Reputation for retreat: Casualty intolerance and the declining credibility of democracies' threats

This dissertation asks whether democracies have developed a reputation for casualty intolerance that has affected the credibility of their threats to use force. It is generally accepted in the discipline that democracies' threats to use force are more credible than those of other regime types. It is also generally accepted that states do not get "reputations for resolve." I argue that these views are flawed. I present theoretical and empirical challenges to the notion of a structural "credibility advantage" for democracies and explain that reputation has been systematically discounted in importance because scholars studying "reputation for resolve" have conflated three, distinct types of reputations: reputations for reliable signaling, reputations for defending certain interests, and reputations for cost tolerance.; I argue that, in the years since the Vietnam War, democracies have developed a reputation for being highly sensitive to the human costs of war that has caused their "credibility advantage" over other regime types to decline. The dissertation consists of three parts: (1) a brief history tracing the emergence and spread of the perception of democracies as casualty intolerant; (2) a quantitative analysis of over 3000 militarized interstate disputes from 1816--2001 that establishes that democracies have seen a systematic decline in the credibility relative to other regime types of their threats to use force; and (3) four in-depth case studies that ascertain whether perceived casualty intolerance is a primary driver of that decline.; The analyses suggest strongly that states in general can acquire reputations, and that democracies in particular have acquired reputations for casualty intolerance in the last 30 years. Consistent with the reputational hypotheses, democracies' credibility advantage began to decline in the mid-1970s, a trend that accelerated in the early 1990s. Being a democracy is now a "credibility liability" in making threats to use force. This finding remains robust through a series of specification checks. The case studies offer robust support for the notion that states can acquire reputations. Throughout the post-1970s cases, decision-makers consistently cited previous behavior by the democracy as driving their perceptions of its casualty intolerance; these perceptions in turn shaped those decision-makers expectations about how the democracy would behave in a conflict situation.