The (Short-Term) Effects of Large Language Models on Unemployment and Earnings

2025-09-18

AI Summary

This study investigates the short-term labor market impacts of large language models (LLMs), focusing on the causal relationships between occupation-level LLM exposure and earnings and unemployment. To address the coarse granularity of industry-based classifications, we develop a task-based occupational LLM exposure index and employ a synthetic difference-in-differences (Synthetic DID) design. Empirical results show that workers in high-exposure occupations experience statistically significant wage increases, whereas unemployment rates exhibit no significant change—indicating that early LLM adoption drives an “income redistribution effect” rather than displacement-driven unemployment. These findings challenge the prevailing narrative of technological unemployment and reveal an asymmetric dynamic in AI’s labor market effects: skill premiums emerge before job losses materialize. The study thus provides novel empirical evidence and a methodological framework for analyzing the short-run labor economics of generative AI.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Philosophy and Ethics of AI: AI & Jobs/Labor

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Large Language Models have spread rapidly since the release of ChatGPT in late 2022, accompanied by claims of major productivity gains but also concerns about job displacement. This paper examines the short-run labor market effects of LLM adoption by comparing earnings and unemployment across occupations with differing levels of exposure to these technologies. Using a Synthetic Difference in Differences approach, we estimate the impact of LLM exposure on earnings and unemployment. Our findings show that workers in highly exposed occupations experienced earnings increases following ChatGPT's introduction, while unemployment rates remained unchanged. These results suggest that initial labor market adjustments to LLMs operate primarily through earnings rather than worker reallocation.

Problem

• Examining short-term labor market effects of LLM adoption
• Comparing earnings and unemployment across occupation exposure levels
• Assessing impacts through Synthetic Difference in Differences approach

Innovation

• Synthetic Difference in Differences approach
• Comparing earnings across occupation exposure levels
• Analyzing unemployment rates post-ChatGPT introduction