David Autor | Does Automation Replace Experts or Augment Expertise? The answer is YES.
Does job task automation augment or diminish the value of labor in the tasks that remain? We argue that the answer depends on whether removing tasks raises or reduces the expertise required to perform the bundle of tasks that remain. Because the tasks that are expert in one occupation may serve primarily as supporting tasks in another, automating any given task may simultaneously replace experts in some occupations and augment expertise in others. We offer a model highlighting that, opposite to conventional demand forces, changing expertise requirements have countervailing effects on wages and employment: falling expertise requirements reduce wages but permit entry of inexpert workers; rising requirements increase wages but reduce the set of qualified workers. We operationalize these ideas by constructing a content-agnostic measure of task expertise grounded in the efficient (linguistic) coding hypothesis, which is used in conjunction with a method for identifying tasks removed from and added to occupations over four decades.
Our analysis confirms the predicted countervailing effects of changing expertise requirements on occupational wages and employment; documents that changes in task content are empirically distinct from changes in task quantities and, moreover, have countervailing effects on employment; and shows that routine task automation has bifurcated occupational expertise demands by lowering wages in occupations where routine tasks were relatively expert and raising wages in occupations where routine tasks were relatively inexpert. Our framework provides a general tool for analyzing how the removal and addition of specific job tasks may reshape the scarcity value of human expertise within and across occupations.
Details:
Time: 12:00 pm - 1:00 pm PT
Location: Gates Computer Science Building, Room 119, 353 Jane Stanford Way, CA 94503
