As part of the series of the „Finance Research Seminar“, VGSF welcomes Mireia Giné from IESE Business School, University of Navarra to present her research paper.
Personal Webpage
Paper
AI Is Already Eroding Wages: Quasi-Experimental Evidence From Occupational Exposure
joint with José Azar and Javier Sanz-Espín
link to the paper
We provide causal evidence on how generative AI is reshaping the internal wage structure of the firm. Using a large-scale dataset of realized worker-firm matches for 138 million U.S. workers, we document that the arrival of ChatGPT significantly reduced the wages of firms whose ex-ante occupation structure was more exposed to automation relative to less exposed firms. In particular, we find that the average wage of exposed firms fell 4.5% relative to unexposed firms. We also find that starting wages of new positions for junior and mid-level workers in exposed firms fell by 6.3% and 5.9% respectively relative to unexposed firms, while senior wages remained stable or even increased. We show that firms in exposed sectors reduced the share of junior employees, and increased the share of mid-level employees. Finally, we find that firms exposed to AI experienced a reduction in the level of education of new positions for junior and mid-level workers, relative to less exposed firms.
Back to overview
