Effects of extreme weather experience on climate change belief

14:00 Tuesday 28 May


Room S2


Daniel Osberghaus (Germany) 1; Carina Fugger (Germany) 2

1 - Centre for European Economic Research (ZEW); 2 - Center for European Economic Research (ZEW)


There is a persistent polarisation of the public debate on the existence of anthropogenic climate change (ACC). However, the endorsement of ACC amongst lay people is the minimum requirement for getting public support for climate change-related policies and private adaptation. A large body of literature has suggested that peoples’ beliefs about the existence of ACC are related to their personal experience of unusual climatic patterns or extreme weather events. However, people may interpret experience in a way which is congruent to their pre-existing beliefs and opinions (motivated reasoning). If new information is inconsistent with personal beliefs and opinions, it is neglected, dismissed or critically questioned and may even ‘backfire’. In this study, we empirically analyse to which extent motivated reasoning exists in the personal beliefs about the existence of ACC.


We base our empirical analysis on a novel longitudinal household survey, enriched by several external data sources. We identify the causal effect of experience of a particular extreme weather event – a major flood in Germany in 2013 – on the beliefs in the existence of ACC. Consecutively, we analyse whether this effect depends on pre-existing beliefs about climate and environmental issues (measured before the flood event). Data on beliefs about ACC are measured in a panel household survey in Germany (12,217 unique respondents, four survey waves, N=25,117). For combining the survey data with data on flood experience, we use geo-located satellite images, and county-level data on insured damages. The empirical analysis relies on multivariate difference-in-differences (DD) regressions with fixed effects on the household-level. In order to test for heterogeneous effects, we employ DD regressions for sub-samples and triple-differences estimations.

Results and conclusions

We show that experiencing the flood of 2013 has no significant effect on beliefs about ACC in the total sample. However, considering sub-samples of the population, we find that individuals who believed in ACC before the flood do not change their beliefs or are even more convinced of ACC if they experienced the flood. In the contrary, climate-sceptical respondents, once they get affected by the flood, get even more sceptical than their unaffected counterparts. These results lend support to the phenomenon of ‘motivated reasoning’, and suggest that information and awareness campaigns on climate change issues basing on personal damage experience may not be as effective as previously thought.