NAFTA, Stopler-Samuelson, and Wage Dispersion

Is free trade a win-win propositions with lots of benefits as some suggest? Kash and Josh Bivens beg to differ.

AB readers have to go beyond theory and discuss the evidence from NAFTA. I shall address one piece of evidence shortly but I could not resist commenting on this from Kash:

In fact, all economists learn about this theoretical reality, known as the Stolper-Samuelson theorem, in their very first course in international trade, and rehash it in every other course about trade that they ever take or teach during the course of their career. Despite its theoretically interesting result, let me suggest that the Stopler-Samuelson theorem doesn’t really tell us anything about how harmful trade is.

While Stopler-Samuelson – which is grounded in a simple two-good, two factor of production framework – cannot hope to capture all of reality, let’s think about how its prediction compares to the Ross Perot prediction that the passage of NAFTA would be immediately followed by a “giant sucking sound” of job opportunities fleeing the U.S. First of all, Stopler-Samuelson assumes that good macroeconomic policies assure that we stay at full employment – a reasonable assumption given the good fortune that we had an economic policy team that was actually listened to by policy makers when NAFTA was implemented. But given the fact that Mexico’s capital-labor ratio was lower than that of the U.S., standard factor price equalization logic would hold that U.S. wages would fall, while Mexican wages would rise. Here is where Josh might object:

Trade, on the other hand, is fully predictable in its impacts on the US: it leads to losses for the worst-positioned and benefits for the already well-off.

Given the relative size of our two economies, it would seem reasonable to think that the impact of NAFTA would have been felt most strongly in the Mexican labor market. So did all Mexican workers gain from NAFTA? Gordon Hanson has examined the evidence in What Has Happened to Wages in Mexico since NAFTA? Implications for Hemispheric Free Trade (see also NBER):

In this paper, I examine the impacts of trade and investment liberalization on the wage structure of Mexico. Part one of the paper surveys recent literature on the labor-market consequences of Mexico’s economic reforms in the 1980’s. Mexico’s policy reforms appear to have raised the demand for skill in the country, reduced rents in industries that prior to reform paid their workers high wages, and raised the premium paid to workers in states along the U.S. border. These changes have resulted in an increase in wage dispersion in the country. Part two of the paper examines changes in Mexico’s wage structure during the 1990’s. In the last decade, Mexico has experienced rising returns to skill, which mirror closely wage movements in the United States. There is, however, little evidence of wage convergence between the two countries. Regional wage differentials in Mexico have widened and appear to be explained largely by variation in regional access to foreign trade and investment and in regional opportunities for migration to the United States. I discuss implications of Mexico’s experience for the rest of Latin America in the event a Free Trade Agreement of the Americas is enacted.

An increase in wage dispersion does not sound like a “win-win” to me.