Last October 14th, with gloom gathering like a thunderhead over Wall Street, the government reported the trade deficit for August -- an abysmal $15.7 billion. The Dow Jones Industrial Average promptly fell 95 points, a psychological break point that helped set the stage for Black Monday five days later.

The monthly trade figures are the most closely watched numbers in Washington these days. Around them orbit a thousand economic moons, each with its own impact on the business world. Interest rates rise and fall on their announcement, the dollar goes up or down, and talk of hard times ahead reduces spending and hiring.

The trade numbers, however, are so inaccurate and so unrepresentative of the U.S. economy that it is frightening that they yield such important consequences. Consider this: the monthly figures measure only merchandise trade -- and not very accurately at that. They ignore completely all trade in services, despite the fact that the service sector accounts for more than 70% of this country's gross national product. It's doubtful that the monthly figures reflect the true position of U.S. trade within several billions of dollars. But coming as they do with the imprimatur of the government, the only source of these figures, they might as well be carved in stone. To paraphrase a saying in the computer industry, it's garbage in, gospel out.

The garbage going in consists of tabulations of tangible merchandise -- Toyotas, VCRs, petroleum, airliners. In all, the U.S. Customs Service collects data on some 10,000 product categories. Counting tangible goods leaving U.S. gateways would seem to be a rather mechanical task, what with export declarations and all manner of Customs paperwork. But Customs stations, overworked or unmanned, occasionally miss things, such as fully loaded 18-wheelers rolling into Canada. Incredible though it may seem, the government last year reckoned it was underestimating exports to Canada alone by about 25%. Soon the government will begin using Canadian import figures, instead of American export data, in trade calculations.

Moreover, Customs sometimes records erroneous valuations. Seldom captured at all are discounts after the fact, rebates, and the like. Then there are the frantic fluctuations of currency exchange rates. "This gets you into all sorts of problems," says economist Sar Levitan, director of the Center of Social Policy at George Washington University. "You have to adjust for the changes in the dollar value of yen, marks, francs, pounds, and pesos, and that becomes a statistical nightmare."

The most serious fault of the monthly figures, however, is that they grossly misrepresent the nature of the U.S. economy. Devised during the bygone days of America's unchallenged industrial supremacy, the numbers measure only merchandise, or manufacturing, trade. As emblems of America's declining industrial strength, the figures may paint an accurate enough picture. But by ignoring the booming service sector, the statistics greatly exaggerate the nation's trade problem.

The government does track service exports -- they are included in the balance of payments -- but there is a good reason that economists like to call them "the invisibles." For if the government can miss tractor trailers entering Canada, what chance does it have of documenting transponder leasing space on Intelsat satellites, say, or overseas phone calls?

The thankless task of monitoring service exports falls to the Bureau of Economic Analysis (BEA), a division of the Commerce Department. This is the sanctum sanctorum of the government's statistical empire, home of the GNP, the balance of payments, and a publication called Survey of Current Business, the bean-counter's bible. Its 400 staffers, mostly economists, operate from a cramped set of offices several blocks from the White House.

To get a sense of the complexity of their job, consider what is classified as service exports. Start with the 350,000 foreign students in the United States and their expenditures on tuition and housing. Move on to the outlays of all foreign tourists and businesspeople who reach our shores. Tally up medical costs to aliens who come here for health care. And throw in all dividends and interest payments to auslanders who invest in American corporate and government securities. Then, to count service imports, they turn it all around and do the same for Americans buying services abroad.

But that's just the start. The tough part involves overseas transactions of American businesses. Every time an American lawyer, engineer, business consultant, or contractor performs work for foreigners, it is an export of services. Much of it is impossible for the BEA to capture.

Take the case of Thomas Gray, chief economist for the Small Business Administration. Not long ago, he took a leave from his job and flew to Turkey to consult for the Turkish government. For a week's work, he earned about $2,000. "That money is not ever going to get caught in the trade figures," Gray says. "And there are probably thousands of cases like that. We are a brainpower center. We have collected and trained bright people from all over the world, so we ship a lot of services. Information and services are what this country is increasingly about."

Nor will the government measure the service exports of Pacer Systems Inc., in Billerica, Mass. A $33-million company, Pacer is on the verge of booking $4 million to $5 million a year in service business abroad, largely in engineering and computer work. It will be operating as a subcontractor to a native company. "That kind of activity is invisible to the government," says president Jack Rennie. "It just doesn't show up anywhere."

To collect information about service exports, the BEA mails out surveys by the thousands to large service companies. One survey tries to measure freight charges by airlines and ocean shipping firms that transport goods to and from the United States -- the Pan Ams and Lufthansas. Another survey collects data on premiums paid by foreigners to American insurance companies -- but for reinsurance only. A third survey covers U.S. companies selling architecture, engineering, construction, and mining services to foreigners -- big firms such as Bechtel Group. The fourth gathers information on fees, royalties, and licenses paid to U.S. companies for foreign use of copyrights, trademarks, franchises, and manufacturing processes.

When it comes to calculating tourism and travel expenditures, a huge component of service trade, the methodology is largely guesswork. The government regularly hands out questionnaires on airliners to Americans returning to the States and foreigners leaving here. How much did you spend on hotels, car rentals, tickets to Disneyland? Then it estimates how many Americans visited which foreign countries, and how many foreigners came here. "You can't include all the tourists. It's just too big a job," says Obie Whichard, an international economist at the BEA. "So you collect the information from samples. And a lot depends on whether people agree to fill out the forms."

Immense gaps exist in the BEA database. The BEA is largely in the dark about the burgeoning banking and financial services industry. "The Treasury Department and the Federal Reserve Bank of New York have a system of tracking capital flows," says Whichard. "But in truth, in banking and financial services, we don't have much information."

The BEA has no information at all about some of the most dynamic segments of the service economy, having only recently initiated a new survey for them. Completely invisible to the government are exports in data processing, direct insurance, legal services, accounting, advertising, management consulting, and agricultural services.

But even with the new survey, small and midsize companies are unlikely to appear on the BEA's radar screen. It tracks only firms with at least one foreign transaction last year exceeding $250,000. Whichard expects the survey to show about 1,000 service companies reporting business deals of that magnitude, and believes that to be no problem. "We think it's true that a very large share of the transactions are accounted for by the larger service firms," he says, "because that tends to be true of international transactions generally."

That bias causes the BEA to overlook a lot of service exports. "A great many small outfits that are exporting highly valued services are made up of experts -- engineers, architects, lawyers, expediters, you name it," says SBA economist Gray. "Certainly it would be good for the whole economy to recognize that flow explicitly."

Whatever the flaws of its surveys, the BEA regularly attaches a dollar figure to U.S. trade in services. In 1984, it reported that direct exports of private services (excluding investment income) totaled $44.3 billion. It also calculated imported services (again excluding investment income) at $42.3 billion. That gave the United States a small surplus, $2 billion.

But the BEA's calculation seemed so flawed that the Office of Technology Assessment (OTA), a congressional research agency, decided to take a look. It found the statistics to be wildly inaccurate. "As much as half of U.S. service exports may escape the official statistics," the OTA reported. While the government maintained 10,000 categories for trade in goods, it had only 40 for trade in services. "Many of the gaps, conceptual flaws, and uncertainties in the data are carryovers from an era when manufacturing and agriculture overshadowed the service sector," it reported.

By the OTA's reckoning, the 1984 service exports, excluding banking, were in the range of $69 billion to $91 billion, while imports ranged from $57 billion to $74 billion. The U.S. surplus that year, the OTA suggested, was more like $14 billion. It estimated that trade in services ran a surplus of $51 billion for the years 1982 to 1984, three times the official government figure. Because the service sector has grown so rapidly over the past four years, the service trade surplus is now probably much more than $14 billion a year.

A surplus in service trade of $20 billion or more would put a significant dent in the merchandise deficit, which came in last year at some $171 billion. But despite the importance of these figures, the government isn't likely to pin them down anytime soon.

Improving the statistics would be a costly and complicated job, and the money isn't there to do it. Unlike the Pentagon or the Social Security Administration, the numbers bureaucracy lacks a strong constituency, and so when the Reagan Administration chopped government spending, the ax fell particularly hard on such data-collecting agencies as the Bureau of Labor Statistics, the Bureau of the Census, and the Bureau of Economic Analysis.

But the cuts may have turned out to be counterproductive. In 1986, a report of the Joint Economic Committee of Congress warned that the country's faulty economic statistics were impeding economic growth. To confirm that, you don't have to look much further than the reaction of the financial markets to the monthly trade figures. Or to the difficulty encountered by Bernard Ascher, who as part of the U.S. team is helping to negotiate the dismantling of more than 500 protectionist barriers that hinder the selling of U.S. services abroad. "It took six years to convince our major trading partners to even put services on the negotiating table," says Ascher, director of service industry affairs at the office of the U.S. trade representative. "But with our lack of data, people ask, 'How can we negotiate a code for trade in services when you Americans don't have the statistics to know what you're talking about?' When you don't have a basis for estimates, it can really impede progress."

There's a saying in Washington that a shabby product is "close enough for government work." It usually brings a laugh. This time, nobody's even cracking a smile.


The trade figures are only the latest in a parade of statistics the financial markets have used to check the nation's economic well-being. In the booming 1960s, traders studied the volume of commercial and industrial loans. Commodity prices held sway in the "oil crisis" days of the early 1970s. During the Carter Administration, with inflation at a gallop, the market pondered the money supply. Now that we are operating in a full-blown international economy, it is the trade figures that seem to overshadow all else.