Low-skilled Occupations Face the Highest Re-skilling Pressure

Substantial scholarship has estimated the susceptibility of jobs to automation, but little has examined how job contents evolve in the information age as new technologies substitute for tasks, shifting required skills rather than eliminating entire jobs. Here we explore the patterns and consequences of changes in occupational skill contents and characterize occupations and workers subject to the greatest re-skilling pressure. Recent research suggests that high-skilled STEM and technology-intensive occupations have experienced the highest rates of skill content change. Analyzing 727 occupations across 167 million job posts covering the near-universe of the U.S. online labor market between 2010 and 2018, we find that when skill distance is accounted for, re-skilling pressure is much higher for low-skilled occupations, no matter how ``low-skill'' is defined, either by skill number, pay level, or education degree. We investigate the implications of uneven occupational skill change on workers and find that those from large labor markets and large employers experienced less change, while non-white males in low-skill jobs are the most demographically vulnerable. We conclude by discussing the broad potential of our skill embedding model, which learns skill proximity from skill co-presence across job posts and represents it as distance in the high-dimensional space of complex human capital that corresponds with skilling costs for workers. This model offers a fine-grained measure of the extent to which jobs evolve, and also indicates in what direction job are evolving, as illustrated by the decline in demand for human-interface skills and the rise for those at the machine-interface.