Reputation building in hostage taking incidents

We test whether governments attempt to build reputations as not negotiating with terrorists. The test is indirect, relying on the greater incentive of governments to build reputations than exists for non-governments. We also test whether terrorists make errors in their appraisal of the state of the world in hostage taking incidents. We employ Tobit and Heckman's selection model in testing, and find strong evidence that governments do attempt to build reputations. We also find that terrorists do not make statistically large errors in their appraisal of the state of the world, lending support to the use of equilibrium models in the study of terrorism.