For the middle period 1973-1979, the OLS price regressions again mandate a decline in wage inequality for both all industries and just manufacturing industries. Analyzing supporting data on outputs, endowments, and net exports, BC conclude that endowment changes were not as strong an influence this period. The three-labor-type regressions for manufacturing only suggest that trade might have helped lower the warranted wage of high-school dropouts relative to high-school graduates. The weighted regressions for this period look qualitatively similar for all industries together but qualitatively different for just manufacturing. The manufacturing weighted estimates tend to indicate a mandated rise, not fall, in wage inequality–particularly when outputs are the weights. payday loans by phone
Finally, for the period 1979-1991 BC estimate different mandated-wage patterns across the two industry groups. For all industries together BC find a mandated rise in wage inequality. For just manufacturing they find a mandated decline in wage inequality (particularly so for the two cases with only two labor types). These manufacturing results are matched by the regressions using the BLS export and import prices. In light of the fact that the manufacturing results suggest that international trade was generating mandated declines in wage inequality rather than the actual rises in inequality (p. 40), BC argue that the most-likely explanation for the observed price patterns is a combination of skill-biased technological change and demand shifts toward skilled-labor-intensive products. The results from the weighted price regressions look qualitatively similar, although the manufacturing results are weaker when weighted for the cases with only two labor types.