Abstract
Technological transformation during the 2010s reshaped U.S. skill demands, but workers experienced divergent outcomes. Understanding this heterogeneity requires identifying not just whether skills changed, but how they changed. We develop a measurement framework from job postings in 2010 and 2018, constructing aligned skill spaces for 514 occupations covering over 75% of U.S. employment. Two geometric dimensions (transformation magnitude and shifts in occupational distinctiveness) jointly reveal three distinct processes. Cognitive distance arises when new requirements conflict with existing expertise, creating heterogeneous adaptation. Scaffolded learning occurs when new capabilities build cumulatively on existing foundations, enabling collective upgrading. Simplification emerges when technology eliminates specialized requirements. Validating these interpretations against within-occupation wage dispersion, we find each mechanism produces its predicted pattern. Among high-wage occupations with substantial transformation, cognitive distance generates 4.8-percentage-point wage spread widening while scaffolded learning maintains stability. Among low-wage occupations with minimal transformation, simplification produces 4.2-percentage-point compression. The same geometric feature (declining distinctiveness) produces opposite wage effects depending on transformation context. The qualitative nature of skill change determines worker outcomes, not magnitude alone.
Citation
George, Nikhil, Christophe Combemale, Ramayya Krishnan, and Rahul Telang. 2025. “Technology and the Shifting Architecture of Occupational Skills.” Working Paper.
@techreport{george2025technology,
author = {George, Nikhil and Combemale, Christophe and Krishnan, Ramayya and Telang, Rahul},
title = {Technology and the Shifting Architecture of Occupational Skills},
year = {2025},
month = {nov},
institution = {SSRN},
type = {Working Paper},
url = {https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4906323},
note = {Last revised: November 18, 2025}
}