Dual effects of automation on economy and environment: Evidence from A-share listed enterprises in China

The “technology dividend” brought by the development of artificial intelligence (AI) can optimize the allocation of resource within enterprises, provide a new model for enterprises to achieve sustainable development, and create a new power source for enterprise economic development and pollution reduction. Although the advantages and disadvantages of AI have been widely discussed, few studies have explored whether it can play the dual effects on economy and environment from the perspective of enterprises. Considering the difficulty of measuring AI indicators, this paper attempts to explore the impact of automation, as a key underlying technology of AI, on enterprise economic performance and environmental performance, and rationally infer the dual impact of AI. We use panel data from China's Shanghai and Shenzhen A-share listed companies from 2009 to 2021 to reveal the dual effects of automation. The benchmark regression results show that automation can boost both enterprise economic and environmental performance. The result is still credible after a series of robustness tests and causal identification. Moreover, we find that the economic performance is stronger in non-heavily polluted enterprises. The dual effects of automation are more significant in areas with low environmental regulations and areas with high levels of industrial digitalization and digital industrialization. The mechanism analysis results show that automation can play the dual effects through the cost-effectiveness channel, the capital-labor substitution channel, and the energy-saving and emission reduction channel.