Reputations and Change in International Relations

Reputations for resolve are critical in international relations for deterring adversaries and reassuring partners. However, a state’s resolve is unobservable and can change unbeknownst to its audience. How does the possibility of unobserved change impact reputation dynamics? I provide a theory of long-run reputations with changing resolve via a formal model covering conflict and cooperation domains. In the model, the possibility that current reputations are based on outdated information makes the audience extend the benefit of the doubt to states with poor reputations. This leads to states building or spending their reputations depending on their current reputations. Importantly, when damaged reputations can be rebuilt, states with better reputations face stronger temptations to spend them. Thus, reputations constrain states with poor reputations the most. Further, because demonstrations of resolve improve reputations, which, in turn, reduce incentives for future demonstrations of resolve, there is a cyclical rhythm to conflict and cooperation. A major implication is that a state’s behavior changes with its reputation even if its resolve is unchanged and the stakes are identical. Reputational enforcement works, but the price is occasional breaches of trust. These results also settle a few long-standing controversies in the IR-reputation literature.