Abstract: | Preventing climate change and damage from natural disasters typically requires policies with up‐front costs that promise a flow of benefits over time. Why has obtaining such policies in a competitive electoral democracy proved so intractable? We develop a formal model of electoral accountability in this context, in which politicians have private information about their motivations. The model shows why fully rational voters, though certain that incumbents spend less on disaster prevention than is good for them, reelect incumbents at very high rates. In addition, in such equilibria, voters would punish incumbents who spent more on disaster prevention. This equilibrium is consistent with (and implies) some of the major empirical regularities observed in the literature on voting and disaster prevention. We discuss some implications of our analysis for advancing public debates about disaster and climate change mitigation. |