The fragility of deterrence in conflicts

We study deterrence in sequential move conflicts, modeled as a contest. We bias the model in favor of peace by assuming that under complete information deterrence is achieved and peace prevails. We show that under incomplete information about states' types (resolve) the chances of deterrence decrease rapidly. Studying a uniform type distribution, we show that the finer the type space becomes the more resolve a defending state must have to support deterrence in equilibrium. In the limit, as types occur on a continuum, deterrence is possible only under relatively extreme conditions.