How do disasters affect voting? A series of postdisaster studies have sought to answer this question using a retrospective framework through which voters deviate from normal patterns of political support (measured by votes or attitudes) to punish or reward officials for their performance, or lack thereof. Here, we argue that the political effects of disasters can last longer than and be qualitatively different from reactions to the original disaster because postdisaster recoveries generate their own issues, to which voters may respond prospectively, and retrospectively. Local communities affected by disasters are likely sites for this effect because their citizens experience the consequences of a disaster more directly and for longer periods than do national audiences. The case of New Orleans after Hurricane Katrina demonstrates this point. Where most studies of postdisaster politics use partisanship as the baseline against which to measure change, we use race because that has been the overriding division in New Orleans. We show that local political effects of Katrina were much more complex and longer lasting than have been found in prior research based on the retrospective model. In the years following the storm, voters changed the pattern of race-based voting for mayoral candidates, approved major governmental reforms, and responded to prospective issues in their evaluation of the incumbent mayor.

No abstract