A Pragmatic Approach to Estimating Nondiscriminatory Non-tariff Trade Costs

Despite the growing importance of non-tariff measures (NTMs) in trade policy, many common methods for estimating their costs exhibit significant challenges in terms of data requirements or the lack of specificity in the estimates they produce. I propose an extension to existing approaches that mitigates many of these limitations. Following earlier work by Fontagne et al. (2011), I estimate a measure of the aggregate nondiscriminatory trade costs present in each country. Then, in a novel second stage, I decompose the estimated trade costs into individual components such as the costs associated with specific NTMs. Additionally, the second stage allows for the identification and removal of non-cost factors that erroneously contributed to the aggregate cost estimates. The methodology provides a means by which to accurately estimate the effects of specific NTMs using widely available data and standard econometric techniques. I test the methodology using a typical gravity dataset and detailed NTM data. The results provide specific cost estimates for SPS measures and TBTs across many sectors and demonstrate that aggregate measures of non-discriminatory trade costs can significantly reflect non-cost factors.