When do commitment problems not cause war? Turkey and Cyprus, 1964 versus 1974

War is often attributed to the fact that parties are unable to commit to an agreement, or have incentives to renege on it. This cause of war is known as “commitment problems†. Yet commitment problems are not in themselves sufficient for war. Parties to a conflict can experience commitment problems and not go to war. Although current theories suggest conditions under which commitment problems should be more or less likely to cause war, these conditions are rarely empirically specified. The paper rectifies this important omission by empirically evaluating hypothesized scope conditions of commitment problems-driven war. Specifically, I analyze the effects of the extent to which power was redistributed prior to war; the speed with which power was redistributed; and the current and anticipated future relative capabilities of the parties. I compare a positive and a negative case of homeland military intervention in support of co-ethnics abroad, using evidence on homeland leaders’ perceptions, reasoning and behavior. The evidence supports the hypotheses. Commitment problems cause military intervention when leaders anticipate shifts in the distribution of power that are large and rapid, and when their current military capabilities favor intervention. Commitment problems do not cause intervention when leaders anticipate shifts in the distribution of power that are small and incremental, and when their current military capabilities are poor.