Abstract:

This paper extends a recent class of quantitative models of international trade to incorporate factor mobility within countries. We present a model-based decomposition of the variance of economic activity into the contributions of locational fundamentals, market access and their covariance. We show how the standard framework for undertaking model-based counterfactuals in trade can be augmented to obtain predictions for endogenous changes in the distribution of economic activity across regions within countries. A region’s trade share with itself is no longer a sufficient statistic for the welfare gains from trade, which also depend on endogenous changes in the distribution of mobile factors.