The results of liberalization studies can be influenced decisively by the methodology of import tariff aggregation. This article deals with different measures used to aggregate import tariffs. Therefore, the advantages and disadvantages of different aggregation measures are discussed theoretically. In the empirical part of the paper, the simple average and the trade-weighted average of bound and applied tariffs are calculated for different agricultural products of the EU. The analysis considers ad valorem and specific tariffs as well as tariff rate quotas. Additionally, for the applied tariffs, a distinction is made between MFN-tariffs and bilateral preferential tariff rates. Using the same data base, the product specific Trade Restrictiveness Index (TRI) and the Mercantilist Trade Restrictiveness Index (MTRI) are calculated. Because of the poor data availability at the detailed tariff-line level, a simple partial equilibrium application of these indices is implemented, which can be extended to the whole import tariff data base of an economic trade model. The results show a strong correlation between all aggregation methods, but the size of the import tariffs differs highly, depending on the aggregation measure used. Furthermore, high differences between bound and applied tariffs for some EU agricultural products are calculated.
