Researcher ORCID Identifier

0009-0004-9090-0973

Graduation Year

2026

Date of Submission

4-2026

Document Type

Campus Only Senior Thesis

Degree Name

Bachelor of Arts

Department

Economics

Reader 1

Nishant Dass

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Rights Information

© 2026 Luke V Torrey

Abstract

This thesis explores how and what ways artificial intelligence (AI) reshapes labor market outcomes using a dataset of more than 1.8 million worker records collected from PayScale through the years 2021 to 2025. I build a novel Job-Level AI Exposure Index that separates this out into AI exposure, the technical feasibility of applying an AI system to some set of job tasks, and AI adoption, which is where firms actually invest in and deploy these technologies. I apply this framework to estimate industries clustering in four quadrants of relative exposure and adoption, uncovering considerable heterogeneity in the manner AI is influencing work across sectors. The analysis generates three main findings. First, wages in AI-exposed occupations are at a premium, if non-monotonic: controlling for constant-intensity adoption level, high intensity adopters pay shape the highest premium but medium-exposure workers earn the most once higher adoption is controlled for, in support of AI complementing rather than replacing intermediate-skill workers under greater intensity of opening. Second, though wage effects are substantial, differences in job satisfaction and worker well-being by AI exposure tiers can be neglected, a phenomenon I call the satisfaction paradox. Third, firm-level adoption decisions mediate a large share of the aggregate inequality associated with AI: high-adoption firms pay substantially higher wages to comparable workers at all exposure levels. Taken together, these findings imply that AI’s labor market effects are more about firm adoption decisions than occupational task content of work, with significant implications for workforce policy and educational investment.

Download

Available for download on Friday, April 14, 2028

This thesis is restricted to the Claremont Colleges current faculty, students, and staff.

Share

COinS