Determinants of trade margins : insights using state export data

We adapt the heterogeneous firm trade models of Helpman, Melitz, and Rubinstein (2008) and Lawless (2010) to analyze extensive and intensive trade margins using state-level exports to foreign nations. Our theoretical analysis provides definitive predictions for the effects of changes in fixed costs, variable costs, and foreign income on the extensive margin, while for the intensive margin the predictions regarding changes in fixed costs are definitive, but the effects of changes in variable costs and foreign income are not. The number of exporting firms of a state is used to measure the extensive margin, while the intensive margin is approximated by the average firm exports of a state. Various count-data models, such as the standard negative binomial and its hurdle extension, are used to address non-trading pairs and overdispersion in the extensive trade estimations, while a Heckman correction is examined to handle sample selection issues in the intensive margin estimations. As the theory predicts, we find more consistent and statistically significant effects of changes in cost-related variables on the extensive than on the intensive margin of trade. Unlike Lawless (2010), but consistent with a truncated Pareto distribution, empirical findings suggest that variable costs reduce average exports. A noteworthy finding is that U.S. foreign direct investment has a positive effect on both margins.