Employment Effects of Generative Artificial Intelligence: Evidence from Novel Firm-Level Adoption Data

Rigorous evidence on generative artificial intelligence’s (AI) labor market effects remains limited, partly due to substantial constraints in measuring actual firm adoption. Economists Germán Reyes and Joaquín Serrano will use novel proprietary data from Ramp that directly captures AI expenditures across more than 25,000 businesses to answer four research questions: 1) How rapidly is generative AI spreading across industries and labor markets? 2) What are the causal effects of AI adoption on local labor market outcomes such as employment, wages, hiring, and separations? 3) Does AI adoption widen earnings inequality and intensify job polarization across occupations? 4) Which workers bear the greatest adjustment costs? Reyes and Serrano will aggregate the AI adoption data to the commuting-zone level and link them to administrative records on employment, wages, and worker flows. Their identification strategy exploits variation in the timing and intensity of adoption across commuting zones and industries.