Do Maquiladoras Take American Jobs? Some Tentative Econometric Results
Mexico's maquiladora industry has been subject to many controversies over the years, but one controversy has dominated. Opponents of the U.S. legal provisions that facilitate maquiladora profitability argue that the maquiladoras take jobs from U.S. workers. That is, Mexican workers compete with U.S. workers for manufacturing jobs. Advocates of the maquiladoras argue that these jobs would have gone abroad, in any case, and that Mexican workers' real competition is the work forces of other third-world countries.
Even though the two sides of this controversy have argued for much more than a score of years, no one has developed a statistical testing procedure to address the dispute econometrically. The reason is that standard hypothesis-testing techniques cannot be used to treat this issue. The most obvious variables to use in testing the two competing hypotheses are so multicollinear that any conventional model using them always generates highly suspect results.
In this article, I present a testing method that mitigates some of the problems of multicollinearity and allows one to examine the two competing hypotheses about maquiladoras and U.S. jobs. The test is indirect, since it measures the responses of maquiladora employment to changes in Mexico-U.S. and Mexico-Asian third world wage differentials, and does not directly measure losses in U.S. jobs or directly resulting increases in Mexican or Asian employment. Moreover, the results of the testing procedure are highly tenuous, because very few observations with clean data are available. Accordingly, it is hard to know much about the intertemporal stability of the relationships implied by the procedures I use. Nevertheless, the results up to now are striking. They suggest that both sides of the controversy are about equally correct.
All rights reserved
Print ISSN: 0886-5655
Online ISSN: 2159-1229