War is in the Variance/Covariance Structure of the Error Terms: International Networks and the Bargaining Theory of War

In collaboration with: Erik Gartzke and Hugh Ward

Why do states fight if conflict is costly and ineffcient? Bargaining theories of war emphasize the role of asymmetric information as a fundamental cause of conflict. Leaders uncertain about one another’s military strength, preferences, or resolve can err in calculating the expected costs and outcomes of fighting which can lead to war. A critical step in the onset of conflict is thus determining what leaders are likely to know, and where they get their information from. Who one knows may help to shape what one knows; we explore the role of networks in shaping conflict dynamics in situations where fighting results from uncertainty. If networks (e.g. diplomatic exchanges, trade) mitigate uncertainty and private information, then states’ network positions should be an important predictor of the onset of international conflict. We develop a theoretical model that links the variance/co-variance structure of errors to conflict onset and derive hypotheses about how the network positions of individual states and dyads affects uncertainty and conflict onset. We test these hypotheses against data on militarized interstate dispute onsets between 1950 and 2000 using network and conventional econometric models.