Template-type: ReDIF-Paper 1.0
Author-Name: Jan Einhoff
Author-Name: Isabella López Trejos
Author-Name: Caroline Paunov
Title: Cross-country skills-technology policy debates through large language models
Abstract: Using language models, this paper conducts a cross-country comparative innovation policy analysis of skills-technology policy debates across seven OECD member countries (Austria, Canada, Finland, Germany, Korea, Sweden, and the United Kingdom). Results highlight the dominance of STEM (science, technology, engineering and mathematics) and digital skills in these policy debates, the relative neglect of green skills, and the emphasis on soft skills across all technology fields. The analysis also identifies common policy instruments, which include collaborative platforms and direct financial support. Overall, the paper shows how large language models can help policy analysts identify patterns and gaps in extensive policy texts that nonetheless critically demands expert oversight and careful interpretation.
Classification-JEL: J24; O30; O33; O38; C63
Keywords: Artificial intelligence, ChatGPT, Computational text analysis, Digital technologies, Green technologies, Innovation policy, Large language models, Quantum technologies, Soft skills, STEM skills, Technology policy, Topic modelling
Creation-Date: 2025-06-30
Number: 2025/20
Handle: RePEc:oec:stiaaa:2025/20-EN