In line w/ the idea of experience-based accountablity, we find that in areas with less prior insurance instability, voters reward the incumbent, but not in those with high instability. W/ surveys, we also show that voters in these areas are more certain (and more critical) of the Gov's performance.

Using a diff-in-diff around eligible vs. ineligible areas outside fire perimeters, we find precisely estimated null effects on support for both the governor and the insurance commissioner, despite evidence that the policy reduced nonrenewals, raised house prices, and was salient.

Christian Baehr, @hannohilbig.bsky.social and I have a new working paper on the electoral consequences of short-term climate adaptation, where we look at interventions in the insurance market.
doi.org/10.31235/osf...

4/ 💡Key finding:
Green vote share 0.5–1.2 percentage point drop in Green vote share in the local county and t effect is stronger the closer you are to the plant.

We also find higher local GDP post-opening, and people worry less about their finances.

3/ We build a novel dataset with locations & opening dates of nuclear power plants and voting records in Germany and use TWFE and IV designs.