Death, Destruction, and Growth in Cities : Entrepreneurial Capital and Economic Geography After the 1918 Influenza

How does city growth respond to catastrophe? I propose a model in which local entrepreneurs have a comparative advantage in starting local businesses and business creation benefits from business activity. The model predicts that cities stagnate following a reduction in local human capital but, consistent with established facts about city resilience, recover after the destruction of local physical capital. I test for the predicted effects of a shock to human capital using U.S. city populations after the 1918 Influenza Pandemic, which killed 0.5% of residents in the largest U.S. cities. Instrumenting for Flu incidence with local weather at the peak of the epidemic, I show that cities with high influenza mortality had persistently low population levels and growth rates, with estimates implying that a 10% increase in Flu incidence caused a 13% reduction in 2010 population. A calibrated version of the model successfully rationalizes both the direction and magnitude of the observed effects