1.School of Accounting, Nanjing Audit University, Nanjing 211815, China 2.Post-Doctoral Station, Shanghai University of Finance and Economics, Shanghai 200433, China 3.School of Economics and Management, Southeast University, Nanjing 211189, China 4.School of Finance, Shanghai University of Finance and Economics, Shanghai 200433, China 5.Dishui Lake Advanced Finance Institute, Shanghai University of Finance and Economics, Shanghai 201306, China 6.School of Finance, Xinjiang University of Finance and Economics, Urumqi 830012, China 7.Chinese Modernization Institute, Shanghai University of Finance and Economics, Shanghai 200433, China 8.School of Management, Xiamen University, Xiamen 361005, China
As China transitions from an upper-middle-income to a high-income status, traditional growth drivers wane, and productivity-enhancing reforms and digital institutional infrastructure become pivotal. Beyond moving services online (“e-government”), digital government integrates fiscal, financial, innovative, and regulatory resources to empower firms. Local governments, with both meso- and micro-governance functions, are key actors. This paper aims to explore whether digital government has a causal effect on enhancing firm TFP and what the mechanism is?
Using a sample of China’s A-share listed companies from 2013 to 2023 and evaluating the text of prefecture-level government work reports through generative large language models, this paper examines the impact of digital government construction on firm TFP. The findings reveal that digital government construction significantly enhances firm TFP. Mechanism testing indicates that this effect operates primarily through four channels: improving government fiscal transparency, enhancing corporate transparency, optimizing resource allocation efficiency, and boosting corporate innovation efficiency. Cross-sectional testing shows that the positive impact of digital government construction on TFP is more pronounced for firms in high-tech industries, those with shorter establishment histories, and those facing higher financing constraints.
This paper uncovers a new productivity channel by showing that digital government—beyond internal governance or macro environments—can enhance firm TFP, expanding the policy toolbox for cultivating new quality productive forces. Methodologically, it advances measurement by constructing an LLM-based semantics-aware index grounded in authoritative policy dimensions, improving validity and reproducibility over lexicon- or survey-based indices. Empirically, it opens the mechanism “black box”, demonstrating that digital government operates through information (greater public fiscal and corporate transparency) and efficiency (improved resource allocation and higher innovation efficiency) channels, thereby illuminating the micro-governance effect of macro institutional reforms.
To fully leverage the enabling role of transforming government governance models from one-way management to two-way interaction for the economy and the society, and to stimulate the vitality of micro-market entities such as firms, the government should accelerate the digital transformation process. This will expedite the formation of a new digital governance landscape and advance the modernization of the national governance system and governance capabilities.
Total factor productivity calculation of listed companies from 2013 to 2023
2143.8 (KB)
2
Calculation of the level of digital government construction of listed companies from 2013 to 2023
1331.4 (KB)
3
Control variables of listed companies from 2013 to 2023 (company size, net profit margin on total assets, asset-liability ratio, Tobin Q value, cash ratio, inventory turnover ratio, growth capacity, regional economic level, industrial structure, digital transformation of enterprises, R&D expenditure
3546.2 (KB)
4
Heterogeneity Variables of Listed Companies from 2013 to 2023 (High-tech Industries, Enterprise Maturity, Financing Constraints)
888.8 (KB)
5
Mechanism variables of listed companies from 2013 to 2023 (enterprise transparency, resource allocation efficiency, innovation efficiency)
1344.6 (KB)
6
2013-2023 Year instrumental variable
768.6 (KB)
7
Public data is open from 2013 to 2023
676.1 (KB)
8
Regional-level control variables 2013-2023 (regional fiscal capacity, digital infrastructure, secondary industry employment, wage index)
1329.1 (KB)
9
2013-2023 Fiscal transparency of prefecture-level cities
63.4 (KB)
10
Control variables for prefecture-level cities from 2013 to 2023: population size, level of opening up to the outside world, urbanization rate, level of human capital, intensity of fiscal investment, and retail trade in commodities.
204.1 (KB)
11
Data and program code
2655.2 (KB)
Cite this article
Zhang Xiling, Yu Haoping, Xu Longbing, et al. How does Digital Government Construction Empower Firm TFP: Evidence Based on Generative AI[J]. Journal of Finance and Economics, 2026, 52(2): 94-108.