Does AI-oriented place-based policy reshape the functional distribution of income? Evidence from China’s National AI Innovation and Development Pilot Zones

We provide the first quasi-experimental evidence on how AI-oriented place-based policy affects the distribution of income within firms. Exploiting the staggered designation of China’s National AI Innovation and Development Pilot Zones (2019–2021) and a panel of 37,855 firm-year observations of listed companies over 2012–2024, heterogeneity-robust difference-in-differences estimates show that designation raises the labour income share (employee compensation over operating revenue) by 0.67% points, about 4.5% of the sample mean, with effects emerging 3 years after designation. The policy stimulates firm AI adoption and is accompanied by employment internalization: external hiring and flexible employment contract sharply while aggregate employment and total compensation remain stable despite temporarily softer revenue, a stock–flow configuration consistent with greater retention of incumbent workers that insulates labour from the costs of the technological transition. An outsourcing-adjusted measure indicates the rise is genuine rather than an accounting artefact. Effects concentrate in financially unconstrained, high-growth firms, indicating that investment capacity conditions the distributional benefits of AI industrial policy.