Yeats, Alexander J.2014-07-302014-07-302012-11World Bank Economic Review1564-698X10.1093/wber/lhr053https://hdl.handle.net/10986/19083U.N. Commodity Trade (COMTRADE) statistics have major shortcomings for many analyses relating to tariffs and other trade barriers. The use of a cost-insurance-freight valuation base for these data results in an upward (sometimes severe) bias in the implied dutiable value of imports for countries that utilize free-on-board tariffs. This problem can be greatly exacerbated by the “general” trade system procedure used to compile the U.N. statistics, as opposed to the “special” trade practice used for the World Trade Organization Integrated Database. U.S. International Trade Commission statistics show that the combined effects of these biases can reach magnitudes that preclude the legitimate use of COMTRADE for many tariff-trade simulations or related trade negotiations.en-USCC BY-NC-ND 3.0 IGOcustomsequilibriumforeign tradeforeign trade zonesgeneral equilibriumimport statisticsimportsinternational tradeinternational trade oommissionmultilateral trade negotiationstariff changestariff linetrade barriertrade barrierstrade creationtrade datatrade modelstrade negotiationstrade restrictionsworld tradeJournal ArticleWorld Bank10.1596/19083