Analyzing the U.S. Content of Imports and the Foreign Content of Exports (2006)

THE COMMITTEE HAS CONSIDERED two content questions: “What is the foreign content of U.S. exports?” and “What is the U.S. content of U.S. imports?” The committee concluded that trying to address these content issues directly would be impractical. However, proxy estimates using input-output data gathered by the U.S. government is possible, and the calculations described in this report and those carried out by Hummels, Isihii, and Yi (2001) have taken a large step toward answering those content questions. However, the accuracy of the answers is limited because of various problems with the data and limitations in the validity of the assumptions that have to be used. In addition, because of the lack of data from other countries, it is not possible to estimate the U.S. content of imports to the United States.

Given that estimating the foreign content of U.S. exports is based on the data in input-output tables, it is reasonable to suggest that the content measure could be improved with greater detail, accuracy, and granularity in the numbers that the Bureau of Economic Analysis (BEA) gathers for input-output tables. But would the effort to gather these finer details of data be worth it? Answering this question depends on whether the measurement of the U.S. content of exports from and imports into the United States helps to answer the important underlying questions: How has international trade affected the demand for U.S. labor? Would finer data provide much information about other economically important factors, such as the budget and trade deficits?

When thinking about the content question, one needs to firmly keep in mind that a competitive system is constantly looking for ways to lower costs and improve efficiency. When a farmer sells wheat and it comes back to the farm as bread, the farmer has made the decision that it is more efficient to have someone else mill the wheat into flour and bake the bread. What is true for individuals is true for countries as well.

In answering the content question, there appears to be a presumption that foreign content in U.S. exports and U.S. content in U.S. imports are problematic and that the United States is losing out in global competition. But these content measurements do not tell anyone whether the United States is becoming less competitive, and they do not provide information on what is happening to U.S. jobs. To help make this clear, consider the following three scenarios, all of which involve the offshoring of U.S. parts manufacturing to Mexico.

The starting position is that $1,500 of U.S. widget exports are produced completely using U.S. content.

• Scenario 1: Offshore sourcing hurts low-skilled U.S. workers. The price of Mexican parts for widgets declines so U.S. widget makers shift widget parts production to Mexico and stop U.S. parts operations: $1,000 of the $1,500 of U.S. widget exports now come from Mexican parts. In this scenario, the foreign content of widgets rises, and U.S. low-skilled widget workers are displaced.

• Scenario 2: Offshore sourcing maintains wage and employment levels of low-skilled U.S. workers. The tight U.S. economy raises the wages of low-skilled workers and draws low-skilled widget workers out of the widget industry and into other U.S. businesses. In the absence of Mexican parts, the U.S. widget industry would go out of business since the costs of low-skilled labor are too high in the United States and no one wants widgets at such a price. But the U.S. widget industry shifts parts operations to Mexico so that $1,000 of each $1,500 of widgets comes from Mexican parts. In this scenario, there are no adverse effects on low-skilled U.S. workers, and, in fact, other widget workers benefit and keep their jobs because of access to cheap Mexican parts. Workers with similar skills to those of the non-parts widget workers are winners in the United States from offshoring widget parts to Mexico; low-skilled U.S. workers are unaffected.

• Scenario 3: Offshore sourcing changes U.S. production technology and leads to skill upgrading of U.S. labor. The price of Mexican parts for widgets declines, and the U.S. widget industry reorganizes production to take advantage by upgrading the jobs of less-skilled U.S. parts workers to be “coordinators” and “supervisors” of Mexican parts production. The Mexican content of U.S. widget exports rise, and U.S. widget exports expand as a result of the reorganization, with all U.S. workers benefiting.

In all three scenarios, the Mexican content of U.S. widget exports increases, but the implications for U.S. workers are quite different. In making this statement, we express no position with regard to the actual effects that competition from Mexico is having on U.S. jobs, which continues to be a matter of debate. Our point is the narrow one: measuring the content more accurately will not help to determine the effects of the offshoring of widget manufacturing jobs to Mexico.

The increasing exposure of the U.S. economy to international market forces through trade and financial integration has coincided with significant changes in the employment and wage patterns of Americans and also with a widening external trade deficit. Are these merely coincidental or are they related in some important ways? Would more accurate answers to the content question help to answer this important question?

These are important questions that are not easy to answer. Since the external trade deficit has to be financed with loans from abroad, answers to the “why the deficit?” question can be found on either the goods side of the balance of payments accounts or on the financial side. Because offshoring as a phenomenon can take place in an economy that is showing deficits, surpluses, or a balance of trade, and because the drivers for the deficit are many and complex, answering the content question—the measurement of content—does not provide any helpful insight into the causes or the consequences of the trade deficit.

Along with increasing volatility of earnings, disparities in the economic fortunes of Americans have increased significantly in recent years (see Mishel, 2005; Attanasio, 2004). Everyone can agree that better understanding is needed of these important changes in the labor markets, but it is difficult to accurately estimate the effects of offshoring on these evolving

aspects of the U.S. labor market. A whole host of other effects—such as the evolution of shorter product cycles, rapid technological change, the availability of more flexible technologies and an increase in the variety of international supply routes—may also result in the shortening of any given product life-cycle at a production location, potentially leading to greater employment volatility. A simple calculation to answer the content question is not a metric for any of the potential drivers of wage disparities, labor volatility, or the other threats to the standard of living in the United States.