Shifting Power, Commitment Problems, and Preventive War

Abstract

Many scholars argue that leaders' expectations about future shifts in the distribution of power can result in preventive war. If a leader expects her adversary to be significantly stronger in the future, the leader may choose to go to war with that adversary rather than bargain with a stronger adversary in the future. However, quantitative evaluations of this argument prove difficult, as they require a measure of leaders' expectations about future shifts in power rather than simply a measure of observed power. In this paper, we develop an empirical model of future power to create this measure. We then use that measure to evaluate the preventive war hypothesis. Our results support the preventive war hypothesis. Increases in a state's expected future probability of winning in war increase its probability of another state in a dyadic relationship initiating war against it.