Occupational Exposure to Capital-Embodied Technical Change

Much of technological progress is embodied in capital and realized through the availability of new, more efficient, capital goods. How does capital-embodied technical change (CETC) impacts the labor market? We draw on the idea that workers’ occupations are a suitable unit of analysis to address this question and study workers’ exposure to CETC at the occupational level. To do so, we construct a novel dataset of the stocks of different types of capital used for production in each occupation. Our dataset yields the first available estimates of CETC and of the elasticity of substitution between capital and labor in each occupation. The latter determines occupational heterogeneity in workers’ exposure to CETC, as summarized by the cross-price elasticity of labor demand to the cost of capital. We document that exposure to CETC is U-shaped across occupations ranked by their skill requirements, and jointly with our measures of CETC in each occupation, it rationalizes the occupational employment gains and losses in the US over the last 40 years. We evaluate the impact of CETC in a general equilibrium model of endogenous sorting of workers of different demographic characteristics across occupations of different CETC and equilibrium exposure. We find that CETC explains 91% of labor reallocation in the US between 1982 and 2015. It is responsible for almost all of the gains in employment in high-skill occupations and for 74% of the employment losses in middle-skill occupations.