Tools of the trade: AI and the restructuring of the experience premium
Eric Shaunn Mattingly, Steven James Hyde, Andrew Steven Manikas · International Journal of Manpower · 8 Oct 2026
Where this sits in the debate
- Will AI destroy more jobs than it creates?
- Will AI widen inequality? · its finding supports the claim (why studies differ)
Direction of the finding as read from the abstract by an AI model (AI-written).
Key takeaway AI-written
Using 337 million US job postings from 2017 to 2026, the authors find that industries more exposed to AI saw the pay gap between experienced and entry-level postings widen, by about $2,780 per percentage point of exposure. Entry-level posting shares fell only modestly and not more in exposed industries; most of the widening comes from a shift toward higher-paid occupations.
Abstract
Purpose We examine how artificial intelligence (AI) reshapes the relative economic positioning of entry-level and experienced workers within industries. Rather than focusing on job elimination, we investigate whether AI changes the returns to experience. Design/methodology/approach We construct an industry-quarter panel from 337 million US Lightcast job postings across 16 NAICS2 industries, with finer-grained inference and decomposition across 86 NAICS3 industries (Q1 2017 to Q1 2026). Difference-in-differences models use a leave-own-industry-out shift-share measure, with industry and quarter fixed effects and small-cluster-robust inference. Lightcast AI-skills rate and Felten task-exposure provide comparisons, with Census firm-reported AI adoption and BLS occupational data providing triangulation. Findings Entry-level posting shares decline modestly but not differentially with AI-exposure. By contrast, the experienced-entry pay gap widens with greater shift-share AI exposure; a one-percentage-point increase in exposure predicts a significant $2,780 widening. Occupational reallocation toward higher-paid jobs accounts for 61% of the widening, and stronger within-occupation wage growth concentrated among experienced workers accounts for 39%. Research limitations/implications The findings suggest AI and labour-market research should distinguish potential task exposure from realized adoption and examine how occupational restructuring alters within-industry career stage inequality. Originality/value We shift attention from displacement to experience repricing, introducing career-stage-biased technical change as a two-channel account in which AI may alter both occupational demand and the rewards to experience-mediated capabilities.
About this publication
Academic journal. Published in an academic journal — typically peer-reviewed before publication, though indexing alone doesn't guarantee review depth.
- Published
- 2026 (recently)
- Citations
- 0 (OpenAlex)
0 citations recorded by OpenAlex — too few 2026 papers in this collection for a fair comparison.
Related by theme, not necessarily by method or conclusion.
- The Effects of Generative AI on High Skilled Work: Evidence from Three Field Experiments with Software Developers
SSRN · 2024 - GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models
arXiv · 2023 - Experimental evidence on the productivity effects of generative artificial intelligence
Science · 2023 - The simple macroeconomics of AI
Economic Policy · 2024
Overview assembled automatically from collected metadata. How this site is built.