Abstract
Countries involved in conflicts learn about their military strength from the battlefield. We study how a third party intervenes to manipulate this learning. An attacker and a defender engage in a conflict whose outcome conveys information about the attacker’s strength. A third party worries that the attacker becomes more confident about its military strength and can intervene to help the defender. This intervention is risky: if the attacker wins despite the help the defender receives, its confidence increases even further. We show that optimal third-party intervention is non-monotonic in the attacker’s strength. We also show that a high level of patriotism and resolve to defend itself improve the defender’s odds in the conflict by inducing third-party intervention.