Shattered housing

Abstract

Do negative housing shocks lead to persistent changes in household attitudes toward housing and homeownership? We use the residential destruction of Germany during World War II (WWII) as a quasi-experiment and exploit the reasonably exogenous region-by-cohort variation in destruction exposure. We find that WWII-experiencing cohorts from high destruction regions are significantly less likely to be homeowners decades later, controlling for regional differences and household characteristics. Underlying this effect are changes in household attitudes toward homeownership that also extend to preferences for housing consumption, with little or no support for risk preferences, income and wealth effects, or supply-side factors.