Make your own free website on

What's in a Trade Statistic? Less and Less

By Jock O'Connell

This article originally appeared in the Opinion section of the Los Angeles Times on August 19, 2001.

SACRAMENTO -- California's robust export trade has inspired frequent flights of chauvinism in Sacramento, where politicians of both parties enjoy reveling in the conceit that, were it something it's not (i.e., a separate country), the Golden State would be the world's fifth-largest economy.

California exports reportedly rose 13.2%, to $30.2 billion, in the first three months of 2001, the fourth consecutive quarter in which state exports topped $30 billion and the seventh consecutive quarter of positive export growth. That streak ended, however, in 2001's second quarter, as the value of California's exports fell by 6.6 percent from the same quarter a year earlier.

This news shouldn't be surprising. Economic growth in virtually every major market has ebbed during the last several months. But that probably won't stop Sacramento legislators and lobbyists from seizing upon the disappointing trade figures to advance various policy agendas. The predictable result will be a plethora of pithy sound-bites demonstrating, if nothing else, how little is understood about what the trade data really tell us. The fact is, in the emerging global economy, statistics on imports and exports are fast becoming meaningless barometers of international commerce.

This is not to say that cross-border trade is unimportant. By conventional measures, merchandise exports now account for nearly 12% of the U.S. gross domestic product, roughly double what its share was in 1975. But trade data do not tell the whole story of the connective tissue of globalization.

Shipping goods across borders is not the only or even the most important way in which California companies do business overseas. Ever since the laying of the first transatlantic telegraph cable in 1866, U.S. firms have been investing in manufacturing plants abroad to avoid the high costs of shipping their wares from the United States, as well as to skirt tariffs and other import restrictions. This trend has accelerated in recent decades because of a host of international trade agreements that have bolstered legal safeguards for foreign investors. The most recent data, that of 1998, show that sales made by foreign subsidiaries of U.S. firms totaled some $2.4 trillion, while total U.S. merchandise exports that year amounted to a comparatively modest $933 billion.

Separate figures on the sales of California companies' overseas subsidiaries are not kept. However, the 2.5 to 1 ratio of overseas sales to exports for the country as a whole suggests that the foreign subsidiaries of California firms reaped over $260 billion in sales in 1998, a year in which the state's merchandise exports amounted to $105 billion.

As more and more companies learn to manage global supply chains, the meaning of trade statistics will become less and less clear. Ironically, it's conceivable that a decline in a country's merchandise export trade could actually be a favorable sign that the national economy is shifting away from an old-economy model based on mass manufacturing to a new one geared to produce high value-added products and services.

Because of the links between U.S.-based firms and their affiliates abroad, a substantial portion of America's trade can be more appropriately characterized as intrafirm. In 2000, intrafirm trade accounted for $563 billion (47%) of U.S. imports, and $246 billion (32%) of American exports.

Although some intrafirm shipments from U.S. factories are finished goods destined for customers abroad, a sizable share, especially in high-technology sectors, are goods sent overseas for further processing before being re-exported to this country. Indeed, one consequence of the transfer of all or a portion of U.S. manufacturing operations overseas is that component shipments that formerly went to a plant in Georgia or Illinois now travel to a factory in Mexico or China. What had formerly been an unremarkable domestic shipment is now a celebrated export transaction.

Trade data do not merely give a distorted impression of the way international business is being conducted. There are entire sectors of the economy whose products simply elude data collectors. Such is the case with most computer software.

The U.S. software industry was the most dynamic and internationally competitive segment of the U.S. economy during the past decade. With foreign sales currently accounting for at least half its revenues, it is also the sector most likely to confound efforts to measure international commerce and gauge its importance to local state or regional economies. The reason is the Internet.

By 2005, the worldwide market for packaged (non-customized) software is projected to be $150 billion. By then, according to a June 2000 survey of software company CEOs by the Business Software Alliance, as much as 66% of all packaged software will be distributed over the Internet, compared with just 12% today. Yet because there are no customs posts on the Internet, a computer program designed by a company in Santa Monica can be downloaded by a consumer in Sweden from a server located in Ireland without any record of a U.S. export having occurred.

All this vastly complicates the already tricky problem of calibrating just how internationalized the U.S. or the Golden State's economy has become. It also raises some prickly public policy issues about what types of economic and trade-development programs may be appropriate. Historically, public officials have been eager to help exporting companies, the rationale being that exporters are apt to be prime sources of new jobs and tax revenue. But what about highly successful companies that have taken advantage of alternate ways of meeting overseas demand?

There is also a more disconcerting issue involved here. Federal Reserve Board Chairman Alan Greenspan has often fretted about "the challenge of measuring and modeling our dynamic economy." If economic globalization involves a fundamental restructuring of cross-border commerce, many of our conventional economic models are becoming obsolete, making forecasts even more dubious.

This is not merely some academic concern. Businesses use such models to guide investment decisions. Wall Street analysts study them for clues that might herald shifts in interest rates or currency values. Government officials factor them into revenue forecasts. International disputes arise whenever trade data are thought to provide proof of unfair trading practices.

Unhappily, economic models and measurements that no longer capture the realities of how the international marketplace operates invite economic shocks, not to mention ill-informed speculation by politicians and pundits desperately peering into the entrails of desiccated data.


Copyright 2001 By J. A. O'Connell