Tariff evasion in global trade data

We find evidence of pervasive tariff evasion in the global data on trade from 1988 to 2015. Using over 35 million observations of data on import and export flows at the HS6 product category level, we find evidence of substantial underreporting of imports relative to export data on average and particularly when tariffs on product categories are high. These effects are stronger in more corrupt destination countries, as measured by the World Bank's Worldwide Governance Indicators [World Bank. 2016. Worldwide Governance Indicators. September 25, 2016]. In addition, evidence of tariff evasion increases significantly in economic downturns. We document these patterns in the global data and explore the welfare effects of this evasion by (1) putting a lower bound on the extent to which there are revenue losses from tariff evasion, and by (2) estimating the effects of corruption as measured by this indicator on global trade in a simple gravity model. We estimate that in total, revenue losses from tariff evasion are currently likely to exceed 400 to 670 million USD globally per year, and find that the effects of corruption on trade flows are ambiguous overall but change from weakly positive (‘grease the wheels’) to largely negative over the years in our sample.