Abstract
The world has seen a massive increase in wealth and wealth inequality over the last decades. Given the skew in policy making towards the preferences of the wealthy, this raises the question of how individual wealth affects political participation. Approaching this question empirically is complicated by the fact that random variation in wealth is rare, and many factors that can bias the estimation of the relationship between wealth and participation are difficult to measure. We address the question using a Swedish discordant identical twin design with a) register-based wealth data, b) validated election turnout for multiple elections, and c) self-reported civic participation measures. This design allows us to rule out all shared confounders, such as genetics, family background and socialization, and shared networks. We find that even though wealthy individuals descriptively vote more often, the causal effect of wealth is probably zero, and may for civic participation even be negative.
