Das Dokument ist öffentlich zugänglich im Rahmen des deutschen Urheberrechts.

Traditional predictors of election outcomes in Germany are increasingly losing their explanatory power. Rather than new cultural divides, this paper introduces the idea of housing cleavages, i.e., homeownership versus tenancy and high-price versus low-price areas, drawing on macro data for electoral districts and urban neighborhoods from the last three elections (2009-2017) in combination with Immoscout24 ad price data and microdata from the ALLBUS survey (1980-2016). Although, due to its low homeownership rate and conservative house price development, Germany represents a least-likely case for housing to be of importance, we find housing effects beyond traditional predictors. Generally, we find that high house prices, house price increases, and homeownership are positively associated with voting for center-right parties and voter turnout, while social tenancy is associated with votes for the left, but these effects weaken over time due to embourgeoisement effects. Beyond this expected left-right distinction between tenants and wealthier homeowners, we also find outliers along two other dimensions. First, there are center-periphery effects that housing can better capture than simple geographical divisions; second, house prices contain a populist dimension, for example when skyrocketing rents increase votes for the urban left or regions where house prices lag behind benefit the AfD. The paper argues against the more causal self-interest and socialization theories of the influence of housing on voting and instead suggests considering housing as an important socioeconomic proxy to explain political outcomes.