My primary research project examines why leaders fail to reach peaceful agreements when conflicts, particularly interstate wars, are very costly. Using a rationalist approach with statistics and formal modeling, my work advocates greater focus on the negotiation process rather than the costs and benefits of conflict, and specific consideration of dyadic dependencies in conflict data. I examine a theory of military resource division using statistics on international conflicts, using both logistic regression and temporal exponential random graph models (TERGMs). Leaders do in fact consider other wars, and the dynamic nature of the network effects reveals cultural and generational differences. The study is the first to identify specific forms of dyadic dependence for international wars. And it offers an easy modification to existing methods to account for network dependencies. I use event history statistics to predict that democratic improvement after a successful foreign imposed regime change (FIRC) will not last as long as those arising internally. I model conflict negotiation to show that bargaining itself can prevent conflict resolution.

My secondary research project uses selectorate-style leadership models to examine the effect of regime type, particularly winning coalition size, on the provision of various goods. This research investigates how polity type affects particular policies, and therefore how policies can be improved. I model economic barriers to entry as a combined public and private good. When a leader has a relatively large coalition, his supporters benefit less from these barriers. The leader will choose to offer more effective rewards while lowering barriers to increase government revenue. I model human rights restrictions to explain why genocide is more common in dictatorships. In moderate sized coalitions, human rights restrictions can be a cost-effective way to pay off coalition members. Small coalitions prefer fully private goods, while public goods are more efficient for large coalitions. I prove the existence of a crime poverty trap and treat law enforcement as a public good. The low crime equilibrium under the rule of law usually only occurs in strong democracies. If members of the winning coalition are themselves criminals, the state can become a haven for criminal activity, such as when pirates controlled Jamaica.

In both research projects, I ask how best to engineer government to create optimal group decisions. My work on leadership explains for domestic political institutions affect local and international policies. My research on conflict considers why leaders choose to abandon bargaining in favor of costly military coercion. Both projects seek to understand impedances to effective group decisions. In the end, My work is about understanding these power dynamics in order to improve society.

“Military Resource Division and its Effects on International Conflict”
Do leaders make war decisions individually or do they consider other ongoing conflicts? History resoundingly says other conflicts matter, yet large scale statistics of conflict typically assume dyadic independence. This assumption has been shown to be clearly wrong (Poast 2010; Erikson, Pinto, and Rader 2014; Minhas, Hoff, and Ward 2016) for a number of uses. However, no theory directly addresses what form dyadic dependencies will take. I present a theory based on military resource division and contrast it with dyadic independence theory. I investigate these theories using logistic regression and network analysis.

Conflict and International Relations
“War and Negotiation in Two Dimensions”
War has been visualized as a single-dimension constant-sum game. An additional dimension captures the possibility of mutual gains and a more complex negotiation process. I present a new tool for visualizing rational conflict, and show how mutual gains may make peace easier or more difficult to achieve. This method intuitively depicts utility for each player with two relevant sets: the settlements that are possible to achieve under a legitimate division of prizes, and the agreements that would be preferable to war for both players. Offers can be shown as points, with the order of offers illustrated as lines between them. This technique can readily accommodate multiple theories of conflict simultaneously. I show examples of information asymmetry, first strike advantage, indivisible goods, commitment problems, the democratic peace, and sunk costs. War may be caused by a failure of the negotiation process, even when a mutually beneficial settlement exists, because neither player wants to offer too much to the other player. While most of the rationalist explanations focus on limitations to the possible and acceptable sets, existing explanations have proven difficult for various reasons. I suggest more work should focus on the negotiation process as a neglected area of study.

Leadership and Selectorate Models