Nuclear Russian Roulette: A Model of Proliferation and Preventive War
Andrew Coe develops a formal model of bargaining between two states, where one can invest in developing nuclear weapons and the other imperfectly observes its efforts and progress over time, and uses it to analyze the occurrence of proliferation and war and the role of intelligence-gathering and estimates in these. Surprise proliferation, sporadic crises over the uncertain progress of a proliferant's efforts, and "mistaken" preventive wars can all arise endogenously in the model. The costs of war, effects of proliferation, and technological sophistication of the proliferant influence the probabilities of war and proliferation in ways that are often counter-intuitive and non-monotonic. However, much of the variation in behavior is driven, not by these potentially policy-manipulable factors, but by chance elements such as when the proliferant's program will make progress and when the other state will discover this. In the absence of a non-proliferation agreement, the United States and proliferants like Iran are playing what amounts to a long game of Russian roulette.
Social Sciences Building (SSB), Room 107.