| Abstract: | Comparisons of matched US COMTRADE and World Trade Organization IDB statistics reveal numerous instances where the WTO data on dutiable imports exceed both the corresponding UN general import statistics and the reported exports of partner countries by very large margins over extended periods. In cases, the dutiable import statistics report major trade in products for which the partners do not record any corresponding exports (to anyone). Empirical tests indicate these biases result from special (and very unorthodox) invoicing practices for shipments into US foreign trade zones. The paper concludes these problems have the potential to seriously misdirect analytical trade studies which must utilise import statistics for their empirical base. |
