Recommending Career Transitions to Job Seekers Using Earnings Estimates, Skills Similarity, and Occupational Demand

Abstract

We describe a career recommendation algorithm that uses government administrative data to help job seekers discover careers transitions that have been successful for workers who have similar prior experience and other background characteristics. Algorithm development was motivated by workers and contractors who were displaced by the COVID-19 economic crisis. Traditional job boards available through state government websites list all available jobs but do little to remove uncertainty associated with starting a new career. Our algorithm lowers this uncertainty and aims to increase successful career transitions for individuals in declining industries. We use causal machine learning models and government administrative data on the universe of individual-level employment histories and earnings to identify career transitions that have resulted in increased earnings for previous job seekers. We combine these estimates with measures of skill similarity across occupations, derived from natural-language processing of millions of full-text job descriptions, and with occupational demand, as measured by nightly job posting volume. To date, our algorithm has been implemented in production workforce development systems in five U.S. states.

Keywords: Recommendation Systems, Future of Work, COVID-19, Unemployment Insurance