Learning How Hard to Fight

• In Event: Formal Models of Conflict

Abstract

According to canonical models of bargaining while fighting, battlefield losses should cause a state to decrease its estimate of the chance of victory, leading that state to scale back its demands at the bargaining table. But in historical and contemporary conflicts, we instead see states respond to unexpected setbacks by fighting harder --- i.e., by committing additional resources to the effort. I study this phenomenon by developing a formal model of bargaining while fighting with endogenous military effort. In the model, information from battlefield events may affect both the decision to concede and the amount of effort a state devotes to the war. Studying the equilibrium dynamics of the model, I find that it is possible to observe "long wars" in which a state increases its efforts over time. This type of equilibrium path is most likely to occur (1) when there is high prior uncertainty, (2) when the military effect of battlefield effort is high relative to the marginal cost, and (3) when players are patient. These comparative statics help discern my theory from accounts of sunk cost fallacies or gambling for resurrection, which might also explain increases in effort following battlefield losses.

Author

• Brenton Kenkel, Vanderbilt University