Certainty and War The Harvard community has made this article openly available. Please share how this access benefits you. Your story matters Citation Schub, Robert Jay. 2016. Certainty and War. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences. Citable link http://nrs.harvard.edu/urn-3:HUL.InstRepos:33493541 Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available under the terms and conditions applicable to Other Posted Material, as set forth at http:// nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of- use#LAA Certainty and War A dissertation presented by Robert Jay Schub to The Department of Government in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the subject of Political Science Harvard University Cambridge, Massachusetts April 2016 ©2016 — Robert Jay Schub All rights reserved. Dissertation Advisor: Jeffry Frieden Robert Jay Schub Certainty and War Abstract Does greater certainty about an adversary’s attributes cause peace? What states believe they can secure through force dictates the diplomatic settlements they will accept. In prevailing accounts which preclude assessment errors, certainty promotes peace as states can readily identify agreements preferable to war. Yet, empirically, high-certainty assessments often contribute to bargaining failure, rather than success. This dissertation resolves the tension. Assessments are not objectively given; leaders must form them through subjective processes. Consistent with behavioral studies, leaders are often more certain than available information warrants. Incorporating these overprecision errors, I show certainty can increase the risk of war. Hence, the relationship between certainty and war is conditional. Whether estimates are overprecise depends on the information leaders receive from ad- visers who have specialized expertise due to a division of labor. Failure to tap into this ex- pertise generates overprecise estimates. This is particularly likely when leaders fail to gather information pertinent to an adversary’s political (versus military) attributes by marginaliz- ing a state’s diplomats—such as US State Department officials. Bureaucracies affect state behavior through the substantive expertise they provide, not through parochial preference divergences which dissipate during crises. To test the argument I construct a measure of certainty using an original corpus of declassified security documents from US Cold War crises. Quantitative tests using the measure demonstrate that State Department officials provide assessments with less certainty than counterparts and the relationship between certainty and conflict is conditional on the State Department’s role. When State Department officials are heavily involved, certainty iii leads to peace; when marginalized, certainty is likely due to overprecision and leads to war. Case studies of the Bay of Pigs and Iraq War assess implications that elude quanti- tative testing. Presidents marginalized diplomats, privileging CIA estimates in 1961 and Pentagon estimates in 2003. Each agency offered high-certainty estimates over political at- tributes affecting conflict outcomes: popular uprisings in Cuba and stability in post-Saddam Iraq. Overprecision is not a matter of hindsight as marginalized advisers invoked greater uncertainty before hostilities commenced. Integrating behavioralist and rationalist approaches offers greater explanatory power in quantitative tests and provides insights into historical cases that are puzzling for extant theories. Moreover, the dissertation shows that certainty is not strictly welfare enhancing and flags policy conditions conducive to assessment errors and costly foreign policy blunders. iv | Contents Abstract iii Acknowledgments vii 1 Bargaining with Certainty 1 1.1 Fallibility in International Politics . .1 1.2 The Argument . .4 1.3 Why Certainty? . .9 1.4 Plan of the Dissertation . 11 2 Theorizing Certainty, Leaders, and War 17 2.1 Limited Information and Subjective Assessment . 18 2.2 War and Certainty without Error . 22 2.3 War and Certainty with Error . 24 2.4 Causes of Overprecision . 30 2.5 Conclusion and Roadmap for Testing . 43 3 Quantitative Evidence for Overprecision in Bargaining 45 3.1 Testing the Argument . 45 3.2 Observations, Documents, and Measurement . 45 3.3 Results . 56 3.4 Robustness and Alternative Explanations . 63 3.5 Conclusion . 70 4 Bay of Pigs: Prospects for Protests 72 4.1 Castro’s Consolidation and Bargaining Failure . 72 4.2 Communist Influence, Expropriation, and Sugar . 77 4.3 Overprecision at CIA and the Oval Office . 87 4.4 Why Diplomats were Marginalized . 105 4.5 Alternative Explanations . 110 4.6 Conclusion . 117 5 Iraq: Prewar Assessments of Postwar Stability 120 5.1 Post-Saddam Stability and Bargaining Failure . 120 5.2 Regional Sway and Authoritarianism . 125 5.3 Overprecision at the Pentagon and Oval Office . 129 v 5.4 Why Diplomats were Marginalized . 156 5.5 Alternative Explanations . 158 5.6 Conclusion . 164 6 Conclusion 166 6.1 Theoretical Implications . 169 6.2 Policy Implications . 173 A Appendix to Chapter 2 176 B Appendix to Chapter 3 179 Bibliography 192 vi | Acknowledgments I am grateful to many for helping make this research possible. Jeff Frieden was the ideal advisor, offering his time, encouragement, and necessary correctives throughout the research process. I am indebted to his theoretical clarity and his patience with my deficits on this front. I thank Jeff for emphasizing the importance of passion above other considerations in selecting a research topic that demands years of effort. I was fortunate to arrive at Harvard the same year as Dustin Tingley and to benefit from his insights from day one. His wide-ranging expertise, theoretical and empirical rigor, commitment to teaching, and work ethic make him a true role model. This project benefitted from these wide-ranging interests, drawing on his knowledge of bargaining models, bureaucratic interests, political psychology, and text-as-data. Muhammet Bas encouraged and improved my forays into studying certainty and war in the classroom, as an advisor, and as a collaborator. I am incredibly lucky that as a co-author he has shown me the requirements for and difficulty of making a meaningful contribution to our field’s collective knowledge. Iain Johnston’s creativity, open-mindedness, and deep historical knowledge were all pivotal for improving this project. My efforts to take the real world and the process of crafting foreign policy seriously are thanks to him. I am incredibly grateful to my “diss group” peers Julie Faller, Noah Nathan, and Ariel White. I was very fortunate to be surrounded by such outstanding scholars. Their friend- ship, patience, and enthusiasm ensured grad school was frequently fun, for lack of a better word. John Marshall straightened out my thinking on information processing and my errant forehand. George Yin encouraged continual refinement of my theory and always stoked my curiosity. For fruitful discussions and feedback, I thank Erin Baggott, Tyson Belanger, Kara Ross Camarena, Andrew Coe, Jeff Friedman, Michael Gill, Nils Hagerdal, Connor Huff, Josh Kertzer, Andy Kydd, Rakeen Mabud, Rich Nielsen, Steve Rosen, Anne Sartori, Beth Sim- mons, Alex Weisiger, and Yuri Zhukov as well as audiences at Harvard, Michigan, Nebraska, USC, Vanderbilt, ISA, MPSA, APSA, and PSS. This work benefitted from a productive year spent at the Harvard Kennedy School’s Belfer Center. I won the lottery with my parents. I thank them (and my brother) for actively encourag- ing me in unspoken but unsubtle ways to exit a career path with a relatively clear trajectory but that sparked limited excitement in favor a path with the inverse conditions. Finally, I want to return to and thank Julie again. Trivially, I get the benefit of living with a better political scientist without the cost of an academic two-body problem. More importantly, thank you for sharing your life with me and letting me share mine with you. vii 1| Bargaining with Certainty 1.1 Fallibility in International Politics The “power to hurt” through military force shapes international politics by dictating which peaceful settlements are more attractive than war (Schelling 1966, p. 2). However, it is difficult for leaders to discern precisely how much “hurt” states can impose or tolerate and thus how unattractive military conflict is. Information limitations and the estimative challenges they generate permeate international politics. President Kennedy authorized an amphibious assault against Cuba in 1961 believing that anti-Castro popular uprisings would sprout in sympathy with the US-trained Brigade’s landing at the Bay of Pigs. Kennedy’s beliefs about the likelihood and scope of uprisings were pivotal for assessing the extent of “hurt” imposed on the Cuban regime: “ultimate success will depend upon political factors; i.e., a sizable popular uprising,” according to the Joint Chiefs of Staff.1 CIA officials Allen Dulles and Richard Bissell shaped Kennedy’s beliefs, offering high-certainty estimates that a robust opposition lay in waiting with as much as 25% of the Cuban public ready to actively support anti-regime forces when provided a catalyst.2 Reality shocked the President. Contrary to prevailing beliefs, there were no uprisings and few defections from Castro’s militia. Castro’s internal security forces, far more capable than Dulles and Bissell anticipated, arrested up to 200,000 potential dissidents in the 48 hours before the Brigade’s landing. Cuban