Impact of Intelligent Development on the Total Factor Productivity of Firms – Based on the Evidence from Listed Chinese Manufacturing Firms

Abstract

The new industrial revolution featuring artificial intelligence (AI) as its core is flourishing globally. However, there are many controversies surrounding the impact of AI on productivity owing to the different understandings of its development. Thus, this study adopts a text mining method to construct indicators for measuring the intelligent development of enterprises based on the information obtained from the annual reports of listed Chinese manufacturing companies from 2009 to 2019. To explore the impact of intelligent development on the total factor productivity (TFP) of enterprises, fixed-effect regression and panel threshold models are employed to empirically prove its overall and threshold effects. The result reveals that the impact of intelligent development on TFP of enterprises is significantly positive at the aggregate level. Regarding the stage characteristics, “Solow’s paradox” exists in the development of intelligence. The effect of intelligence development on TFP is not significant at its early stage; moreover, the rapid development of intelligence exerts a “promotion effect.” However, at the extreme stage (when intelligent development crosses the critical value), it exerts a negative effect on the TFP of enterprises.