I use formal theory and empirical methods to research international politics, conflict, and political economy. My focus is on applying game theory and structural estimation methods to study models of war. My work features dynamic models, models of strategic information transmission, and games on networks.
Conflict Management and Peace Science, 2021
The preventive motive for war arises because states cannot commit to limit the use of their growing power. This commitment problem can lead to war when there are not enough resources available today to compensate the declining state for their expected losses. In this article, we show how capital markets affect preventive war incentives by introducing a profit-maximizing bond market to the canonical bargaining model of war. We find that the nature of the power shift and fundamentals of the market for debt interact to determine when a preventive motive is more likely to lead to war. Two main results show that (1) less probable but more extreme power shifts are most dangerous and (2) unlike the direct effect of interest rates on the cost of war, higher interest on sovereign debt makes war more likely. We present evidence for the latter effect by extending Lemke's (2003) study of preventive war for major-power dyads between 1816 and 1992.
Maintaining peace is costly. To understand what this implies about war outbreak and frequency, this article provides a dynamic crisis bargaining model where peace is costly and countries can take diplomatic action to compete over the bargaining surplus. Against conventional wisdom, repeated interactions of patient countries can destabilize peace. The likelihood of war relies on fundamentals of the international order, such as the persistence of war outcomes and peace agreements, as well as the severity of competition in diplomatic affairs. Even when countries prefer to cooperate, inadvertent wars are inevitable in the long run due to a coordination problem induced by costly peace. Diplomatic competition reduces the gains from peace and affects the probability of inadvertent war, but does not directly instigate attacks. The model offers new explanations for war, highlights the importance of institutional design in averting conflict, and sheds light on which international orders are most likely to fare well over time.
Social structure is an important determinant of political outcomes, yet analyses of vote buying do not account for interdependence between actors. We provide a formal model in which candidates can improve their electoral performance by providing private transfers to policy-motivated voters connected on a social network at the expense of a public good. Advances in spectral graph theory allow us to analytically derive comparative statics explicitly in terms of deep parameters governing social structure. Contrary to much existing theory, equilibrium transfers are not determined by network density, but primarily by group fractionalization and homophily, and are driven by a disproportionate targeting of minorities. In addition, we extend the model to account for heterogeneous information structure, demonstrating density still does not affect candidate strategies on average.