The mystery of misreported trade: A novel method to identify distorted trade figures

This research shows that misreporting of trade is rampant, which markedly reduces the quality of reported bilateral trade data. We devise a novel methodology to correct for misreported bilateral trade while accounting for the composition of product types and for trading partners. Based on bilateral data for 196 countries in the world at the product level over the period 1996–2017, we find that (i) our corrected trade data are significantly more reliable than reported data; (ii) our corrected data are significantly more reliable than data based on conventional misreporting correction methods; (iii) corrected trade balances are often quite distinct from their reported counterparts; (iv) capital flight is the most important source of misreporting and (v) the misreporting propensity tends to cluster geographically, with the main offenders often deviating from common perceptions in the literature. Based on estimates of the gravity model, we show that geographical characteristics are more influential than institutional variables in explaining misreporting.