Job Decentralization and Residential Location

Abstract

How does the spatial distribution of employment opportunities influence residential location? We revisit this classic question in urban economics by exploiting a natural experiment generated by the history of state capitals. Many state employees in capital cities work in centrally located government buildings that were constructed in the nineteenth century, while state workers elsewhere mirror the decentralization of the private sector. We compare the work and residential locations of state workers in capital and non-capital cities relative to other workers in their metropolitan areas. Our results suggest that assigning 1,000 jobs to the central city would attract approximately 250 working residents to the city. Evidence from other industries with historically-determined locations, including the postal service and defense contractors, corroborates our basic finding.