In both equations (L-T) and (L-G) Leamer interprets the parameter estimates as “mandated” factor-price changes. “These are the changes in factor costs that are needed to keep the zero profits condition operative in the face of changes in technology [i.e., Bit] and product prices [i.e., Big]” (p. 23). Stated differently, the regressions estimate the factor-price changes mandated by changes in technology and/or product prices to maintain zero profits in all sectors. The error term in each equation allows the zero-profit conditions not to bind exactly (for whatever unspecified reasons). Because the econometric fit is not perfect, the mandated factor-price changes can be interpreted as the changes “consistent with the least change in profits in the economy” (p. 29). These mandated factor-price estimates Bit and Big can then be compared with actual factor-price changes. “If the two conform adequately, we will argue that we have provided an accurate explanation of the trends in wages” (p. 23). Thus the parameter estimates are “tested” by directly comparing them with actual data. Overall, these regressions can be interpreted as an accounting exercise. What changes in factor prices are mandated from the observed changes in technology and/or product prices–and what share of actual changes do these mandated changes explain? debit card payday loans
In his regressions intermediate inputs are measured as materials purchased. For primary factors he uses capital and either labor or labor disaggregated between more-skilled and less-skilled workers. Leamer measures skills two different ways. One is the classification between nonproduction and production workers. Concerned that this measure excessively misclassifies actual skills, Leamer constructs an alternative skill ranking of industries based on their average earnings per worker (higher average earnings are translated into a higher mix of more-skilled workers). Changes in product prices and technology are the annualized changes over the decades defined above, and cost shares are measured for the first year of each decade. In equation (L-T) Leamer assumes that technological progress affects the prices of primary factors only. In equation (L-G) Leamer constrains the parameter estimate on intermediate inputs to equal the actual observed price change for intermediate inputs; thus, he accounts for inputs by following equation (3’’). Finally, Leamer weights industries by either employment or value-added.