UC San Diego

This dissertation introduces an agent based model driven by the behavioral assumptions of the bargaining theory of war. The model is applied to explain why wars cluster geographically and why democracies do not fight other democracies. Simulation results suggest new explanations for both of these phenomena. The emergence of regionally clustering conflict can be explained by the tendency of shifting power to motivate renegotiation when agents pay costs for projecting power and select their bargaining partners. Regions of democratic peace occur when groups of agents share information more effectively than their competitors. The dissertation develops and validates these theories with statistical analysis of simulation results and case studies