Occupational Exposure to Capital-Embodied Technical Change

Abstract

We study differences in exposure to factor-biased technical change among occupations by providing the first measures of capital-embodied technical change (CETC) and of the elasticity of substitution between capital and labor at the occupational level. We document sizable occupational heterogeneity in both measures, but quantitatively, it is the heterogeneity in factor substitutability that fuels workers’ exposure to CETC. In a general equilibrium model of worker sorting across occupations, CETC accounts for almost all of the observed labor reallocation in the US between 1984 and 2015. Absent occupational heterogeneity in factor substitutability, CETC accounts for only 17 percent of it (JEL I26, J16, J24, J31, O33)

Technology and the Task Content of Jobs Across the Development Spectrum

“…Beaudry et al (2016) showed that the decline of middle-skill jobs was first mirrored by the rise of high-skill jobs but after 2000 increasingly by the rise of low-skill jobs. Autor et al (2003), Aum (2017) and Aum et al (2018) point to the role of computerization and Caunedo et al (2019) to the technological change embodied in a broader set of tools that workers use.…”

“…For the US, the O*NET lists tool requirements for each occupation. When merged with data on capital stock and prices, this can be used to measure workers' exposure to capital-embodied technological change (Caunedo et al, 2019). It is desirable to collect harmonized data across countries on tool requirements by occupation, as well as time series of capital stock and quality-adjusted prices of capital goods.…”

The Cost of Regulatory Compliance in the United States

“…Nevertheless, alternatives have been devised in the literature based on tools and capital equipment use. Here we follow Caunedo et al (2023) and construct an approximate measure of capital cost per employee for each occupation, based on the occupation's use of tools and equipment from the O*NET Tools and Technology module and from the NIPA quantity and price tables for each equipment item. With this additional information available, one is able to sum up capital user costs per employee and annual wages per employee for each occupation, obtaining the total input cost (in terms of labor and capital) for each employee in each occupation.…”

“…As discussed in Section 3, precisely estimating firms' use of capital for regulatory compliance purposes is challenging, mostly due to the lack of detailed data on various capital user costs at the micro-level in the OEWS. Nonetheless, following Caunedo et al (2023) and focusing on O*NET tools and capital equipment requirements of occupations, we are able to construct an approximate measure of share of regulatory compliance costs out of the total labor and capital costs in Section 3. This measure, RegIndex tot i,t , is employed in replicating Figure 5.…”

“…After discussing labor costs, we extend the analysis by including establishments' capital equipment and tools expenditure related to regulatory compliance. Capital equipment costs are measured following the approach presented in Caunedo et al (2023). We used this methodology to create an extended version of regulatory costs including capital to generalize and assess the robustness of our findings.…”