Artificial Intelligence and Labour Productivity by Skill Stratification: Empirical Evidence from Chinese Municipal Provinces, 2000–2020

Artificial intelligence has a significant impact on labour productivity, primarily by enhancing production efficiency, changing work patterns, and reshaping the labour market. However, its effect on productivity is limited by factors such as adoption rates and the alignment between AI advancements and workforce skill levels. As a result, these technological advancements have not yet led to substantial productivity gains for businesses and regions. The main objective of this paper is to analyse the impact of AI on the productivity of high-skilled, medium-skilled and low-skilled workers in municipal provinces in China between 2000 and 2020. Using Ordinary Least Squares (OLS) on a 21-year panel dataset from four Chinese municipal provinces, this study evaluates how three proxies of AI adoption affect labour productivity across high-, middle-, and low-skilled workers. Findings show divergent impacts depending on skill tier, with AI patents and research investments strongly benefiting high-skilled workers, while infrastructure-based AI investments support low- and medium-skilled workers. The findings from this study provide valuable insights for policymakers in China and other countries. Our results suggest that collaboration among research institutions, industries, and local governments is essential to foster a culture of continuous learning and adaptation. This collaborative approach ensures that workers across different skill levels are equipped with the necessary skills to thrive in an evolving labour market, making them more competitive and better prepared to face future challenges.