I make two related claims: (1) assessments of stability made by political actors and analysts are largely hit or miss; and (2) that leader responses to fear of fragility or confidence in robustness are unpredictable in their consequences. Leader assessments are often made with respect to historical lessons derived from dramatic past events that appear relevant to the present. These lessons may or may not be based on good history and may or may not be relevant to the case at hand. Leaders and elites who believe their orders to be robust can help make their beliefs self-fulfilling. However, overconfidence can help make these orders fragile. I argue that leader and elite assessments of robustness and fragility are influenced by cognitive biases and also often highly motivated. Leaders and their advisors use information selectively and can confirm tautologically the lessons they apply.

We review our theoretical claims in light of the empirical chapters and their evidence that leader assessments matter, are highly subjective, and very much influenced by ideology and role models. They are also influenced by leader estimates of what needs to be done and their political freedom to act. This is in turn shows variation across leaders. The most common response to fragility is denial, although some leaders convince themselves – usually unrealistically – they can enact far-reaching reforms to address it.