Is Technological Change Exacerbated in Recessions? Evidence from Vacancy Postings

Is Technological Change Exacerbated in Recessions? Evidence from Vacancy Postings

We test whether recessions exacerbate routine-baised technological change (RBTC), whereby firms replace routine-task workers with machines and overseas labor, and hire more-skilled workers that complement them, accelerating the job polarization of the U.S. labor market. In doing so we establish a new fact about the demand for skill over the business cycle. Using a new database containing the near-universe of electronic job vacancy postings over the period 2007 and 2010-2015, we find evidence of upskilling - firms demanding more-skilled workers when local employment growth is slower. In a large recession compared to a boom, our estimates imply that firms demand roughly two-thirds of a year more experience and are 50% more likely to ask for a college degree. We find that the vast majority of these effects can be attributed to changes in skill requirements for the same firms and occupations, with only small roles for shifting distributions of the firms that hire and the occupations they seek. We propose upskilling is driven primarily by firm restructuring of production towards more-skilled workers. We show that (1) upskilling is persistent after local economies recover from the great recession adn this elevated skill demand is driven by the same firms that upskilled early in the recover; (2) among publicly trader firms in our data, those that upskill more also increase capital stock by more over the same time period, and (3) upskilling is concentrated within routine-task occupations - those most vulnerable to RBTC. We also rule out a number of cyclical explanations for upskilling. For example our result is unlikely to be driven by firms' opporunistically seeking to hire more-skilled workers in a slack labor market. We thus present the first direct evidence that the Great Recession exacerbated technological change.

  1. How good is this research?

    Assessing the quality of research can be a tricky business. This blog from our editor offers some tools and tips.