Trade statistics are like baseball statistics. Nobody really understands them, but everyone uses them to support their arguments. Should Mark McGwire or Sammy Sosa have won the 1998 MVP award? Let's talk about slugging percentages, whatever they are. Is America doing well in the global economy? Let's wave those trade deficit numbers, so long as they support our point of view.

But what if baseball fans had to consult several different and differing sources to figure out how many times McGwire and Sosa went long last year? What if official scorers refused to deal with sacrifice flies because it was too complicated for their scoresheets? And what if every game had to be stopped for several minutes after every play to make sure the right data was collected? That is how we keep up with statistics in the game of trade.

Much of the problem derives from a statistical service that was set up to measure the industrial economy, in which all goods moved by truck, ship, and rail. Goods moving by air are rarely counted, even though they include most high-tech products. The measurement of new and increasingly important transactions such as electronic commerce is crude at best. Small value exports (under $1,000) are not properly measured, understating the importance of small-business trade and small exporters. One Census Bureau economist calculates that U.S. exports are consistently undercounted by $62 billion -- or one-third of the overall trade deficit.

Data collection and the processing of exports and imports are also inefficient and archaic. In 1997, the trade community spent $3.2 billion filing forms with the 104 different federal agencies involved in trade transactions. Many government agencies also publish trade data, resulting in a myriad of different, confusing sources. Finally, the process of data collection and publication takes far too long. Trade statistics are currently published with a six-week lag, leaving policymakers in the dark for long periods, as happened during the recent debate over steel imports.

Such data problems will only get worse as the economy increasingly shifts from an industrial to an information-age structure. We need a trade data system that imposes minimal costs and delays on exporters and importers, and properly measures this New Economy, if we are to effectively monitor, understand, and react to its new realities. Business also need to adopt new technologies, accept shorter filing deadlines, and embrace change in the old system.

Part of the solution has already been proposed by Vice President Gore's Reinventing Government Initiative. Its recently unveiled "International Trade Data System" (ITDS) is a single, integrated, government-wide system for the electronic collection, use, and dissemination of trade data. The trade community would send in just one standardized and harmonized set of information -- electronically where possible -- and trade data users would have access to all published data through a single web site. The system would eliminate repetitive and confusing processing, speed the flow of goods and services, increase access to data, and speed up data publication. (For a description of the proposed system, see http://www.itds.treas.gov).

Implementing ITDS would cost an estimated $256 million. That's about one-fourth of what the Customs Service alone is asking for to create a new stand-alone computer system for tracking transactions along the border in isolation from other federal agencies. Congress should appropriate the funds for ITDS this year. Its debates over who's winning and who's losing in the game of trade could at least proceed with a common set of stats.