“Spatial Linkages, Global Shocks, and Local Labor Markets: Theory and Evidence” by Dr. Costas Arkolakis
We analyze and measure the impact of economic shocks on local labor markets in the presence of rich cross-market links, arising from trade, agglomeration spillovers and labor mobility. We model these links within a generalized perfectly competitive framework and characterize the key aggregate elasticities that fully determine both the differential and aggregate impact of trade and productivity shocks. In this environment, we show that the presence of spatial links across markets generates indirect spillover effects and thus reduced form empirical analysis may not identify the correct elasticities. We instead propose an empirical methodology that relies on a Model-Implied Instrumental Variable, which we show to be optimal, in the sense that it yields the most efficient estimator of the elasticities. These insights are general since our framework is observationally equivalent to a large class of existing trade and geography models with cross-market links. We apply our estimator using disaggregated regional trade data to measure the impact of trade shocks on US local labor markets. Our counterfactual exercises reveal that the importance of general equilibrium indirect effects are directly related to the magnitude of the migration elasticities.