Explaining Trade Flows: Traditional and New determinants of Trade Patterns

International audienceAn empirical tradition in international trade seeks to establish whether the predictions of factor abundance theory match with the data. The relation between factor endowments and trade in goods (commodity version of Hecksher-Ohlin) provide mildly encouraging empirical results. But in the analysis of factor service trade and factor endowments (factor content version of H-O), the results show that it performs poorly and reject strict H-O-V models in favor of modifications that allow for technology differences, consumer's preferences differences, increasing returns to scale, or cost of trade. In this paper, we test if these new determinants help us improve our estimation of trade patterns in commodities. Since the commodity version allows obtaining a large panel data, we also compare two periods, pre and post 1980. We use a Heckman procedure to allow for non-linearity in the relation between factor endowments and net exports and between trade intensity and net exports. The results show that adding the new determinants of factor content studies help us improve the prediction of specialization in different manufactured products. However specialization according to factor endowments is stronger than ever, especially concerning the specialization according to human capital endowment. Trade patterns are also determined by trade intensity. Here, differences in technology, trade policy, transport and transaction costs, explain the difference in trade intensity