Dynamic Screening in International Crises

In this paper, I challenge costly signaling theory and present a rival theoretical framework for understanding how states behave and assess resolve in crises. Modeling crisis diplomacy as a war of attrition, I ask how long countries will be willing to participate in costly negotiations and invest in sunk costs and audience costs before going to war or conceding. I show that more resolved states display greater impatience with diplomacy, preferring to fight instead. In turn, the least resolved states prefer to concede quickly to avoid having to fight. Finally, moderately resolved states negotiate longer, pay more sunk costs, and accumulate more audience costs. Consequently, moderately resolved states are more likely to obtain concessions, not because belief in their resolve increases, but because they grant their rival more time to concede. The model also features stalemated negotiations, providing new microfoundations for a common crisis outcome.