Political Science Speaker Series: Michael Joseph

Can states use unattributable attacks to coerce others? Since Schelling (1960), the conventional wisdom among scholars of reputation, resolve, grey zone conflict, and covert action is no. They reason that Targets must know the logic behind attacks for coercive success, and attackers must develop a reputation as resolved. Using formal analysis and intuitive argumentation, I argue that unattributable coercion is rationalizable. Targets can learn from a history of harm, that harm will come to them in the future, without learning who chose to harm them. Attackers can coerce a Target without cultivating a reputation as resolved. Unattributable coercion is easiest when the harm inflicted is large, the risk of an accident is small, and—most critically—there is more than one potential attacker. I argue that the US faces these conditions in modern times, and illustrate the mechanism with a case analysis of still-unattributed Havana Syndrome incidents. Anecdotes of military occupations, insurgent attacks, and Arab-Western relations at the origins of the Cold War illustrate the general theory's portability to other substantive domains.