Using Tariff Indices to Evaluate Preferential Trading Arrangements: An Application to EU

Abstract

This work provides a revised method of aggregation of tariff line data that is consistent with a common objective of the preferential policies. Using constant-elasticity-of-substitution aggregators we compute the uniform preferential tariff reduction that would leave the volume of imports constant. Our preference index is in the spirit of Anderson and Neary protection index theory and builds on the Mercantilistic Trade Preference Index (MTPI) by Cipollina and Salvatici (2011). The new index definition allows to overcome a well-known problem in the protection indexes computation, namely the necessity to incorporate a reference (i.e., untaxed) good due to the homogeneity properties of import demand functions. We use the MAcMapHS6 version 2 (2004) to assess the intensity of trade preferences granted European Union (EU) to different sectors and country groups. We find that results using the preference aggregator may differ substantially from those derived using simple averages. The main message resulting from the paper is that ignoring the information available in tariff line data or aggregating it without a sound theoretical model, could lead to a substantial bias in trade policy analysis.