| Over the last twenty years, scholars of all persuasions have become increasingly convinced of the need to incorporate the role of perceptions and beliefs into our theories of alliance formation. We have come to understand military alliances as signaling devices that states employ to enhance the credibility of their promises and threats and thereby deter potential enemies from initiating disputes. The three theoretical papers that compose this dissertation extend the logic of the signaling model to address overlooked issues of alliance design.; The European alliance system prior to the First World War has been characterized as "rigid" and "inflexible" because the great powers formed alliances in anticipation of possible future disputes rather than in reaction to disputes as they emerged. In "Standing and Reactive Alliances," I argue that the timing of alliance formation has important implications for crisis bargaining behavior that rational states will take into account when forming alliances. I contrast two simple game-theoretic models showing that states should favor "standing" alliances when deterrence is their highest priority, but that "reactive" alliances allow states to avoid the costs of forming alliances that may prove to be unnecessary.; If we understand alliances as signaling devices, the prevalence of secret defensive alliances in nineteenth-century Europe seems rather puzzling. In "Silent Partners," I develop a model in which states form secret defensive alliances in order to prevent their enemies from determining their optimal means of attack. Somewhat surprisingly, I identify conditions under which the potential for secret alliance formation deters attackers who could not be deterred if all alliances were publicly observable.; Finally, in "Entrapment in Alliance Politics," I argue draw on insights from game theory to show that the risk of entrapment---being dragged into war involuntarily by one's ally---is not inherent in alliance formation and that it is not directly related to the risk of abandonment. States can be entrapped only when they rely upon commitment mechanisms rather than costly signals to establish the credibility of their alliance promises. States design their alliance treaties to avoid the risk of entrapment when they mistrust their partners' incentives. |
