Multinationals Do Not Appear to Export U.S. Jobs

09/01/1997
Summary of working paper 5958

... employment at affiliates in developing countries is very sensitive to wages in other developing countries, but employment at the parent responds very little when foreign affiliate wages fall.

Critics have alleged that U.S. multinational corporations (MNCs) export U.S. jobs when they expand abroad. For instance, several groups campaigned against NAFTA on the grounds that it would encourage U.S. plants to locate south of the border and to substitute cheap Mexican labor for U.S. workers. Now, new NBER studies by Lael Brainard and David Riker find that employment at foreign affiliates of U.S. MNCs substitutes only modestly for employment at the U.S. parent, and that there is much stronger substitution between workers at affiliates in low wage locations. The authors observe that the pace of activities performed by affiliates with different workforce skill levels within the same geographic region tend to increase or decrease together. This suggests that companies divide production and other activities among countries whose workforces have different skill levels: workers in developing countries, for example, compete with each other to perform the activities most sensitive to labor costs. In other words, employment shifting takes place predominantly between offshore affiliates in less developed countries.

In Are U.S. Multinationals Exporting U.S. Jobs? (NBER Working Paper No. 5958), Brainard and Riker estimate that when wages fall 10 percent in developing countries like Mexico, employment at the U.S. parent falls 0.17 percent, while affiliates in other developing countries, such as Malaysia, lay off 1.6 percent of their workforce. That is, employment at affiliates in developing countries is very sensitive to wages in other developing countries, but employment at the parent responds very little when foreign affiliate wages fall, they conclude.

The data in this study come from the Annual Survey of U.S. Direct Investment Abroad, which is administered on a mandatory basis and audited by the Bureau of Economic Analysis. Each firm's production activities in up to 90 countries are tracked over a ten-year period ending in 1992. The authors include in their sample all firms whose parent industry is in the manufacturing sector.

Using the same data, Brainard and Riker ask whether the globalization of production has led to a diminishing demand for less skilled workers in industrialized countries. Although the relative growth of aggregate employment in U.S.-owned MNCs appears to support this concern (from 1983-92, employment increased 11 percent in developing countries, decreased by 3.5 percent in industrialized countries outside the United States, and fell 12.5 percent in the United States) these aggregates misrepresent reality. Only one third of U.S.-owned MNCs have any production activities in developing countries; of that subset of firms, employment in industrialized country affiliates actually increased 13 percent between 1983 and 1992.

In U.S. Multinationals and Competition from Low Wage Countries (NBER Working Paper No. 5959), Brainard and Riker note that although the global scale of production of these multinational firms has fluctuated over time, the proportion of employment located at industrialized and developing country affiliates has remained relatively fixed. Workers at locations with different levels of development are likely engaged in complementary activities, the authors suggest, and there is evidence of this complementarity in the electronic components, food products, glass products, plastic products, and other industries. But the authors find that the labor demands of affiliates in the resource-intensive chemical industries, for example, are not linked internationally.

They conclude that within MNCs, labor demand in each affiliate is related to the cost and demand conditions of other affiliates owned by the same firm. Output levels of affiliates in developing and industrialized countries tend to move together, possibly reflecting a division of production activities by skill requirements. For instance, in the electronic components industry, a 10 percent decline in affiliate wages in a developing country leads to a 1.9 percent increase in affiliate employment in industrialized countries, and a 3.7 percent increase in local affiliate employment in that country, while reducing affiliate employment in other developing countries by 6.3 percent.