Artificial Intelligence and Green Total Factor Productivity in Manufacturing: An Empirical Analysis

Research hypotheses: H1: The development level of AI has a significant positive direct effect on the green transition of manufacturing. H2a: AI indirectly drives the green transition of manufacturing through green technology innovation. H2b: AI indirectly drives the green transition of manufacturing through improved digital infrastructure support. H3a: The impact of AI on the green transition of manufacturing exhibits a threshold effect contingent on government R&D investment. H3b: The impact of AI on the green transition of manufacturing exhibits a threshold effect contingent on economic development. H4: The effect of AI on the green transition of manufacturing exhibits a positive spatial correlation effect. The data are sourced from the China City Statistical Yearbook, provincial statistical yearbooks, the EPS database, and the National Enterprise Credit Information Publicity System. Missing values for individual observations are imputed using linear interpolation. To mitigate the influence of extreme values, the raw data are winsorized at the 1% level. The study sample comprises 281 prefecture-level and above cities in China, and it covers the period from 2009 to 2023. This yields a total of 4,215 observations.