PRESIDENT ZEDILLO'S visit to California last month churned up numerous references to what would appear to be a rather damning factoid: that Texas exports two to three times as much to Mexico as does the Golden State, even though its economy is only half of the size of California's. The obvious implication is that our brethren in Texas have been doing a better job developing business ties with our mutual amigos south of the border.

Inevitably, fingers get pointed at former Governor Pete Wilson, whose willingness to demonize all things Mexican in order to advance his own career is seen to have dampened the eagerness of Mexican businesses to trade with California companies.

While Wilson's campaign rhetoric did no favors for California's commercial relations with Mexico, there is a major problem with this canard. It is, more than anything else, an artifact of grotesquely misleading trade statistics.

What is not widely appreciated is that the process for compiling the so-called state-of-origin export data -- the figures most commonly cited in government reports and newspaper articles -- is remarkably indifferent to the question of where the exported merchandise was actually manufactured, grown or processed.

Instead of the state of origin of the goods, the critical variable for government statisticians is the location of the party identified as the exporter of record on the official shipper's export declaration. It is this bit of information that determines which state receives credit for the export.

Now the exporter of record and the product's manufacturer may be the same. But that is becoming increasingly rare. More large corporations are cutting costs by outsourcing transportation and distribution. Smaller companies generally have no choice but to ship their products through an intermediary. As a result, the likelihood is growing that the exporter of record will be a party wholly unconnected to the company responsible for producing the goods. Indeed, the two parties may not even be in the same city or even in the same state.

This distinction between the export's producer and the exporter of record is scarcely trivial. If the actual manufacturer were located in the Bay Area but the exporter of record was a warehouse manager in Laredo, the shipment will almost certainly be officially credited as an export of Texas.

Consider now the logistical problem of transporting goods from California to Mexico. Although there is a considerable amount of trade across the border through San Diego and Imperial counties, most of these shipments are bound for maquiladoras in places such as Tijuana and Mexicali. However, for cargoes destined for Mexico's industrial and population centers, transshipment via Texas offers decided advantages and cost-savings to shippers.

A glance at a map will reveal why. Mountain ranges in Mexico run north-south, severely inhibiting east-west traffic. Due to these natural barriers and to the historically limited trade between the country's northern population centers, Mexico's east-west transportation corridors have not been highly developed in the northern border region.

In addition, Mexico's principal population and industrial centers are due south of Texas along the nation's most extensive overland transportation network. As a result, overland shipments from California to, say, Monterrey or Mexico City are apt to be routed east along the U.S. Interstate highway system or on parallel rail lines before crossing the Texas-Mexico border and entering this corridor.

Further contributing to the distortion of trade data, overland shippers have been making greater use in recent years of a practice more commonly associated with airlines -- the ""hub-and-spoke'' system -- that allows for several clients located over a large geographic area to be served by one centralized distribution or warehouse facility. Other large or frequent shippers send containers to the border region where loads are consolidated, broken or held in inventory for shipment beyond the border to a retail outlet or other buyer.

As a bi-national report on transportation infrastructure issues observed in 1998, the relatively high cost and low-availability of warehouse space in Mexico has led to a significant investment in new space being added north of the border. ""These large, centralized storage points, located north of the border along east-west trade routes such as I-10, fit the distribution needs of both U.S. and Mexican shippers for both U.S. domestic and U.S.-Mexico international trade, particularly in Texas.''

The end result: Texas receives credit for an unknown but probably very significant volume of goods bound for Mexico not merely from California but from much of the rest of the United States. In short, Texas' exports to Mexico are grossly overstated.