Quantitative Comparison of Ad-Valorem Equivalents for Non-Tariff Trade Measures: WTO Versus TRAINS

Abstract

Non-tariff measures (NTMs) play a growing role in trade policy. Economists quantify NTMs by calculating their Ad Valorem Equivalents (AVE), typically employing one of two global NTM datasets: TRAINS from UNCTAD or WTO notifications. While TRAINS data better measure NTMs, their limited coverage over countries and time is inadequate for many types of analyses, and thus researchers often resort to using WTO data, which is not as accurate as the TRAINS but has better coverage. We compare the two datasets to assess whether AVEs from WTO notifications align with those from TRAINS, given similar country and time coverage. We find significant AVE discrepancies between datasets, sometimes reversing signs, warranting caution when using WTO notifications in the absence of TRAINS data to construct NTMs.