Local scars of the US housing crisis

Abstract

The 2006–09 US housing crisis had scarring local effects. For a given county, a housing shock generating a 10% reduction in housing wealth from 2006 through 2009 led to a 4.4% decline in employment by 2018 and a commensurate decline in value added. This persistent local effect occurred despite the shock having no significant impact on labor productivity. The local labor market adjustment to the housing shock was particularly costly: local wages did not respond, and long-run convergence in the local labor market slack instead took place entirely through population losses in affected regions. Moreover, the 2002–06 housing boom does not generate significant employment gains, indicating that the employment losses relative to 2006 are also losses relative to the counterfactual case in which there was no housing cycle.