Pay Equity Pressures and the Organization of AI Talent and Innovation

Rising pressures for pay equity create a dilemma for firms competing for scarce talent: how can they attract workers who command high market wages while keeping internal pay differences socially defensible? We argue firms respond by changing where AI talent works. When pay equity pressures fall unevenly across a firm’s units, firms may place AI talent where high pay is easier to justify, shifting from headquarters to subsidiaries. We study this in China, where a 2018 reform sought to curb excessive pay gaps in state-owned enterprises (SOEs), constraining headquarters more tightly than subsidiaries. Linking data on corporate hierarchies, patents, job postings, and inventor careers, we track how AI activities differ from non-AI activities within the same firm, and how that difference changes for SOEs after the reform. SOEs posted more of their AI R&D jobs through subsidiaries, hired more of their new AI inventors there, and were more likely to move existing AI inventors into subsidiaries. SOE headquarters also became less likely to state salaries in AI job postings. Innovation followed: SOEs became more likely to develop AI patents to subsidiaries. This R&D decentralization is associated with lower patent novelty. Pay equity pressures change where AI innovation occurs and outcomes.