Column (1) shows the regressions without industry dummies; column (2) those with industry fixed effects. Overall the estimates confirm the patterns suggested by the separate regressions. A reduction in the industry tariff or in quota coverage increases the skill gap. But this reflects primarily relative shifts in wages across industries. When industry dummies are included in the regression, the effects of tariffs and quotas become insignificant. The coefficients on productivity per worker and on capital per worker, on the other hand, are positive and significant in both specifications. Because these two variables are fairly correlated, they are probably picking up the same thing: capital and technology (or technology-induced productivity growth) are more complementary to skilled labor than to unskilled labor.
In this chapter we have examined recent shifts in relative wages within Mexico and between Mexico and the United States and have attempted to link them, directly and indirectly, to the effects of trade liberalization.
Overall, the results suggest that by itself, the removal of trade protection had a negative impact on the wage of Mexican workers relative to their U.S. counterparts. A 10 percent reduction in tariff and quota protection appears to have been associated with reductions in relative wages of 3 to 4 percent. However, to the extent that trade liberalization allowed for imports of new capital and technology and induced productivity growth, it may have yielded increases in the relative wage. The impact of productivity growth seems particularly important in explaining increases in relative wages over time, within industries.
These two facts combine to provide a plausible explanation for rising skill differentials within Mexico. On the one hand, the reduction of trade barriers seems to have had a larger negative impact on demand for unskilled labor than on demand for skilled workers, pushing unskilled wages down relative to skilled ones. At the same time, productivity growth has been concentrated in more-skill-intensive sectors, raising the relative demand for skilled labor and hence its relative wage. Within industries too, productivity growth appears to have been associated with rising skill differentials.
The data used in the chapter come from three different sources: