ is the logistic function. I implement a Markov chain Monte Carlo (MCMC) algorithm using R and a Bayesian software and R plugin called STAN. For the latent value θ, I take the mean value per third-party–conflict–year across four chains of 3,000 sampling iterations preceded by 800 initial iterations which calibrate the model to the data and then are discarded (usually called “burn-ins” or “warmup draws”). I turn now to a brief demonstration of the behavior of the model in a few aspiring and third-party cases before moving on to quantitative tests comparing this model of recognition to that used in prior work. 2.4 TRENDS IN RECOGNITION Because the model takes a standard normal distribution as the prior for the estimate of latent recognition (θ), the estimates generally fall between -2 and 2. In this context, the most suitable interpretation of the estimated value 0 is that the weight of favorable actions in the data is roughly equal to those that are unfavorable, as far as the aspiring state is concerned. Figure 2-9. Visual aid to understand latent recognition scores 2 - 1 - 0 - -1 - -2 - Aspiring State Host state More Recognition Less Recognition Year = … 1994 1995 1996 1997 1998 1999 … The dyadic relationship Third-Party State 31 Figure 2-9 serves as a visual aid for understanding these scores. It is roughly analogous to the net approval ratings one often sees in polls of opinion on the president, wherein 0 means that the same percentage of constituents approve as disapprove. Note that, within the triad, the aspiring and host state actors are collapsed into one “half” of a dyad, and that the score applies to the third party’s evaluation of the conflict overall. Also, a new score is generated for each third party for each year the conflict continues. Figure 2-10 is a plot of all third-party positions on all self-determination conflicts across all years in the data. The black band just below 0 shows that for most third parties, the estimated baseline towards these movements is in the negative, but still close to favoring no party. Figure 2-10. All estimated third-party scores over all conflict-years These scores are updated when more data is added for each third-party–conflict–year. Moving away from 0, there is substantial variation in this data, enough to provide new answers to important questions. It should be noted that the density of cases in the subzero band makes statistical tests for relationships difficult. The majority of cases do not see much variation. This is part of why the coefficients reported shortly are relatively small. It is also why any significant relationships found between conflict-level variables and movement towards and away from recognition are of substantive importance. This measure incorporates a broader swath of decisions than just the one difficult decision of recognition, allowing for the detection in smaller shifts in overall orientation. A useful example of what this looks like over time is the Palestinian case. Figure 2-11 displays the positions of two sets of third parties: the medium sized powers Mexico, Brazil, Sweden, Turkey, Egypt, Israel, Saudi Arabia, India, and Indonesia; and the major powers USA, UK, France, Germany, Italy, Russia, China, and Japan. Several features are evident in this graph. 32 Figure 2-11. Latent recognition of Palestine First, since 1967, most countries have converged on a similar score towards Palestine. They tend to favor Palestine overall, with positive scores throughout most of the period. Second, this score has been, on average, increasing steadily over time. Third, medium sized powers have tended to favor Palestine in their positions more than major powers have. It is also helpful to discuss briefly what goes into these scores for this case to get a sense of what state behaviors these scores reflect. Figure 2-12 breaks down the scores a bit further, with special attention paid to the scores of Mexico, Sweden, Saudi Arabia, the US, Russia, and China. Figure 2-12. Latent recognition of Palestine by six powers Looking first at the medium powers, Mexico’s score starts off near 0, voting in the UN with Israel or abstaining before switching to mostly pro-Palestine votes through the 70s and beyond. Sweden has a similar trajectory, but voted with Israel more often and transitioned later. Saudi Arabia starts off voting against Israel about 80% of the time (with some abstentions), but quickly reaches a point where it is voting against Israel a nearly every possible point. Additionally, Saudi Arabia also participates in sanctions against Israel. For the major powers, the US starts out as a Mexico, Brazil, Sweden, Turkey, Egypt, Israel, Saudi Arabia, India, Indonesia US, UK, France, Germany, Italy, Russia, China, Japan 33 having a more pro-Palestinian stance, participating in sanctions against Israel and voting with Israel only a part of the time, but as time advances, it no longer participates in sanctions (except in the early 2000s for a few years), and it sends more and more military aid and lends more and more support to Israel in the UN. Both China and Russia’s scores are basically determined by UN voting patterns; they abstain for several years, but gradually vote more with Palestine. To speculate briefly, it may not come as a surprise to anyone that while other countries have steadily tended towards lending more credibility to Palestinian claims of statehood, the level of latent recognition of Palestine by the US has dropped steadily throughout the period. Figure 2-13. Latent recognition of Western Sahara by six powers Looking now to the Western Sahara case in Figure 2-13, Saudi Arabia has provided arms and UN support to Morocco throughout this period, while Mexico and Sweden have tended to vote against Morocco regarding Western Sahara. Mexico has also officially recognized Western Sahara and has had diplomatic channels with the Saharawi government throughout the period. The estimates for Russia and the US reflect the fact that Russia and the US have funneled weapons to the Saharawi (and Algerian) and Moroccan governments, respectively, as well as generally voting along those lines in the UN, even while neither has never taken a position other than neutrality in the conflict. 2.5 REVISITING HYPOTHESES ABOUT RECOGNITION With evident trends in recognition at both the case and the third-party level, it is possible to compare the θ estimates of the latent recognition model with official diplomatic recognition as a determinant dichotomous measure. To do this, I revisit a few of Coggins’ 85 hypotheses regarding major powers’ decisions to officially recognize aspiring states. For the purpose of this chapter, I maintain a neutral stance regarding expectations of whether tests using this latent measure of recognition should yield the same findings as those using official recognition. There is one exception: with this measure, third-party concern about precedent should not be a factor. No 85 Coggins 2011; Coggins 2014. 34 single decision captured in the measure should implicate the creation of a precedent regarding either other self-determination groups in the world or those in the third party’s own territory. Thus, while tests of official recognition show a strong association between a third party’s number of domestic challengers and an aversion to recognition, if my measure is valid, I should see a null finding. Furthermore, rather than using my entire set of estimated scores, I only analyze major power latent recognition in these cases, so that my sample matches that of Coggins. I examine the following hypotheses from Coggins (2014): Domestic-Level Hypotheses H1: Institutionally empowered groups are more likely to receive Great Power recognition. H2: Materially stronger groups are more likely to receive Great Power recognition. International-Level Hypotheses H3: Great Powers with secessionist challengers of their own will be less likely to recognize secessionists in other states. (I hypothesize a null finding here) H4: Great Powers will be more likely to recognize secessionists when another Great Power or Powers have already done so. H5: Great Powers with a conflictual relationship with a home state will be more likely to recognize its secessionists. H6: Great Powers with a friendly relationship with a home state will be less likely to recognize its secessionists. Coggins (2014) tests these hypotheses through several domestic and international level variables. “Ethnic federation” is a dichotomous variable indicating whether a given separatist group is an ethno-federal unit. These units should have stable boundaries and certain administrative capacities already stable and well developed. Hypothesis 1 predicts a positive relationship between this variable and levels of recognition. The two dichotomous variables “Violence Level” and “War Victory” were chosen to capture material capability of the separatists. Violence Level indicates that at least 1000 battle deaths occurred in that year of conflict, 86 and War Victory indicates that the separatists have wrested control of the territory through a military victory. Hypothesis 2 would predict a positive relationship for both these variables. The first international level variable tests Hypothesis 3: “Number of Challengers” indicates the number of secessionist or separatist challengers the third-party state faces at home in that year, and sees a negative relationship with official recognition. However, to reiterate, the theoretical justification for my measure of recognition would lead to an expected miniscule or null finding, if my measure avoids the concern for precedent-setting that official recognition is subject to. 86 UCDP/PRIO 2008. 35 “Prior recognition” is a dichotomous variable indicating that at least one other great power has officially recognized the separatists that year. It should be associated positively with recognition to test Hypothesis 4. For Hypothesis 5, the variable “MID” is a dichotomous variable indicating that either the third party or the host state started a militarized dispute with the other (note that the aspiring state is not party in this variable). This variable should have a positive relationship with recognition. Finally, the two variables “Mutual Democracy” and “Mutual Autocracy” are both dichotomous variables indicating the same regime type (and theorized shared security concerns) between the third-party state and the host state; both should see a negative relationship with recognition to lend support to Hypothesis 6. These variables are all inherited from Coggins’ (2014) dataset. Table 2-2 displays side by side results of the two tests. The first column shows the results of a fixed effects regression model with robust standard errors clustered at the dyad level, using the latent recognition variable explained above. The second column displays the Cox Hazard Ratio tests from Coggins (2014). In this kind of test, numbers above 1 are associated with a k- fold increase in likelihood of recognition; numbers below 1 are associated with proportionally lower likelihood of recognition. For the purposes of examining these hypotheses, Cox hazard rations above and below 1 are analogous to positive and negative coefficients, respectively. I have included signs in parentheses alongside the Cox estimates to make comparison across the two models more intuitive. Table 2-2. Comparing effects on latent recognition in a dyadic fixed effects model and Coggins' (2014) Cox hazard model Model: Dyadic Fixed Effects Coggins (2014) Cox Hazard Ratio a Domestic Ethnic Federation -.12 (.02) *** (+) 5.35 *** Violence level >999 deaths .09 (.02) *** (-) .7 *** War Victory .12 (.04) ** (+) 5.33 *** International Number of Challengers .002 (.000) ** (-) .76 b Prior recognition by other great power .003 (.002) (+) 28.16 *** MID b/t 3 rd party and host state .03 (.01) * Mutual autocracy -.02 (.01) *** (-) .2 *** Mutual democracy -.02 (.007) (+) 1.22 Constant -.15 *** N c 1947 1334 p-values are * .05, ** .01, *** .001; standard errors in parentheses a Ratios >1 are interpreted as an increased likelihood of recognition; <1 a decreased likelihood b Coggins’ was a dummy variable for “unusually high number” of challengers c Difference in Ns due to different clustering techniques between Cox and fixed effects models. Beginning with the three domestic level variables, I find strong countervailing evidence against H1 and equally strong evidence in favor of H2. When a self-determination group is an ethnic federation, we see strong movement away from recognition by major power third parties. However, high levels of violence and victory in civil war are associated with movement towards 36 recognition, by my measure. Interestingly, Coggins did not find the effect she expected for violence, but with this measure we do see the expected effect. 87 I find a weak positive relationship between number of domestic challengers and movement towards recognition. This finding provides a glimpse of the utility of this measure of recognition as a complement to official recognition. If precedent-setting were a concern of third- party states dealing with self-determination movements of their own, we would expect, as Coggins does, a negative effect. I do not find the predicted null effect here, but a weak positive effect. The explicit concern for precedent held by many states 88 does not inform these other foreign policy decisions. 89 Prior official recognition by another major power does not have the expected positive effect on latent recognition. This make sense. Coggins theorizes a cascade effect with official recognition, in which one major power “breaking the ice” by recognizing a new state (e.g. when Germany recognizes Croatia and Slovenia in December 1991) 90 spurs other governments to make decisions to recognize or withhold recognition. It would seem inappropriate to expect a similar effect for the diffusely constructed measure I use here, built as it is on a broad range of diverse foreign policy decisions. When the third party and the host state are in dispute with each other, we see movement towards recognition of aspiring states, a finding predicted by Coggins (but tested elsewhere in her work). Additionally, I find, as Coggins predicts, that third-party autocracies are less likely to move towards recognition when both they and the host state share regime type and assumed security interests. In this model, I thus find support for Hypotheses 2, 5, and 6, and no support for 1, 3, or 4. 87 As a quick aside, the next chapter deals with my theoretical expectations for the relationship between violence and latent recognition, and provides an explanation why the findings diverge between these two measures. 88 Coggins 2011, 433. 89 I offer the conjecture in the next chapter that this is due to a general preference for regional stability by third-party states facing domestic challengers, though I have not tested for this mechanism yet. 90 Libal 1997. 37 Table 2-3. Test of latent recognition—dyadic fixed effects by major power 3 rd parties Major Party: (Dyadic Fixed Effects) USA UK France Russia China Domestic Ethnic Federation -.17 (.04) *** -.22 (.04) *** .22 (.06) *** -.21 (.04) *** -.36 (.03) *** Violence level >999 deaths .18 (.04) *** -.03 (.04) .1 (.07) .07 (.04) * .06 (.03) * War Victory .19 (.05) -.1 (.1) .27 (.2) .18 (.1) .15 (.07) * International Number of Challengers .002 (.002) -.001 (.002) .005 (.003) .002 (.002) .003 (.001) ** Prior recognition -.09 (.06) -.08 (.06) -.15 (.11) -.02 (.05) .07 (.04) MID b/t 3 rd party and host state .01 (.03) -.01 (.04) .09 (.05) -.01 (.03) .05 (.02) * Mutual autocracy --- --- --- -.08 (.02) *** -.06 (.01) *** Mutual democracy .001 (.02) -.09 (.02) *** .04 (.01) * -.02 (.02) --- Constant -.33 *** -.21*** -.04 -.09 *** -.07*** N of dyad 273 219 242 256 269 p-values are * .05, ** .01, *** .001 Table 2-3 expounds on these results, breaking down these estimates by individual observations of five major powers: The United States, the United Kingdom, France, Russia, and China. All cases generally follow the results discussed earlier, but there are a few exceptions. France diverges from the rest of the major powers concerning ethnic federations, supporting Coggins’ finding. The US, Russia, and China carry the effect of high levels of violence on movement towards recognition. 2.6 DISCUSSION AND CONCLUSION How do self-determination groups move from being considered subversive actors to welcome members in the international system? This project aims to provide a new way to answer a part of this question, filling in a gap in the literature on secessionist, separatist, self-determination movements and all sorts of unrecognized and aspiring states. There is rich work on the origins and characteristics of unrecognized states, which have some of the trimmings of the modern state but still lack the general character of international sovereignty. And there is a separate body of work on the domestic and international determinants of diplomatic decision-making by third parties towards both separatist groups and the states that unwillingly play host to them. The other factors involved in third parties’ foreign policy towards separatist groups—military, economic, diplomatic—have generally been set aside, or least not considered as crucial to the advancement and success of the causes of aspiring states. 38 Moving towards a measure of recognition that takes these other factors into account, in addition to examining formal recognition, explores the vast foreign policy gray area between third parties who see only a distant aspiration on the part of a weak aspiring state and those who see international sovereignty as so inevitable that they jump aboard the bandwagon once someone makes the first move. In this chapter, I push for a conceptualization of “recognition” that is closer to its standard definition: not an easily observed, formally prescribed decision, but a reaction on the part of foreign policy markers. I situate prior work within this conceptualization, showing them to have covered important questions but relatively limited in scope, given the breadth of the aspiring state problem. With this in mind, I develop a new measure of recognition based on several kinds of international exchange between host states, aspiring states, and the third parties with whom international sovereignty lies. The primary interest is not the origin of a discrete decision, but movement towards and away from favorability towards separatist groups and the evolution of their “eligibility” for recognition, measured through formal act. With a brief slate of tests using this new measure, I compare this measure to one based only on official recognition. Most notably, I find support for the hypothesis that victory in battle and levels of violence should be associated with movement towards recognition, and the hypothesis that countries that share regime types should tend not to support each other’s self- determination groups. Additionally, hypotheses on official recognition by major powers and number of domestic challengers in the third-party territory have minimal effects on recognition. International responses to separatist, secessionist, and self-determination movements are still poorly understood, with many unasked questions. Do smaller third parties behave similarly to major powers? Do conflict-level variables like nonviolent movements or human rights abuses affect the movement of third parties along this continuum? Conceptualizing recognition as a process deeper than diplomacy will allow future work on the topic to buttress prior work by taking account of the effects of foreign policy decisions subtler and less politically volatile than official recognition. 39 FOULWEATHER FRIENDS VIOLENCE AND THIRD-PARTY SUPPORT IN SELF-DETERMINATION CONFLICTS 3.1 HALF MEASURES Self-determination movements have been making demands of increased autonomy or independent statehood on every continent since the end of World War II. These movements remain, as Callahan put it, “among the most important factors driving international politics,” 91 a predominant source of global instability to this day. With nationalist parties on the rise across the globe, there is little reason to expect any decrease in the salience of self-determination any time soon. Studying the effects third parties have on these conflicts thus persists as a crucial research agenda. Recalling the Kurdish example from Chapter 2, it is puzzling that a third party should support a self-determination movement to this extent, abetting the development of so many trappings of the modern state—decent civilian infrastructure, a professional military, permanent channels of diplomatic exchange—while remaining steadfast in a policy of nonrecognition. Yet, this is hardly the only case with such a trend. Third parties routinely send military and civilian aid to separatists, support them on the floor of the UN General Assembly, and levy sanctions against their opposing governments. However, diplomatic recognition is exceedingly rare. Thus, although these conflicts may result in the de facto cleft of the state—the loss of control of the 91 Callahan 2002, 2. 40 central government over the disputed territory—there it remains, in the no man’s land of partial statehood. The field has a robust understanding of how major international players like the US and Russia treat the decision to confer diplomatic recognition on self-determination groups. Both domestic politics and international context affect this political decision, and a concern for setting precedent rules in both cases. This explains why the decision to officially recognize a self- determination group is so rare. However, we have a less robust understanding about the aims of states that support self-determination groups in many other ways. This chapter begins to fill in this gap, broadly exploring the other foreign policy decisions states make towards these conflicts—military aid, sanctions, IGO voting, and other bilateral decisions. I expand on prior theories utilizing stability-seeking behaviors as an explanation of third-party support, 92 arguing that even as third parties oppose official entry of aspiring states into the system at the end of a long chain of events progressing towards independence, they still assist self-determination groups in other stages of this progression, and they do so in predictable ways. A persistent desire to foster and maintain stability can explain both their support and opposition at separate stages. Partial states like Kurdistan, Taiwan, and Somaliland can be remarkably stable as they are, and their status is relatively more stable than either the violent stages that lead up to territorial capture or the final push into official statehood, which requires a renegotiation of boundaries and the international accommodation of new states. There is a stability “sweet spot” that exists post-conflict and pre-recognition, and third parties assist self- determination groups in reaching it. To demonstrate this trend of partial support, I utilize a latent variable model of recognition that incorporates several unilateral decisions states make vis-à-vis both parties in these conflicts. This measure allows me to detect shifts in policy orientation towards self- determination conflicts and differences in policy between third parties. Moreover, incorporating these other activities allows for the estimation of levels of support from non-major powers, allowing this question to become truly comparative. To my knowledge, this is the first study of third-party recognition or international sovereignty that investigates the support of smaller countries in these conflicts. I show that high levels of violence and rebel victory precipitate movement towards recognition, that third parties punish violence against civilians, that domestic self-determination challenges have no effect on other policy decisions towards these conflicts, and that official recognition by major powers may trigger other forms of support. 3.2 CURRENT UNDERSTANDING Most empirical work on international sovereignty ascribes Krasner’s definition, which hinges on states “recognizing” each other as juridically independent units. 93 Official recognition has been taken to be the main, perhaps the only, empirically observable political act that captures 92 Paquin 2010; Saideman 2002; Saideman 2001. 93 Krasner 1999. 41 international sovereignty. As such, our understanding of international sovereignty and official recognition are largely one and the same. If a self-governing territory is not universally recognized by all other states, it is not considered a state. In the post-WWII order, diplomatic recognition has emerged as one of the most costly and effectual single political statements a state makes in its foreign policy. This foreign policy decision has been subject to extensive scholarly scrutiny, and as a result, we have a well-developed understanding of how modern states—at least the most powerful—make the decision to officially recognize self-determination movements as sovereign. One of the clearest facts about diplomatic recognition is its relative rarity. It would seem that new states are generally only able to garner diplomatic recognition when some other international trend or crisis is ongoing—such as the global push to end colonialism through the 1960s and 1970s or the fall of Soviet Union in the late 1980s and early 1990s. Even those movements whose progress was quickened by such massive international shifts have had no guarantee for universal recognition and sovereign statehood, as evident in the many current “partial states” left from these eras—Western Sahara, Kosovo, Abkhazia, or South Ossetia—or those successful cases that nevertheless took decades of negotiations or thousands of deaths to achieve independence—East Timor, Namibia, or Eritrea. Once drawn, international borders are a remarkably durable institution, 94 bringing stability to interstate relationships while making revisions nearly impossible. When states have unilaterally extended recognition to self-determination movements, they have done so in predictable ways along two dimensions. The first is the international context. How secessionist and separatist conflicts progress and how other states respond both largely determine the likelihood of a given third party’s official recognition. 95 When a separatist military is able to fight its way to de facto capture and control of the territory in question, third parties are more likely to recognize, owing to the demonstrated capability to establish and maintain order. 96 States also tend to coordinate in recognizing new states. 97 To defy the status quo through unilateral recognition can destabilize relations between the recognizer and other states: the compromised host state may respond by cutting off diplomatic relations or even retaliating with military action, 98 and other third parties may also have strong negative reactions, either unilaterally or within IGO procedures. 99 There is a high diplomatic cost to being the first to recognize, so states tend to do so together, especially if they do so without the host state’s consent. Furthermore, although states sometimes utilize unilateral recognition as a sort of diplomatic proxy to punish rival states, this has been constrained to a few cases, primarily former 94 Simmons 2005. 95 Coggins 2014. 96 Ibid. 97 Ibid. 98 Goebel 1915; Cole 1928; Coggins 2011. 99 Libal 1997, cited in Coggins 2011 42 Soviet entities following the collapse of the USSR. 100 Even facing enemies, states have not been willing to upset the still waters of the international sovereign system. This is perhaps because states’ domestic political stability is, in many ways, bound to the international. Many states—including Russia, China, the US, the UK, and France—face one or more domestic claims of self-determination to this day. In such cases, they are doubly unwilling to confer recognition, 101 for fear they might inspire reciprocal recognition of their own rebels. They are still marginally more likely to recognize self-determination groups conflicting with rival states, 102 but in general, each state behaves as if any precedent it sets can come around and result in domestic instability. All these motivations can be understood as coming from a profound concern for stability in the international system, 103 which requires maintaining the established international institutions as they are to the extent it is possible. In the current system, state death is much less likely than it was in centuries past; 104 borders, once drawn, are durable. In liberalist descriptions of the world order, these international institutions, and evident respect for a status quo of territorial integrity and diplomatic non-intervention, are ultimately responsible for both the relative state of peace the world now enjoys and the unprecedented levels of international economic, diplomatic, and cultural exchange. 105 Paradoxically, the longer our borders endure, the easier it is to cross them. The corollary paradox introduced in the Kurdish case is the proliferation of political entities that look and behave very much like states—with well-trained militaries, robust civilian infrastructure, and deep channels of diplomatic, and perhaps economic, exchange. The “organized hypocrisy” 106 of international legal sovereignty has set the bar for state entry high, and has made it easy for relatively few players to veto any revisions entirely. As a result, it is possible for cohesive political groups to obtain a high degree of autonomy and pursue relatively peaceful relations with neighbors and trading partners with little or no chance of invitation to join the club of states. Although gaining de facto control of the disputed territory does make recognition more likely, it is by no means a guarantee. The international accumulation of these “partially recognized states,” “unrecognized states,” “pseudo-states,” and “quasi-states” has, to a minor extent, mirrored the steady growth of the number of UN members. In the international legal sense, this could be characterized as follows: states seem to favor a “constitutive” approach to recognition over a “declaratory” approach. 107 The constitutive theory of recognition puts forth that aspiring states only become states when the other states recognize them as such. Statehood is not automatic, and international legal sovereignty does not 100 Rich 1993. 101 Coggins 2011; Coggins 2014. 102 Coggins 2014. 103 Paquin 2010. 104 Fazal 2007; Fazal 2004. 105 Russett and O’neal 2001. 106 Krasner 1999. 107 Grant 1999. 43 necessary follow from domestic sovereignty. A declaratory theory of statehood is based primarily on successful autonomous assertion of authority over a territory. Statehood is independent of recognition, based on fact, not the discretion of third parties. Legal scholars tend to favor this view, 108 this idea was enshrined in the Montevideo Convention of 1933, 109 and the International Court of Justice has ruled along these lines. 110 Still, many states, including and especially aspiring states, treat recognition as if it is an integral part of statehood. 111 Further evidence of a predominance of constitutive approaches on the ground is the maintenance of “statehood” for governments that have lost all domestic authority, inside whose boundaries there may be many factions competing for authority. In a predominantly declaratory world, states should go extinct. That is, once a government loses the components of statehood—a monopoly on violence, legitimately applied within a conscribed territory—a declaratory theory of recognition would imply withdrawal of recognition of certain governments (though not necessarily the borders). In a constitutive world, once boundaries are drawn, they remain legally determined as long as the international community enforces them. Where the declaratory theory of statehood becomes useful is in thinking about unrecognized states. Perpetually unrecognized statehood can be a remarkably stable outcome. 112 In these conflicts, the self-determination group typically seeks nothing less than sovereign independence, and the central government seeks full legal control of the territory in question. However, self-determination groups fighting for statehood often stall upon becoming “contested states”, and “states of exception.” 113 Mandel 114 points to a pattern of non-state groups providing better stability and more complete provisions of public goods to their localities than the central government in their respective states. Caspersen 115 echoes this pattern in her exhaustive treatment of unrecognized states, calling into question conventional notions of sovereignty and asking whether these situations are necessarily doomed to relapse into violence. She outlines the functions of the state that can be fulfilled without recognition, details the power uncertainty often in place, and makes policy recommendations for how third parties should treat these entities. Many others’ treatments of these territories 116 seem all to subscribe to the same pattern of classification. They informally place these movements on a continuum ranging from weak unrecognized states to strong, mostly autonomous regions. These judgments have been made informally based on a variety of intuitive measures: levels of conflict, public goods provision, trade with other countries, popular legitimacy of security forces, education levels, and many others. Taiwan, for example, has a strong democratic government, a well-trained military, high levels of international trade, and a highly educated populace. Nagorno-Karabakh may have well 108 Brierly 1963; Chen 1951; Brownlie 2008; Harris 1979. 109 Montevideo Convention 1933. 110 Oda et al. 1996. 111 Coggins 2011. 112 Buzard, Graham, and Horne 2017; Caspersen 2012. 113 Caspersen 2012; Geldenhuys 2009; Biswas and Nair 2010. 114 Mandel 2013. 115 Caspersen 2012. 116 Harvey and Stansfield 2011; Closson 2011; Geldenhuys 2009. 44 maintained roads and schools, but has decidedly higher levels of crime and is considered at greater risk of relapsing into conflict. While this subcategorization of unrecognized states is useful for understanding the problem, it is also very static in its view of the state. The notion that there are real, material benefits to being a full-fledged member of the state system 117 is universal throughout this literature, but those benefits have been treated as all coming automatically with statehood, and all only after that state is accepted by all. That these movements have relations with states is not ignored—it is impossible to discuss Taiwan’s status without pointing to a robust foreign policy apparatus—but the effects of the in-between stages of legitimization and recognition are generally unexplored. Meanwhile, the international community generally seeks status quo stability, the control of the territory by the host state. In their formal model of unrecognized statehood equilibrium Buzard et al. 118 claim third parties will not support self-determination groups unless they are “patron states” that have direct reason to see the group succeed: to impose costs on the host state; ethnic solidarity with the self-determination group; or hope of eventual annexation. The international community wants to avoid “the likelihood that one of the inside actors chooses to fight,” 119 and so will only invest resources to maintain the status quo or cause the secessionists to cede their demands. Self-determination groups and established governments agree that “without external legitimacy, an actor cannot really be considered a state.” 120 Universal diplomatic recognition and an open invitation to join international organizations constitute the end goal sought by most secessionist and separatist movements. 121 It is perhaps for this reason that the “success” of these movements is defined—by secessionists and by scholars of secessionism—as the attainment of this ultimate goal. 122 The literature organized around the study of “unrecognized states” prioritizes international sovereignty as the defining characteristic of these problematic states of exception. 123 They do not have international sovereignty because they do not have recognition. 124 What they often do have is a degree of domestic sovereignty, and they often have external actors to thanks for it. Third-party states often lend aid to self-determination groups during conflict. Heraclides’ 1990 study found bilateral involvement of at least 72 countries in just seven secessionist conflicts. 125 (As an aside, his study also finds that third parties are not dissuaded by concerns for precedent from involving themselves in foreign conflicts of secession in manners short of diplomatic recognition, a key prediction of the theory developed here.) 117 Fazal and Griffiths 2014. 118 Buzard, Graham, and Horne 2017. 119 Ibid., 592. 120 Coggins 2011, 461. 121 Coggins 2014. 122 Heraclides 1990. 123 Fabry 2010; Griffiths 2016; Fazal and Griffiths 2014; Harvey and Stansfield 2011; Caspersen 2012. 124 Krasner 1999. 125 Heraclides 1990, 352. 45 Third parties often get involved to militarily and diplomatically support co-ethnic host or aspiring state actors across borders, 126 constrained to varying degrees by domestic institutional structure and by the ethnic makeup of their own populations. 127 Democracies have also been found to support each other militarily when facing separatists, compared to autocratic governments. 128 Secessionist groups in “dangerous neighborhoods” have been found to be more likely to receive support than those in more stable regions. 129 And rebel groups who are moderately strong—not considered weak or strong—are the most likely candidates to receive aid from third parties. 130 Several features of intervention have been shown to affect conflict termination. First, the type of support is a key influence. Sawyer et al. show that highly fungible forms of external support, such as money, have a deleterious effect on the probability of conflict termination, relative to less fungible forms, like foreign troops. 131 When neighboring states provide the use of adjacent cross-border territory, it can sustain insurgency indefinitely, 132 something that might be considered deterministic for conflicts like that involving Western Sahara. 133 The scale of the intervention can also have its own effects on the duration of conflict. 134 When states get involved in foreign conflicts for reasons unrelated to the goals held by the host or aspiring state actors, it has an independent negative effect on conflict termination. 135 The field has a fairly well-developed understanding of when and why third parties, especially major powers, involved themselves in self-determination conflicts militarily. However, aside from Heraclides, 136 that understanding has not been expanded to include the broader range of nonmilitary forms of advocacy third parties sometimes resort to. Changes in overall policy stance towards these conflicts are not well understood; when they are explored, it is primarily major powers that are studied. Despite the rich work we have on diplomatic recognition and foreign intervention in secessionist conflicts, it is fair to say that a persistent gap in our understanding of third-party roles in these conflicts remains. This is true for the questions of how and when third parties create and change policy towards self-determination conflicts; how they influence self- determination groups’ domestic sovereignty, legitimacy, and eligibility for diplomatic recognition; and what are the effects of the conflict itself on third-party policy. The rich discussion on the attainment of international sovereignty by self-determination groups may be complimented by a shift of empirical focus on the contributions third parties 126 Carment and James 1996; Carment and James 1997; Saideman 2002. 127 Carment and James 1996. 128 Bélanger, Duchesne, and Paquin 2005. 129 Saideman 2002. 130 Salehyan, Gleditsch, and Cunningham 2011. 131 Sawyer, Cunningham, and Reed 2017. 132 Salehyan 2009. 133 Jensen 2012; Zunes and Mundy 2006. 134 Baev 1999. 135 Cunningham 2010. 136 Heraclides 1990; Heraclides 1991. 46 make to aspiring states’ sovereignty by other means. To borrow Krasner’s terminology, 137 it is worth asking how third parties contribute to the domestic sovereignty of separatist governments even while they refuse to confer diplomatic recognition upon them, filling in an acknowledged deficit in the study of self-determination. 138 In so doing, the line between international and domestic sovereignty becomes increasingly blurred. The exclusive focus on diplomatic recognition to date has been a natural choice in understanding international sovereignty in the international order. It has been the primary way of thinking about international sovereignty because the relationship between diplomatic recognition and sovereignty is obvious. When a state is recognized and invited to join the UN as a voting member, international institutions prescribe that state a clear role as an equal among other states. There is no ambiguity on the importance of diplomatic recognition to international sovereignty. However, prior literature has also shown how difficult, costly, and rare is the decision to confer international sovereignty through diplomatic recognition. In a sense, it may be considered the hardest test for international sovereignty; it demands the coordination of many members of the international community, and it implies potentially lethal consequences for bucking the status quo. 139 Few other diplomatic, “soft power” decisions warrant the threats of military retaliation that recognition has earned in the past. 140 3.3 STABILITY-SEEKERS It is safe to say that as a concept, international sovereignty is not entirely captured by diplomatic recognition as a measure. To conceptualize it as such would render it impossible to study the substantive phenomenon I explore here, which is the pattern by which third parties involve themselves in these disputes in other means. There are two reasons to expand the concept of international sovereignty. The first is that these actions short of diplomatic recognition do assist self-determination groups in reaching many of their goals. They enable them to endure conflict longer, 141 and they contribute to building the legitimacy of their claims in the eyes of the world. 142 The second is that, as a concept, diplomatic recognition cannot possibly capture the range of foreign policy stances third parties take towards self-determination groups. It is not as if states go directly from no interaction with or acknowledgement of the aspiring state to full recognition overnight. As in US policy towards Kosovo, there are often many intervening steps that show a gradual increase in perceived legitimacy or sensibility of the establishment of a new state. The question then is, how can these interim steps be mapped onto the underlying concept of recognition—not the political decision, but the reaction to a shifting reality on the ground? In Chapter 2, I have endeavored an answer, and I carry on here. I take external support of domestic 137 Krasner 1999. 138 Cunningham 2014, 14. 139 Coggins 2014. 140 Goebel 1915. 141 Salehyan, Gleditsch, and Cunningham 2011; Baev 1999; Cunningham 2010. 142 Heraclides 1990; Heraclides 1991. 47 sovereignty as a reflection of some underlying recognition of the right to or the legitimacy of the self-determination claim of international sovereignty. Other foreign policy decisions—the enactment of sanctions against states facing these movements, military intervention with boots on the ground or provisions of weapons, votes for elevated status at the UN, or any number of other possible foreign policy decisions in relation to self-determination movements—are actions that make meaningful contributions to the progress of these movements towards statehood. Often, as states militarily engage secessionist groups, they face consequences from other international actors, even while none diplomatically recognize a new state. It is not as if, absent recognition, third parties pay no costs or have no effect on the developments that take place. The US and many European states were heavily involved in the conflict over Kosovo for over a decade before any of them ever made any official acts of recognition. Likewise, the Palestinian Authority has seen gradual progress in legitimation of its claim of independence over the years, sometimes amounting to diplomatic recognition of the 1967 boundaries, but more often taking the form of widely favored UN resolutions and international boycotts of Israel. Such acts of intervention have been treated as unrelated to these aspiring states’ official statuses. To theorize about and measure how states enact smaller shifts in policy towards these groups, I set aside this precedent and take any unilateral decisions that help establish the domestic sovereignty of separatist movements as a lesser manifestation of international sovereignty. Like some prior work, 143 I take status quo stability to be the driving interest for third parties looking on in these conflicts. A few important premises underlie my theory. First and most central, states value international stability. A stability based on the status quo means predictability, and predictability means plannability. Ideally, foreign policy decision-making happens in a stable system; stability is the precursor to open trade and diplomatic relations, reliable military cooperation, and the free movement of people and things. States will generally orient their foreign policy to preserve the stability of the system when everything is well, and will orient it to return to stability when something is off. 144 In the context of self-determination conflicts, there are two main sources of instability states will respond to. The first is the threat to the stability of a state’s bilateral relations with other states and to the system as a whole. This threat comes from adding new members and redrawing borders, and avoiding this kind of instability is the cause of extreme reluctance to recognize new states. The second is the manifestation of instability that comes with violent or protracted conflicts, which cause uncertainty both regionally and internationally. To weigh the first, consider this. One of the most efficient custodians of international stability has been the international institution. The United Nations, and the various smaller diplomatic and economic unions of states, have resolved many of the problems in the anarchic international system, making coordination not only feasible, but even compulsory in some ways. By codifying the rights and privileges of states in the international community, they have downplayed the historical ability of states to accrue power through military conquest—a means 143 Paquin 2010; Grieco 1990; Buzard, Graham, and Horne 2017; Saideman 2001; Saideman 2002. 144 Theoretically, there may be situations more stable than the current status quo. For the purposes of this theory, however, the two concepts should generally align. 48 both unsavory and uncertain—in favor of diplomatic and economic influence. The UN, in its ideal form, was founded to eradicate war and enable peace through cooperation and prosperity. It its actual form, it attracted members not just because it was a shining beacon of peace but because it lays out paths of communication between states and makes for nonmilitary methods of managing conflict. At the same time, by implementing rules and codifying state behavior, international institutions have tied the hands of member states. International laws yield policy-making that is concerned with argument, precedent, and legal justification before third parties. 145 Since “self- determination” first became a part of the terminology surrounding the idea of international institutionalization and eventual decolonization, government leaders have been duly concerned with establishing precedents that any minority group can point to when they make self- determination claims. These concerns are not ill-founded. The norm in the international sphere has been to scale up the rights of “all men created equal” to all groups created equal, deserving the same rights and privileges of inclusive representation in the government. 146 A result of this tendency has been for states and self-determination activists to gaze upon the success and failures of other similar groups and draw lessons for themselves. After all, we are now all subject to the same international laws, rules, and norms. The Chinese government has been openly hostile to most separatists around the world following this line; it is concerned that any proliferation of the application of liberal interpretations of self-determination rights will eventually have strong repercussions for its claims of sovereignty over Tibetan, Taiwanese, and Uighur territories. 147 Likewise, Saharawi students study the South Sudanese and East Timorese cases closely to see if there are any lessons to be drawn about convincing the international community to take their side in their 40-year-old conflict. 148 Another, more direct source of concern for precedent is fear of reciprocation, and the destabilizing consequences that may follow. For a third party to recognize a separatist group against the wishes of the central government is to damage whatever relationship they have with that government. If the US, for example, recognizes any Kurdish sovereign claims, it compromises relations with one of its most valued allies, Turkey. Likewise, for the US to recognize any Chechen claims would be considered an act of provocation of Russia. In either case, it is a bold move that would be considered destabilizing and would likely demand a reciprocal act. It is also destabilizing at the macro level. One of the chief benefits of borders is their immutability, 149 which closes the conversation on where one state ends and another begins. To diplomatically recognize is to encourage revision of this valuable institution and to precipitate challenges and counter-challenges to these revisions that may involve other countries. If that is not enough, each additional state in the UN adds complexity. Thus, in this context of rules and 145 Stone Sweet 1999. 146 van Sickle and Sandholtz 2008. 147 Coggins 2011. 148 Personal interview, 2011 149 Simmons 2005. 49 institutions, reciprocation, concerns over precedent, and stability, to recognize a self- determination group as sovereign is perhaps one of the costliest diplomatic decisions a state can do in the “soft” foreign policy realm. For a self-determination group to attempt to persuade third parties to do so is to ask them to sign on to instability in their own relations with other regional parties, deal with the resulting instability between the host and aspiring state parties, and to add complexity to the international institutions that maintain stability. It may therefore be no surprise that third parties, both individually and as a group, are loath to recognize newcomers. Yet, as Buzard et al. 150 explain, separatist and secessionist conflicts can and sometimes do reach a point of equilibrium characterized neither by full authority of the host state’s central government nor by the establishment of a new, internationally recognized state. Exploring this point brings me to the second threat of instability. First, it is helpful to understand a basic framework of how self-determination groups get off the ground. Take Figure 3-1 to visualize the relationship between any state and any of its self-identified or externally identified minority groups, based on ethnicity, religion, language, or other discrete factors. Theoretically, any of these groups might be seen to constitute a potential self- determination movement, so long as they are somewhat geographically cohesive. This is part of why so many established powers are concerned with the precedent of self-determination. 151 One can identify two planes of risk for the government concerning minority groups: one capturing level of their contentment with the government, and one capturing the relative strength of that group. 152 The lowest risk for the government (and the lowest instability) is in the third quadrant, where it has a content minority with little means or lower numbers with which to challenge the government. The highest risk comes, of course, with a large or strong minority group, malcontent 150 Buzard, Graham, and Horne 2017. 151 Hechter and Borland 2001. 152 One might add size, in terms of population, but I take strength in numbers as part of strength. Highest unrest potential Lowest unrest potential Malcontent Content Strong Weak 1 3 4 2 Low unrest potential Low unrest potential Figure 3-1. Dimensions of minority unrest 50 with the government, shown in the first quadrant. Weak, malcontent groups, and strong, content groups also constitute relatively little risk, since there is some minimum threshold of both strength and malcontentment that must be met to inspire unrest. A government has two basic ways to maintain stability when facing these groups: increase their contentment or decrease their strength. The former can be accomplished through improving or ensuring quality of representation in the central government, increasing autonomy for particularly cohesive, geographically bound regions, or bettering the economic welfare and social infrastructure of the region. These activities can simultaneously decrease group strength by making recruitment for separatist activities less appealing. The central government can also weaken the group less through the carrot and more through the stick. Violent crackdowns on assembly, protest, and other political-symbolic acts—such as Catalonia’s independence referendum in October 2017 153 —are another way to make group identification and participation in political activities less appealing, an attempt to make further recruitment difficult. If these efforts fail, and conflict escalates, instability increases, in the state itself and in the surrounding region and beyond. This is the second threat to stability that third parties pay attention to. One can think of separatist conflicts as having several stages of development, in a way that is somewhat analogous to Figure 2-1 in Chapter 2, depicting the various orientations third parties can have of self-determination movements and the conflicts they engender. Figure 3-2 illustrates the relationship between the character of these conflicts and the levels and sources of instability they engender, as perceived by third parties. Figure 3-2. Self-determination conflict status and perceived international instability Quiet Minority Vocal, but few in number Organized protests Repression, two-sided conflict Protracted conflict, high violence Territorial capture by SD group Sustained relations, nonrecognition Recognition Note: This figure serves to illustrate the theory, and is not based on empirical data. At the far left end of the x-axis are those stable situations in which the minority group is either sufficiently repressed or content so as to be quiet, not vying for revision in the system nor a change in boundaries or governance. Both well-represented minority groups, such as the Faroese 153 Huddleston 2017. Instability Local minimum Escalation of conflict, violence, destruction Revision of regional and international order Source of instability: 51 in the Danish society, and harshly repressed minority groups, such as the Tamils in Sri Lanka are in states of relative stability. As malcontentment and strength of the minority group increase, calls for self- determination become stronger and louder. This effect may be particularly pronounced when combined with a marked history of nationalist ideas and identification. Failures to respond adequately lead to further recruitment and protest. When self-determination group leadership reaches this level of strength and precipitates crisis through showy demonstrations or challenges of the state’s strength, the central government must respond, normally relying on a mixture of repression of group activity and accommodation of group demands. Most self-determination groups that have reached either unrecognized or recognized statehood underwent a period of transition from protest to taking up arms against the host state government. 154 This period of two- sided conflict is the most destabilizing, both within and beyond the state. Protracted conflicts with high levels of violence are known to interrupt social and economic activities, destroy infrastructure, destabilize politics in involved parties, erode trust, and lead to general instability. 155 When separatist military forces best those of the state and manage to capture the territory in question, a relatively more stable situation arises. For some groups that have reached this point, it has been accompanied by ceasefire agreements, limited exchange with third parties, reconstruction and repair of infrastructure, and a relative return to stable social and economic activity. According to Coggins, 156 after a group reaches this point, garnering official recognition of their achievement of gaining de facto control of the claimed territory becomes more likely, though still far from guaranteed. This is a point on the road to statehood when self-determination groups can afford to shift more of their efforts from the military to the diplomatic (though diplomatic efforts are often present earlier on). From the perspective of the group, only recognized statehood is a more sustainable, more desirable outcome. However, third parties generally do not share this view. Once violence is abated, they can reorient policy to address the first threat to stability, that of bilateral relations and the international system. Recognition by a given third party creates several new sources of instability that it is unwilling to face. Perhaps most important are the dire implications for relations with the host state government, which is a voting member of the United Nations and may be ally or rival. Second, many third parties themselves have religious, ethnic, and linguistic minorities, and many of those groups already have outstanding claims of self-determination. 157 It is widely perceived that recognition of aspiring states that achieved this status within a territory constitutes a harmful precedent for third party states’ own real and potential claims of self-determination. Third, incorporating new members into international organizations like the United Nations and regional IGOs can engender instability within; it can cause host state members to take actions against 154 Cunningham 2014. 155 Carment and James 1996; Carment and James 1997; Saideman 2002; Saideman 2001. 156 Coggins 2014. 157 Ibid.; Heraclides 1990. 52 other members or to withdraw completely, as Morocco withdrew from the Organization for African Unity (now the African Union) in 1984 after a majority voted to admit the Saharawi Arab Democratic Republic (SADR). 158 This can undercut the credibility and effectiveness of these organizations and poison the political waters in other ways. Finally, if enough third parties officially recognize an aspiring state, it requires new or renewed negotiation of borders. Since borders are a stability-creating institution, to move or remove them necessarily increases instability. Additionally, for separatist groups that span recognized international borders, as the Kurds do, the recognition of that minority group within one state’s land may engender all these forms of instability to a much greater extent. What this means, and what Figure 3-2 is meant to illustrate, is that even though full recognition and membership in the international system may represent the most stable point in the progression overall, the transition from de facto control and nonrecognition to full recognition carries too many opportunities for renewed domestic and international instability to win the support of many third parties. For the international community, this status of de facto territorial control without recognition constitutes a local minimum of regional and systemic instability. Unrecognized statehood is a stability sweet spot. Both a return to conflict with the host state and recognition and integration into the international community require high tolerance of new and increasing regional instability. If states in the international arena are stability-seekers, a few important expectations about their behavior follow. First, we should expect third parties to lend their support to the host state by default, especially in the beginning stages of these conflicts. The point of stability closest at hand for third parties is that of the silent minority. If and when these conflicts escalate into repression and violence, third parties will still view the former status quo as the preferred situation. However, if violence continues to escalate or if conflict drags on, that new local minimum of unrecognized statehood becomes the closest stable point. A state driven primarily by preference for stability will begin lending support to the self-determination group in order help it gain de facto control and bring a new stability to the situation. This is very much in line with Paquin’s evaluation of American acts of official recognition in Croatia, Slovenia, Macedonia, Kosovo, and Eritrea. 159 Put more simply, when self-determination conflicts get too violent, stability-seeking third parties will start trying to encourage territorial capture by the aspiring state, even when they have no intention of recognizing it once it gets there. Moreover, when a self-determination group successfully seizes control, holds off the host state, and begins governing at some level, by Figure 3-2 it may considered on the downward slope towards the local minimum. At this point, stability seeking third parties have an incentive to assist in this process, establishing a new equilibrium. By the same token, the moment the rebels “begin” winning might be considered just past the peak of the instability curve, but in practice such a moment would be impossible to recognize from any distance. It may be helpful to think of this notion as related to prior work finding third parties prefer to aid secessionist groups 158 Jensen 2012. 159 Paquin 2010. 53 that are of moderate strength, promising to prevail to some extent but not completely overturn the status quo. 160 There may be some exceptions to this trend in third party support. One notable exception is the targeting of civilians. The targeting of civilians in these conflicts is destabilizing on a completely different level, representing more permanent forms of damage that are harder to come back from. For example, if urban infrastructure such as hospitals, schools, roads and bridges, and markets are targeted, the basic means of economic pursuit and societal progress are destroyed and must be rebuilt. If children and families are targeted, it can have the effect of entrenching people much more deeply in the conflict. 161 Note that this applies in ongoing two- sided conflicts, and less so with police repression of protest and political expression, unless it is very severe. Thus, we would expect third parties to primarily punish those who target civilians and civilian infrastructure. Official recognition is still an important component during all stages of these conflicts. Coggins’ most prominent finding 162 was that once a major power like the US, China, or Germany enacts a policy of official recognition, other major powers become much more likely to follow suit, so that there is a sort of cascade effect to recognition. With movement towards recognition, but short of it, we would still expect there to be an effect of diplomatic recognition on lesser forms of support. When great powers start recognizing a given self-determination group, it is a sign that the closest stable point lies in either full or partial recognition of a group that has managed to gain territorial control. Thus, we should expect third parties to move towards recognition when the group garners official diplomatic recognition from a major power. One last important expectation sets this study up as an important complement to existing studies of recognition. 163 Unlike official recognition, lesser forms of support hold much milder implications for precedent, both on the international stage and in a third party’s own domestic politics. 164 Official recognition is a bold, discrete diplomatic act that invites immediate reciprocation and sets a clear example for movements that see themselves similarly. Participation in sanctions, the lending of aid, support in the UN general assembly, and contact with separatists all constitute less impactful forms of recognition that do not invite the same kind of backlash. Therefore, the concern for precedent should not be a factor in these third-party decisions, and the presence of separatist claims in the third party’s domestic arena should have no effect on movement towards and away from recognition. Together, these expectations, derived from a stability-seeking theory of movement towards recognition, yield five primary hypotheses: 1. Violence Hypothesis: When levels of violence in a separatist or secessionist conflict are higher, third parties will move towards recognition. 160 Salehyan, Gleditsch, and Cunningham 2011. 161 Downes 2006, 153; Carr 2002. 162 Coggins 2014. 163 Fabry 2010; Grant 1999; Griffiths 2016; Coggins 2014. 164 Heraclides 1990. 54 2. Victory Hypothesis: When self-determination groups successfully gain control of territory, third parties will move towards recognition. 3. Civilian Hypothesis: When the host state targets civilians, third parties will move towards recognition; when the aspiring state targets civilians, third parties will move away from recognition. 4. Diplomatic Recognition Hypothesis: When a great power officially recognizes a self- determination group, other third parties will move towards recognition. 5. Precedent Hypothesis: The number of self-determination challenges a third party faces should have no effect on movement towards recognition. 3.4 ANALYSIS Official recognition, as a measure of international sovereignty, has yielded valuable insights in the literature on self-determination groups and sovereignty up to this point. Research consistently finds that self-determination groups make international recognition one of their top priorities, 165 and contrarily, host states make prevention of recognition one of theirs. 166 The laundry list of benefits of official membership in the international state system could fill volumes, 167 so focusing on international legal sovereignty exclusively through the lens of diplomatic recognition has been a crucial first step in understanding how it works. However, a more inclusive measure of third-party stance towards self-determination movements may reveal yet new insights that are undetectable through that lens, primarily because official recognition is such a rare, difficult, and risky decision for states to make. As described in Chapter 2, I developed a new, continuous measure of third-party recognition. Rather than treating “recognition” as a political decision made by third parties, I treat it as an underlying trait that determines a number of unilateral political decisions states make towards these conflicts, of which official recognition constitutes only one. Figure 3-3, repeated from Chapter 2, illustrates this intuition. I conceptualize these third-party actions as a form of declaratory recognition, meaning they admit to what already is rather than manifest what the third party’s leadership thinks should be. 165 Caspersen 2012; Buzard, Graham, and Horne 2017; Coggins 2014. 166 Ker-Lindsay 2014. 167 Fazal and Griffiths 2014. 55 Figure 3-3. Latent variable model of recognition illustrated The sample of cases was formed by combining Coggins’ sample 168 with Griffiths’ sample, 169 the combination of which included secessionist, self-determination, and separatist movements back into the mid-19 th century. 170 I took all the cases from 1945-2015, and merged 10 datasets on unilateral foreign policy decision by third parties. My final full dataset has each conflict-year paired with eight major powers and ten medium sized powers. For this model, the observation count was about 67,000. With this dataset, I estimate a “latent variable” model of recognition, building on Bayesian item response theory (IRT) models. 171 As a brief reminder of how exactly this measure works, I present the Palestinian case a second time in Figure 3-4 to show the variation at the third-party-year level. It illustrates the positions of Mexico, Brazil, Sweden, Turkey, Egypt, Israel, Saudi Arabia, India, Indonesia, USA, UK, France, Germany, Italy, Russia, China, and Japan towards Palestine starting in 1967. Figure 3-4. Latent recognition of Palestine 168 Coggins 2014. 169 Griffiths and Butcher 2013. 170 I discuss the shortcomings of taking this sample in Chapter 5. 171 For full description of data used to develop this model, see Chapter 2. Underlying Recognition Military Intervention UN Voting Sanctions What we can see Assumed Causal Relationship Latent Recognition We can’t observe What we estimate Mexico, Brazil, Sweden, Turkey, Egypt, Israel, Saudi Arabia, India, Indonesia US, UK, France, Germany, Italy, Russia, China, Japan 56 Most countries have converged on a similar score towards Palestine, coming to favor Palestine more over time, with positive scores throughout most of the period. Additionally, medium sized powers have favored Palestine more than major powers have. Figure 3-5. Latent recognition of Palestine by six powers Figure 3-5 shows the positions of six third parties towards Palestine to restate what informs these scores. The scores of Russia, China, Mexico, Sweden, and Saudi Arabia are informed primarily by their proportions of pro-Palestinian votes in the UN and their participation in sanctions against Israel. The US’ score is informed by these variables, in addition to data on levels of military aid to Israel. Because this model detects variation both between third parties’ scores and within parties’ scores over time, it renders these measures useful for exploring changes in policy orientation towards the self-determination conflicts. I now turn to a slate of tests of stability- seeking hypotheses. To test the Violence Hypothesis, I use PRIO’s data on battle deaths, an ordinal variable with three categories: no deaths, low violence (25-999), and high violence (1000 or more). 172 Unfortunately, for the full period in question, this is the best data available on deaths in battle. For the Victory Hypothesis, I use the COW intrastate war data’s dichotomous variable indicating separatist victory in battle. 173 Peak instability is probably best conceived of as falling somewhere in between these two points—high violence and rebel control—but together, they should provide a decent proxy for the point in conflict when concerns for stability would implicate support of the self-determination group. A positive association between latent recognition and each of these variables would support a stability-based theory of movement towards recognition. For the Diplomatic Recognition and Precedent hypotheses, I use Coggins’ data. 174 The “prior recognition” variable is a dichotomous variable indicating that at least one major power has diplomatically recognized the self-determination group as a sovereign state. For precedent, I 172 Gleditsch et al. 2002. 173 Sarkees 2010. 174 Coggins 2014. 57 use her variable on the number of self-determination challenges the third party faces in its own territory in a given year. To test the Civilian Hypothesis, I use the UCDP GED data. 175 Using this data entails one major benefit and one major drawback. The benefit is that its level of detail on battle deaths and civilian deaths is much better than the PRIO data dating back to 1945. It includes substate dyadic conflict data on: sides 1 and 2 identities, deaths per side, civilians killed by either side, geocoded locations, and identification of sources for each datapoint. Its drawback is that it only dates back to 1989, severely curtailing my ability to test most of my hypotheses. For this reason, I test the Civilian Hypothesis separately from the rest. I estimate a fixed effects model with robust standard errors clustered at the dyad level. Table 3-1 displays the results of this test in three columns: one model for major powers, one model for medium sized powers, and one pooled. The purpose of fixed effects models is to control for innate unobservable characteristics within each unit and estimate the correlation between the dependent variable’s movement across years and the movement of unit and system level variables that change over time. Because there is no year-to-year variation in major power status, a “major power” control variable would be collinear with the fixed effects country level controls. Therefore, I run the major, medium, and pooled models to understand the difference between major and medium sized third parties in their foreign policy orientation towards these conflicts. Table 3-1. Effects on latent recognition of violence, victory, domestic politics, and international factors Model: Major Powers 1945-2015 Medium Powers 1945-2015 Pooled 1945-2015 Violence: 25-999 deaths (PRIO) .01 (.02) -.02 (.007) * -.006 (.01) Violence: >999 deaths (PRIO) .08 (.02) *** .03 (.009) *** .05 (.01) *** SD group victory .13 (.04) ** .06 (.02) *** .09 (.02) *** Number of Challengers .002 (.001) *** .001 (.000) *** .002 (.000) *** 3rd – Host MID .03 (.01) .01 (.01) .02 (.006) ** Prior recognition by great power .005 (.02) .05 (.01) *** .03 (.01) ** Mutual autocracy -.03 (.01) *** -.02 (.003) *** -.02 (.004) *** Mutual democracy -.02 (.01) ** -.01 (.003) ** -.01 (.003) *** Constant -.18 (.01) *** -.08 (.003) *** -.12 (.003) *** N 1947 2635 4582 p-values are * .05, ** .01, *** .001 (SEs) 175 Croicu and Sundberg 2017. 58 Two findings stand out above the rest. The first is a clear, positive association between high violence and latent recognition, and between rebel victory and latent recognition. When the number of deaths in conflict have passed the 1000 mark, third parties of every sort move towards recognition. Moreover, major powers, which have more opportunity to influence the conflicts have had about triple the movement as medium sized powers, according to this measure. When these conflicts reach a high level of violence, it would seem as if third parties do indeed predominantly move towards supporting an unrecognized de facto state as preferable to current levels of violence. Self-determination group victory is also associated with substantial increases in latent recognition. With this variable, there is a potential concern for endogeneity. Does latent recognition cause group victory, or does group victory cause movement towards recognition. In fact, either causal relationship provides support of a stability-based argument. Either third parties are helping them win, or they are helping them once they do. Both of these are consistent with the theory put forth here. Insofar as third parties make efforts across the spectrum of possible unilateral foreign policy actions constituting this measure, they do seem to push these conflicts towards domestic juridical sovereignty once violence reaches a certain point. This is consistent with Salehyan et al.’s finding 176 that rebel groups that secure territorial control are more likely to receive external support from third parties. The medium powers model contains further proof of this general trend of response to violence. At lower levels of violence, we find evidence that third parties move away from recognition, supporting host state efforts to silence separatist voices when levels of violence are lower. This too falls in line with the expectations of the stability-seeking model of third party recognition. Initially, third parties want to see a return to status quo, but as levels of violence increase, they perceive progress towards independence (but without recognized status) as an increasingly promising means of stability. This apparent relationship between violence, victory, and latent recognition lends support to the Violence and Victory Hypotheses. Second, there is no negative relationship between the number of self-determination challenges a third party faces domestically and changes in recognition. Surprisingly, a positive relationship is estimated for both major and medium power third parties. This finding supports the Precedent Hypothesis to an extent, but it also reveals a puzzling relationship that is worth interrogating in the future. One possible explanation might be that countries facing more separatist claims place greater value on international stability, so are willing to push self- determination movements toward territorial capture at earlier stages. Another important finding lends limited support to the Recognition Hypothesis. When a major power goes out on a limb to diplomatically recognize a self-determination group, medium sized third parties tend to move towards recognition as well. Diplomatic recognition from powers like the US or Russia sends a strong signal that a new status quo is at hand, and stability-seekers add to that signal by lending other kinds of support. It is somewhat puzzling that major powers do not follow this pattern to the same extent, especially given Coggins’ finding that prior official 176 Salehyan, Gleditsch, and Cunningham 2011. 59 recognition by other powers has the strongest effect on likelihood of new acts of major power official recognition. There is one major caveat with this inclusion of this variable in the model, closely related as it is to my concept of latent recognition. Diplomatic recognition by a great power should be considered a post-treatment variable in my model, and as such, it has the potential to affect the other relationships I estimate. To deal with this problem, I have also run the same models with prior recognition excluded; these are reported in the appendix, and the findings are consistent with those reported above. The most important controls from Coggins’ estimated model of recognition—current conflict between third party and host state, and shared regime type between third party and host state—both echo her findings. There is some evidence that when a third party is engaged in a militarized dispute with a given host state, it moves towards recognition of that state’s self- determination movements. Additionally, democratic third parties tend to move away from recognizing self-determination movements in democratic states, as authoritarian third parties do for authoritarian host states. These findings are more or less inconsequential for the stability- seeking theoretical expectations reported herein, but they do lend a degree of convergent validity to this measure. To explore the relationship between civilian deaths and recognition, I turn to the UCDP GED dataset, 177 which includes conflict level data that specifies actor, victim, date, and location. This data distinguishes between civilians killed by host state and aspiring state forces. This data only goes back to 1989, so the findings reported from here are curtailed from the full sample. Additionally, because the N is so much smaller, I lose a degree of statistical power. Table 3-2. Effects on latent recognition of violence against civilians Major Powers 1989-2011 Medium Powers 1989-2011 Violence: 25-999 deaths (PRIO) .03 (.08) -.01 (.00) *** Violence: >999 deaths (PRIO) .13 (.13) .01 (.00) *** SD group victory .24 (.17) .003 (.001) *** # Civilians killed by host state .00 (.00) .00001 (.000) ** # Civilians killed by aspiring state .00 (.00) -.0001 (.000) Constant -.12 (.06) -.09 (.002) *** N 355 442 p-values are * .05, ** .01, *** .001 (SEs) Table 3-2 shows that the same basic findings hold true as in the larger model for the Violence and Victory Hypotheses. Medium sized powers seem to follow a curvilinear pattern in their patterns of recognition. Additionally, we see limited support for the Civilian Hypothesis, but 177 Sundberg and Melander 2013; Croicu and Sundberg 2017. 60 only for medium powers, and only on the host state side. When the host state kills civilians, third parties move towards recognition. Finally, I use the curtailed dataset to further investigate the apparently curvilinear relationship between violence and medium power recognition. I estimate separate coefficients of deaths and deaths-squared. The first coefficient gives the trend in the data crossing the y- intercept; the second tells us about whether and where the trend is curving. Table 3-3. Quadratic model to test the curvilinear relationship between violence and recognition Medium Powers 1989-2011 Deaths (UCDP) -.00003 (.0000) * Deaths Squared .00009 (.0000) * SD group victory .008 (.002) *** Number of Challengers .0006 (.001) 3rd – Host MID -.006 (.002) * Prior recognition by great power -.02 (.03) Mutual autocracy -.006 (.003) * Mutual democracy -.006 (.001) *** Constant -.11 (.002) *** N 422 p-values are * .05, ** .01, *** .001 (SEs) Table 3-3 shows a negative relationship between deaths and recognition at the intercept, but it also shows a convex relationship, so that recognition levels curve upward as violence increases. Figure 3-6. Marginal effect of deaths on latent recognition 61 Using this model, I estimate the marginal effects of deaths on recognition. Figure 3-4 illustrates this relationship. As the number of deaths in the conflict rises, medium sized third parties start moving towards recognition of self-determination groups more and more. An important caveat to point out is that it illustrates only a weak effect, if significant. Much flatter curves can be fitted into the confidence intervals shown here. Therefore, this curvilinear finding does not allow one to conclude any strong points about the precise nature of the relationship between violence and recognition. However, between the three models reported here, there is consistent enough evidence of a curved relationship between violence and recognition that closer exploration of the relationship between types and levels of violence and third-party movement towards recognition would make for a worthwhile endeavor. One other issue with both the utilized data and my tests should be discussed. My data is coded at a yearly interval, like most IR data. Theoretically, there could be a lag in third-party states’ shifts in policy that could extend into the following year and beyond, just as the response could be within the same year. As such, by estimating the models with both dependent and independent variables within the same year, I risk some level of misspecification. As a robustness check, I estimate the model with a one-year lag on all independent variables. This is reported in the appendix. The effect of SD group victory disappears in the lagged model, but every other effect is consistent with those reported in Table 3-1. Overall, these fixed effects models have two strong takeaways. First, across all estimated models, high levels of violence and rebel victory have a strong positive relationship with third party movement towards recognition of self-determination groups. This is true of major and medium sized powers and evident throughout the post-WWII period. Second, the precedent- oriented strategy that characterizes foreign policy decisions about diplomatic recognition seems to be constrained only to those decisions. In this measure, based on lower forms of recognition— UN voting, military aid, sanctions, and others—the number of self-determination challenges a third party faces has either no relationship with latent recognition or a positive relationship. 3.5 DISCUSSION There are distinct, nearly impassable hurdles to recognizing new states, even when the domestic sovereignty of an aspiring state equals that of most recognized states. As an example, Somaliland is almost entirely self-governing, with relatively well-developed infrastructure, a democratic government, and a prosperous economy with its own currency, and international agreements on usage of its port at Berbera. 178 In a region characterized by governments with woefully low juridical sovereignty, 179 it is a bright spot of stability, both in the otherwise “failed state” of Somalia and in the Horn of Africa in general, and it has established permanent channels of diplomatic exchange with several countries. However, no state recognizes Somaliland. For most 178 Fitch 2016; Al Jazeera 2018. 179 Jackson 1992; Jackson 1990; Jackson and Rosberg 1986. 62 third parties, recognition is only partially about granting international legal sovereignty to match the domestic sovereignty a group has achieved. For most states, the challenge of granting international legal sovereignty is greater than a simple acknowledgement of domestic sovereignty achieved. Stated more concretely, no matter how well established an aspiring state may be in quality of governance and enduring stability, the costs of official diplomatic recognition—provocation of friends and foes and the destabilization of borders and international institutions—are high enough that few third parties reach for it as a tool to resolve these conflicts. However, third parties frequently take stances and get involved in self-determination conflicts in other ways, and these activities can be characterized as pointing to a lesser form of recognition. A measure of recognition that incorporates these activities allows me to answer questions that complement prior work focusing on official diplomatic recognition. This concept and instrument for international sovereignty discerns smaller changes in foreign policy orientation towards these conflicts, with predictable patterns at the country level and over time. I put forth the theory that stability is the strong preference driving third-party policy towards self-determination groups, and I test a few hypotheses derived therefrom. I show that when self-determination conflicts become sufficiently violent (and unstable), third parties move towards recognition, seemingly to reach a point of stability along the state-progression commonly known as “unrecognized statehood.” When these rebels successfully capture territory, third parties move towards recognition. Furthermore, at lower levels of violence, there is some evidence that states move away from recognition—meaning their actions are oriented towards assisting the host state in its efforts of repression. These three findings are entirely consistent with a state that seeks the closest available solution of stability. Official recognition by major powers also has a strong positive association with latent recognition among medium sized powers. We should also expect this finding, given the strong signal that great power recognition holds. There is also some further evidence backing findings in prior research: first, that third parties move towards recognizing the self-determination movements within the countries they fight with; and second, that both autocracies and democracies tend not to support the self- determination groups other autocracies and democracies face, respectively. This pair of findings complements Coggins’ slate of tests on major power recognition. Overall, my model, theory, and findings here provide a useful complement to prior work on international sovereignty, diplomatic recognition, and international intervention in foreign conflicts. Future work might try to explore some of the patterns I find from a much closer perspective, exploring how important third-party intentions are and the extent to which they know that they contribute towards enduring anomalies in the international system that may eventually need to be addressed. 63 BOOED OFF STAGE SYRIA, CONSTRUAL LEVEL THEORY, AND TESTING AUDIENCE COSTS 4.1 SYRIA SETS THE STAGE In an August 2012 press conference, the president of the United States of America took one of the clearest foreign policy stances of his tenure, issuing what can only be called a threat: “We have been very clear to the Assad regime…that a red line for us is we start seeing a whole bunch of chemical weapons moving around or being utilized. That would change my calculus. That would change my equation…We have communicated in no uncertain terms with every player in the region that that’s a red line for us and that there would be enormous consequences if we start seeing movement of the chemical weapons front or the use of chemical weapons. 180 The following day, the White House principal deputy press secretary reiterated: “…We're watching very closely the stockpile of Syrian chemical weapons; that any use or proliferation efforts related to those chemical weapons is something that would be very serious and it would be a grave mistake. There are important international obligations that the Syrian regime must live up to in terms of the handling of their chemical weapons. And the officials who have that 180 Obama 2012. 64 responsibility will be held accountable for their actions and will be held accountable for living up to those international obligations.” 181 The statements of the Obama administration appealed to international norms and to the “national honor” of the US for justification. Obama seemed to have been concerned with a domestic and an international audience, and he made specific commitments to action with clear conditions. It is as if Obama had read the international security literature about the critical role of domestic audiences in international security crises, 182 and knowingly attempted to use candor to signal resolve. Predominant international relations theories make clear predictions about what happens when the conditions of a military threat issued by a democratic leader are met. 183 Domestic audiences take interest in leadership decisions that “engage the national honor,” 184 binding leaders’ hands, expecting them to follow through on threats, and punishing them if they do not. This cornerstone mechanism has been developed formally 185 and tested experimentally 186 and observationally. 187 It enables democracies to issue credible threats, eliminate costly confusion, and avoid bargaining breakdown, ultimately resulting in fewer conflicts altogether for democratic actors. When chemical weapons were used in Syria a year later, the Assad government was blamed, 188 and the Obama administration soon began beating the drums of war, deploying cruisers and imploring support in internationally televised speeches. 189 He had every reason to expect the American public to back him up. Several polls throughout the prior year had indicated broad support for Obama’s threat to intervene in the Syrian conflict if the Assad government ever used chemical weapons. Yet, as the administration moved forward with intervention— president, secretary of state, and defense secretary all urging the nation’s support—the public turned against the president’s plan. This apparently abrupt swing in opinion makes the Syrian case a curious one. It seems to defy two generally recognized features of the public in the foreign policy context. Insofar as Americans give foreign policy any thought, they are expected to follow the cues of elites and to prefer consistent foreign policy. In this case, the public defied the pleas of the most visible foreign policy elites and mounted opposition to something it had recently supported. However, there had also been signs that support for involvement in another Mid- Eastern conflict would be lacking. Those same polls had shown that Americans thought the US was overcommitted in the region. Perhaps early polls indicating support for intervention in Syria were too good to be true. 181 Earnest 2012. 182 Fearon 1994. 183 Ibid. 184 Ibid. 185 Ibid.; Schultz 2001. 186 Tomz 2007; Trager and Vavreck 2011; Levendusky and Horowitz 2012; Chaudoin 2014a. 187 Snyder and Borghard 2011. 188 Gladstone and Chivers 2013; Reuters 2013a. 189 Shalal-Esa 2013; Logiuratto 2013. 65 For years, psychologists and survey methodologists have warned that people give the benefits of hypothetical decisions more weight than the costs, and they may not consider cost at all. In this chapter, I argue that survey experimental research relying on vignettes as a treatment assignment strategy is likely falling victim to this same tendency. First, through the Syrian case I show that the security field has neglected several keys characteristics of the public in the development of ACT, and therefore has been inchoate in its empirical tests. I emphasize two key points: it cannot be taken as given that the audience values consistency over specific policy choices, because the audience itself may be far from consistent. So far, the standard audience cost set up has assumed the audience is concerned mainly with consistency when judging its leaders. 190 More importantly, and thus far unquestioned, it has also assumed consistency over time within the public, neglecting the potential discrepancy between preferences stated in public opinion surveys and those revealed months later during a crisis. I consider the possibility that the apparent swing in opinion reflected a common bias people exhibit when thinking about hypothetical situations, the tendency to “construe” distant decisions at a higher level, thereby neglecting to consider the costs of an action. I turn to a selection of survey methodological literature in psychology, economics, and political science to explain why this behavior should come as no surprise. Experimental tests of opinion in foreign crises have repeatedly shown that survey respondents punish inconsistency in presidential actions. 191 Some have directly manipulated the cost of the hypothetical conflict so that all subjects have similar expectations, 192 but most audience cost experiments have not explored respondent expectations about cost when they are unmanipulated. From the Syrian case and this interdisciplinary literature review, I derive an empirical hypothesis for survey experiments: subjects primed to think about the costs of intervention will be more critical of it. I present the design and results of a survey experiment that shows that merely asking respondents about their expectations of American casualties both drastically reduces their rates of disapproval of inconsistency in leadership and increases their disapproval of hawkishness in hypothetical scenarios. I provide evidence of several important points: (1) respondents have different assumptions about expected casualties of conflict across treatment groups—their assumptions are not orthogonal to treatment assignment; (2) prompting cost-thinking reduces support for intervention in key treatment groups, and (3) cuts the estimated absolute audience cost by more than half; and (4) while respondents do have different assessments of the effect on international reputation of hypothetical actions, approval levels are not affected by this prime in the same way. I argue that these construal level biases may represent a unique threat to external validity in surveys and survey experiments utilizing vignettes and hypotheticals to study public opinion on foreign affairs, and potentially any other areas where cost-thinking might be a concern. I conclude that improving survey experimental methods that test ACT requires incorporating these insights from other fields into our experimental tests of public opinion during foreign policy crises. Additionally, I recommend some alterations to the experimental agenda to 190 Chaudoin 2014a. 191 Ibid.; Levendusky and Horowitz 2012; Potter and Baum 2014; Tomz 2007. 192 Kertzer and Brutger 2016; Trager and Vavreck 2011. 66 refine our understanding of when and how domestic audiences express and enforce their preference for consistency, and whether they do so consistently. 4.2 AUDIENCE COST THEORY AND ITS CRITICS Before expounding on the events in Syria, it is helpful to lay out the predictions and scope of audience cost theory (ACT) in the current literature. Schelling 193 was the first to suggest “that democracies could improve their bargaining leverage by marshaling public opinion in support of a particular negotiating stance.” 194 Fearon developed a model to show the generation of audience costs on the “notion that troop movements and public demands or threats ‘engage the national honor.’” 195 In Fearon’s model, the audience prefers foremost that its leader follows through on military threats made vis-à-vis other world actors. This generates an incentive for democratic leaders to follow one of two paths: (1) not to make threats; or (2) to make threats sincerely and follow through when their conditions are met. When they fail to follow the audience’s preferences, democratically elected leaders face potential sanctions: they can be voted out if they fail to uphold the national honor. 196 Thus, decision-makers only escalate when they are serious about following through on their threats. Publicly broadcasted escalation of a dispute constitutes a “relatively informative and credible signal of willingness to fight over the issue.” 197 Tomz identifies the same mechanism, “Citizens…believe that hollow threats and promises undermine the country's reputation…that inconsistency is evidence of incompetence.” 198 Fearon’s model has proven inherently challenging to test. After all, based on his model, explicit threats should be as rare as international crises, even more so. Schultz 199 and Tomz 200 explain that the power of the theory is in comparing “threat made” cases with “threat not made” cases. Snyder and Borghard’s findings seem to support this framework, even while they conclude that the theory performs poorly in their selected cases. 201 Whether democratic threats are relatively credible is not as important as whether they are treated as such by rivals in crises. The qualities of the audience are not important as a mechanism but as a context, a force that leaders utilize to indicate intention. The difficulty of testing audience costs with observational data has inspired innovative and persuasive experimental studies, pioneered by Tomz. 202 He tests the theory’s fundamental assumption that audiences punish empty threats, yielding a substantive extension of the theory, 193 1960 194 Downes and Sechser 2012. 195 Fearon 1994, 581. 196 Fearon 1994. 197 Ibid. 198 Tomz 2007. 199 2001 200 Tomz 2007. 201 Snyder and Borghard 2011. 202 Tomz 2007. 67 that leaders might make empty threats if they do not face “the prospect of losing domestic support—or even office.” 203 Other experimental tests of audience costs have been able to isolate other key variables in the causal logic by which the audience operates: the urgency of the foreign policy crisis; 204 the ambiguity of the threat and party reputation; 205 potential reasons for backing down; 206 and other inconsistency costs. 207 The established territory of ACT and its tests has not been without its critics. Kertzer and Brutger critique survey experimental tests of audience costs by pointing to a recurring problem of double-barreled treatment groups and showing that ACT tests have failed to differentiate between two treatment effects. 208 Chaudoin points to ACT’s tendency to assume a public preference for consistency within leadership to plain old policy positions, and he finds that policy matters more to the public in the international trade context. 209 Audiences often support defections, and institutions cannot easily facilitate actions beyond the preference of the audience. 210 Berinsky has made a similar argument in the American Politics literature, noting that across decisions of intervention since WWII, US presidents have kept domestic opinion close to heart. 211 However, like others, 212 he finds that public opinion is primarily informed by elite cues. 213 Those elite cues come through media outlets, and thus, the levels of political independence and competition in the media are key variables in determining the timing and extent of public opinion on specific foreign policy decisions. 214 Even when the public is attentive and opinionated concerning political issues, its preferences do not necessitate corresponding actions. 215 Framing effects can yield wide variation in stated preferences, 216 but it is important to keep context of the frame in mind. 217 For example, as an event gets closer at hand, people tend to give the more weight to frames encountered recently than to those further in the past. 218 The ease with which stated public preference can be manipulated by reframing is evidence enough for some to abandon the idea of stable policy preferences and attitudes altogether. 219 203 Ibid., 821. 204 Tomz 2007. 205 Trager and Vavreck 2011. 206 Levendusky and Horowitz 2012. 207 Levy et al. 2015. 208 Kertzer and Brutger 2016. 209 Chaudoin 2014b. 210 Ibid., 880. 211 Berinsky 2009. 212 Zaller 1999; Baum 2004; Potter and Baum 2014; Potter and Baum 2010; Baum, Potter, and Baum 2015. 213 Berinsky 2009. 214 Baum 2004; Potter and Baum 2014; Slantchev 2006. 215 Sears, Huddy, and Jervis 2003. 216 Tversky and Kahneman 1981; Levin, Schneider, and Gaeth 1998. 217 Druckman 2004. 218 Chong and Druckman 2010. 219 Bartels 2003; Zaller and Feldman 1992. 68 While common throughout the public opinion and political psychology literature, these insights about potential differences between stated and revealed preferences have generally been left unaddressed in audience cost models and tests. Experimental tests have focused uniformly on gathering and comparing stated preferences, 220 and though the vignettes placed before experimental subjects have seen extensive variation, that has been the limit of innovation in survey experiments on audience costs. Important modifications made to survey research, such as introducing respondents to costs, testing the effects of framing and default options, and comparing immediate and future consequences, have made no appearance in the international security survey experimental research. These developments for survey research may offer us some valuable tools for honing our understanding of public opinion during crises. The Syrian crisis offers a useful case to recognize the public’s potential preference for policy over consistency, as well as its own inconsistency over time, and to begin to think about where the survey experimental literature, is it now stands, still has room to develop. 4.3 A CRUCIAL CASE: OBAMA’S PUBLIC CHANGES COURSE Eckstein 221 establishes a crucial case as one “that must closely fit a theory if one is to have confidence in the theory’s validity…” Gerring elaborates, “The key proviso is that the theory under investigation must take a consistent form, even if its predictions are not terrifically precise, well elaborated, or broad…A theory that is understood to be deterministic may be disconfirmed by a case study, properly chosen.” 222 Fortunately for my purposes, ACT has very precise conditions, making it a straightforward candidate for a crucial case test. Taking the most common interpretations of audience costs, 223 we have a few ironclad conditions that must be present, and a few more that are implied. Two foreign policy decisions set the scene: the explicit conditional threat issued publicly by the democratic leader; and the adversary’s defiance of the threat’s conditions. The theory predicts that upon such an act of defiance the democratic leader will escalate with support of his public, especially if it supported him in his first threat, or the leader will backpedal on the threat, facing public disapproval and blame for disgracing national credibility. This general reaction is why democratic threats are understood to be more credible; publics hold their leaders to their military commitments and “bind their hands.” Democratically elected leaders can thus “generate” audience costs because of this consistent reaction of the public. Near perfect fulfillment of the establishing conditions makes the US-Syria crisis an apt crucial case. Table 4-1 outlines how audience cost theory adheres to Gerring’s requirements to make for a good crucial case test of the Syrian case. It also shows how the events that played out 220 Tomz 2007; Trager and Vavreck 2011; Levendusky and Horowitz 2012; Chaudoin 2014a; Potter and Baum 2014. 221 Eckstein 1975, quoted in Gerring 2001, 118. 222 Gerring 2007, 120–121. 223 Fearon 1994; Tomz 2007; Snyder and Borghard 2011. 69 serve to highlight potential questions about audience cost theory’s tendency to oversimplify the way a domestic audience interact with the state leadership during an international crisis. Table 4-1. Syrian crisis as a crucial case under audience cost theory Note: Regular font denotes theoretical predictions fulfilled by case; Underlined denotes unfulfilled predictions; Bold denote crucial case’s deviation from theory. Returning to the case, in April 2013, after significant escalation in the Syrian civil war, White House officials reiterated the administration’s position: “We go on to reaffirm that the President has set a clear red line as it relates to the United States that the use of chemical weapons or the transfer of chemical weapons to terrorist groups is a red line that is not acceptable to us, nor should it be to the international community.” 224 224 White-House-Official 2013. Gerring (2007) Crucial Case Clear context Clear conditions Precise, determinate predictions Predicts mechanism Predicts outcome Audience Cost Theory  International Military Crisis  Leader makes conditional threat  Threat observed by audience  Conditions of threat met by adversary  Democratic leaders make only threats they intend to keep  Domestic audience pressures or expects leader to follow through on threat  Upon met conditions, leader escalates militarily with domestic support  If leader backs down from threat, domestic disapproval follows  Democratic leaders are concerned with domestic support  Domestic concern for credibility greater than concern for military engagement  Empty threat erodes domestic support  Democratic leader will follow through on clear threat  If not, credibility will be questioned Syria Crisis  Crisis over international intervention in Syria over use of chemical weapons  Pres. Obama threatens intervention if chemical weapons used  Threat is broadcasted nationally  Polled Americans support it  Syrian government found to have used chemical weapons  Pres. Obama confirms intent to intervene several times  Pres. Obama sends several destroyers to Syrian coast after news of chemical weapons usage  Domestic audience opposes escalation; pressures president to back down  Disapproval increases as president escalates  Obama administration is concerned with domestic support  Concern for military engagement greater than concern for credibility  Escalation erodes support  President Obama backs down; resolves dispute through Russia  Credibility questioned by international actors and by media, but not by voter-base 70 Opinion polls throughout the year had showed that Americans generally understood and supported the threat. In December 2012, an early poll showed that 63% of Americans supported American military involvement as a response to chemical weapons usage in Syria. 225 Soon after the April 2013 reiteration of the “red line,” a CNN poll showed that 66% believed that military action in Syria would be “justified” if chemical weapons were used there, 226 and a Pew poll had Americans supporting conditional US military action by a 45% plurality, with 31% opposed to intervention. 227 This level of support for a Syrian intervention only concerned a possible US reaction to the regime’s use of chemical weapons, not to a general intervention in Syria. On August 21, 2013, rockets containing the chemical agent sarin struck the opposition- controlled city of Ghouta, Syria. Estimates of the number dead ranged from 281 to 1429. 228 Within three weeks the United Nations confirmed the use of sarin, 229 and several prominent governments, NGOs, and independent investigations concluded the culpability of the Syrian government. 230 The Obama administration’s initial reaction to chemical weapons usage in Syria was consistent with audience cost theory. By August 30, six US Navy destroyers were stationed off the coast of Syria, 231 and the Obama administration appeared to try to generate support for US military intervention in Syria. On August 31, he broadcast a solemn speech to an attentive domestic audience: “This attack is an assault on human dignity…a serious danger to our national security…a mockery of the global prohibition on the use of chemical weapons. It endangers our friends and our partners along Syria’s borders…In a world with many dangers, this menace must be confronted. Now, after careful deliberation, I have decided that the United States should take military action against Syrian regime targets…Our military has positioned assets in the region…we are prepared to strike…And I’m prepared to give that order.” 232 The president announced that he would seek US congressional approval for military intervention, 233 and he urged congressional members to support military intervention, suggesting that US credibility was on the line: 225 “Post-ABC poll” 2012 226 CNN 2013a. 227 Pew-Research 2013. 228 Daily-Star 2013; White-House 2013. 229 UN 2013. 230 Gladstone and Chivers 2013; Reuters 2013a. 231 Shalal-Esa 2013. 232 Obama 2013a. 233 Geman 2013. 71 “If we won’t enforce accountability in the face of this heinous act, what does it say about our resolve to stand up to others who flout fundamental international rules?” 234 The president reiterated all of these ideas in another speech 10 days later, 235 all the while he and administration officials met with 85 senators and 165 congressmen, urging them to vote in favor of US air strikes on Syrian regime targets, 236 The message broadcast to the public was that US intervention was needed for national and international security. Figure 4-1. Net popularity of intervention in Syria and Obama foreign policy in 2012 and 2013 (Sources: 53 representative polls from 15 polling agencies. Line is fit using a standard LOWESS model.) A clear threat was made by an unmatched military power with the support of its domestic audience, the acknowledged adversary breached an international rule, and that democratic leader escalated conflict with military display and strong domestic rhetoric. US credibility was on the line. However, the American public did not back the president’s call for military action. While ACT expects support for the president and concern for US credibility, we saw escalating opposition to US intervention in Syria. Figure 4-1 shows a dot plot of foreign policy opinion ratings of 40 randomly drawn polls from 15 polling agencies; the bar chart shows support and opposition from 13 randomly drawn polls that specifically asked respondents for 234 Obama 2013b. 235 Obama 2013a. 236 Bellantoni and Burlij 2013. 72 opinions of approval, disapproval, or no opinion on military intervention in Syria if and when chemical weapons were used. 237 The American public’s stark change of heart concerning a Syrian intervention and Obama’s foreign policy decisions is clear in this graph. Before the Assad government employed chemical weapons, polls showed an average of 58% support for and 30% opposition to military intervention as a response to chemical weapons usage. The flurry of polls through August and September 2013 showed average levels of 25% support and 54% opposition, a complete reversal in American opinion. 238 Throughout the month of September 2013, the president faced increasing negative reaction from the American public. Disapproval of Obama’s foreign policy climbed from an average 39% at the beginning of 2013 to one of 53% in September 2013. One poll showed that, by a 15-point margin, Americans believed Obama’s handling of the Syrian situation, including military escalation, weakened rather than strengthened “US global leadership.” 239 From the news of the Ghouta attack on August 21, 2013 to the end of September 2013, the president’s general job approval rating slid from 49% to 44%, and disapproval rose from 43% to 50%. 240 During this time, alongside personal and administrative meetings with senators and congressmen, Obama conducted a major media campaign for support of military intervention, making prime-time appearances on six major news networks, 241 calling upon international norms and US credibility to generate support for follow-through on the “red line” threat. Obama’s defense secretary Chuck Hagel appeared before the Senate, making the same argument, “A refusal to act would undermine the credibility of America's other security commitments—including the president's commitment to prevent Iran from acquiring a nuclear weapon…The word of the United States must mean something.” 242 As Congress prepared to meet in September, 58% of Americans wanted Congress to oppose military involvement. 243 Voters called congressional representatives in large numbers, urging them to vote no on an American intervention limited to air strikes by a margin of up to 499:1. 244 Formerly undecided congressmen came out in opposition to military intervention, 245 and congressional approval looked highly unlikely with a 6:1 ratio of anticipated no votes. 246 237 All questions asked about whether the US should engage in “intervention,” “military action,” or “military force” in Syria if/when chemical weapons are/were involved. There is one exception: one poll asked if military action was “justified,” not whether the US should do anything. All polling sources are in a separate reference list in appendix B. 238 This 25% figure rises to 27% if you include a poll which found that 39% would support congress if it authorized military action for 60 – 90 days, barring the employment of troops on the ground CNN 2013b. 239 Langer 2013. 240 Gallup 2014b; Gallup 2014a. 241 Logiuratto 2013. 242 Quoted in Reuters 2013b. “U.S. credibility on Iran at stake in Syria decision: Hagel says” 243 Vitali 2013. 244 Kant and Schilling 2013. 245 Davis 2013. 246 Page and Kelly 2013. 73 Russia’s President Putin made the case for a diplomatic solution in a New York Times op- ed, 247 and Obama’s foreign policy path became increasingly and significantly constrained. His “hands were tied,” but not as ACT would predict. Naturally, the administration changed course, asking Congress to postpone its vote, 248 withdrawing its commitment to the “red line,” 249 and reaching out to its Russian counterpart for a way out. Ultimately, the Russian government acted as an intermediary, brokering a deal that would avoid US intervention and would see Syria dismantle its chemical weapons stockpile. 250 This plan witnessed 79% American support. 251 Even though the president’s own public forced him to backpedal on an internationally witnessed commitment, it was still his administration’s credibility that was called into question. 252 As Hagel predicted, leaders of other governments expressed doubt about the seriousness of any “red line” drawn by the Obama administration, explicitly naming the Syrian case as their evidence. 253 Schultz’s crisis game tree (Figure 4-2) is helpful in summarizing this case. 254 247 Putin 2013. 248 Davis 2013. 249 Check 2013. 250 Irish and Nichols 2013. 251 “ Post-ABC News poll - Syria” 2013 252 At least one argument exists that may challenge this characterization of the Syrian case as a “failure” of ACT. It is that because the Syrian government ultimately did give up its chemical weapons, the threat worked, and military enforcement was never necessary. My account addresses the behavior of the audience, not the outcome of the crisis. The audience forced Obama to turn to Russia for help, something his administration tried to avoid. That the audience would prevent fulfillment of a policy promise is well outside ACT’s general characterization of audience behavior during foreign policy crises. 253 Bremer 2013; Singh 2012. 254 2001, 37 74 Figure 4-2. The US-Syrian crisis with audience costs Source: Figure 1 from Schultz (2001), modified to fit case. Arrows follow decisions in this case. Obama’s threat and the Syrian government’s alleged fulfillment of its conditions set up the Obama administration for a decision to follow through on or back down from the threat. Fearon and Schultz theorize about the difference between situations in which a democratic challenger does and does not make a threat (nodes A and B, resp.). This case is situated after node D, at which point the president must decide to follow through on or back down from his threat. This is the same branch on the tree that experimental tests have tended to feature, measuring variation in disapproval of hypothetical “empty threats” as a result of a number of components, such as structure of the conflict, 255 presidential party, 256 and presidential claims of updated information. 257 Fearon and Schultz’s model, assuming an unconditional and one-dimensional preference of the audience, leads to a strong expectation for the president move to node E, attack. As expected by the theory, the president tries to move to node E, stationing destroyers off Syria’s case and engaging in a media campaign to win American support for intervention. However, when Obama tries to stand firm on his threat, he is punished by his audience for diverging from its preference, and the effect is great enough that he is persuaded to back down. This behavior is consistent with Chaudoin’s characterization of audience costs, 258 but with few other versions of the theory. 255 Tomz 2007. 256 Trager and Vavreck 2011. 257 Levendusky and Horowitz 2012. 258 2014a, 2014b 75 What is still missing in this literature is an explanation for why polled Americans would support the intervention as a hypothetical, but then turn so decidedly against once the dice are rolled. Fortunately, work has been done in other fields to help understand this phenomenon. It is possible that support for intervention genuinely reversed in the three months between May and August 2013, but fluctuations this extreme are far from common. 259 It may be equally likely that polled Americans simply failed to estimate their opinions on intervention accurately. In hindsight, perhaps Americans’ concern about the US already being overly committed to intervention 260 was a better indicator of American preference than the hypothetical question hinged upon chemical weapons usage. Indeed, nearly 90% of Americans polled in the midst of the crisis were concerned that military action would be too costly. 261 The glimpse provided by these few polls is not enough to determine whether this reversal in support was due to a genuine trend or weakness in survey instruments. Still, an explanation based primarily on elite cues 262 would have to grapple with at least a few key problems. First, the opposition to intervention seemed mostly driven by the public itself. The intensity of public preference against intervention was perhaps best illustrated in public messages by dozens of congressional members, in both parties, who reported hundreds of calls per day to their offices, at a ratio of up to 499:1 opposing intervention. 263 Second, the country’s foremost foreign policy elites—president, secretary of state, defense secretary—pushed publicly for intervention. The vast majority of the Senate and House were undecided throughout the affair. 264 This was not an elite-driven opposition. Another possible explanation is that the public was tacitly opposed to intervention the whole time, and early polls failed to capture it. This could happen if early respondents considered costs differently from those later on. Survey methodological work has found that respondents often make different decisions with cost on their minds and that costs are less salient to survey respondents judging hypothetical scenarios than those considering crises in real time. Concerning intervention, in addition to pecuniary costs, casualties are known to be one of the most salient costs of war. From this retrospective look at the Syrian crisis, it is not possible to ascribe the outcome to popular “delay discounting” one way or another, only to consider the rapid change of mind polled individuals expressed as American intervention moved from hypothetical to imminent. This odd movement of public opinion throughout the Syrian case inspires closer consideration of how sensitive hypotheticals concerning intervention might be to the delay discounting phenomena others have found to affect survey instruments. This inspires a question: how differently do respondents facing hypothetical scenarios in survey experiments respond when they are prompted to think generally about costs? I turn now to a discussion of how prior work 259 Clawson and Oxley 2013, 106. 260 Pew-Research 2013. 261 New York Times 2013. 262 Berinsky 2009. 263 Kant and Schilling 2013 264 Page and Kelly 2013. 76 should lead researchers to expect respondents considering hypothetical scenarios about foreign policy to be relatively biased away from considering costs the way they would in the real world. 4.4 SURVEY EXPERIMENTS ARE SURVEYS Survey experiments occupy a privileged position across the literature of political processes and international security. They have the distinct advantages of experiments. Randomization allows the researcher to narrowly identify the effects of manipulated variables, informed by theory. Survey experiments can be less costly than other methods, interpretation of their results is straightforward, and internal validity faces relatively few challenges. However, they also suffer all the same problems as surveys, including the common shortcuts survey respondents take as they humor the researcher and select answers to questions they may never have considered before the moment they encounter it on the survey. 265 Often, survey experiments expose the respondents to hypothetical scenarios or candidates before surveying their opinions, taking their responses as the dependent variable of interest. This method has been used to tackle research on voter candidate choice, 266 support levels for domestic political arrangements, 267 and foreign policy preferences. 268 The treatment is embedded in one of two or more randomly assigned vignettes, and comparisons of opinion made across the vignette groups provide evidence of the treatment effect. When researchers rely on hypothetical scenarios as their treatment medium, they expose themselves to the tendency of survey respondents to consider the benefits, but not the costs, of the fictional decisions made in these vignettes. In the audience costs context, for example, this could mean considering the symbolic value of strong, consistent American foreign policy, but not the blood-and-treasure costs of intervention. In the ideal setup, we expect respondents behave the same way as they would in a real survey; they do not know they have been assigned to one of multiple treatment groups. By virtue of randomization, the researcher can therefore test for effects on survey responses due to their manipulation of elements of the survey. This makes survey experiments a powerful tool. However, it also exposes them to all the same weaknesses of real surveys. Good critiques of survey experiments have been around nearly as long as the method itself. Gaines et al. 269 explain several problematic practices common in the design, including not measuring duration of effects, presenting one-shot treatments, overlooking mutual causation, and failing to include control groups. Dafoe et al. 270 and Huddleston and Weller 271 point to internal validity problems arising when experimenters unintentionally manipulate respondents’ beliefs about unmentioned aspects in the vignettes; the former recommend utilizing placebo tests to 265 Zaller and Feldman 1992. 266 Adida, Davenport, and McClendon 2016; Brader and Tucker 2009; Hainmueller and Hiscox 2010; Visalvanich 2017. 267 Gibson, Caldeira, and Spence 2005; Gibson et al. 2011; Haselswerdt and Bartels 2015. 268 Isani and Silverman 2016; Levy et al. 2015; Quek 2016; Trager and Vavreck 2011. 269 2007 270 2016 271 Huddleston and Weller 2017 77 better control for this tendency, while the latter recommend directly testing the extent to which theoretically important causal mechanisms mediate treatment effects. Some have also critiqued the external validity of survey experiments themselves, questioning whether survey experimental situations come across like the real-world scenarios they are meant to emulate. Hainmueller et al. 272 provide some validation of the use of vignettes and conjoint analyses by showing that the same attributes explain the choices of Swiss voting age survey experimental respondents as those explaining real world choices on the same issues. In the context of survey experiments, few have addressed the general problems that come with surveys. Respondents are expected to follow the same paths of thought as they do in surveys, so the critique of these experiments needs to be expanded to address concerns outlined in the survey methods literature as well as those in the experimental literature. As an example, avoiding inconsistency between hypothetical and realized preferences in survey research has been a topic of rich discussion in economics and psychology for several decades. Economists have long studied the problems in judging the behavior of the “farsighted planners and myopic doers” 273 that often respond to surveys. Bertrand and Mullainathan tackle the accuracy of subjective survey responses on a number of issues in their meta-analysis of the economic survey methods literature. 274 Survey respondents often “make little mental effort” in answering questions and are poor at “forecasting their behavior and understanding why they did what they did.” 275 Answers to subjective survey questions have proven to be such poor predictors of behavior that some argue there should be “serious doubts [about] attempts to use subjective data as dependent variables.” 276 In psychology, this is often discussed in the context of construal level theory (CLT), 277 which claims that people discount the long-term consequences of decisions, and think about distant or hypothetical events more abstractly than they do immediate scenarios. In “distal” scenarios, people engage in “high-level construal,” considering the situation abstractly, overlooking detail and focusing on general themes and desirability of outcomes. 278 By contrast, in more immediate scenarios, people engage in “low-level construal,” with increased consideration of the specific details and feasibility of outcomes. 279 For example, contemplating a vacation six months from now may involve thinking about “relaxing” and “having fun,” whereas doing so a week beforehand may involve thinking about making dinner reservations and the stress of driving a rented car on foreign roads. This disparity is known to cause some of the poor planning habits and inaccurate predictions about self-behavior and abilities discussed earlier. 280 272 2015 273 Thaler and Shefrin 1981. 274 2001 275 2001: 68 276 2001: 71 277 Bar-Anan, Liberman, and Trope 2006; Liberman and Trope 1998; Trope 2012; Trope and Liberman 2010. 278 Trope 2012. 279 Trope and Liberman 2000. 280 Liberman and Trope 1998. 78 Costs are associated with low-level construal. This is part of why people fall victim to what is known as delay discounting, 281 the tendency to discount (or neglect entirely) the costs and rewards in the distant future or in hypothetical situations, giving them less weight than costs and rewards closer at hand. Surveys have a hard time circumventing these flaws in the way people think about the future, and costs in particular are inadequately considered. Respondents’ preferences are not consistent over time; they procrastinate costs and “preproperate” benefits; 282 they are naïve about their ability to predict their own behavior; 283 and they have “self-control problems” that cause a persistent “gap between long-run intentions and short-run actions”; 284 when faced with complicated choices, people have a hard time estimating value. 285 When prompted for opinions or preference statements, many survey respondents have not given them any thought until that moment. “Preferences for and beliefs about objects or events of any complexity are often constructed—not merely revealed—in the generation of a response…” 286 Furthermore, people’s choices may “stem from affective judgments that preclude evaluation of options.” 287 Among others, a lesson that can be drawn generally from the survey methods literature across different fields is that people do not think about costs until they have to. Some of these problems are already well understood by some political scientists. Zaller and Feldman argue for a model that assumes people do not have crystallized policy opinions but instead, “respond to survey questions on the basis of whatever ideas are at the top of their heads.” 288 Berinsky’s results support such a view and “underscore the importance of attending to the effects of the social environment” when explaining divergence between stated and revealed preferences of survey respondents. 289 Funk explores differences in the breadth of these survey discrepancies based on issue type. 290 Especially relevant to research on public opinion on foreign intervention has been work on the effects of costs on support for policy. Jacoby has found specificity of cost to be particularly important in determining a framing effect: “…specific formulation of the issue…moves public opinion toward greater support for government spending.” 291 Government expenditures affect presidential approval, 292 and Geys shows that pecuniary costs particularly affect support for presidential belligerence. 293 A few recent experimental studies have also tested 281 Frederick, Loewenstein, and O’Donoghue 2002; Green and Myerson 2004; Murphy, Vuchinich, and Simpson 2001; Odum 2011. 282 O’Donoghue and Rabin 1999. 283 DellaVigna and Malmendier 2006. 284 Angeletos et al. 2001, 47. 285 Gabaix, Li, and Laibson 2005. 286 Payne, Bettman, and Johnson 1992, 89. 287 Shafir, Simonson, and Tversky 1993, 32. 288 1992: 579 289 Berinsky 1999. 290 Funk 2013. 291 Jacoby, 2000: 758; emphasis original 292 Geys and Vermeir 2008. 293 2010 79 this point using questions on taxation. 294 Finally, Lergetporer et al. show that furnishing information on current levels of defense spending tempers support for increases in spending. 295 The effect of casualties on support for war has also been extensively studied, both observationally 296 and experimentally. 297 Boettcher’s experimental findings suggest that the expected number of American casualties and probability of success have particularly strong influence on public opinion of intervention. 298 Gartner’s similar survey experiment confirms Boettcher’s findings: “Critical are military fatalities—the most salient cost of war.” 299 These survey experiments confirm the conclusions of a long line of research on wartime presidential approval: public support is affected by casualty count. 300 However, it should be emphasized that it is not the consensus that casualties represent the main or the most important cost of conflict to domestic audiences. Gelpi et al. 301 themselves argue that casualties are less consequential on public opinion than prospects for success or original justifications for war, and Berinsky makes the case that elite consensus matters far more than casualties. 302 Still, casualties remain an important feature of conflict that is far from irrelevant to domestic audiences. Taken together, these bodies of work—one on survey response shortcuts and a common bias in evaluating abstract situations, the other on cost perceptions’ effects on support for intervention—suggest that although cost-thinking constitutes some part of real-world opinion formation, respondents to the vignettes in these experiments may tend not to think about costs associated with intervention. Instead, they look superficially at whatever information is contained within the prompt, construing at the high level, considering the symbolic weight (or desirability) of it rather than treating it like an actual decision to approve or disapprove. If they do this to a greater extent in the experimental context than they would during live surveys, it would present an external validity problem; people surveyed about an actual impending act of US military deployment would be more likely to engage in low-level construal, consider the costs, and form their opinions thusly. The public is rationally ignorant, 303 and likely remains so facing the abstract vignettes common in these experiments. If this bias is at play in these experiments, it would cause a larger proportion of their respondents not to consider costs than would in a survey on a real impending intervention. The now standard audience costs survey experimental test offers a useful, well-trodden path towards testing the effects of construal level bias on survey experimental findings in 294 Kriner, Lechase, and Zielinski 2015; Flores-Macías and Kreps 2015. 295 2016 296 Eichenberg 2005; Kriner 2006; Karol and Miguel 2007; Gelpi, Feaver, and Reifler 2009. 297 Boettcher and Cobb, 2006; Boettcher, 2004; Gartner, 2008, 2011; Kriner and Shen, 2012; Walsh, 2015 298 2004 299 Gartner 2008, 105. 300 Mueller 1973; Kernell 1978; Ostrom and Simon 1985. 301 Gelpi et al. 2009 302 Berinsky 2009. 303 Popkin 1994. 80 international security studies. Costs—in terms of human life and national expenditures—are a salient consideration of citizens forming opinion on their nation’s decision to intervene. If it can be shown that respondents who consider costs when they answer opinion questions have higher disapproval for intervention and lower disapproval for presidential inconsistency, it opens an avenue for improving external validity in survey experiments. As established, when it comes to understanding the effects of costs on opinion, researchers have generally directly manipulated cost within the vignette, inserting some quantity within the mind of the respondent and measuring the effect of that number. 304 This study attempts to get at another effect of cost, namely the mere prompt to consider it, on survey experimental results. In the low information context of the survey hypothetical, respondents may form their answers with whatever expectations they already hold, or else neglect to consider relevant costs altogether. There are good empirical reasons to look for these biases in the context of foreign policy crises. However, most survey experiments have generally left the effects of survey respondents’ assumptions about costs unmeasured and untested. Take Tomz as an example. 305 He directly manipulates the economic cost of the US decision between “low” and “high,” and he varies the effort of the US between “a major effort” and “without major effort.” And to his credit, he adds a fourth “partial intervention” treatment group reporting that 20 Americans died before the president decided to back down. That group levied the highest audience cost of any group, at 9 percentage points higher than the next closest group. From this design, one cannot conclude whether it was the information on cost or the partial intervention itself that moved opinion. However, that result alone is enough to inspire questions in the audience costs context about how people treat costs when they are not given specifics. They might fill in that blank themselves with an unknown value. For example, Huddleston and Weller find that, across treatment groups, survey experimental respondents have systematically different expectations about unstated aspects of the survey scenarios, and that their conceptions mediate a part of the experimental effect. 306 Or, perhaps they ignore the question of cost entirely. Both of these possible tendencies would heighten already common concerns about the external validity of these experiments. 307 To fill in this gap, a good place to start would be to explore how the audience itself thinks about costs of military action. Some important questions to answer might be: do survey respondents generally consider costs when they are not prompted? What do they assume about costs? To what extent does variation in expectation about costs determine effect sizes in terms of disapproval? How do respondents’ assumptions about monetary costs compare with those about military casualties? As this line of questioning has generally not been explored in the literature, I chose to test the effects of prompting survey respondents to consider the costs of intervention. 304 Kriner and Shen 2012; Kriner, Lechase, and Zielinski 2015; Flores-Macías and Kreps 2015; Walsh 2015. 305 2007 306 2017 307 Barabas and Jerit 2010. 81 4.5 RESEARCH DESIGN I designed an experiment to establish whether respondents in survey experimental tests of foreign policy opinion discount the costs associated with hypothetical acts of intervention. The basic design was to induce low-level construal in the treatment group by introducing an open- ended question to consider casualties before measuring approval, and then compare the size of the effect within that group to that of the control, which received no such prime prior to indicating approval. The research question here was straightforward: Do respondents facing hypothetical foreign intervention scenarios respond differently when primed to consider casualties? Rather than planting a specific casualty level in the vignette, the design used here allowed the respondents to maintain whatever expectations they already held about foreign intervention. This better reflects what might happen in a poll in which survey respondents are asked their opinion on an impending act of intervention, such as polls of American opinion about intervention in Syria in 2013, 308 when respondent assumptions about casualties were undiscussed. 309 My treatment was a single question asking how many casualties they expected in the hypothetical scenario. This prime made them consider casualties for a moment before answering the approval question. In this way, the design addresses work showing that real-world opinion formation on military intervention is affected by concern for casualties. Primed with only an abstract scenario, respondents in the control group likely do not consider these costs. Treated respondents are made to consider them. With this inducement of low-level construal of the scenario, one should expect to see some mitigation of audience costs. Following this logic, the key hypotheses are: Cost Hypothesis: Respondents primed to consider American casualties before indicating approval will: 1. Indicate greater disapproval of intervention. 2. Indicate lower disapproval of backing down from threats. 3. Generate a lower estimated audience cost. The experimental test was simple. In October 2015, I recruited 1512 Amazon Mechanical Turk users for an experiment modelled after Tomz 2007. 310 I also ran a second test with an open reputation prompt (reported below). I did not test the Follow Through prompt in the second experiment because the traditional treatment effect is calculated by taking the difference in approval between the Stay Out and Empty Threat groups. MTurk is rightly criticized for the non- representative samples it produces. Still, it has repeatedly been defended as superior to undergraduate samples (Berinsky et al., 2011, 2012; Buhrmester et al., 2011), and it has been found to yield results similar to samples collected by Time-Sharing Experiments for the Social 308 CNN 2013c; The Economist 2013. 309 As an aside, it is interesting to note that the American public was vehemently opposed to Obama’s push to intervene in Syria, despite his apparent attempt to follow through on an explicit threat. 310 The sample was limited to Mechanical Turk users in the USA. 82 Sciences (TESS) (Mullinix et al., 2015). More broadly, the external validity question here is whether survey experiments reflect the non-experimental survey context of polling opinion during ongoing situations. Validity questions concerning sampling are certainly valid, but they are separate, and a nonrandom sample is unlikely to undermine the results reported herein. All respondents were shown the same introductory statement and one vignette from Tomz, in which a dictator invades its neighbor to get more power and resources, it has a strong military, and its victory hurts the safety and economy of the United States. 311 This group garnered the highest disapproval in his study and should represent the most difficult test for my hypotheses. The respondents were assigned randomly to one of six groups in a 3x2 arrangement. 312 They each encountered one of three presidential actions now standard among audience cost experiments: the president stays out of the conflict (the Stay Out condition); threatens intervention and backs down (Empty Threat); or threatens and intervenes (Follow Through). They also encountered an attention check. Respondents were asked for their approval/disapproval following the scenario: Do you approve, disapprove, or neither approve nor disapprove of the way the US President handled the situation? o Approve o Disapprove o Neither approve nor disapprove I followed up with a question of whether they “very strongly” or “somewhat” approve or disapprove of the president’s action, or, if they chose “neither”, a question of whether they “lean” one way or the other. The answer to the initial question and the follow-up question created a 7-point ordinal measure: strongly disapprove, disapprove, lean disapprove, neutral, lean approve, approve, strongly approve. The treatment group received the following question before answering the approval question: How many US troops would you expect were [to be] killed when [if] the president [had] decided to stop the invasion with military force? o Fewer than 50 o 50–99 o 100–499 o 500–1999 o 2000–4999 o 5000–10000 o More than 10000 311 Full survey instruments are in Appendix D. 312 The N for each group is in Appendix B. 83 Respondents reported what level of fatalities they expected in the scenario. This prompted them to give the human cost of intervention at least a few seconds of thought just before they indicated their level of approval of the president’s action. The control group answered the approval question first, providing the measure of the dependent variable before indicating their expectations about casualties. In the control group, the casualties question serves the role of a data collection instrument; its presence has no effect on the findings discussed below. This design does not so much improve on previous designs as it does utilize a common design to test whether purposefully prompting cost-thinking reduces treatment effects in survey experiments designed to simulate the relevant real-world scenarios, as well as to compare the effect of that cost prompt with that of the important mechanism of reputation. The goal is to nudge treated respondents to consider the costs of intervention momentarily before forming their opinions, but unlike past designs, the cost in the scenario is not itself manipulated, only the order of the cost and approval questions. Respondents are allowed to apply their own perceptions of cost, “filling in the blank” with either preconceived notion or rational calculation. To reiterate, my intention was to give the cost procrastination hypothesis a difficult test by: (1) choosing Tomz’s most compelling scenario; (2) testing the effect of a question about casualties that is easy for respondents to think about, but perhaps less impactful than other factors; and (3) giving respondents as weak a prompt as possible, as opposed to strong, specific manipulation of cost like those in other survey experiments; and (4) comparing the results from that test to those of a second test similarly designed to prompt concern for reputation. 4.6 EXPERIMENTAL RESULTS Figure 4-3 displays the differences in respondent estimates of casualties across vignettes. The Stay Out group expected higher casualties than the Empty Threat group, who expected more than the Follow Through group. Since casualties are not discussed or even alluded to in any of the three vignettes, one would expect pretreatment perceptions about casualties to be orthogonal to vignette assignment. However, vignette assignment clearly alters expectations, even without mentioning casualties. 84 Figure 4-3. Effect of vignettes on expected casualties Dafoe et al. 313 call this kind of variance in assumptions about undiscussed details in the vignettes an “information equivalence violation.” This kind of systematic variation in background assumptions can sometimes mediate experimental effects through atheoretical pathways. This possibility does not affect the validity of the comparisons made in this chapter— which are based on assignment to a casualties prime, not to a specific scenario—but is nonetheless helpful to consider. I analyzed differences in presidential approval across treatment groups in two ways: independent two-group t-tests using the 7-point measure of approval to obtain the initial comparison across the treatment groups; 314 and the better suited ordered logit test. 315 313 Dafoe, Zhang, and Caughey forthcoming. 314 Approval is treated as continuous. Full table of t-test results in Appendix B. 315 The parallel regressions assumption was tested with likelihood ratio, score, Wald, Wolfe-Gould, and Brant tests; and was found not to be violated. See Appendix B. 85 Figure 4-4. Difference in approval from casualties prime Figure 4-4 shows that the casualties prime significantly increased average approval of the president in the Empty Threat condition, by about .4 points on our 7-point scale. The other two groups move in the expected direction, but are insignificant. Experimental studies of audience costs generally hinge upon changes within the Empty Threat group, so this result is important. 86 Figure 4-5. Change in predicted assignment with casualties prime The ordered logit model provides an even clearer picture. 316 The difference for the Stay Out group was negligible, but the Empty Threat and Follow Through groups saw a similarly oriented effect on approval of intervention, significant at the .05 level. Figure 4-5 shows the difference caused by the casualties prime on the predicted probability of assignment to each approval category. Respondents in the Empty Threat group, primed to consider casualties were: 5.9% less likely to strongly disapprove, 1.9% less likely to somewhat disapprove, 3.3% more likely to somewhat approve, and 2.7% more likely to strongly approve. Conversely, Follow Through respondents were 3.1% more likely to strongly disapprove, 1.9% more likely to somewhat disapprove, 2.1% less likely to somewhat approve, and 4.9% less likely to strongly approve. In short, primed subjects were generally more critical of intervention. The casualties prime also altered the “absolute audience cost.” This was defined by Tomz as the “surge in disapproval” caused by not following through on the commitment to intervene, as compared to a commitment to stay out of the conflict altogether. 317 316 These results are robust to demographic controls and an ordered probit test. See Appendix B for all models. 317 Tomz 2007, 829. 87 Table 4-2. The effect of casualties priming on absolute audience costs Public reaction to empty threat (%) – Public reaction to staying out (%) = Difference in opinion (%) Summary of differences (%) Control Group Disapprove strongly 29 (24–35) 19 (14–24) 10 (2–18) 17 Disapprove somewhat 26 (20–31) 18 (13–23) 7 * (0–15) Casualties Prompt Disapprove strongly 23 (17–28) 16 (12–20) 7 (0–13) 6 Disapprove somewhat 20 (15–25) 21 (16–25) -1 (-8–6) Change in Absolute Audience Cost: -11 95% confidence interval in parentheses. * Discrepancy due to rounding error Table 4-2 replicates Tomz’s method of calculating absolute audience cost, 318 and demonstrates the difference in absolute cost caused by the casualties prime. With an absolute cost of 17 for the control group, my estimate approximates Tomz’s estimate of 16. However, the casualties prime resulted in an 11-point drop in absolute audience cost, leaving it at just 6 for the treated group. Some discounting of casualties may be expected in these experiments, but that priming casualty consideration should alter the standard result so much is surprising. Prior work indicates expectations about casualties inform opinion on military engagement, so this result may indeed indicate a gap between how subjects respond to real and hypothetical acts of military intervention. 4.7 PRIMING REPUTATION It might be argued that the effect of the prime is independent of the content of the prime, that any prime could produce this effect. To account for this possibility, and to show that casualties constitute a particular concern for these experiments, I also examine the subjects’ responses to 318 Ibid., 827. 88 priming of concern for international reputation. The second design was similar. As an abstract rather than a concrete idea, reputation is likely to be construed at the high-level. Construal level theory instructs us that “high-level construals are more likely than low-level construals to remain unchanged as one gets closer.” 319 That is, people do not vary as much in their assessment of abstract implications between the proximal and the distal as they do assessing concrete details. As such, whatever negative effects on reputation people expect should be accounted for in their formation of their opinion, so a prime of reputation should alter the experimental effect very little. In December 2016 and March 2017, I recruited 1323 more Mturkers, ensuring both times that I would have no repeat respondents. The experimental design was the same as that reported in the main text, except that I did not include the “Follow Through” group. For this comparative test, in place of costs, I inserted a prime about the US’ reputation, a key mechanism in ACT. 320 The “open reputation” group received the following question: What effect, if any, would the President’s actions in the prior scenario have on the international reputation of the US? o It would harm the US’ reputation o It would not affect the US’ reputation o It would improve the US’ reputation The order of the choices was randomized. I followed up with a question about whether they believed it would harm/improve US reputation “somewhat” or “a great deal,” yielding a five- point estimate. 319 Trope and Liberman 2010. 320 Fearon 1994; Schultz 2001. 89 Figure 4-6. Difference in approval from reputation prime Figure 4-6 displays differences in approval caused by that intervention. In both groups, the difference caused by the reputation prime is minimal. Unlike the casualties prime, the reputation prime caused no movement in the treatment effect in either direction. It appears respondents already account for reputation without being primed. This is consistent with ACT, and with Brutger and Kertzer’s finding that varying the richness of the vignette does not much affect audience costs via concern over national reputation. 321 Overall, the findings for the reputation prompt can be summarized simply. While respondents’ estimates of the effect on reputation varies more widely between vignette groups than do estimates of casualties, the effect of the prompt itself is not equivalent. It seems respondents are already consciously or unconsciously accounting for reputation in their responses to the approval question, and the nudge to consider reputation does not have much effect. It does not significantly weaken approval of the empty threat, nor does it significantly alter the overall treatment effect (the difference in approval between Stay Out and Empty Threat). 4.8 DISCUSSION Psychology research associated with construal level theory suggests that people tend to give less consideration to low-level concerns in abstract scenarios than they do in those closer at hand. Costs are subject to low-level construal and consideration of the feasibility of outcomes. However, abstract situations encourage primarily high-level construal and prioritizing concerns of desirability. In the survey experimental context, this pattern would seem to suggest the existence of a small but specific external validity problem: that subjects responding to the low 321 Brutger and Kertzer forthcoming. 90 information vignettes common in these designs would give relevant costs less consideration than they would if they were facing the analogous situation of impending American intervention in the real world. Two findings reported herein affirm cause for this concern. First, I find an effect consistent with expectations derived from construal level theory and delay discounting. Respondents asked to estimate casualties support a hypothetical act of intervention to a significantly lesser degree than those who are not. Inducing them to think about casualties increases approval of backing down from commitments and decreases approval of following through on threats. This is despite the fact that they reported a relatively lower expected casualty level in the Follow Through scenario. Second, this difference in deliberation substantially decreases estimated audience costs. This might imply that some prior experimental results on audience costs are less reflective of real world opinion formation on foreign intervention than they are taken to be. Citizens assessing real military intervention consider casualties. Inducing them to do so in the hypothetical is shown here to strongly mute the size of the estimated audience cost the president faces. 322 To improve external validity, researchers may benefit from considering possible gaps in construal level between their experimental design and the analogous political environment, and then design features to encourage the right level of consideration. Several useful avenues for future work are implied in these findings. A project aimed at fully understanding these findings might identify whether and to what extent a similar prime to consider casualties would affect opinion on a real international military engagement. Another useful direction might be to explore untested respondent assumptions about the benefits of intervention, and test whether those are construed similarly to costs. Overall, there is much room to expand our understanding of respondent behavior in the experimental context within the audience costs research agenda. A smaller finding of interest is that respondents have different assumptions about casualties across treatment groups, even though they are mentioned nowhere in the initial scenario. Further work might explore the extent to which information equivalence violations are present in or responsible for common experimental findings. As a slight aside, it is not presumptuous to conclude that audience cost theory falls short in explaining public opinion in the Syrian case. It is, of course, only one case, and no genuinely substantive theoretical modification can be made based upon this case alone. However, it should have been an easy case for ACT, given the clarity of the threat, the early support of the public, and the relatively low cost of intervention. Thus, while one might resist drawing theoretical conclusions, there is good reason in this case to push for new approaches to one common method utilized to test the bounds of this theory, survey experiments. The Syrian case provides a strong example of how surveys attempting to capture public opinion on something fluid like foreign crises can be found lacking. My experimental results show evidence that a construal bias towards high-level processing of abstract scenarios may affect the results of some survey experiments. Scholars ought to push to discover the scope of the problem, begin measuring the size and 322 Furthermore, there is no equivalent effect from priming consideration of national reputation. See Appendix A. 91 direction of its effect on experimental results, and improve validity by crafting designs that minimize cost procrastination. 92 CONCLUSION: SEPARATION ANXIETY 5.1 SEPARATION ANXIETY This study joins a body of scholarship 323 supporting to the same unfortunate conclusion. Despite a nominally peace-driven purpose to our international institutions, created to foster ties and curtail conflict, when it comes to self-determination, chaos and instability are the conditions that convince third parties to consider and support the bids of aspiring state actors. The key finding to this end is that outside states move towards supporting aspiring states when violence reaches high levels and when rebels successfully capture territory by force. There are many trade-offs involved in crafting policy towards these groups. Quasi-state entities with various degrees of juridical sovereignty create long term regional instability, since the boundaries on the map do not reflect the reality on the ground, so revisionist groups are encouraged to keep striving generation after generation. However, the precedent set by official admission into the international system makes for instability of the system itself. Diplomatic recognition is a costly action for a state to take, especially given that is a mere declaration of support, words on paper, rather than material or economic support. Third-party states may not lend support via diplomatic recognition, but during conflict they still may move towards 323 Coggins 2014; Paquin 2010; Buzard, Graham, and Horne 2017; Chenoweth and Stephan 2012. 93 providing other means of support, encouraging a temporary kind of stability, but not likely sustainable for very long. That aspiring states are effectively put into a position in which nonviolent, political means of reaching their goals is an ineffective strategy is equally unfortunate. The post-WWII channels of exchange between states makes up a system that prevents the types of conflict seen in the past, so that Tilly’s characterization of the relationship between war-making and state- building is no longer so apt. However, for an aspiring state to garner the international legal sovereignty that comes with universal recognition is very difficult through those same channels. Minority groups still sometimes perceive themselves as so poorly represented by their respective central governments, that stirring for greater autonomy seems to be the only path forward. Those actors are still in the land of Tilly. International legal standards for official recognition call for the establishment of the monopoly on force, and third parties will assist rebels in getting there in the right conditions, even if they do not follow the “declaratory” practice towards recognition from there. Hence, the question of self-determination is characterized by ambiguity at every level. No clear answer lies in international law or the norms of international institutions, and as I show here, taking diplomatic recognition as the only reflection of sovereignty does leave something of a gap in our understanding of the role third parties play in these conflicts. I have attempted to add a degree of nuance to the question by conceptualizing international legal sovereignty, and recognition, as a quality that has many foreign policy consequences. Stated plainly, a state cannot send arms to or build diplomatic channels with actors it does not recognize, in the conventional sense of the term. To manifest the broadened characterization of recognition, I model recognition as a latent trait or underlying quality, rather than as a discrete foreign policy decision (which happens to be very difficult for a government to make.) In so doing, I provide a novel complement to the excellent prior work on diplomatic recognition. 5.2 LIMITATIONS AND FUTURE DIRECTIONS This study has established a foundation for a larger research agenda on international sovereignty. Moving forward, several limitations will need to be addressed. By nature, my measure and approach to this question “black boxes” the state, incapable of distinguishing the state level processes that lead to the slate of foreign policy decisions upon which I base my analysis. I have offered no insights into the question of whether third parties’ intentions matter at all in these processes. My findings are consistent with a stability-seeking theory of third-party support of self-determination movements, but that does not mean I can explain the processes that generate these decisions. Moving forward, a chief task to solidify the evidence shown here will be to approach a few cases much more closely and break down the decision-making processes of the leadership. A few useful questions stand out. Do leaders discuss decisions like intervention, sanctions, and diplomatic support in terms of stability and conflict resolution? Are these decisions even made by the same parties, or are they spread across the many institutions that make up the state? Do leaders discuss the implications of their decisions on long term stability of the international system? 94 My fourth chapter accomplishes some understanding of the inputs into the black box of the state, but what goes on within the state is left undiscussed. A good way to answer these concerns, and a natural next step for this project, would be interviews with elites from several past and present third-party governments, as well as archival exploration and other qualitative study. The quantitative nuance I bring to the discussion is still somewhat shallow and open to empirical interpretation, even though most of the predictions stemming from a stability-seeking theory of latent recognition are born out in the statistical tests. A selection of interviews with elites on the inside of these foreign policy decisions would provide a much more concrete view of the mechanisms at play in these statistical relationships, especially concerning the intentions guiding governments to shift their foreign policy. For that matter, my study black boxes the conflict as well. This plays out on two levels. The first is at the theoretical level. I have not given any attention to the strategies of either the self-determination groups or the host state governments. Although these topics have been explored at length elsewhere, 324 the insights from those works have not played a large role theoretically or empirically in this study. For example, self-determination groups often have high competition within their ranks, 325 so that host state governments find points where they can drive a wedge between groups or highlight the domestic chaos as a means of convincing third parties to take certain positions. My use of rough measures of violence and victory as independent variables may miss this kind of nuance in evaluating these conflicts. As I have with the third party, I have simply taken the “outputs” of the conflict, and there are drawbacks to doing so. The second is at the level of the data itself. I use a combination of Coggins’ and Griffiths’ samples 326 of active self-determination and secessionist groups. These observations begin when self-determination groups formally declare independence, have a national flag, claim a territory, and are active for more than a week. They end when they formally concede, go five years without public activity, or reach some resolution with the host state government. The problem with this data structure for my purposes is that the births and deaths of some of these movements depend on their expectations of how third parties, especially major powers, will respond. The decision to “begin” a separatist or self-determination claim is not random—it may depend on the characteristics of the leadership in important governments (e.g. the difference in stances between Roosevelt and Truman vis-à-vis the Soviet Union following WWII). As such, it is conceivable that certain relationships detected in my analysis here may also apply to the appearance and disappearance of the data itself. This will be an important issue to address moving forward in this project, both with further development of the model and with the kind of case study research outlined earlier. Another shortcoming that, if mended, would make a stronger case for the way I characterize stability in conflict is the type of data I use to analyze both violence and instability. The ordinal variable used to measure violence is a rough measure, and the better continuous measure simply does not cover enough cases over a long enough time frame to be very useful. 324 Cunningham 2014; Ker-Lindsay 2014. 325 Cunningham 2014. 326 Coggins 2014; Griffiths and Butcher 2013. 95 Although a weak curvilinear relationship between violence and latent recognition is apparent in the data, that finding is far from conclusive. Likewise, other stages of the progression of conflict—organizing, political protest, repression, and early stages of two-sided conflict—are left out of the tests reported herein. A complete test of a stability-seeking theory would include analysis of third party actions towards governments cracking down on separatist protests, and would pay closer attention to the timing of these events. Often, these kinds of conflicts often move much more quickly than t =1-year data can reflect. In this case, this constraint serves in some ways to make my findings that much stronger, since I show that year over year, foreign policy trends are consistent with this theory. However, it does mean I cannot speak authoritatively on the implications of governments facing nascent separatist stirrings. Finally, there is the issue of fully validating the measure developed here. Building upon my findings here will require a robust framework to prove validity of my measurement strategy. I have offered a degree of face validity in my demonstration of the measure’s performance in the cases of Palestine and Western Sahara. Likewise, I have shown it to perform as expected alongside Coggins’ conventional measure of likelihood of diplomatic recognition, converging with and diverging from that measure in theoretically sensible ways. However, for a methodological revision as ambitious as this study, what I have presented here is still imperfect. The next step towards robust validation of this measure will involve seeking and selecting some data on international phenomena that are theoretically related to underlying recognition but should nevertheless be uncorrelated with my measure, and then proving it to be so. To revert to the student ability example presented in Chapter 2, an instrument designed to evaluate ability or college preparedness through multiple choice questions has to be checked to ensure it is not simply measuring unintended qualities, such as cultural knowledge. So, there are several challenges to surmount moving forward in the research agenda established by this study. Nevertheless, the theory, measure, and findings presented here provide a novel and useful complement to the literature on intervention and international sovereignty. 5.3 POLICY IMPLICATIONS The clearest policy implications of this research are general. Several are clear for states facing separatist and other self-determination groups. The first—the most easily said and least easily done—is to avoid letting the crises that precipitate these conflicts lapse into violence. When conflict escalates into civil war, host states lose their position as the guardian of the status quo, and they risk becoming internationally seen as interlopers, falling gradually out of the favor of third parties with interests in the region. Governments can avoid the crisis by making separatist recruitment more difficult. Often, agitation along identity-group lines on the part of the government is a key part of the process of escalation. Minority groups with good channels of representation in the government—able to voice and address their grievances, and not unfairly situated in the economy or society— naturally have a harder time recruiting followers and building strength. This may be part of why, although there have been dozens of separatist groups in European democracies active in the last 96 20 years, 327 only a very small minority of them have managed to garner the local support for even something as basic as a referendum. Of those who have, none have seen a compelling majority both participate in the referendum and vote for independence. A more general policy prescription to address the incentive to instigate or invite violence during crisis would be to lower the bar somewhat for international legal sovereignty. As Keating puts it, “It might be helpful if there were more precedents for peaceful, orderly, democratic separations, rather than violent, chaotic ones.” 328 Currently, the concern for precedent held by so many states precludes granting international legal sovereignty in the form of bilateral recognition. As UN policy and international law remain ambivalent on the question of when it is appropriate to extend recognition, another helpful change might be to codify the criteria for state entry on a basis other than one that requires seizure of the monopoly on force. In his “thorough and unapologetic defense of the right to secede,” 329 Wellman argues that states ought to start laying out the criteria for peaceful secession, based on preconditions such as size and governing capability. 330 A select few countries have experimented along these lines. One example might be the constitution of St. Kitts and Nevis, which lays out a legal, peaceable path for the separation of Nevis: a two-thirds majority in a referendum. 331 This provision sets the bar for democratic separation high, but still within the attainable limits of historical cases of strong separatist movements. Other states, such as Denmark, 332 Ethiopia, 333 and Liechtenstein 334 have similar provisions in their constitutions. It is conceivable that a peaceable path towards separation could be laid out in international law in a similar way. Perhaps a two-thirds majority in an independence referendum could also be codified as a new standard for bilateral and international recognition, serving to replace the current standard, which makes no democratic provisions and even implies a path of force. There may be more practicable remedies for the chaos surrounding self-determination. Breaking these trails starts with admitting to the messy string of contradictions self- determination movements encounter: we say we will only recognize places that look and act like states, but you shouldn’t get to that point violently, except we are more likely to recognize the ones who do get there violently, and we’ll even help them, but only sometimes, and we may still not recognize them once they get there. In this realm, the usual channels for finding compromise and encouraging peaceable resolution of conflict, the purpose of our international institutions, are benighted by anarchy and practices rooted in bad faith. Baev may have put it best: “The state in question cannot ignore the democratically expressed preference of its province and should in good will negotiate the conditions and timetable for a separation. This compromise between the 327 Noack 2017. 328 Keating 2017. 329 Wellman 2005, 1. 330 Ibid., 38. 331 Constitution of Saint Kitts and Nevis 1983. 332 Constitution of Denmark 1953. 333 Constitution of Ethiopia 1994. 334 Constitution of Liechtenstein 1921. 97 basic principles of territorial integrity and self-determination needs solid international guarantees to be applicable to conflict situations.” 335 It may just be that chaos is innate in an international order based on a patchwork of borders inherited from war and domination. When that same system sets a premium on nationalism and touts high standards for descriptive and substantive representation in government, the incentive for separatism endures. A system oriented to encourage systemic stability, free exchange, and ties that bind should encourage better ways of addressing (rightly or wrongly) aggrieved minorities than the double-speak that currently rules the conversation. As long as the current order endures unchanged, the international community should not expect these conflicts to cease to occur. Like the Long Peace following WWII, this peace must be made deliberately. 335 Baev 1999, 47. 98 BIBLIOGRAPHY Adida, Claire L., Lauren D. Davenport, and Gwyneth McClendon. 2016. Ethnic Cueing across Minorities: A Survey Experiment on Candidate Evaluation in the United States. Public Opinion Quarterly 80 (4): 815–836. Al Jazeera. 2018. Ethiopia to join Somaliland, UAE port development venture. Al Jazeera. Available at <https://www.aljazeera.com/news/2018/03/ethiopia-join-somaliland-uae- port-development-venture-180301151054778.html>. Accessed 3 March 2018. Al Jazeera. 2010. Obama renews Sudan sanctions. Al Jazeera. Available at <http://www.aljazeera.com/news/africa/2010/11/2010111191135816881.html>. Accessed 18 October 2017. Anderson, Benedict. 1991. Imagined communities: reflections on the origin and spread of nationalism. Rev. ed. London; New York: Verso. Angeletos, George-Marios, David Laibson, Andrea Repetto, Jeremy Tobacman, and Stephen Weinberg. 2001. The Hyperbolic Consumption Model: Calibration, Simulation, and Empirical Evaluation. The Journal of Economic Perspectives 15 (3): 47–68. ARSO. 2002. On the Establishment of Diplomatic Relations between the Democratic Republic of East Timor and the Sahrawi Arab Democratic Republic. Communique. Bir-Lahlou, Western Sahara: Association de soutien à un réféndum libre et régulier au Sahara Occidental (ARSO). Available at <http://www.arso.org/timor200502.htm>. Accessed 28 April 2014. ARSO. 2014. States which recognize the SADR. Association de soutien à un réféndum libre et régulier au Sahara Occidental (ARSO). Available at <http://www.arso.org/03-2.htm>. Accessed 28 April 2014. Baev, Pavel K. 1999. External interventions in secessionist conflicts in Europe in the 1990s. European Security 8 (2): 22–51. Barabas, Jason, and Jennifer Jerit. 2010. Are Survey Experiments Externally Valid? The American Political Science Review 104 (2): 226–242. Bar-Anan, Yoav, Nira Liberman, and Yaacov Trope. 2006. The association between psychological distance and construal level: Evidence from an implicit association test. Journal of Experimental Psychology: General 135 (4): 609–622. Bartels, Larry M. 2003. The Political Psychology of Party Identification. In Electoral democracy, edited by Michael MacKuen and George Rabinowitz, 83. Ann Arbor: University of Michigan Press. Baum, Matthew A. 2004. Going Private: Public Opinion, Presidential Rhetoric, and the Domestic Politics of Audience Costs in U.S. Foreign Policy Crises. The Journal of Conflict Resolution 48 (5): 603–631. Baum, Matthew, Philip B. K. Potter, and Matthew Baum. 2015. War and democratic constraint: how the public influences foreign policy. Princeton, New Jersey: Princeton University Press. Bayer, Reşat. 2006. Diplomatic Exchange Data set, v2006.1. Available at <http://correlatesofwar.org/data-sets/diplomatic-exchange>. Bélanger, Louis, Érick Duchesne, and Jonathan Paquin. 2005. Foreign Interventions and Secessionist Movements: The Democratic Factor. Canadian Journal of Political Science/Revue canadienne de science politique 38 (2): 435–462. 99 Bellantoni, Christian, and Terence Burlij. 2013. Congress Returns as President Mounts Major Syria Lobbying Effort. PBS Newshour. Available at <http://www.pbs.org/newshour/rundown/2013/09/congress-returns-as-president-mounts- major-syria-lobbying-effort.html>. Berinsky, Adam J. 2009. In time of war: understanding American public opinion from World War II to Iraq. (Chicago studies in American politics); Chicago studies in American politics. Chicago: University of Chicago Press. Berinsky, Adam J. 1999. The Two Faces of Public Opinion. American Journal of Political Science 43 (4): 1209–1230. Bertrand, Marianne, and Sendhil Mullainathan. 2001. Do People Mean What They Say? Implications for Subjective Survey Data. The American Economic Review 91 (2): 67–72. Biswas, Shampa, and Sheila Nair. 2010. International relations and states of exception: margins, peripheries, and excluded bodies. London; New York: Routledge. Boettcher, William A. 2004. Military Intervention Decisions regarding Humanitarian Crises: Framing Induced Risk Behavior. The Journal of Conflict Resolution 48 (3): 331–355. Boettcher, William A., and Michael D. Cobb. 2006. Echoes of Vietnam? Casualty Framing and Public Perceptions of Success and Failure in Iraq. The Journal of Conflict Resolution 50 (6): 831–854. Brader, Ted A., and Joshua A. Tucker. 2009. What’s left behind when the party’s over: survey experiments on the effects of partisan cues in Putin’s Russia. Politics and Policy 37 (4): 843–869. Bremer, Ian. 2013. Iran is America’s real Middle East priority. Reuters. Available at <http://blogs.reuters.com/ian-bremmer/2013/09/06/iran-is-americas-real-middle-east- priority/>. Brierly, James Leslie. 1963. The law of nations; an introduction to the international law of peace. 6th ed. edited by Humphrey Waldock. New York: Oxford University Press. Brownlie, Ian. 2008. Principles of public international law. 7th ed.. Oxford; New York: Oxford University Press. Brutger, Ryan, and Joshua D. Kertzer. forthcoming. A Dispositional Theory of Reputation Costs. International Organization. Buchanan, Allen. 1998. Democracy and Secession. In National Self-Determination and Secession, edited by Margaret Moore. Oxford; New York: Oxford University Press. Buchanan, Allen. 1991. Secession: The Morality of Political Divorce from Fort Sumter to Lithuania and Quebec. Boulder; San Francisco; Oxford: Westview Press. Buzard, Kristy, Benjamin A. T. Graham, and Ben Horne. 2017. Unrecognized States: A Theory of Self-Determination and Foreign Influence. The Journal of Law, Economics, and Organization 33 (3): 578–611. Callahan, David. 2002. The Enduring Challenge: Self Determination and Ethnic Conflict in the 21st Century. Carnegie Corporation of New York. Carment, David, and Patrick James. 1995. Internal Constraints and Interstate Ethnic Conflict: Toward a Crisis-Based Assessment of Irredentism. The Journal of Conflict Resolution 39 (1): 82–109. Carment, David, and Patrick James. 1997. Secession and Irredenta in World Politics: The Neglected Interstate Dimension. In Wars in the midst of peace: the international politics of ethnic conflict, edited by Patrick James and David Carment. Pittsburg, Pa.: University of Pittsburgh Press. 100 Carment, David, and Patrick James. 1996. Two-Level Games and Third-Party Intervention: Evidence from Ethnic Conflict in the Balkans and South Asia. Canadian Journal of Political Science / Revue canadienne de science politique 29 (3): 521–554. Carr, Caleb. 2002. The lessons of terror: a history of warfare against civilians: why it has always failed, and why it will fail again / Caleb Carr. New York: Random House. Carter, David B., and H. E. Goemans. 2011. The Making of the Territorial Order: New Borders and the Emergence of Interstate Conflict. International Organization 65 (02): 275–309. Caryl, Christian. 2015. The World’s Next Country. Foreign Policy. Available at <https://foreignpolicy.com/2015/01/21/the-worlds-next-country-kurdistan-kurds-iraq/>. Accessed 9 September 2017. Caspersen, Nina. 2011. Unrecognized States: The Struggle for Sovereignty in the Modern International System. Malden, MA: Polity Press. Caspersen, Nina. 2012. Unrecognized states: the struggle for sovereignty in the modern international system. Cambridge; Malden, MA: Polity. Chaudoin, Stephen. 2014a. Audience Features and the Strategic Timing of Trade Disputes. International Organization 68 (04): 877–911. Chaudoin, Stephen. 2014b. Promises or Policies? An Experimental Analysis of International Agreements and Audience Reactions. International Organization 68 (1). Check, Fact. 2013. Obama’s Blurry Red Line. FactCheck.org. Available at <http://www.factcheck.org/2013/09/obamas-blurry-red-line/>. Chen, Ti-chiang. 1951. The international law of recognition, with special reference to practice in Great Britain and the United States. The Library of world affairs, no. 16. New York: Praeger. Chenoweth, Erica, and Maria J. Stephan. 2012. Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. Reprint edition. New York: Columbia University Press. Chong, Dennis, and James N. Druckman. 2010. Dynamic Public Opinion: Communication Effects over Time. The American Political Science Review 104 (4): 663–680. Clawson, Rosalee A., and Zoe M. Oxley. 2013. Public opinion: democratic ideals, democratic practice. 2nd ed. Los Angeles: Sage/CQ Press. Closson, Stacy. 2011. What do unrecognized states tell us about sovereignty? In Unrecognized states in the international system, edited by Nina Caspersen and Gareth R. V. Stansfield. (Exeter studies in ethno politics). Milton Park, Abingdon, Oxon; New York: Routledge. CNN. 2013a. CNN | ORC Poll. Poll. Available at <http://i2.cdn.turner.com/cnn/2013/images/05/28/syriapoll.pdf>. Accessed 9 June 2015. CNN. 2013b. CNN | ORC Poll. Poll. Available at <http://i2.cdn.turner.com/cnn/2013/images/09/09/6a.poll.syria.pdf>. Accessed 9 June 2015. CNN. 2013c. CNN | ORC Poll. CNN. Available at <http://i2.cdn.turner.com/cnn/2013/images/05/28/syriapoll.pdf>. Accessed 9 June 2015. Coggins, Bridget. 2011. Friends in High Places: International Politics and the Emergence of States from Secessionism. International Organization 65 (03): 433–467. Coggins, Bridget. 2014. Power Politics and State Formation in the Twentieth Century: The Dynamics of Recognition. New York, NY, USA: Cambridge University Press. Cole, Taylor. 1928. The recognition policy of the United States since 1901. Baton Rouge: Department of government, Louisiana state university. Available at <https://hdl.handle.net/2027/mdp.39015062379675>. 101 Constitution of Denmark. 1953. Part 3, Section 19. Constitution of Denmark. In Constitution of Denmark. Available at <http://pdba.georgetown.edu/Constitutions/Kitts/kitts83.html>. Accessed 12 March 2018. Constitution of Ethiopia. 1994. Chapter 3, Part 2, Article 39. Constitution of Ethiopia. In Constitution of Ethiopia. Constitution of Liechtenstein. 1921. Chapter 1, Article 4. Constitution of Liechtenstein. In Constitution of Liechtenstein. Constitution of Saint Kitts and Nevis. 1983. Chapter 10, Section 113. Constitution of Saint Kitts and Nevis. In Constitution of Saint Kitts and Nevis. Available at <http://pdba.georgetown.edu/Constitutions/Kitts/kitts83.html>. Accessed 12 March 2018. Cook, Steven A. 2016. Who Exactly Are ‘the Kurds’? The Atlantic. Available at <https://www.theatlantic.com/international/archive/2016/02/kurds-turkey-pkk- ypg/470991/>. Accessed 9 September 2017. Croicu, Mihai, and Ralph Sundberg. 2017. UCDP GED Codebook version 17.2. Department of Peace and Conflict Research: Uppsala University. Cunningham, David E. 2010. Blocking resolution: How external states can prolong civil wars. Journal of Peace Research 47 (2): 115–127. Cunningham, Kathleen. 2014. Inside the politics of self-determination. New York, NY: Oxford University Press. Cunningham, Kathleen Gallagher. 2011. Divide and Conquer or Divide and Concede: How Do States Respond to Internally Divided Separatists? American Political Science Review 105 (02): 275–297. Dafoe, Allan, Baobao Zhang, and Devin Caughey. forthcoming. Informational Equivalence in Survey Experiments. Political Analysis. Available at <http://www.allandafoe.com/ie>. Accessed 9 February 2018. Daily-Star. 2013. France says ‘at least 281’ killed in Syria chemical attack. The Daily Star. Available at <http://www.dailystar.com.lb/News/Middle-East/2013/Sep-02/229697- france-says-at-least-281-killed-in-syria-chemical-attack.ashx#axzz2mGk8gMxW>. Davis, Susan. 2013. Senate delays Syria vote as Obama loses momentum. USA Today. Available at <http://www.usatoday.com/story/news/politics/2013/09/09/obama-congress-syria-vote- in-doubt/2788597/>. DellaVigna, Stefano, and Ulrike Malmendier. 2006. Paying Not to Go to the Gym. The American Economic Review 96 (3): 694–719. Downes, Alexander B. 2006. Desperate Times, Desperate Measures: The Causes of Civilian Victimization in War. International Security 30 (4): 152–195. Downes, Alexander B., and Todd S. Sechser. 2012. The Illusion of Democratic Credibility. International Organization 66 (3): 457–489. Druckman, James N. 2004. Political Preference Formation: Competition, Deliberation, and the (Ir)relevance of Framing Effects. The American Political Science Review 98 (4): 671– 686. Earnest, Josh. 2012. Press Gaggle by Principal Deputy Press Secretary Josh Earnest en route Columbus, OH, 8/21/2012. Available at <http://www.whitehouse.gov/the-press- office/2012/08/21/press-gaggle-principal-deputy-press-secretary-josh-earnest-en-route- colu>. 102 Eckstein, Harry. 1975. Case Studies and Theory in Political Science. In Handbook of Political Science, vol. 7. Political Science: Scope and Theory, edited by Fred I. Greenstein and Nelson W. Polsby, 94–137. Reading, MA, USA: Addison-Wesley. Eichenberg, Richard C. 2005. Victory Has Many Friends: U.S. Public Opinion and the Use of Military Force, 1981-2005. International Security 30 (1): 140–177. European Community. 1992. European Community: Declaration on Yugoslavia and on the Guidelines on the Recognition Of New States. International Legal Materials 31 (6): 1485–1487. Fabry, Mikulas. 2010. Recognizing states: international society and the establishment of new states since 1776. Oxford; New York, N.Y.: Oxford University Press. Fazal, Tanisha. 2007. State Death: The Politics and Geography of Conquest, Occupation, and Annexation. Princeton: Princeton University Press. Fazal, Tanisha M., and Ryan D. Griffiths. 2014. Membership Has Its Privileges: The Changing Benefits of Statehood. International Studies Review 16 (1): 79–106. Fazal, T.M. 2004. State Death in the International System. International Organization 58 (02): 311–344. Fearon, James D. 1994. Domestic Political Audiences and the Escalation of International Disputes. American Political Science Review 88 (3): 577–592. Fitch, Asa. 2016. DP World to Manage Somaliland Port of Berbera. Wall Street Journal, sec. Business. Available at <http://www.wsj.com/articles/dp-world-to-manage-somaliland- port-of-berbera-1473086050>. Accessed 10 December 2017. Flores-Macías, Gustavo A., and Sarah E. Kreps. 2015. Borrowing Support for War The Effect of War Finance on Public Attitudes toward Conflict. Journal of Conflict Resolution: 0022002715600762. Frederick, Shane, George Loewenstein, and Ted O’Donoghue. 2002. Time Discounting and Time Preference: A Critical Review. Journal of Economic Literature 40 (2): 351–401. Funk, Patricia. 2013. How Accurate are Surveyed Preferences for Public Policies? Evidence from a Unique Institutional Setup. Available at <https://ideas.repec.org/p/bge/wpaper/657.html>. Gabaix, Xavier, Hongyi Li, and David Isaac Laibson. 2005. Extreme Value Theory and the Effects of Competition on Profits. Available at <http://econpapers.repec.org/paper/clalevrem/784828000000000656.htm>. Accessed 1 June 2015. Gaines, Brian J., James H. Kuklinski, and Paul J. Quirk. 2007. The Logic of the Survey Experiment Reexamined. Political Analysis 15 (1): 1–20. Gallup. 2014a. Gallup Daily: Obama Job Approval. Available at <http://www.gallup.com/poll/113980/gallup-daily-obama-job-approval.aspx>. Accessed 14 March 2014. Gallup. 2014b. Poll: Presidential Approval Ratings. Gallup, Inc. Available at <http://www.gallup.com/poll/121199/Obama-Weekly-Job-Approval-Demographic- Groups.aspx>. Accessed 22 February 2014. Gartner, Scott S. 2011. On Behalf of a Grateful Nation: Conventionalized Images of Loss and Individual Opinion Change in War. International Studies Quarterly 55 (2): 545–561. Gartner, Scott Sigmund. 2008. The Multiple Effects of Casualties on Public Support for War: An Experimental Approach. The American Political Science Review 102 (1): 95–106. 103 Geldenhuys, Deon. 2009. Contested states in world politics. Houndmills, Basingstoke, Hampshire; New York, NY: Palgrave Macmillan. Gellner, Ernest. 1983. Nations and nationalism. (New perspectives on the past.). Ithaca: Cornell University Press. Gelpi, Christopher, Peter Feaver, and Jason Aaron Reifler. 2009. Paying the human costs of war: American public opinion and casualties in military conflicts. Princeton: Princeton University Press. Geman, Ben. 2013. Obama breaks from precedent in asking Congress for approval to strike Syria. The Hill. Available at <http://thehill.com/homenews/administration/319827- obama-breaks-from-precedent-on-syria>. Gerring, John. 2007. Case Study Research: Principles and Practices. New York, NY, USA: Cambridge University Press. Geys, Benny. 2010. Wars, Presidents, and Popularity: The Political Cost(s) of War Re- Examined. Public Opinion Quarterly 74 (2): 357–374. Geys, Benny, and Jan Vermeir. 2008. Taxation and Presidential Approval: Separate Effects from Tax Burden and Tax Structure Turbulence? Public Choice 135 (3/4): 301–317. Gibler, Douglas M. 2007. Bordering on Peace: Democracy, Territorial Issues, and Conflict. International Studies Quarterly 51: 509–532. Gibler, Douglas M., and Jaroslav Tir. 2010. Settled Borders and Regime Type: Democratic Transitions as Consequences of Peaceful Territorial Transfers. American Journal of Political Science 54 (4): 951–968. Gibson, James L., Gregory A. Caldeira, and Lester Kenyatta Spence. 2005. Why Do People Accept Public Policies They Oppose? Testing Legitimacy Theory with a Survey-Based Experiment. Political Research Quarterly 58 (2): 187–201. Gibson, James L., Jeffrey A. Gottfried, Michael X. Delli Carpini, and Kathleen Hall Jamieson. 2011. The Effects of Judicial Campaign Activity on the Legitimacy of Courts: A Survey- based Experiment. Political Research Quarterly 64 (3): 545–558. Gladstone, Rick, and C. J. Chivers. 2013. Forensic Details in U.N. Report Point to Assad’s Use of Gas. New York Times. Available at <http://www.nytimes.com/2013/09/17/world/europe/syria-united-nations.html>. Gleditsch, Nils Petter, Peter Wallensteen, Mikael Eriksson, Margareta Sollenberg, and Håvard Strand. 2002. Armed Conflict 1946-2001: A New Dataset. Journal of Peace Research 39 (5): 615–637. Goebel, Julius. 1915. The recognition policy of the United States. Columbia University. Studies in history, economics and public law, v. 66, no. 1. New York: Columbia University. Goldman, Russell. 2017. Kurds Voted for Independence. Here’s Who Else Has a Say. The New York Times, sec. Middle East. Available at <https://www.nytimes.com/2017/09/29/world/middleeast/kurds-independence- referendum-explainer.html>. Accessed 18 October 2017. Grant, Thomas D. 1999. The recognition of states: law and practice in debate and evolution. Westport, Conn.: Praeger. Green, Leonard, and Joel Myerson. 2004. A Discounting Framework for Choice With Delayed and Probabilistic Rewards. Psychological Bulletin 130 (5): 769–792. Grenadan Government. 2010. Talking diplomatic ties and cooperation with Morocco. Available at 104 <http://www.gov.gd/egov/news/2010/aug10/11_08_10/item_1/diplomatic_ties_morocco. html>. Accessed 28 April 2014. Grieco, Joseph M. 1990. Cooperation Among Nations: Europe, America, and Non-Tariff Barriers to Trade. Ithaca, NY: Cornell University Press. Griffiths, Ryan D. 2016. Age of Secession: The International and Domestic Determinants of State Birth. Cambridge [U.K.] ; New York, N.Y.: Cambridge University Press. Griffiths, Ryan D., and Charles R. Butcher. 2013. Introducing the International System(s) Dataset (ISD), 1816–2011. International Interactions 39 (5): 748–768. Guguen, Christophe. 2013. L’Etat marocain construit un hôtel de luxe aux Caraïbes. Lakome. Dominique. Available at <https://fr.lakome.info/index.php/maroc/362-l-etat-marocain- construit-un-hotel-de-luxe-aux-caraibes>. Accessed 28 April 3014. Hainmueller, Jens, Dominik Hangartner, and Teppei Yamamoto. 2015. Validating vignette and conjoint survey experiments against real-world behavior. Proceedings of the National Academy of Sciences of the United States of America 112 (8): 2395–2400. Hainmueller, Jens, and Michael J. Hiscox. 2010. Attitudes toward Highly Skilled and Low- skilled Immigration: Evidence from a Survey Experiment. The American Political Science Review 104 (1): 61–84. Hale, Henry E. 2008. The foundations of ethnic politics: separatism of states and nations in Eurasia and the world. (Cambridge studies in comparative politics). Cambridge; New York: Cambridge University Press. Harbom, Lotta, Erik Melander, and Peter Wallensteen. 2008. Dyadic Dimensions of Armed Conflict, 1946-2007. Journal of Peace Research 45 (5): 697–710. Harris, D. J. 1979. Cases and materials on international law. 2nd ed. London: Sweet & Maxwell. Harvey, James, and Gareth Stansfield. 2011. Theorizing unrecognized states: sovereignty, secessionism, and political economy. In Unrecognized states in the international system, edited by Nina Caspersen and Gareth R. V. Stansfield. (Exeter studies in ethno politics). Milton Park, Abingdon, Oxon; New York: Routledge. Haselswerdt, Jake, and Brandon L. Bartels. 2015. Public Opinion, Policy Tools, and the Status Quo: Evidence from a Survey Experiment. Political Research Quarterly 68 (3): 607–621. Hechter, Michael, and Elizabeth Borland. 2001. National Self-Determination: The Emergence of an International Norm. In Social Norms, 186-233]. New York, NY, USA: Russell Sage Foundation. Heraclides, Alexis. 1990. Secessionist Minorities and External Involvement. International Organization 44 (3): 341–378. Heraclides, Alexis. 1991. The self-determination of minorities in international politics. Totowa, N.J.: FCass. Huddleston, R. Joseph. 2017. Spain’s Flawed Response to the Catalonian Referendum. Foreign Affairs. Available at <https://www.foreignaffairs.com/articles/spain/2017-10-03/spains- flawed-response-catalonian-referendum>. Accessed 11 March 2018. Huddleston, R. Joseph, and Nicholas Weller. 2017. Unintended Causal Pathways: Probing Experimental Mechanisms Through Mediation Analysis. SSRN Scholarly Paper. Rochester, NY: Social Science Research Network. Available at <https://papers.ssrn.com/abstract=2964336>. Accessed 6 May 2017. 105 Irish, John, and Michelle Nichols. 2013. U.S., Russia agree on Syria U.N. chemical arms measure. Reuters. Available at <http://www.reuters.com/article/2013/09/26/us-un- assembly-syria-resolution-idUSBRE98P1AJ20130926>. Isani, Mujtaba, and Daniel Silverman. 2016. Foreign Policy Attitudes toward Islamic Actors: An Experimental Approach. Political Research Quarterly 69 (3): 571–582. Jackson. 2009. 09Rabat195: Morocco’s Tally of Countries Rejecting/Supporting Recognition of Independent Sahara. US State Dept. Diplomatic Cable. Rabat, Morocco: US State Department. Available at <http://www.cablegatesearch.net/cable.php?id=09RABAT195>. Accessed 2 May 2014. Jackson, Robert H. 1992. Juridical Statehood in Sub-Saharan Africa. Journal of International Affairs 46 (1): 1–16. Jackson, Robert H. 1990. Quasi-States: Sovereignty, International Relations and the Third World. New York: Cambridge University Press. Jackson, Robert H., and Carl G. Rosberg. 1986. Sovereignty and Underdevelopment: Juridical Statehood in the African Crisis. The Journal of Modern African Studies 24 (1): 1–31. Jacoby, William G. 2000. Issue Framing and Public Opinion on Government Spending. American Journal of Political Science 44 (4): 750–767. James, Patrick, and Michael Lusztig. 2001. Assessing the reliability of Predictions on the Future of Quebec. In The International Politics of Quebec Secession: State Making and State Breaking in North America, edited by David Carment, John F. Stack, and Frank P. Harvey. Westport, Conn: Praeger. Jenne, Erin K., Stephen M. Saideman, and Will Lowe. 2007. Separatism as a Bargaining Posture: The Role of Leverage in Minority Radicalization. Journal of Peace Research 44 (5): 539–558. Jensen, Erik. 2012. Western Sahara: anatomy of a stalemate? 2nd ed. Boulder, Colo.: Lynne Rienner Publishers. Kant, Garth, and Chelsea Schilling. 2013. Calls to Congress 499 to 1 against Syria war. World Net Daily. Kaplan. 2010. 10Rabat126: Commandant of the Marine Corps Focuses on Exercises and Peacekeeping During Morocco Visit. US State Dept. Diplomatic Cable. Rabat, Morocco: US State Department. Available at <http://www.cablegatesearch.net/cable.php?id=10RABAT126>. Accessed 2 May 2014. Karol, David, and Edward Miguel. 2007. The Electoral Cost of War: Iraq Casualties and the 2004 U.S. Presidential Election. The Journal of Politics 69 (3): 633–648. Keating, Joshua. 2017. Learning to Live With a Changing World Map. The New York Times, sec. Opinion. Available at <https://www.nytimes.com/2017/09/22/opinion/sunday/kurdistan- catalonia-independence.html>. Accessed 12 March 2018. Keohane, Robert O. 1986. Reciprocity in International Relations. International Organization 40 (1): 1–27. Ker-Lindsay, James. 2014. The Foreign Policy of Counter Secession: Preventing the Recognition of Contested States. Oxford: OUP Oxford. Kernell, Samuel. 1978. Explaining Presidential Popularity. How Ad Hoc Theorizing, Misplaced Emphasis, and Insufficient Care in Measuring One’s Variables Refuted Common Sense and Led Conventional Wisdom Down the Path of Anomalies. The American Political Science Review 72 (2): 506–522. 106 Kertzer, Joshua D., and Ryan Brutger. 2016. Decomposing Audience Costs: Bringing the Audience Back into Audience Cost Theory. American Journal of Political Science 60 (1): 234–249. Kisangani, Emizet F., and Jeffrey Pickering. 2015. International Military Intervention, 1989- 2005. University of Michigan, Ann Arbor. Kolstø, Pål. 2006. The Sustainability and Future of Unrecognized Quasi-States. Journal of Peace Research 43 (6): 723–740. Krasner, Stephen D. 1999a. Sovereignty: organized hypocrisy. (Princeton paperbacks). Princeton, N.J.: Princeton University Press. KRG. 2012. KRG and USAID embark on joint $18 million dollar health initiative. Kurdish Regional Government. Available at <http://cabinet.gov.krd/a/d.aspx?s=040000&l=12&a=45402>. Accessed 9 September 2017. Kriner, Douglas L. 2006. Examining Variance in Presidential Approval: The Case of FDR in World War II. The Public Opinion Quarterly 70 (1): 23–47. Kriner, Douglas L., and Francis X. Shen. 2012. How Citizens Respond to Combat Casualties The Differential Impact of Local Casualties on Support for the War in Afghanistan. Public Opinion Quarterly 76 (4): 761–770. Kriner, Douglas, Breanna Lechase, and Rosella Cappella Zielinski. 2015. Self-interest, partisanship, and the conditional influence of taxation on support for war in the USA. Conflict Management and Peace Science: 0738894215611133. Langer, Gary. 2013. With Poor Ratings on Handling Syria, Obama’s Approval Worst in Over a Year. ABC News. Available at <http://abcnews.go.com/blogs/politics/2013/09/with-poor- ratings-on-handling-syria-obamas-approval-worst-in-over-a-year/ http://www.langerresearch.com/uploads/1151a2SyriaandPolitics.pdf>. Lauterpacht, Hersch. 2012. Recognition in International Law. Cambridge University Press. Lergetporer, Philipp, Guido Schwerdt, Katharina Werner, and Ludger Woessmann. 2016. Information and Preferences for Public Spending: Evidence from Representative Survey Experiments. SSRN Scholarly Paper. Rochester, NY: Social Science Research Network. Available at <https://papers.ssrn.com/abstract=2805851>. Accessed 7 November 2016. Levendusky, Matthew S., and Michael C. Horowitz. 2012. When Backing Down Is the Right Decisions: Partisanship, New Information, and Audience Costs. The Journal of Politics 74 (2): 323–338. Levin, Irwin P., Sandra L. Schneider, and Gary J. Gaeth. 1998. All Frames Are Not Created Equal: A Typology and Critical Analysis of Framing Effects. Organizational Behavior and Human Decision Processes 76 (2): 149–188. Levy, Jack S., Michael K. McKoy, Paul Poast, and Geoffrey P.R. Wallace. 2015. Backing Out or Backing In? Commitment and Consistency in Audience Costs Theory. American Journal of Political Science 59 (4): 988–1001. Libal, Michael. 1997. Limits of Persuasion: Germany and the Yugoslav Crisis, 1991-1992. Westport, Conn.: Praeger. Liberman, Nira, and Yaacov Trope. 1998. The role of feasibility and desirability considerations in near and distant future decisions: A test of temporal construal theory. Journal of Personality and Social Psychology 75 (1): 5–18. 107 Logiuratto, Brett. 2013. Obama Is Going On A Furious Media Push To Sell A Syria Strike. Business Insider. Available at <http://www.businessinsider.com/syria-obama-tv- interviews-fox-news-cnn-nbc-pbs-abc-2013-9>. Mandel, Robert. 2013. Global Security Upheaval : Armed Nonstate Groups Usurping State Stability Functions. Palo Alto, CA, USA: Stanford Security Studies. Available at <http://site.ebrary.com/lib/alltitles/docDetail.action?docID=10685395>. Accessed 28 February 2015. McCleary, Paul. 2017. With Referendum Approaching, Kurds Wait for More U.S. Military Aid. Foreign Policy. Available at <https://foreignpolicy.com/2017/08/28/with-referendum- approaching-kurds-wait-for-more-u-s-military-aid/>. Accessed 9 September 2017. McNair, Arnold, Basdevant, Hackworth, Zoricic, Klaestad, Badawi, Read, et al. 1952. Case Concerning Rights of Nationals of the United States of America in Morocco. Millard. 2009. 09Casablanca10: Morocco, India, Phostphates and Beyond. US State Dept. Diplomatic Cable. Casablanca, Morocco: US State Department. Available at <http://www.cablegatesearch.net/cable.php?id=09CASABLANCA10>. Accessed 2 May 2014. Montevideo Convention, i. 1933. Montevideo Convention on the Rights and Duties of States. Available at <http://www.cfr.org/sovereignty/montevideo-convention-rights-duties- states/p15897>. Accessed 28 April 2014. Moore, Margaret. 2010. Ethics of Secession and Political Mobilization in Quebec. In Secession as an international phenomenon: from America’s Civil War to contemporary separatist movements, edited by Don Harrison Doyle. Athens: University of Georgia Press. Morgan, T. Clifton, Navin Bapat, and Yoshiharu Kobayashi. 2014. Threat and imposition of economic sanctions 1945–2005: Updating the TIES dataset. Conflict Management and Peace Science 31 (5): 541–558. Mueller, John E. 1973. War, presidents, and public opinion. New York: Wiley. Murphy, James G., Rudy E. Vuchinich, and Cathy A. Simpson. 2001. Delayed reward and cost discounting. The Psychological Record; Heidelberg 51 (4): 571–588. New York Times. 2013. American Views on Intervention in Syria. New York Times/CBS. Available at <http://www.nytimes.com/interactive/2013/09/10/world/middleeast/american-views-on- intervention-in-syria.html>. Accessed 25 October 2016. Noack, Rick. 2017. Europe has plenty of secessionist movements like Catalonia. Most don’t want full independence, though. Washington Post, sec. WorldViews Analysis Analysis Interpretation of the news based on evidence, including data, as well as anticipating how events might unfold based on past events. Available at <https://www.washingtonpost.com/news/worldviews/wp/2017/10/11/europe-has-many- independence-movements-apart-from-catalonia-few-of-them-want-full-self- determination/>. Accessed 12 March 2018. Obama, Barack. 2013a. FULL TRANSCRIPT: President Obama’s Sept. 10 speech on Syria. Washington Post. Available at <http://articles.washingtonpost.com/2013-09- 10/politics/41939044_1_chemical-weapons-poison-gas-sarin-gas/3>. Obama, Barack. 2012. Remarks by the President to the White House Press Corps. Available at <http://www.whitehouse.gov/the-press-office/2012/08/20/remarks-president-white- house-press-corps>. 108 Obama, Barack. 2013b. Statement by the President on Syria | The White House. Available at <http://www.whitehouse.gov/the-press-office/2013/08/31/statement-president-syria>. Accessed 21 March 2014. Oda, Shi, Vereshchetin, Lauterpacht, Shahabuddeen, Weeramantry, Parra-Aranguren, and Kreca. 1996. Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v. Serbia and Montenegro). O’Donoghue, Ted, and Matthew Rabin. 1999. Doing It Now or Later. The American Economic Review 89 (1): 103–124. Odum, Amy L. 2011. Delay Discounting: I’m a k, You’re a k. Journal of the Experimental Analysis of Behavior 96 (3): 427–439. Ostrom, Charles W., Jr., and Dennis M. Simon. 1985. Promise and Performance: A Dynamic Model of Presidential Popularity. The American Political Science Review 79 (2): 334– 358. Page, Susan, and John Kelly. 2013. Uphill battle for Obama: Where Congress stands on Syria. USA Today. Available at <http://www.usatoday.com/story/news/politics/2013/09/08/usa- today-network-survey-of-congress-on-syria-as-obama-lobbies-for-support/2781739/>. Paquin, Jonathan. 2010. A Stability-Seeking Power: U.S. Foreign Policy and Secessionist Conflicts. Montréal Que.: Mcgill Queens Univ Pr. Payne, John W., James R. Bettman, and Eric J. Johnson. 1992. Behavioral Decision Research: A Constructive Processing Perspective. Annual Review of Psychology 43 (1): 87–131. Pearson, Frederic S., and Robert A. Baumann. 1993. International Military Intervention, 1946- 1988. 1993rd-05–13th ed. (ICPSR 6035); ICPSR (Series) 6035. Ann Arbor, Mich. Inter- university Consortium for Political and Social Research [distributor] 1993. Pew-Research. 2013. Modest Support for Military Force if Syria Used Chemical Weapons. Available at <http://www.people-press.org/2013/04/29/modest-support-for-military- force-if-syria-used-chemical-weapons/>. Philpott, Daniel. 2001. Revolutions in Sovereignty: How Ideas Shaped Modern International Relations. Princeton, NJ: Princeton University Press. Philpott, Daniel. 1998. Self-Determination in Practice. In National Self-Determination and Secession, edited by Margaret Moore. Oxford; New York: Oxford University Press. Popkin, Samuel L. 1994. The Reasoning Voter: Communication and Persuasion in Presidential Campaigns. 1 edition. Chicago: University Of Chicago Press. Potter, Philip B. K., and Matthew A. Baum. 2010. Democratic Peace, Domestic Audience Costs, and Political Communication. Political Communication 27 (4): 453–470. Potter, Philip B. K., and Matthew A. Baum. 2014. Looking for Audience Costs in all the Wrong Places: Electoral Institutions, Media Access, and Democratic Constraint. The Journal of Politics 76 (01): 167–181. Putin, Vladamir. 2013. A Plea for Caution From Russia. New York Times. Available at <http://www.nytimes.com/2013/09/12/opinion/putin-plea-for-caution-from-russia-on- syria.html>. Quek, Kai. 2016. Nuclear Proliferation and the Use of Nuclear Options: Experimental Tests. Political Research Quarterly 69 (2): 195–206. Ramos, Jennifer. 2013. Changing norms through actions: the evolution of sovereignty. New York: Oxford University Press. Ranneberger. 2007. 07Nairobi1221: Kenya’s Foreign Policy Priorities, Diplomatic Presence And Expansion Plans: Promoting Economic Interests. US State Dept. Diplomatic Cable. 109 Nairobi, Kenya: US State Department. Available at <http://www.cablegatesearch.net/cable.php?id=07NAIROBI1221>. Accessed 2 May 2014. Reuning, Kevin, Michael R. Kenwick, and Christopher J. Fariss. 2016. Exploring the Dynamics of Latent Variable Models. SSRN Scholarly Paper. Rochester, NY: Social Science Research Network. Available at <https://papers.ssrn.com/abstract=2828703>. Accessed 3 August 2017. Reuters. 2013a. Arab League blames Syria’s Assad for chemical attack. Reuters. Available at <http://www.reuters.com/article/2013/08/27/us-syria-crisis-league- idUSBRE97Q0NI20130827>. Reuters. 2013b. U.S. credibility on Iran at stake in Syria decision: Hagel says. Reuters. Available at <http://www.reuters.com/article/2013/09/03/us-syria-crisis-usa-hagel- idUSBRE9820YY20130903>. Rich, Roland. 1993. Recognition of States: The Collapse of Yugoslavia and the Soviet Union Symposium: Recent Developments in the Practice of State Recognition. European Journal of International Law 4: 36–65. Roeder, Philip G. 2007. Where nation-states come from: institutional change in the age of nationalism. Princeton: Princeton University Press. Rolin-Jaequemyns, Rostworowski, Fromageot, Anzilotti, Adatci, Schücking, Negulesco, Van Eysinga, and Wang. 1934. Lighthouses Case between France and Greece, France v. Greece, Judgment. Russett, Bruce, and John R. O’neal. 2001. Triangulating Peace: Democracy, Interdependence, and International Organizations. New York: W. W. Norton. Saideman, Stephen. 2002. Discrimination in International Relations: Analyzing External Support for Ethnic Groups. Journal of Peace Research 39 (1): 27–50. Saideman, Stephen M. 2001. The Ties That Divide: Ethnic Politics, Foreign Policy, and International Conflict. Columbia University Press. Salehyan, Idean. 2009. Rebels without borders: transnational insurgencies in world politics / Idean Salehyan. Ithaca: Cornell University Press. Salehyan, Idean, Kristian Skrede Gleditsch, and David E. Cunningham. 2011. Explaining External Support for Insurgent Groups. International Organization 65 (04): 709–744. Sarkees, Meredith Reid. 2010. Resort to war: a data guide to inter-state, extra-state, intra-state, and non-state wars, 1816-2007 / Meredith Reid Sarkees and Frank Whelon Wayman. Correlates of war series. Washington, D.C.: CQ Press. Sawyer, Katherine, Kathleen Gallagher Cunningham, and William Reed. 2017. The Role of External Support in Civil War Termination. Journal of Conflict Resolution 61 (6): 1174– 1202. Schelling, Thomas C. 1960. The Strategy of Conflict. Cambridge, MA, USA: Harvard University Press. Schnakenberg, Keith E., and Christopher J. Fariss. 2014. Dynamic Patterns of Human Rights Practices. Political Science Research and Methods 2 (01): 1–31. Schultz, Kenneth A. 2001. Looking For Audience Costs. The Journal of Conflict Resolution 45 (1): 32–60. Sears, David O., Leonie Huddy, and Robert Jervis. 2003. Oxford Handbook of political psychology. New York: Oxford University Press. 110 Shafir, Eldar, Itamar Simonson, and Amos Tversky. 1993. Reason-based choice. Cognition 49 (1–2): 11–36. Shalal-Esa, Andrea. 2013. Sixth U.S. ship now in eastern Mediterranean ‘as precaution’. Reuters. Available at <http://www.reuters.com/article/2013/08/31/syria-crisis-ships- idUSL2N0GW01620130831>. van Sickle, Alix, and Wayne Sandholtz. 2008. The Emerging Right to Democracy. In International Norms and Cycles of Change, 1–26. New York, NY, USA: Oxford University Press. Simmons, Beth A. 2005. Rules over Real Estate: Trade, Territorial Conflict, and International Borders as Institution. The Journal of Conflict Resolution 49 (6): 823–848. Singh, Michael. 2012. A red line Iran would take seriously. Washington Post. Available at <http://articles.washingtonpost.com/2012-10-05/opinions/35500925_1_red-line-nuclear- weapon-iran>. SIPRI. 2016. SIPRI Arms Transfers Database. Stockholm International Pease Research Institute. Available at <https://www.sipri.org/databases/armstransfers>. Siroky, David S. 2009. Secession and survival: Nations, states and violent conflict. Ph.D., United States -- North Carolina: Duke University. Available at <http://search.proquest.com.libproxy.usc.edu/docview/304883720/abstract?accountid=14 749>. Accessed 1 March 2015. Slantchev, Branislav L. 2006. Politicians, the Media, and Domestic Audience Costs. International Studies Quarterly 50 (2): 445–477. Snyder, Jack, and Erica D. Borghard. 2011. The Cost of Empty Threats: A Penny, Not a Pound. American Political Science Review 105 (3): 437–456. SPSRASD. 2011. Saharawi Republic formally recognizes Republic of South Sudan. Sahara Press Service. Bir-Lahlou, Western Sahara. Sterio, Milena. 2013. The right to self-determination under international law: ‘selfistans’, secession and the rule of the great powers. Milton Park, Abingdon, Oxon ; New York: Routledge. Stone Sweet, Alec. 1999. Judicialization and the Construction of Governance. Comparative Political Studies 32 (2): 147–184. Sullivan, Patricia L., and Michael T. Koch. 2009. Military Intervention by Powerful States, 1945—2003. Journal of Peace Research 46 (5): 707–718. Sundberg, Ralph, and Erik Melander. 2013. Introducing the UCDP Georeferenced Event Dataset. Journal of Peace Research 50 (4): 523–532. Thaler, Richard H., and H. M. Shefrin. 1981. An Economic Theory of Self-Control. Journal of Political Economy 89 (2): 392–406. The Economist. 2013. YouGov | Economist/YouGov Poll. Available at <https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/dtqk62b7do/econTo plines.pdf>. Accessed 9 June 2015. Tilly, Charles. 1990. Coercion, Capital, and European States, AD 990-1990. Cambridge, MA: Basil Blackwell. Toft, Monica Duffy. 2003. The Geography of Ethnic Violence: Identity, Interests, and the Indivisibility of Territory. Princeton, NJ: Princeton University Press. Tomson, Chris. 2017. US Army trains 8,500 Kurdish fighters for battle of Raqqa. AMN - Al- Masdar News | . Available at 111 <https://www.almasdarnews.com/article/us-army-trains-8500-kurdish-fighters-battle- raqqa/>. Accessed 9 September 2017. Tomz, Michael. 2007. Domestic Audience Costs in International Relations: An Experimental Approach. International Organization 61 (4): 821–840. Trager, Robert F., and Lynn Vavreck. 2011. The Political Costs of Crisis Bargaining: Presidential Rhetoric and the Role of Party. American Journal of Political Science 55 (3): 526–545. Trope, Yaacov. 2012. Construal Level Theory. In Handbook of theories of social psychology, edited by Paul A. M. Van Lange, Arie W. Kruglanski, and E. Tory Higgins, 118–134. Los Angeles: SAGE. Trope, Yaacov, and Nira Liberman. 2010. Construal-level theory of psychological distance. Psychological Review 117 (2): 440–463. Trope, Yaacov, and Nira Liberman. 2000. Temporal construal and time-dependent changes in preference. Journal of Personality and Social Psychology 79 (6): 876–889. Tversky, Amos, and Daniel Kahneman. 1981. The Framing of Decisions and the Psychology of Choice. Science 211 (4481). New Series: 453–458. UCDP. 2017. UCDP Conflict Encyclopedia (UCDP database) - Uppsala University, Sweden. Available at <http://www.pcr.uu.se/research/ucdp/program_overview/current_projects/ucdp_database/ >. Accessed 29 August 2017. UCDP/PRIO. 2008. UCDP/PRIO Armed Conflict Dataset Codebook, Version 4-2008. Uppsala Conflict Data Program/International Peace Research Institute. Available at <http://www.pcr.uu.se/research/ucdp/datasets/ucdp_prio_armed_conflict_dataset/>. UN. 1945. Charter of the United Nations. Available at <http://www.un.org/en/documents/charter/chapter1.shtml>. Accessed 5 May 2014. UN. 1960. General Assembly, 15th session : 864th plenary meeting (opening meeting), Tuesday, 20 September 1960, New York. Available at <http://dag.un.org/handle/11176/275209>. Accessed 17 March 2018. UN. 2013. United Nations Mission to Investigate Allegations of the Use of Chemical Weapons in the Syrian Arab Republic: Report on the Alleged Use of Chemical Weapons in the Ghouta Area of Damascus on 21 August 2013. The Hague, Netherlands: United Nations. Available at <http://www.un.org/disarmament/content/slideshow/Secretary_General_Report_of_CW_ Investigation.pdf>. Unterberger, Betty Miller. 1996. The United States and National Self-Determination: A Wilsonian Perspective. Presidential Studies Quarterly 26 (4): 926–941. USAID. 2017. Development Data Library (DDL) | U.S. Agency for International Development. Available at <https://www.usaid.gov/data>. Accessed 29 August 2017. Visalvanich, Neil. 2017. Asian Candidates in America: The Surprising Effects of Positive Racial Stereotyping. Political Research Quarterly 70 (1): 68–81. Vitali, Ali. 2013. 58% of Americans want Congress to vote No on Syria. MSNBC News. Available at <http://www.msnbc.com/morning-joe/58-americans-want-congress-vote- no>. Voeten, Erik. 2012. Data and Analyses of Voting in the UN General Assembly. SSRN Scholarly Paper. Rochester, NY: Social Science Research Network. Available at <http://papers.ssrn.com/abstract=2111149>. Accessed 22 August 2016. 112 Walsh, James Igoe. 2015. Precision Weapons, Civilian Casualties, and Support for the Use of Force. Political Psychology 36 (5): 507–523. Weber, Max. 1946. Politics as a Vocation. In From Max Weber: Essays in Sociology, edited by H. H. Gerth and C. Wright Mills, 77–79. New York, NY, USA: Oxford University Press. Wellman, Christopher. 2005. A theory of secession: the case for political self-determination. Cambridge; New York: Cambridge University Press. Wellman, Christopher. 2010. The Morality of Secession. In Secession as an international phenomenon: from America’s Civil War to contemporary separatist movements, edited by Don Harrison Doyle. Athens: University of Georgia Press. White House Archives, t. 2005. President Bush Meets with President Barzani of Kurdistan Regional Government of Iraq. White House Archives. Available at <https://georgewbush- whitehouse.archives.gov/news/releases/2005/10/20051025-7.html>. Accessed 9 September 2017. White-House. 2013. Government Assessment of the Syrian Government’s Use of Chemical Weapons on August 21, 2013. Available at <http://www.whitehouse.gov/the-press- office/2013/08/30/government-assessment-syrian-government-s-use-chemical-weapons- august-21>. White-House-Official. 2013. Background Conference Call by White House Official on Syria. Available at <http://www.whitehouse.gov/the-press-office/2013/04/25/background- conference-call-white-house-official-syria>. Worster, William Thomas. 2009. Law, Politics, and the Conception of the State in State Recognition Theory. Boston University International Law Journal 27 (1): 115–171. Zaller, John. 1999. Governing with the News: The News Media as a Political Institution by Timothy E. Cook. American Journal of Sociology 104 (6): 1819–1821. Zaller, John, and Stanley Feldman. 1992. A Simple Theory of the Survey Response: Answering Questions versus Revealing Preferences. American Journal of Political Science 36 (3): 579–616. Zucchino, David. 2017. Iraqi Forces Sweep Into Kirkuk, Checking Kurdish Independence Drive. The New York Times, sec. Middle East. Available at <https://www.nytimes.com/2017/10/16/world/middleeast/kirkuk-iraq-kurds.html>. Accessed 18 October 2017. Zunes, Stephen, and Jacob Mundy. 2006. Western Sahara: War, Nationalism, and Conflict Irresolution. Syracuse Studies on Peace and Conflict Resolution. Syracuse: Syracuse University Press. 2012. Post-ABC poll: U.S. involvement in Syria. Washington Post. Available at <http://www.washingtonpost.com/page/2010- 2019/WashingtonPost/2012/12/20/National-Politics/Polling/release_187.xml>. 2013. September 2013 Washington Post-ABC News poll - Syria. Washington Post. Available at <http://www.washingtonpost.com/page/2010- 2019/WashingtonPost/2013/09/17/National-Politics/Polling/release_261.xml>. 113 APPENDICES APPENDIX A. SUPPLEMENTARY MATERIAL FOR CHAPTER 3 Robust to removal of PREC Table A-1. Effects on latent recognition of violence, victory, domestic politics, and international factors with prior recognition dummy variable removed Model: Prior Recognition Dummy Removed Major Powers 1945-2015 Medium Powers 1945-2015 Pooled 1945-2015 Violence: 25-999 deaths (PRIO) .01 (.02) -.02 (.007) * -.006 (.01) Violence: >999 deaths (PRIO) .08 (.02) *** .03 (.009) *** .05 (.01) *** SD group victory .13 (.04) ** .08 (.02) *** .1 (.02) *** Number of Challengers .002 (.001) *** .001 (.000) *** .002 (.000) *** 3rd – Host MID .03 (.01) .01 (.01) .02 (.006) ** Mutual autocracy -.03 (.01) *** -.02 (.003) *** -.02 (.004) *** Mutual democracy -.02 (.01) ** -.01 (.003) ** -.01 (.003) *** Constant -.18 (.01) *** -.08 (.003) *** -.12 (.003) *** N 1947 2635 4582 p-values are * .05, ** .01, *** .001 (SEs) A few changes with lagged IVs 114 Table A-2. Effects on latent recognition of violence, victory, domestic politics, and international factors with all independent variables lagged on year Model: All Independent Variables Lagged One Year Major Powers 1945-2015 Medium Powers 1945-2015 Pooled 1945-2015 Violence: 25-999 deaths (PRIO) .01 (.02) -.01 (.008) a -.003 (.01) Violence: >999 deaths (PRIO) .06 (.02) *** .02 (.008) *** .04 (.01) *** SD group victory -.02 (.07) -.03 (.03) -.03 (.03) Number of Challengers .002 (.001) *** .001 (.000) *** .002 (.000) *** 3rd – Host MID .01 (.01) .01 (.006) * .02 (.007) Prior recognition by great power .07 (.04) * .08 (.02) *** .08 (.02) *** Mutual autocracy -.04 (.01) *** -.02 (.003) *** -.03 (.004) *** Mutual democracy -.02 (.01) ** -.004 (.003) -.01 (.003) ** Constant -.19 (.01) *** -.08 (.003) *** -.13 (.003) *** N 1643 2277 3920 p-values are * .05, ** .01, *** .001 (SEs) a: significant at p=.056 APPENDIX B. POLL DATA FOR CHAPTER 4 Poll Data for Figure 4-1. Intervention opinion polls bolded. ABC/Post. 10/1/2012, http://www.langerresearch.com/uploads/1141a1IntotheDebates.pdf ABC/Post. 12/18/2012, http://www.langerresearch.com/uploads/1145a2PoliticsandtheFiscalCliff.pdf ABC/Post. 12/20/2012, http://www.washingtonpost.com/page/2010- 2019/WashingtonPost/2012/12/20/National-Politics/Polling/release_187.xml ABC/Post. 9/16/2013, http://s3.documentcloud.org/documents/797514/2013-09-15-trend-for- release.pdf ABC/Post. 9/19/2013, http://www.washingtonpost.com/page/2010- 2019/WashingtonPost/2013/09/17/National-Politics/Polling/release_261.xml CBS. 12/19/2012, http://www.scribd.com/doc/117298635/CBS-News-Poll-Fiscal-Cliff CBS. 11/20/2013, http://www.scribd.com/doc/185695292/Nov13c-Morning CBS/Times. 9/14/2012, http://s3.documentcloud.org/documents/432864/latest-cbs-news-new- york-times-poll.pdf CBS/Times. 10/30/2012, http://big.assets.huffingtonpost.com/OctQuin.pdf CBS/Times. 1/22/2013, http://s3.documentcloud.org/documents/557713/the-new-york-times- cbs-news-poll.pdf CBS/Times. 5/2/2013, http://s3.documentcloud.org/documents/695362/april13-nytimes-cbs- poll.pdf 115 CBS/Times. 9/26/2013, http://www.nytimes.com/interactive/2013/09/26/us/politics/26poll- results.html CNN. 4/8/2013, http://i2.cdn.turner.com/cnn/2013/images/04/08/rel4a.pdf CNN. 5/19/2013, http://i2.cdn.turner.com/cnn/2013/images/05/28/syriapoll.pdf CNN. 6/17/2013, http://i2.cdn.turner.com/cnn/2013/images/06/17/rel7a.pdf CNN. 9/8/2013, http://i2.cdn.turner.com/cnn/2013/images/09/09/6a.poll.syria.pdf CNN. 9/9/2013, http://i2.cdn.turner.com/cnn/2013/images/09/09/rel8c.pdf FOX. 6/25/2013, http://www.foxnews.com/politics/interactive/2013/06/25/fox-news-poll-voters- weigh-in-on-obama-nsa-surveillance/ FOX. 9/9/2013, http://www.foxnews.com/politics/interactive/2013/09/09/fox-news-poll-voters- say-us-less-respected-since-obama-took-office/ FOX. 11/13/2013, http://www.foxnews.com/politics/interactive/2013/11/13/fox-news-polls-half- voters-think-president-lied-about-obamacare-majority-want/ Gallup. 9/10/2013, http://www.gallup.com/poll/164336/syria-yet-affect-americans-rate-obama- key-issues.aspx Marist/McClatchy. 1/11/2013. http://maristpoll.marist.edu/wp- content/misc/usapolls/US121204/Obama/Complete%20December%2012. %202012%20USA%20McClatchy-Marist%20Poll%20Release%20and%20Tables.pdf McClatchy/Marist. 4/4/2013. http://maristpoll.marist.edu/wp- content/misc/usapolls/us130325/Obama/Complete%20April%204. %202013%20USA%20McClatchy-Marist%20Poll%20Results%20and%20Tables.pdf McClatchy/Marist. 7/23/2013. "http://maristpoll.marist.edu/wp- content/misc/usapolls/us130715/Obama/Complete%20July%2023. %202013%20USA%20McClatchy-Marist%20Poll%20Results%20and%20Tables.pdf" McClatchy/Marist. 9/9/2013, http://s3.documentcloud.org/documents/784977/mcclatchy-marist- poll-syria.pdf NBC News. 8/30/2013, http://msnbcmedia.msn.com/i//MSNBC/Sections/A_Politics/_Today_Stories_Teases/13336_ NBC_Syria_Poll.pdf NBC/WSJ. 8/21/2012, http://msnbcmedia.msn.com/i/MSNBC/Sections/A_Politics/_Today_Stories_Teases/August_ NBC-WSJ_Int_Sched.pdf NBC/WSJ. 9/18/2012, http://msnbcmedia.msn.com/i/MSNBC/Sections/A_Politics/_Today_Stories_Teases/Septemb er_WSJ_NBC_Poll.pdf NBC/WSJ. 10/2/2012, http://msnbcmedia.msn.com/i/MSNBC/Sections/A_Politics/_Today_Stories_Teases/NBCNe ws-WSJPoll_9-12.pdf NBC/WSJ. 10/21/2012, http://msnbcmedia.msn.com/i/MSNBC/Sections/A_Politics/nbc_oct_poll.pdf NBC/WSJ. 12/13/2012, http://msnbcmedia.msn.com/i/MSNBC/Sections/A_Politics/_Today_Stories_Teases/NBCPO LL1212.pdf 116 NBC/WSJ. 4/11/2013, http://msnbcmedia.msn.com/i/MSNBC/Sections/A_Politics/13127%20APRIL%20NBC- WSJ%20%284-11%20Release%29.pdf NBC/WSJ. 7/24/2013, http://msnbcmedia.msn.com/i/MSNBC/Sections/A_Politics/_Today_Stories_Teases/130724- July-NBC-WSJ-poll.pdf Pew. 4/29/2013, http://www.people-press.org/2013/04/29/modest-support-for-military-force- if-syria-used-chemical-weapons/ Pew. 11/8/2013, http://www.people-press.org/files/legacy-questionnaires/11-08- 13%20Obama%20Topline%20for%20Release.pdf Pew/USA Today. 9/9/2013, http://www.people-press.org/files/legacy-questionnaires/9-9- 13%20Syria%20Topline%20for%20Release.pdf Politico/GWU/Battleground. 8/13/2012, http://images.politico.com/global/2012/08/politico_gw_bg_48_questionnaire.pdf Politico/GWU/Battleground. 9/24/2012, http://images.politico.com/global/2012/09/battlegroundpoll.html Quinnipiac. 5/1/2013, https://poll.qu.edu/national/release-detail?ReleaseID=1890 Quinnipiac. 7/11/2013, https://poll.qu.edu/national/release-detail?ReleaseID=1920 Quinnipiac. 8/2/2013, https://poll.qu.edu/national/release-detail?ReleaseID=1931 Quinnipiac. 10/1/2013, https://poll.qu.edu/national/release-detail?ReleaseID=1958 Quinnipiac. 11/12/2013, https://poll.qu.edu/national/release-detail?ReleaseID=1975 Reason/Rupe. 9/10/2013, http://reason.com/assets/db/13787963538251.pdf Reuters/Ipsos. 8/23/2013, http://www.ipsos-na.com/download/pr.aspx?id=12979 Reuters/Ipsos. 8/30/2013, http://www.ipsos-na.com/download/pr.aspx?id=12988 Reuters/Ipsos. 9/9/2013, http://www.ipsos-na.com/download/pr.aspx?id=13008 Reuters/Ipsos. 9/10/2013, http://www.ipsos-na.com/download/pr.aspx?id=13013 Time. 6/17/2013, http://swampland.time.com/2013/06/16/new-time-poll-americans-believe- country-heading-in-wrong-direction/ YouGov/Economist. 6/12/2013, http://cdn.yougov.com/cumulus_uploads/document/iqn03vbeyl/20130610econToplines.pdf YouGov/Economist. 6/19/2013, http://cdn.yougov.com/cumulus_uploads/document/etf70dw36d/20130615econToplines.pdf YouGov/Economist. 6/26/2013, http://cdn.yougov.com/cumulus_uploads/document/z9pr64pqs6/Economist%20Toplines%20 June%2024%202013.pdf YouGov/Economist. 7/3/2013, http://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/0rd3x7rgj2/20130701econ Toplines.pdf YouGov/Economist. 7/10/2013, http://cdn.yougov.com/cumulus_uploads/document/gstmy2b6e8/20130708econToplines.pdf YouGov/Economist. 9/3/2013, https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/dtqk62b7do/econTo plines.pdf 117 YouGov/Economist. 9/9/2013, http://cdn.yougov.com/cumulus_uploads/document/dtqk62b7do/econToplines.pdf YouGov/Economist. 9/11/2013, https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/aclx0wp1pc/econTo plines.pdf YouGov/Economist. 9/16/2013, https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/m5y7odcf5f/econTo plines.pdf YouGov/Economist. 9/25/2013, https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/3ccim7h98g/econTo plines.pdf YouGov/Economist. 10/2/2013, http://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/5q7kz4jllo/econToplines.p df Example questions on foreign policy from each pollster Polling operations tend to use the same question repeatedly across different polls. Below is a list of selected polling questions on Obama’s foreign policy.  ABC News/Washington Post http://s3.documentcloud.org/documents/797514/2013-09-15-trend-for-release.pdf o Do you approve or disapprove of the way Obama is handling [international affairs]? o Do you approve/disapprove strongly or somewhat?  CBS/Times http://s3.documentcloud.org/documents/432864/latest-cbs-news-new-york-times-poll.pdf o Do you approve or disapprove of the way Barack Obama is handling foreign policy?  CNN http://i2.cdn.turner.com/cnn/2013/images/04/08/rel4a.pdf o Do you approve or disapprove of the way Barack Obama is handling [foreign affairs]?  FOX News http://www.foxnews.com/politics/interactive/2013/06/25/fox-news-poll-voters-weigh-in- on-obama-nsa-surveillance/ o Do you approve or disapprove of the job Barack Obama is doing on the following issues? [foreign policy]  Gallup http://www.gallup.com/file/poll/164345/Obama_Issue_Approvals_130910.pdf o Do you approve or disapprove of the way Barack Obama is handling [foreign policy]? 118  McClatchy/Marist http://s3.documentcloud.org/documents/784977/mcclatchy-marist-poll-syria.pdf o Do you approve or disapprove of how President Barack Obama is handling foreign policy?  NBC/WSJ http://msnbcmedia.msn.com/i/MSNBC/Sections/A_Politics/_Today_Stories_Teases/Aug ust_NBC-WSJ_Int_Sched.pdf o Do you generally approve or disapprove of the job Barack Obama is doing in handling foreign policy?  Pew http://www.people-press.org/files/legacy-questionnaires/11-08- 13%20Obama%20Topline%20for%20Release.pdf o Do you approve or disapprove of the way Barack Obama is handling [foreign policy]?  Politico http://images.politico.com/global/2012/08/politico_gw_bg_48_questionnaire.pdf o Do you approve or disapprove of the job that President Obama is doing on [foreign policy]?  Quinnipiac https://poll.qu.edu/national/release-detail?ReleaseID=1975 o Do you approve or disapprove of the way President Obama is handling [foreign policy]?  YouGov/Economist http://cdn.yougov.com/cumulus_uploads/document/iqn03vbeyl/20130610econToplines.p df o Do you approve or disapprove of the way Barack Obama is handling foreign policy? 119 APPENDIX C. SUPPLEMENTARY MATERIAL FOR CHAPTER 4 Table C-1. Mean approval level for each cost comparison group (ordinal measure 1 – 7) Cost Category: Tomz Default Open Cost Prompt Vignette: Empty threat 3.11 3.51* Stay out 4.03 4.1 Follow through 4.94 4.66 Results of independent two group t-tests. Asterisks indicate significance level (*.05, **.01) of difference from the “Tomz Default” category. The ordinal measure ranges from 1 (strongly disapprove) to 7 (strongly approve). Table C-2. Mean approval level for each reputation comparison group (ordinal measure 1 – 7) Reputation Category: Tomz Default Open Reputation Prompt Vignette: Empty threat 3.11 2.98 Stay out 4.06 4.19 Results of independent two group t-tests. Asterisks indicate significance level (*.05, **.01) of difference from the “Tomz Default” category. The ordinal measure ranges from 1 (strongly disapprove) to 7 (strongly approve). 120 Table C-3. Ordered logit coefficients for three comparison groups Variable Empty Threat Stay Out Follow Through Costfirst .36* (.16) .05 (.16) -.32* (.16) Cutpoint 1 -.88 (.15) -1.5 (.14) -2.2 (.16) Cutpoint 2 .13 (.12) -.5 (.12) -1.45 (.14) Cutpoint 3 .53 (.12) -.22 (.12) -1.11 (.13) Cutpoint 4 .92 (.13) .09 (.12) -.8 (.13) Cutpoint 5 1.39 (.14) .48 (.12) -.34 (.12) Cutpoint 6 2.58 (.18) 1.58 (.14) 1.09 (.12) N 500 502 510 Note: Cell entries are ordered logit coefficients, with associated standard errors in parentheses. * p , .05, ** p , .01 121 Table C-4. Ordered logit coefficients for two comparison groups with all demographic controls Variable Empty Threat Follow Through Age 25-34 0.45 (0.27) -0.34 (0.28) 35-44 0.31 (0.29) 0.06 (0.31) 45-54 -0.35 (0.35) -0.25 (0.37) 55-64 -0.27 (0.45) 0.77 (0.41) >64 -1.14 (0.75) 0.16 (0.63) Female 0.44* (0.18) -0.17 (0.17) Party strong dem -0.1 (0.29) 0.16 (0.28) dem -0.16 (0.28) 0.37 (0.28) lean dem -0.1 (0.33) 0.45 (0.3) lean repub -0.89 (0.39) 0.82 (0.39) repub -1.46* (0.35) 1.47* (0.34) strong repub -2.01* (0.45) 1.68* (0.43) Education HSgrad -0.35 (1.28) some college 0.12 (1.27) -0.08 (0.31) associate -0.05 (1.28) -0.01 (0.34) bachelors -0.4 (1.27) -0.06 (0.29) phd 1.46 (1.57) 0.37 (0.71) grad/prof degree -0.61 (1.31) 0.15 (0.38) Race black 0.51 (0.31) 0.38 (0.32) native-amer -0.05 (0.84) -1.68 (0.94) asian 0.15 (0.36) 0.45 (0.31) other -0.72 (0.71) -1.82 (1.18) Costfirst 0.41* (0.17) -0.32* (0.17) Cutpoint 1 -1.26 (1.31) -2.1 (0.42) Cutpoint 2 -0.05 (1.3) -1.3 (0.41) Cutpoint 3 0.39 (1.3) -0.91 (0.4) Cutpoint 4 0.83 (1.31) -0.57 (0.4) Cutpoint 5 1.33 (1.31) -0.07 (0.4) Cutpoint 6 2.62 (1.31) 1.52 (0.41) Note: Cell entries are ordered logit coefficients, with associated standard errors in parentheses. * p , .05, ** p , .01 122 Table C-5. Tests of the parallel regression assumption: Empty Threat Test χ 2 Degrees of Freedom p > χ 2 Wolfe Gould 6.495 5 .261 Brant 6.524 5 .258 score 6.797 5 .236 likelihood ratio 6.485 5 .262 Wald 6.524 5 .258 Table C-6. Tests of the parallel regression assumption: Follow Through Test χ 2 Degrees of Freedom p > χ 2 Wolfe Gould 5.498 5 .358 Brant 5.254 5 .386 score 5.417 5 .367 likelihood ratio 5.231 5 .388 Wald 5.254 5 .386 Note: These tests are designed to test the assumption that the relationship between each pair of outcome groups is the same. Thus, the desired statistic for these tests would be insignificant, failing to reject the null hypothesis of a difference between groups. 123 Table C-7. Demographics for each round Total 10/15 12/16 3/17 Party strong dem 22.46% 22.79% 21.39% 25.21% dem 21.93% 23.13% 20.66% 21.85% lean dem 14.21% 14.36% 14.38% 12.89% ind/no pref 13.71% 14.90% 13.37% 10.08% lean repub 7.41% 6.54% 8.24% 7.84% repub 12.31% 11.40% 13.01% 13.45% strong repub 7.97% 6.88% 8.96% 8.68% Race white 82.40% 83.90% 81.51% 79.67% black 7.53% 6.96% 7.97% 8.24% native-amer 0.61% 0.66% 0.43% 1.10% asian 7.20% 6.83% 7.18% 8.79% pacific islander 0.09% 0.07% 0.07% 0.27% other 2.17% 1.59% 2.84% 1.92% latino (any) 6.60% 5.98% 7.40% 6.06% Sex male 53.88% 57.85% 50.00% 52.34% female 45.82% 41.75% 49.79% 47.38% other 0.31% 0.40% 0.21% 0.28% Marital Status single 49.41% 50.03% 47.94% 52.47% married 40.68% 40.89% 41.39% 37.09% separated 1.37% 1.52% 1.35% 0.82% widowed 1.25% 1.13% 1.28% 1.65% divorced 7.29% 6.43% 8.04% 7.97% Age 18-24 12.24% 11.81% 12.54% 12.91% 25-34 43.33% 44.66% 41.67% 44.23% 35-44 23.73% 24.29% 23.79% 21.15% 45-54 11.57% 10.68% 11.97% 13.74% 55-64 6.93% 6.50% 7.62% 6.04% >64 2.20% 2.06% 2.42% 1.92% Education <HS 0.31% 27.00% 0.29% 0.55% HSgrad 10.90% 12.83% 9.27% 9.32% some college 25.24% 24.58% 26.16% 24.38% associate 11.91% 12.96% 11.33% 9.86% bachelors 39.75% 39.75% 39.77% 39.73% phd 1.26% 1.20% 1.28% 1.37% grad/prof degree 10.63% 8.42% 11.90% 14.79% Income <25k 18.25% 18.32% 17.98% 19.01% 25k-35k 15.48% 16.14% 14.36% 17.08% 35k-50k 18.53% 19.25% 18.69% 14.88% 50k-75k 22.30% 22.35% 22.74% 20.39% 75k-100k 13.74% 12.96% 14.71% 13.22% 100k-150k 9.14% 8.60% 9.03% 11.85% >150k 2.56% 2.38% 2.49% 3.58% 124 Table C-8. N for each comparison group across rounds Cost Category: Tomz Default Open Cost Prompt Open Reputation Prompt 10/1 5 12/1 6 3/17 10/15 12/16 3/17 10/15 12/16 3/17 Vignette: Empty Threat 256 219 244 257 198 Stay Out 239 205 263 244 200 Follow Through 247 263 October 2015 subjects saw only cost, not reputation. December 2016 subjects saw both cost and reputation prompts together in random order. March 2017 subjects saw only the reputation prompt. 125 APPENDIX D. SUPPLEMENTARY ANALYSIS FOR CHAPTER 4 Here, I consider the conditional average treatment effect by partisan identification. The primary tests in the paper came from the October 2015 experimental round, so I consider only that round. This also avoids the potential confounding problems associated with the newly inaugurated president in January 2017. I use the initial party ID question, which gave three values: Republican, Democrat, and Independent. Although a follow-up question was asked of Independents to ascertain whether they “lean” towards one party or another, I include them in the same category for this test (the results are consistent, as can be seen in the Stata replication file). Table D-1 shows the variation in the effects of treatment by party and vignette. Because the sample is now split 18 ways, most of the standard errors are too large to maintain statistical significance in the reported effects, but effect directions are consistent those reported in the main body of the text. Respondents in the Empty Threat groups show an increase in approval when primed to consider casualties, regardless of partisan affiliation. Respondents in the Follow Through groups show a decrease in approval with treatment, regardless of partisan affiliation. It is also interesting to note that Republicans and Independents cancel each other out in the Stay Out group: Republicans respond to the cost prime with a 1.2 point decrease in approval, and Independents respond with a 1 point increase in approval. Table D-1. Separate Cost tests by partisanship Tomz Prompt Open Cost Prompt Effect of Treatment p N Republicans Empty Threat 2.07 2.5 .42 .12 109 Stay Out 4.02 2.78 -1.24** .01 96 Follow Through 5.78 5.14 -.64* .05 95 Democrats Empty Threat 3.5 3.79 .29 .13 226 Stay Out 4.47 4.4 -.07 .6 224 Follow Through 4.62 4.57 -.05 .6 231 Independents (including “lean”) Empty Threat 3.19 3.88 .7* .02 165 Stay Out 3.58 4.51 .93** .00 182 Follow Through 4.84 4.57 -.27 .19 184 Total: 1512 Note: * p , .05, ** p , .01 126 As an additional check, table D-2 shows the base ordered logit model with control for party ID (with party=“lean neither way” as the reference category). Though Republicans are more bellicose overall, the estimates for the effect of the cost prime are consistent with those reported in Table C-4 and displayed in Figure 4-3 in the main text. Table D-2. Ordered logit coefficients for two comparison groups with party control Variable Empty Threat Follow Through Party strong dem -0.05 (0.29) 0.04 (0.26) dem -0.12 (0.27) 0.29 (0.26) lean dem -0.06 (0.32) 0.34 (0.29) lean repub -1.08** (0.38) 0.68 (0.38) repub -1.45** (0.33) 1.33** (0.33) strong repub -2.28** (0.43) 1.26** (0.41) Costfirst 0.41** (0.16) -0.31* (0.16) Cutpoint 1 -1.40 (0.25) -1.92 (0.25) Cutpoint 2 -0.26 (0.25) -1.13 (0.24) Cutpoint 3 0.17 (0.25) -0.78 (0.23) Cutpoint 4 0.60 (0.25) -0.44 (0.23) Cutpoint 5 1.08 (0.25) 0.02 (0.23) Cutpoint 6 2.36 (0.28) 1.53 (0.24) N 489 497 Note: Cell entries are ordered logit coefficients, with associated standard errors in parentheses. * p , .05, ** p , .01 127 APPENDIX E. EXPERIMENTAL PROMPT FOR CHAPTER 4 Respondents in all three treatment groups read the following: Initial Prompt This survey will ask you some questions about US relations with other countries around the world. It will take 5-8 minutes to complete. You will read about a situation our country has faced many times in the past and will probably face again. Different leaders have handled the situation in different ways. We will describe one approach US leaders have taken, ask you a few questions to make sure you understood the situation, and then ask your opinion of the leader's actions. Then the respondents are randomly placed in one of the following three treatment groups. Vignette 1: Follow Through with Intervention A country sent its military to take over a neighboring country. The attacking country was led by a dictator, who invaded to get more power and resources. The attacking country had a strong military, so it would take a major effort for the United States to help push them out. A victory by the attacking country would hurt the safety and economy of the United States. The US president said that if the attack continued, the US military would push out the invaders. He sent troops to the region and prepared them for war. The attacking country continued to invade. The president then ordered US troops to destroy one of the invader’s military bases. US troops destroyed the base, but the invasion still continued. So the president sent in more US military forces, which fought the aggressive government until it halted its invasion. Vignette 2: Threaten and Back Down A country sent its military to take over a neighboring country. The attacking country was led by a dictator, who invaded to get more power and resources. The attacking country had a strong military, so it would take a major effort for the United States to help push them out. A victory by the attacking country would hurt the safety and economy of the United States. The US president said that if the attack continued, the US military would push out the invaders. The attacking country continued to invade. In the end, the US president did not send troops, and the attacking country took over its neighbor. 128 Vignette 3: No Threat or Intervention A country sent its military to take over a neighboring country. The attacking country was led by a dictator, who invaded to get more power and resources. The attacking country had a strong military, so it would take a major effort for the United States to help push them out. A victory by the attacking country would hurt the safety and economy of the United States. The US president said the United States would stay out of the conflict. The attacking country continued to invade. In the end, the US president did not send troops, and the attacking country took over its neighbor. The respondents then take an attention check. They are allowed to go back and reread the scenario if they need to. If they fail to answer any of these questions correctly, they are excused from the experiment and not allowed to complete it. Attention Check In order to make sure you understood the situation, we will ask you three questions about the scenario. If you answer any of them incorrectly, the survey will end and you will not be allowed to complete the survey. You may go back to reread the scenario if necessary. What kind of government does the country have?  Communist  Dictator  Democratic Government  I don't know What does the country aim to do?  Incite civil war in a neighboring country  Take over a neighboring country  Buy nuclear weapons from a neighboring country  I don't know What did the US president do in the end?  Sent troops to stop the country  Stayed out of the conflict  Enacted economic sanctions  I don't know The respondents are then asked the following two sets of questions to establish the primary treatment and control groups of concern for this study. 129 Questions After the Vignette Casualties Prime Question (randomly assigned to be before or after approval question) How many US troops would you expect were [to be] killed when [if] the president [had] decided to stop the invasion with military force?  Fewer than 50  50-99  100-499  500-1999  2000-4999  5000-10000  More than 10000 Reputation Prime Questions What effect, if any, would the President’s actions in the prior scenario have on the international reputation of the US?  It would harm the US' reputation.  It would not affect the US' reputation.  It would improve the US' reputation. If the subject chooses harm:  Would it harm the US' reputation a lot or only a little? o It would harm the US' reputation a lot. o It would only harm the US' reputation a little. If the subject chooses improve:  Would it harm the US' reputation a lot or only a little? o It would improve the US' reputation a lot. o It would only improve the US' reputation a little. Self-Report of Presidential Approval Do you approve, disapprove, or neither approve nor disapprove of the way the US president handled the situation?  Approve  Neither Approve nor Disapprove  Disapprove If subject chooses Approve:  Do you approve very strongly, or only somewhat? o Very strongly 130 o Somewhat If subject chooses Disapprove:  Do you disapprove very strongly, or only somewhat? o Very Strongly o Somewhat If subject chooses: Neither Approve nor Disapprove o Do you lean toward approving of the way the US president handled the situation, lean toward disapproving, or don't you lean either way? o Lean toward Disapproving o Lean toward approving o Don’t lean either way Then a number of demographic questions are asked, they are thanked, and their participation is complete. Demographic Questions Generally speaking, do you think of yourself as a...  Republican (1)  Democrat (2)  Independent (3)  No preference (4)  Other - please specify (5) ____________________ If Republican is Selected  Would you call yourself...  A strong republican (1)  Not a very strong republican (2) If Democrat is Selected  Would you call yourself...  A strong democrat (1)  Not a very strong democrat (2) If no preference is Selected  Do you think of yourself as closer to the...  Republican party (1)  Democratic party (2)  Neither (3) 131 If Independent is Selected  As an independent, do you think of yourself as closer to the...  Republican party (1)  Democratic party (2)  Neither (3) Please tell us how much you agree or disagree with this statement: The use of military force only makes problems worse.  Agree strongly (1)  Agree somewhat (2)  Neither agree nor disagree (3)  Disagree somewhat (4)  Disagree strongly (5) Please tell us how much you agree or disagree with this statement: The United States needs to play an active role in solving conflicts around the world.  Agree strongly (1)  Agree somewhat (2)  Neither agree nor disagree (3)  Disagree somewhat (4)  Disagree strongly (5) Did you vote in the 2012 election?  Yes (1)  No (2)  I'm not sure or I don't remember (3) Which region of the country do you live in?  Midwest - IA, IL, IN, KS, MI, MN, MO, ND, NE, OH, SD, WI (1)  Northeast - CT, DC, DE, MA, MD, ME, NH, NJ, NY, PA, RI, VT (2)  Southeast - AL, AR, FL, GA, KY, LA, MS, NC, SC, TN, VA, WV (3)  Southwest - AZ, NM, OK, TX (4)  West - AK, CA, CO, HI, ID, MT, NV, OR, UT, WA, WY (5) 132 What was your total household income before taxes during the past 12 months?  Less than $24,999 (1)  $25,000 to $34,999 (2)  $35,000 to $49,999 (3)  $50,000 to $74,999 (4)  $75,000 to $99,999 (5)  $100,000 to $149,999 (6)  $150,000 or more (7) Are you of Hispanic, Latino, or Spanish origin?  Yes (1)  No (2) What is your race? For purposes of this question, persons of Spanish/Hispanic/Latino origin may be of any race.  White (1)  Black or African American (2)  American Indian and Alaska Native (3)  Asian (4)  Native Hawaiian and Other Pacific Islander (5)  Other (6) ____________________ What is your age?  18 to 24 years (1)  25 to 34 years (2)  35 to 44 years (3)  45 to 54 years (4)  55 to 64 years (5)  Age 65 or older (6) What is the highest degree or level of education you have completed?  Less than high school (1)  High school graduate (includes equivalency) (2)  Some college, no degree (3)  Associate's degree (4)  Bachelor's degree (5)  Ph.D. (6)  Graduate or professional degree (7) 133 What is your marital status?  Single (never married) (1)  Married (2)  Separated (3)  Widowed (4)  Divorced (5) What is your gender?  Male (1)  Female (2)  Other (3) ____________________ Is there anything else you would like us to know?  Yes (1) ____________________  No (2)

Abstract (if available)

Abstract This dissertation examines foreign policy towards states facing active separatist or self-determination movements. It develops a new measure of international recognition, shows how foreign policy decisions affect the international sovereignty of separatists, and focuses on the impact of normative pressures on state relations with countries embroiled in conflict.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

Information, public opinion, and international relations

PDF

Like father, like son? A succession-based explanation for conflict initiation by authoritarian regimes

PDF

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

PDF

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

PDF

Perspectives on state capacity and the political geography of conflict

PDF

Anti-foreign boycotts as a tool of economic coercion: the case of China

PDF

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

PDF

Private interests in American government institutions

PDF

Circuit breakers: how policy entrepreneurs interrupted the electric flow with Peru’s first renewable energy legislation for the grid

PDF

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

PDF

Securitizing the democratic peace: democratic identity and its role in the construction of threat

PDF

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

PDF

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

PDF

A two-level analysis of foreign policy decision making: an empirical investigation of the case of China-Taiwan

PDF

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

PDF

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

PDF

The politics of transformation: mass media and the Northern Ireland peace process

PDF

The paradox of personalism: internal political constraints on nuclear posture in authoritarian regimes

PDF

Things fall apart: the unraveling of international institutions through withdrawal

Asset Metadata

Creator Huddleston, Robert Joseph (author)

Core Title Legitimizing self-determination: advancing the sovereignty of separatist movements

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Political Science and International Relations

Publication Date 07/19/2018

Defense Date 04/18/2018

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

Tag Bayesian analysis,Foreign policy,foreign policy opinion,hierarchical modeling,latent variable analysis,military intervention,OAI-PMH Harvest,secession,self-determination,separatism,Sovereignty,survey experiments,United Nations,Western Sahara

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor James, Patrick (committee chair), Cull, Nicholas (committee member), Sandholtz, Wayne (committee member), Weller, Nicholas (committee member)

Creator Email joeyhuddleston@gmail.com,rhuddles@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-19366

Unique identifier UC11672271

Identifier etd-Huddleston-6423.pdf (filename),usctheses-c89-19366 (legacy record id)

Legacy Identifier etd-Huddleston-6423.pdf

Dmrecord 19366

Document Type Dissertation

Format application/pdf (imt)

Rights Huddleston, Robert Joseph

Type texts

Source 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 a...

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