CONCEPT
Job Polarization
The empirical pattern — discovered by Autor and his collaborators — of hollowing-out in the wage distribution: employment growing at the top and the bottom while shrinking in the middle, driven by the automation of routine middle-skill tasks.
Job polarization is the empirical regularity that defined Autor's early career and reshaped labor economics. Beginning in the 1980s, employment and wage growth in advanced economies no longer followed the smooth skill-upgrading pattern that standard models predicted. Instead, a U-shaped pattern emerged: employment grew at the top of the distribution (professional, technical, managerial) and at the bottom (personal services, caregiving, food preparation) while collapsing in the middle (clerical, administrative, production). The mechanism was the task-based framework's central prediction: middle-skill jobs were disproportionately composed of routine tasks susceptible to automation, while jobs at the top and bottom were composed of non-routine tasks that machines could not perform. AI threatens to extend this polarization upward, hollowing out portions of the professional class that were previously protected.
In The You On AI Field Guide
The pattern was first documented in US data by Autor, Lawrence Katz, and Melissa Kearney in a landmark 2006 paper, then confirmed across Western European labor