Pierre-Daniel G. Sarte, Esteban Rossi-Hansberg, Fernando Parro and Lorenzo Caliendo
We study the impact of regional and sectoral productivity changes on the U.S. economy. To that end, we consider an environment that captures the effects of interregional and intersectoral trade in propagating disaggregated productivity changes at the level of a sector in a given U.S. state to the rest of the economy. The quantitative model we develop features pairwise interregional trade across all 50 U.S. states, 26 traded and non-traded industries, labor as a mobile factor, and structures and land as an immobile factor. We allow for sectoral linkages in the form of an intermediate input structure that matches the U.S. input-output matrix. Using data on trade flows by industry between states, as well as other regional and industry data, we obtain the aggregate, regional, and sectoral elasticities of measured TFP, GDP, and employment to regional and sectoral productivity changes. We find that such elasticities can vary significantly depending on the sectors and regions effected and are importantly determined by the spatial structure of the US economy. We highlight the role of these elasticities by tracing out the effects of productivity gains in California in the Computers and Electronics industry between 2002 and 2007 on all other U.S. sectors and regions.