Extreme weather and risk preference : panel evidence from Germany

Individual risk preference may change after experiencing external socio-economic or natural shocks. Theoretical predictions and empirical studies suggest that risk taking may increase or decrease after experiencing shocks. So far the empirical evidence is sparse, especially when it comes to developed countries. We contribute to this literature by investigating whether experiencing financial and health-related damage caused by storms affects risk preference of individuals in Germany. Using unique panel data, we find that households who report storm damage increased their risk taking. We do not find evidence of exposure to storm per see (regardless of damage experience), which suggests that households have to suffer damage for their risk preference to be affected. These results are robust across a battery of alternative model specifications and alternative storm damage measures (magnitude of financial damage). We rule out other potential explanations such as health-related and economic shocks. The self-reported storm damage data is broadly confirmed by regional storm damage data provided by the insurance industry. While we cannot identify the channels through which experiencing storm damage affects risk preference from our data, we suggest and discuss some potential channels. The results may have important policy implications as risk preference affects, for instance, individuals' savings and investment behaviour, adoption of self-protection and self-insurance strategies, and technology adoption.