SSP Wednesday Seminar with speaker Danielle Gilbert, the David and Cindy Edelson Fellow in U.S. Foreign Policy and International Security at Dartmouth College. The talk will be broadcast live on the MIT Security Studies Program Youtube channel. 

Kidnappings of soldiers, journalists, aid workers, and other civilians by armed groups happen every day. Yet, the politics of hostage recovery remains relatively understudied by international relations scholars. Whether and how governments choose to recover their citizens varies widely, as does public sentiment about bringing hostages home. To explain this variation, we develop a theory of hostage deservingness. We argue that deservingness is determined by the circumstance of capture—particularly whether hostages are perceived to be to blame for their capture—and that more deserving hostages receive more public support for risky rescue operations. Our multimethod approach to hypothesis testing includes the use of survey experiments embedded in two different nationally representative surveys of Americans, a behavioral experiment fielded on Facebook using Meta’s a/b testing platform, and 20 elite interviews with current and former senior hostage recovery personnel. Across our experiments, we find that when hostages are described as not to blame for their capture, support for recovery is at its highest. However, when capture occurs under circumstances that suggest the hostage bears responsibility, support for recovery decreases. Despite official U.S. government policy that the circumstances of capture are irrelevant to recovery decisions, our interviews show that notions of deservingness (and the public’s approval of hostages) matter a great deal in the hostage recovery process. We demonstrate that policymakers are similarly susceptible to notions of deservingness, yielding substantial internal debate about recovery options. Moreover, policymakers adapt their messaging to avoid backlash from recovering less sympathetic hostages or failing to recover those seen as more deserving.