Merchandise Trade Statistics--Some Observations
OCE-89-1BR: Published: Apr 21, 1989. Publicly Released: Apr 21, 1989.
GAO analyzed the nation's foreign trade statistics system, focusing on: (1) the timeliness, accuracy, and comprehensiveness of merchandise trade statistics; (2) possible shortcomings in the trade statistics collection and reporting system; and (3) possible areas of improvement.
GAO found that: (1) variations in processing lags from month to month, from port to port, and between exports and imports add to fluctuations in the overall monthly trade balance and affect data accuracy at the commodity level; (2) because the monthly trade balance represents the difference between the large numbers of exports and imports, even small changes can produce large percentage changes; (3) past merchandise trade statistics may point to dramatic growth in U.S. trade and substantial change in U.S. trade patterns in recent decades, but do not necessarily indicate changes in trade performance; (4) although the United States published its monthly trade statistics on a less timely basis, the 2-week delay in release of the monthly data allowed the U.S. Customs Service and the Bureau of the Census more time for processing and reduced carryovers, which enhanced trade data accuracy; (5) lack of an adequate mechanism to ensure complete and timely reporting of exports diminished the accuracy of export data and resulted in overstated U.S. merchandise trade deficits for several years; (6) the Census Bureau further compromised the accuracy of trade data by errors in commodity classification, quantity, value, country-of-origin/destination, and arbitrary imputations where documentation was incomplete; and (7) Census Bureau publication of 40 different monthly and annual reports at various levels of detail meets the diverse needs of many users, but does not reflect increased global trade development, since the present statistical system is 30 years old. GAO believes that automating data collection and reporting and tightening administrative and statistical controls would improve the quality of trade data.