Slicing the Pie: Quantifying the Aggregate and Distributional Effects of Trade

We develop a multi-sector gravity model with heterogeneous workers to quantify the aggregate and group-level welfare effects of trade. The model generalizes the specific-factors intuition to a setting with labour reallocation, leads to a parsimonious formula for the group-level welfare effects from trade, and nests the aggregate results in Arkolakis, Costinot and Rodriguez-Clare (2012, "New Trade Models, Same Old Gains?", American Economic Review, 102, 94-130). We estimate the model using the structural relationship between China-shock driven changes in manufacturing employment and average earnings across US groups defined as commuting zones. We find that the China shock increases average welfare but some groups experience losses as high as four times the average gain. However, adjusting for plausible measures of inequality aversion barely affects the welfare gains. We also develop and estimate an extension of the model that endogenizes labour force participation and unemployment, finding similar welfare effects from the China shock.