Co-author(s): Dominick Bartelme, Zhimin Li

Presented in: 17th North American Meeting of the Urban Economics Association

Abstract: This paper examines the impact of WWII on U.S. economic geography. We approach this question empirically, theoretically, and quantitatively. Our empirical results show positive long run effects of WWII spending on overall population and manufacturing activity only in counties with initially low manufacturing employment. Motivated by these findings, we develop a multisector economic geography model that features external economies of scale in manufacturing with non-constant elasticity. We estimate the model and show that an S-shaped scale function in which sizable scale economies are limited to locations below about 1500 manufacturing workers matches the data well. We use the estimated model to re-examine how WWII shaped the evolution of U.S. postwar economic geography. Despite the presence of big push dynamics and multiple steady states, the aggregate effects of WWII spending on long run outcomes are quite limited due to the small domain over which coordination failures operate as well as the untargeted nature of the spending.