Are the EU trade preferences really effective? A generalized propensity score evaluation of the Southern Mediterranea countries’ case in Agriculture and Fishery

The aim of this work is to assess the trade impact of preferential schemes in agriculture and fishery granted by the European Union (EU) to the Southern Mediterranean Countries (SMCs). This analysis presents several methodological improvements to previous works. First of all, we rely on a continuous treatment - i.e., preferential margins - to capture the ``average treatment effect'' of trade preferences rather than on a binary treatment based on dummy variables. Second, we apply non parametric matching techniques for continuous treatment, namely a generalized propensity score matching (GPS) technique, to assess the average causal effects of preferences on trade flows. Third, we use highly disaggregated data at sectoral level in order to evaluate properly the preferential treatment which is conceived to be applied at the product level. Our results show how the impact of EU preferences in agriculture and fishery granted to SMCs is positive and significant and better evaluated using impact evaluation techniques. We also assess the functional form of the relationship between EU-SMCs preferences and bilateral trade flows as well as the optimal level of preferential margin above which the marginal impact decreases.(This abstract was borrowed from another version of this item.)