Trade and Welfare: Does Industrial Organization Matter

Many contemporary theoretic studies of trade over geography reduce to an examination of constant-elasticity reactions to changes in iceberg trade costs. These impacts are readily analyzed in simple constant-returns models based on the Armington (1969) assumption of regionally differentiated goods. Following the line of reasoning suggested by Arkolakis et al. (2008) one can reach the surprising conclusion that industrial organization does not matter. In the present paper, we show that this finding is fragile, and with a minor elaboration of their model, the rich industrial-organization features of the popular Melitz (2003) model do, in fact, generate important differences for trade and welfare.