BC describe aggregate price movements between more-skilled and less-skilled categories in Table 1. For each of the Census of Manufactures years plus the year 1980, BC calculate the domestic price ratio of industries intensively using more-skilled workers to industries intensively using less-skilled workers. They construct this price ratio for two industry samples: all sectors and just manufacturing. Within each sample, each industry is placed in either skill category based on whether its total employment ratio (i.e., ratio of direct employment plus indirect calculated from input-output tables) of workers with 13-plus years of education to workers with 12 or fewer years is above or below the median industry’s ratio. From 1967 to 1972 both price ratios fell. From 1972 to 1980 the price ratio for all sectors continued to fall but the manufacturing price ratio rose. Finally, from 1980 to 1992 the price ratio for all sectors rose sharply while the manufacturing price ratio declined. Notice these descriptive price movements within manufacturing generally match those found by Leamer: during the 1970s the relative price of unskilled-labor-intensive products fell, but during the 1980s this relative price did not fall further. payday loans lenders
BC report their main OLS estimation results in Tables 3 and 7; Appendix 1 reports analogous results for the weighted regressions. Again, their overall methodology is first to estimate equation (BC) for each time-period/factor set combination and then to discuss what combinations of changes in trade, technology, and endowments most plausibly explains the observed price patterns and implied mandated factor-price changes indicated by bi.
For the period 1968-1973, OLS price regressions for both all industries and just manufacturing industries and all three factor sets mandate a decline in wage inequality. BC argue that this implied decline in the relative price of skilled-labor-intensive products was caused primarily by an expanding relative endowment of skilled labor which in turn expanded the relative output of skilled-labor-intensive products. The weighted regressions yield the same qualitative result.