Brown Bag Seminar: Backward-Engineering Trade Protection: A Method to Estimate Worldwide Industry-Level Trade Barriers

Abstract

Non-tariff barriers to trade are widely recognized as the primary form of trade protection in the modernworld economy. Yet reliable information on these measures remains elusive due to the complexity of theunderlying regulations and political incentives for non-transparency. In this paper, I present a penalizedregression method that allows me to infer the size and evolution of trade barriers from observable tradefrictions, rather than having to rely on inventories of legislation or official notifications. As a consequence, myestimates are not affected by the coverage and selection problems prevalent in existing compilations.

Based on cross-validated time-series smoothing, my estimates also go beyond the status-quo by exhibitingreliable time-variation. Overall, the resulting data cover 152 countries and 212 industries over the thirty-yearperiod 1989-2018. Expressed in terms of ad valorem tariff-equivalents, these data provide the mostcomprehensive picture of applied trade protection currently available. I demonstrate that my estimates notonly display a host of intuitive empirical patterns, but also outperform existing inventories data in variousways. In a replication study of Kono (2006), I show that my estimates make it possible to uncover hithertounknown temporal trends in the pattern of protection across democracies and autocracies.

To join this seminar virtually, please use this link.
Meeting ID: 892 7642 8187
Passcode: 790947