Broadness summary and data

Occupational broadness

We know from various studies that more human capital is occupation specific than industry specific. In Occupation-industry mismatch in the cross-section and the aggregate, I define an occupation's broadness by how many industries it is employable in. I show that there is significant heterogeneity: some occupations are specialized in few industries, while some are very dispersed. If indeed human capital is more occupation specific than industry specific, the same industry-specific shock will have different implications for workers in broad and specialized occupations. While the latter are exposed to the industry shock and can only change occupation (which is costly), the former can change industry while remaining in their occupation (less costly). I find evidence for this in the United States.

Broadness as a mismatch indicator

The general equilibrium model developed in that paper suggests that the average broadness of the unemployed (as indicated by their current/last occupation) is a useful measure of mismatch unemployment. Intuitively, when the pool of the unemployed consists more of workers from specialized occupations, the possibility of mismatch is high: these workers are less able to respond to industry-specific shocks. As a consequence, the risk of mismatch across industries (misallocation of labor demand vis a vis labor supply) is higher.

The figure above displays the average broadness of the unemployed for the United States. Note that it is much lower during the Great Recession than during previous recessions, suggesting a heightened risk of mismatch.

Data

• 3 digit industries and higher aggregation of industries
• Broadness computed each year, once over all years,