How artificial intelligence applications affect the total factor productivity of the service industry: Firm-level evidence from China

Purpose: The 21st century has witnessed the emergence of the Fourth Industrial Revolution, bringing forth a new technological revolution worldwide. Artificial intelligence (AI), as cutting- edge general-purpose technology within the information technology (ICT) sector, is increasingly gaining influence and integrating much more closely with various industries. With the service economy becoming a critical component of national economies, the rapid development of AI presents new opportunities for enhancing the total factor productivity (TFP) of the service industry. In light of the convergence between the "service economy" and the "digital economy", this paper aims to assess the level of AI application in the service industry using firm-level data and explore the specific mechanisms through which AI impacts the TFP of the service industry from both theoretical and empirical perspectives. Approach: Firstly, this paper conducts the theoretical modeling with the CES-type production function to analyze the empowering effect and structural effect of AI on service firms. Next, using the data of China A-share listed service companies between 2010 and 2020 as a sample, the study constructs an index to measure the level of AI application of service firms by textual analysis and calculates TFP of service firms. Empirically, the paper investigates the impact of AI application on TFP of service firms and examines its heterogeneous characteristics using a two-way fixed-effects model. In addition, this work explores the mediating role of R&D investment and human capital in contributing to TFP. Findings: The results show a positive relationship between the level of AI application in service firms and their TFP. Significantly, the impact of AI application on TFP promotion varies across different industries, firm sizes, and ownership. Heterogeneity tests reveal that modern service firms, small and medium-sized firms, and state-owned firms can achieve more significant TFP gains from AI implementation. Furthermore, the empirical model successfully passed both the endogeneity and the robustness tests. Mechanism analysis further shows that AI drives TFP improvements by boosting R&D investment and optimizing human capital allocation within service firms. Value: The paper contributes to the existing literature by examining the relationship between AI application and TFP in the service industry. The findings hold significant implications for policymakers and service firms, guiding their decisions on promoting the widespread adoption of AI in the service industry. Theoretically, this work systematically summarizes the mechanism through which AI influences the TFP of service firms. Empirically, it constructs an indicator of AI