"Relative Prices and Wage Inequality: Evidence from Mexico"
Robertson, Raymond

"The Impact of Trade Liberalization on Wage Inequality: Evidence from Argentina"
Galiani, Sebastián and Sanguinetti, Pablo

"Inequality and Its Remedies in an Age of Integration"
Larudee, Mehrene

The papers in this panel seek to address the most basic of integration questions: Does trade hurt or harm ordinary workers? Although standard trade theory predicts overall benefits to an integrating economy, there is no question that opening to free trade creates winners and losers. If the losers are those who were doing less well to begin with, then one consequence of trade will be greater inequality. It is not at all clear, though, why lower paid workers would be at a disadvantage under free trade. In fact, trade theory also predicts that unskilled workers in low wage countries should see their wages rise with increased trade, since the factor which the country uses intensively stands to gain. It is something of an anomaly, therefore, that the period of trade liberalization in Latin America coincided with a deterioration in the wages of unskilled workers relative to skilled workers. One explanation, mentioned in the papers by Raymond and Mehrene, is that Latin American countries are not intensive in unskilled labor, at least compared with developing countries in other parts of the world, and, in particular, with China. Another possibility raised in the papers by Pablo and Mehrene, and central to discussions of patterns of wage inequality in the United States, is that trade may not be the only, or even the crucial, factor determining inequality.

I wish to commend all of the panelists for their careful and thoughtful research. The papers complement each other nicely and I encourage the authors to borrow each other's ideas. In particular, Raymond's presentation would benefit from presenting direct evidence of import penetration by skill content, mirroring Pablo's calculations of sectors most exposed to trade competition. In turn, Pablo might consider the variation in skill intensity among manufacturing sectors, as does Raymond. The distinction between Pablo's and Raymond's more narrowly focused studies and Mehrene's more wide-ranging research suggests another area for cross-fertilization: it is important for those of us concerned with equity issues to recognize that wage inequality is just one of many welfare measures. I thought it was instructive that the audience pressed for more general conclusions about trade and equity: did trade increase or decrease poverty? Are absolute wages (as opposed to relative wages) higher or lower than they would be under a different trade regime? With Pablo's and Raymond's (and my own) methodological approach, each of these questions requires its own extensive research effort and a narrow focus allows us to draw inferences that are more certain and less speculative than a broader approach. It is important, however, to produce careful empirical work in the context of a broad research agenda.

I now take up each paper in turn. Raymond asks whether relative prices determine relative wages and if so, when? As dramatically illustrated in figure 8b, he finds a close correlation between prices and wages, which emerges after a three to five year lag. The implication is that when the price of a certain good rises, the type of labor used to produce that good (i.e., skilled or unskilled labor) will benefit.

Raymond also asks whether trade liberalization can explain movements in relative prices. Raymond's thesis is that in joining GATT in 1985, Mexico exposed its economy to low wage competition from other developing countries. Compared to these countries, Mexico was relatively skilled, and prior to 1985, had a tariff structure to protect relatively unskilled industries. Upon joining GATT, tariffs (and, therefore prices) fell more for unskilled goods. These points are well illustrated in figures 3-5.

Raymond then argues that NAFTA represented an opening to the United States, compared to which Mexico is certainly less skill-intensive. Mexico's more skill-intensive industries presumably faced more competition under NAFTA and this would lead to a decline in their relative prices. The data, however, do not support this story. In figure 3, relative prices appear to level off "too early"- in 1993. And, in fact, other Latin American countries also appear to level off, at least in terms of relative wages (Londońo and Székeley 1997). Furthermore, table 2 shows contradictory results on price changes by skill category under NAFTA. Thus while the data support the connection between prices and GATT, the connection between prices and NAFTA is doubtful. My hunch is that after a dramatic liberalization like GATT, further tariff reductions just don't matter as much. It is possible that the pattern of rising and then leveling relative prices traces the adjustment path after trade liberalization. Trade theory offers little in anticipating this adjustment path. If, for example, the main price changes are driven by tariff reduction, why wouldn't we expect to move to a new equilibrium with relatively stable prices, consistent with what we see in figure 3? In other words, could the 1980s rise in relative prices be the adjustment period and the 1990s leveling off be the new equilibrium? These considerations also suggest that, perhaps, NAFTA, and by extension, FTAA, don't matter all that much, at least in terms of relative prices, and in the wake of a dramatic (GATT) liberalization.

Sanguinetti also asks what role trade liberalization has played in wage inequality, for the case of Argentina. Sanguinetti argues that because unskilled workers may not be perfectly mobile among manufacturing sectors, those in industries that faced more import competition should have experienced worse wage outcomes. Skilled workers, on the other hand, have more portable skills and can cushion themselves from a declining sector by finding work elsewhere. Thus we would expect that trade competition would hurt the unskilled more than the skilled, and lead to more inequality. Sanguinetti tests this hypothesis by estimating the effect of import penetration by industry (measured as the ratio of imports to domestic gross value added) on wages of skilled and unskilled workers. Table 7 shows that relative wages of skilled workers rose 50% between 1992 and 1997. Curiously, wages rose more for both skilled and unskilled workers in industries that faced more import competition, as shown in table 8. Since skilled wages rose more, Sanguinetti concludes that trade widened inequality. The contribution to overall inequality, however, is small: Sanguinetti finds that only 10% of the overall increase in inequality results from trade. The finding that trade can explain only a small portion of overall inequality is common in similar studies using United States data. And while I believe that it is important to investigate other potential explanations of inequality (for example, my own work suggests the importance of labor supply), it is also highly likely that trade liberalization affects the wage structure in indirect and hard to measure ways. Rodrik (1997), for example, points out that workers may face competition even in industries that face little explicit import competition, if, for example, wages are suppressed to keep prices low and, correspondingly, keep importers at bay. Similarly, the threat that companies can relocate to lower-wage countries may also restrain the wage demands of unskilled workers.

Mehrene's paper takes a broader approach than the others in considering equity and integration. Equity, she argues, encompasses much more than wage inequality among workers: we need to consider inequality between labor and capital and power relations between local jurisdictions and multi-national corporations. Although Mehrene does not investigate these topics directly, they deserve careful consideration. It is a failing of mainstream economics that we do not often apply our painstaking methodology to questions concerning economic power. We should. Mehrene covers several other topics that are more readily measured, in particular, the impact of trade on manufacturing and the maquiladora sector. A tension runs through her discussion: sometimes it seems as if NAFTA doesn't matter; other times it seems that it does. For example, Mehrene points out that compared with the dramatic trade liberalization surrounding GATT, tariff reductions in NAFTA were small potatoes. Moreover, an acceleration in maquiladora employment after NAFTA can be explained by "the traditional sensitivity of maquila employment both to fluctuations in US GDP and changes in the dollar value of the maquila wage" (p. 8). At the same time, Mehrene argues that Mexico's preferential status compared to other developing countries was responsible for growth in apparel and electronics. The evidence for this line of reasoning is the decline in apparel employment which coincided with the extension of Mexico's preferential status vis a vis the US to Caribbean Basin countries in 2000. However, since electronics also declined at this time, it is unlikely that increased competition from the CB only can explain the decline in maquiladora employment. It seems more likely that NAFTA, as I argued earlier, on the heels of GATT does not have much effect in the trade arena. But, as Mehrene points out, the non-trade provisions of NAFTA and the proposed FTAA may matter more, especially in terms of the balance of economic power.

References

Londońo, Juan Luis and Miguel Széleky. 1997. "Persistent Poverty and Excess Inequality: Latin America 1970-1995." Inter-American Development Bank Office of the Chief Economist Working Paper #357.

Rodrik, Dani. 1997. Has Globalization Gone Too Far? Washington DC: Institute for International Economics.