The Role of Digital Governance in Fostering County-Level Digital Entrepreneurial Vitality: A Quasi-Experimental Analysis of China’s Information Accessibility Pilot

Abstract

Digital entrepreneurship, driven by digital technology and the digital economy, is a vital pathway for high-quality economic development. In China, the “National Pilot Policy of Information Accessibility” has been a key initiative to advance digital governance. Using enterprise registration data from 1206 counties during 2010–2020 and employing a difference-in-differences framework, this study evaluates the policy’s impact on county-level digital entrepreneurial vitality. Results show that the pilot significantly stimulates entrepreneurship, with robust evidence from multiple tests including placebo, PSM-DID, and controls for other policies and the pandemic. Mechanism analysis reveals three main channels: enhancing regional innovation capacity, optimizing employment structure, and strengthening infrastructure. Heterogeneity tests indicate stronger effects in western counties with weaker governance and resources, while impacts in eastern regions are less pronounced. These findings highlight the importance of government digital governance in narrowing regional gaps and fostering inclusive digital transformation. The study provides theoretical and empirical insights into how government-led digital initiatives can energize local entrepreneurship, support structural upgrading, and promote high-quality development in the digital era.

1. Introduction

Humanity has entered the era of the digital technology revolution, with the digital economy becoming the main theme of current economic development. Countries worldwide are making digital economy a national development strategy through top-level design and policy mechanisms [1]. For instance, France was the first to release its digital economy strategy in 2008, followed by the United States with its data-driven innovation policy, and the European Digital Agenda proposed new guidelines on digital technology standards and compatibility. China also attaches great importance to digital entrepreneurship as a major strategic issue. As China enters the digital economy era, activating the digital economy’s vitality in counties is of great significance for innovation-driven development and promoting urbanization with counties as important carriers. To promote this process, in 2014, the National Development and Reform Commission of China issued a notice on accelerating the implementation of the Information Accessibility Project, aiming to promote digital infrastructure construction, facilitate the digital transformation of industries, cultivate new forms of digital economy, and enhance digital service capabilities, thereby fostering county-level digital industry development and stimulating market vitality and social creativity at the county level. The 2018 China Digital Enterprise White Paper emphasized the critical role of entrepreneurial elements such as digital strategies, digital technologies, and digital talents in digital entrepreneurial enterprises. These national-level digital strategies can revolutionize traditional entrepreneurship, ignite digital entrepreneurship, and promote rapid development of the national digital economy [2]. Digital entrepreneurship has emerged in response to the rapid advancement of digital technologies and the digital economy, becoming a significant force driving social transformation. As a key catalyst for digital economic development, it is fundamentally reshaping economic growth models, industrial structures, entrepreneurial paradigms, and lifestyles. From a practical standpoint, while many still celebrate the accomplishments of traditional entrepreneurship, the convergence of “intelligence and entrepreneurship” is redefining conventional approaches. Numerous digital ventures that embrace adaptability and innovation have overcome traditional temporal and spatial limitations, forging new entrepreneurial paths and notable success stories. For example, companies like Alibaba Cloud and Tencent have effectively embedded emerging digital technologies into their services and products, exemplifying “comprehensive digital entrepreneurship.” Taobao, by creating a digital information platform, pioneered a “hybrid digital entrepreneurship” model that seamlessly integrates online and offline resources, playing a key role in advancing China’s digital economy. Thus, whether viewed through the lens of national strategic planning or entrepreneurial evolution, digital entrepreneurship represents a dynamic and challenging frontier that continues to garner attention from governments, industries, and academia alike.

Digital entrepreneurship, as an integral component of the digital economy, has emerged as a global trend in entrepreneurial activity. It is not only a result of the application of information technology but also a process in which entrepreneurs leverage digital tools to identify and exploit market opportunities, thereby creating innovative business models. Compared with traditional forms of entrepreneurship, digital entrepreneurship is characterized by lower costs, reduced entry barriers, and greater flexibility. These features make it a vital driver of innovation and economic development. Digital entrepreneurship refers to the process by which entrepreneurs leverage digital technologies—particularly the Internet and mobile Internet platforms—to develop new products and services and establish innovative business models. Compared with traditional forms of entrepreneurship, digital entrepreneurship is marked by lower entry barriers, greater flexibility, and stronger innovativeness, making it a vital force in driving high-quality economic development. It is not merely the outcome of technological application, but rather the result of a deep integration between digital technologies and market demand. This integration has facilitated the emergence of new forms of products and services and transformed the operational models of traditional industries. In China, the development of digital government has played a pivotal role in promoting digital entrepreneurship. The digital transformation of government governance is a dynamic process in which governments, driven by digital technologies, achieve the digitalization of social governance and public service delivery. It plays a crucial role in reshaping the mechanisms of economic and social operations and advancing the modernization of government governance [3]. By enhancing digital infrastructure, streamlining entrepreneurial procedures, and optimizing the policy environment, the government has significantly facilitated the rapid growth of digital entrepreneurship. Notably, at the county level, digital government has not only improved the entrepreneurial ecosystem but also supported the initiation and growth of numerous digital ventures through incentive mechanisms. China’s “National Pilot Policy of Information Accessibility”, launched in 2014, provides a valuable quasi-natural experiment for identifying how the digital transformation of government governance influences county-level digital entrepreneurial vitality.

Theoretical research on digital entrepreneurship is still lagging and insufficient to guide practice. While scholars have mainly focused on traditional entrepreneurship, digital entrepreneurship is rapidly flourishing, meeting user demands, driving digital economic development, and becoming central to smart city construction and digital life. Existing studies have examined entrepreneurial elements [4,5,6], outputs [7,8], influencing factors [9,10,11], and business models, but these findings remain fragmented and lack depth. As the origin of digital economic and social transformation, digital entrepreneurship plays a crucial role in employment creation and innovation. Enhancing county-level entrepreneurial vitality can strengthen the digital and platform economy, generate flexible jobs, and enable counties to integrate into the digital economy. Yet, more than half of China’s entrepreneurial activities remain concentrated in wholesale and retail, with relatively low shares in knowledge-intensive industries such as information and communication, showing a gap compared with developed countries [12]. Moreover, most research on government digital governance and entrepreneurship focuses on urban contexts [13], with limited attention to underdeveloped counties. Thus, how government resources and digital infrastructure stimulate county-level entrepreneurial vitality remains an important but underexplored issue.

To address this research gap, this paper uses panel data from 1206 counties across China from 2010 to 2020, with the 2014 implementation of the “National Pilot Policy of Information Accessibility” as a quasi-natural experiment, to systematically examine the impact of the information accessibility pilot project—a specific measure of government digital governance—on county-level digital entrepreneurial activity and its mechanisms using the difference-in-differences method. The results indicate that the implementation of the “National Pilot Policy of Information Accessibility” significantly stimulates county-level digital entrepreneurial vitality. The primary mechanisms through which the policy promotes entrepreneurship include enhancing regional innovation capacity, optimizing employment structures, and strengthening infrastructure. Moreover, the policy demonstrates a more pronounced effect in western regions, where resources and governance efficiency are relatively lagging.

Compared with existing studies, this paper makes three main contributions. First, it examines the impact of government digital governance on county-level digital entrepreneurship, thereby extending research on the economic effects of digital governance and complementing analyses of less-studied county areas. Second, using data from 2010 to 2020, it provides empirical evidence on the development of county-level digital entrepreneurship in China. Third, it theoretically and empirically identifies the mechanisms through which digital governance influences entrepreneurship, offering new policy insights for stimulating market vitality, advancing economic transformation, and promoting high-quality development.

The remaining structure of the paper is as follows: Section 2 presents the literature review and policy background; Section 3 outlines the theoretical framework and research hypotheses; Section 4 introduces the research design, variable descriptions, and data sources; Section 5 provides the empirical results and mechanism tests; Section 6 explores the regional heterogeneity of the pilot policy; Section 7 discusses the findings; and Section 8 concludes with the main results and policy recommendations.

2. Literature Review

2.1. Government Digital Governance

Government digital governance is an important aspect of modernizing national governance and is key to modernizing public services and improving government efficiency. Digital governance involves not only the application of technology but also the formulation of corresponding policies and management systems to ensure and optimize the effective use of these technologies. Existing studies have sufficiently discussed government digital governance from a theoretical perspective. Zhang Tao et al. (2023) proposed a comprehensive digital government governance framework including technology governance, business governance, and operational management systems [14]. This framework aims to promote coordinated development of digital government through centralized sharing of technical governance architecture, open and integrated business governance structure, and collaborative development of operational management systems. H. Scholl (2020) [15] discussed the evolution of digital government from its early stages to the present and made predictions about future development directions over the next twenty years. Scholl pointed out that modern information technology brings transformation possibilities to the public sector, but technology alone is not enough to fully describe the various dimensions and impacts of digital government [15]. E. D. Madyatmadja et al. (2016) studied the IT governance practices of the Jakarta provincial government in the digital era, emphasizing the importance of good IT governance for government strategy, especially in decision-making and oversight of high-level meetings, which is particularly important for preventing the abuse of digital assets [16]. Virgílio A. F. Almeida et al. (2019) emphasized that governance is key to aligning public interest with societal benefits in digital services [17]. By establishing appropriate governance principles, digital technology can become a powerful tool for improving public services.

2.2. Digital Entrepreneurial Vitality

Digital entrepreneurial vitality reflects the growth and innovation capability of entrepreneurial activities in the context of the digital economy. Yuming Zhai et al. (2022) used keyword co-occurrence and citation clustering analysis to propose a 3E (Empowerment, Evolution, and Ecosystem) framework to promote interdisciplinary dialogue and support evidence-based policy-making and practice [18]. Catarina Fernandes et al. (2022) analyzed the link between digital entrepreneurship and sustainability through bibliometric analysis of literature in the Scopus database, revealing three main thematic clusters: innovation and entrepreneurship, digital transformation: strategy and business models, and sustainability and sustainable development goals [19]. Gedas Baranauskas and A. G. Raišienė (2022) proposed a new conceptual framework combining sustainability and digital entrepreneurship, exploring how traditional business models are transformed into networked and integrated digital platform business models in a rapidly digitalizing economy, and highlighting sustainable management challenges in digital entrepreneurial and digital business ecosystems, such as vulnerabilities in organizational and social networks, sociotechnical pressures, and asymmetries in digital information and resources [20]. In 2021, the National Bureau of Statistics officially released the Statistical Classification of the Digital Economy and Its Core Industries (2021) (hereinafter referred to as the Classification), which provides an authoritative definition of the conceptual connotation and scope of core industries within the digital economy. This has gradually unified the understanding of digital industries across various sectors of society. Dai Ruichen et al. (2022) were among the first to use the industry correspondence table provided in the Classification to match industry information from business registration data [21]. By identifying the number of newly established enterprises, they used this as a sub-indicator and, together with other data sources, constructed a city-level digital economy development index. Similarly, Feng et al. (2023) used business registration data to assess digital entrepreneurial vitality across Chinese provinces [22]. Overall, existing literature has yet to shift the research perspective to the county level, which represents the most fundamental regional unit within the national economic system.

2.3. Government Digital Governance and Entrepreneurial Activity

Government digital governance has an important impact on promoting entrepreneurial activity. By implementing effective digital policies and providing a supportive entrepreneurial environment, the government can significantly increase entrepreneurial activity and success rates. Tian Gan et al. (2023) [13] found that the number of newly registered enterprises increased significantly after the implementation of China’s digital government policy. This study shows that digital government strategies positively promote entrepreneurial activity by improving government efficiency, reducing local government risk behavior, and improving access to financial resources [13]. B. Hansen (2019) found that the digitalization and entrepreneurship promotion strategies advocated by the government significantly increased social trust and support for entrepreneurship [11]. R. Tassabehji et al. (2016) [23] studied how digital era governance (DEG) promotes organizational change in the public sector. Through case analysis, the study showed that chief information officers (CIOs) and other key figures in the public sector took on the role of institutional entrepreneurs, driving technological governance and policy shifts through proactive community mobilization and legitimization strategies [23]. This shows that the government plays a key role in leading and supporting digital transformation.

In summary, as the digital economy era emerges, an increasing number of scholars have affirmed the driving role of digital economy development in fostering entrepreneurship. Unfortunately, this part of the research mainly focuses on the role of the openness of digital technology/platforms in identifying business opportunities and forming business models [24,25]. Existing research has not fully considered the promotional effect of digital government construction on digital entrepreneurship from the perspective of digital governance, especially in economically underdeveloped county areas. The government is one of the important subjects that enterprises face in the entrepreneurial process, and the level of government governance is an important part of the entrepreneurial environment. This research is of important policy significance for exploring new ways to enhance county-level digital entrepreneurial vitality and stimulate market vitality.

2.4. Policy Background

In 2017, the 19th National Congress of the Communist Party of China first explicitly raised “Digital China” as a national strategy, marking a new height of attention being paid to digital transformation in China. Subsequent policy documents, such as the “14th Five-Year Plan” in 2020, further detailed the implementation direction of the “Digital China” strategy, emphasizing the coordinated development of digital government, digital economy, and digital society, with digital government playing a leading and coordinating role in promoting overall digital transformation.

In June 2014, the National Development and Reform Commission of China, together with 12 departments including the Ministry of Industry and Information Technology, jointly issued a notice on accelerating the implementation of the Information Accessibility Project, marking the official launch of the “National Pilot Policy of Information Accessibility.” This policy selected 80 cities as national pilots, aiming to promote the modernization of public services and social management through government informationization. The selection of these pilot cities was not random but based on a comprehensive set of criteria, reflecting the government’s strong emphasis on the feasibility, representativeness, and scalability of the policy. First, candidate cities were required to possess a solid foundation in informatization, including well-developed broadband networks, data centers, and digital platforms as essential infrastructure. Second, cities were expected to have prior experience in smart city initiatives, along with a certain level of institutional foundation and technological accumulation. Additionally, the enthusiasm and governance capacity of local governments were critical considerations, such as whether dedicated implementation mechanisms were in place and whether matching financial resources were available. The selection process also paid special attention to regional balance and the diversity of city types, ensuring representation from the eastern, central, and western regions, as well as from coastal and inland areas, and including both megacities and prefecture-level cities. This diversity was intended to facilitate the generation of differentiated pilot experiences. Although all 80 pilot cities were incorporated into a unified national policy framework, there were notable variations in implementation during the execution phase. These differences were primarily reflected in the content of policy execution, areas of emphasis, implementation strategies, and supporting measures. For instance, economically developed eastern cities often focused on the intelligentization of public services and the improvement of governance efficiency. Cities like Shenzhen and Hangzhou were early adopters in developing smart healthcare and government service platforms. In contrast, cities in the central and western regions prioritized addressing infrastructure deficiencies and expanding access to digital public services. For example, Guiyang and Lanzhou emphasized the equalization of urban and rural information services. Each city tailored its pilot initiatives to its specific industrial structure and social needs: some focused on “unified city cards” or “convenient payment platforms,” while others promoted “government big data platforms” or “rural informatization service systems.”

As an important measure to advance the modernization of the national governance system and governance capacity, the “National Pilot Policy of Information Accessibility” aims to break information silos and achieve interconnection and information sharing among government departments through digital technology. This policy focuses on exploring how to optimize the allocation of public resources through big data and other modern information technologies, and to innovate new mechanisms and models for social management and public services. In terms of funding allocation, the policy adopted a strategy of “initial allocation and subsequent rewards and punishments,” encouraging pilot cities to actively implement informationization projects and ensuring the effectiveness of projects through subsequent assessments. In 2015, the National Development and Reform Commission and the Ministry of Finance, together with relevant departments, issued opinions on evaluating the “National Pilot Policy of Information Accessibility,” fully deploying the evaluation and acceptance work of pilot cities, and clarifying the evaluation indicator system, procedures, and requirements. This evaluation mechanism not only ensured the quality of policy implementation but also provided a basis for continuous policy optimization. With the rapid development of digital technology, in 2016, the General Office of the State Council further issued an implementation plan for promoting “Internet + Government Services” and implementing the “Information Accessibility Pilot Project,” requiring pilot cities to deepen “Internet + Government Services” and promote the realization of “one application, one window acceptance, one network handling.” These measures not only improved the quality and efficiency of government services but also promoted the modernization of the government governance system and governance capacity.

It is worth noting that the Information Benefits to the People policy adopted a fiscal mechanism of “initial allocation combined with subsequent rewards and penalties”, which reinforced performance orientation and resource-based incentives in policy design. However, this mechanism may have introduced selective effects on both the selection of pilot cities and the intensity of policy implementation. On one hand, regions with stronger fiscal execution capacity, data management infrastructure, and institutional responsiveness were more likely to be included in the initial pilot list. On the other hand, counties with greater initial capacity may have been more motivated to allocate additional resources to enhance policy implementation due to the incentive structure. As a result, the observed policy effects may partially reflect a combination of “policy intervention effects” and “self-selection effects”. Although current data limitations prevent us from fully disentangling these mechanisms, we have controlled for key baseline characteristics in the main models and employed propensity score matching to mitigate potential sample selection bias. Future research could further develop a more comprehensive identification strategy to uncover how fiscal incentive structures shape the endogenous pathways of policy implementation.

3. Theoretical Analysis and Research Hypotheses

3.1. The Impact of Government Digital Governance on County-Level Digital Entrepreneurial Vitality

Studies have shown that the establishment of digital government platforms and process reengineering effectively reduce the time and financial costs associated with business registration, approval, and taxation procedures, thereby directly lowering the threshold for entrepreneurship and enhancing entrepreneurial willingness [13]. At the same time, the openness of government information and data sharing improve institutional transparency, reduce information asymmetry in policy and market access, and help entrepreneurs make decisions more efficiently [26]. In addition, local governments’ digital procurement activities (such as smart agriculture and e-government outsourcing) create direct market demand for entrepreneurs, providing them with initial clients and funding sources [27]. Furthermore, government-led construction of e-commerce, financial, and agricultural digital service platforms lowers market entry barriers for entrepreneurs, directly enhancing the feasibility and vitality of digital entrepreneurship in county regions [28].

3.2. Mechanism Analysis

The National Pilot Policy of Information Accessibility aims to improve the level of government digital governance, which affects county-level digital entrepreneurship through the following three mechanisms:

1. Enhancing Regional Innovation Capability: Digital governance, as an innovation in government governance, promotes the application of new technologies and fosters a regional innovation atmosphere, leading to increased regional innovation levels. The theory of knowledge spillover entrepreneurship suggests that regions with more innovative knowledge are more likely to generate entrepreneurial activities [29]. The improvement of regional innovation capability means that local innovation activities are active, new technologies, new business models, and new business formats are continuously emerging and applied, which will bring opportunities for digital entrepreneurship, thus promoting digital entrepreneurial vitality.

2. Improving Employment Structure: Government digital governance can indirectly improve the entrepreneurial environment by changing the employment structure. For example, by supporting governance institutions to help enterprises improve their performance, digital transformation of enterprises has a positive impact on the productivity level of listed companies, creating a more favorable environment for entrepreneurial activities [30]. Additionally, by investing in technological infrastructure, digital governance can drive economic growth and innovation, providing equal opportunities for different genders. On one hand, the construction of “Information Accessibility Pilot Cities” aims to optimize the supply of public services through effective online and offline linkage of service resources, thereby enhancing the level of public services such as education, healthcare, and culture, which attracts high-quality digital technology talents. On the other hand, digital governance improves the satisfaction and participation experience of the public in governance [31], providing talent support for entrepreneurial activities. Digital governance significantly enhances public satisfaction with government services by improving transparency, administrative efficiency, and responsiveness. This institutional optimization not only increases trust in regional policies but also helps cultivate a more attractive environment for talent development. In the context of digital entrepreneurship, workers with digital skills and platform management capabilities are particularly sensitive to institutional stability, convenience of living, and policy certainty. A sound digital governance environment thus contributes to attracting and retaining such skilled digital talent. At the same time, entrepreneurs are more likely to register and establish businesses in regions with high-quality governance and efficient services, thereby promoting a local employment structure that shifts toward high-skilled and information-oriented occupations. Therefore, improved public satisfaction can be interpreted as a mediating manifestation of institutional effects, reflecting how digital governance enhances the business and talent environment and, in turn, indirectly optimizes the employment structure within the digital entrepreneurship ecosystem. Talent agglomeration not only contributes to the accumulation of innovation and entrepreneurship knowledge but also helps the “Information Accessibility Pilot Cities” attract more digital entrepreneurship talent. This process leads to a self-reinforcing employment structure, which in turn enhances the entrepreneurial vitality at the county level.

3. Strengthening Infrastructure: Government digital governance has played a positive role in improving county-level infrastructure, thereby promoting the prosperity of digital entrepreneurial activities in these areas. Key indicators of a region’s entrepreneurial environment include infrastructure elements such as transportation, healthcare, water and power supply, commercial services, and cultural education. In rural areas, the quality of infrastructure, including healthcare, sanitation, and transportation, significantly impacts the likelihood of migrant workers returning to start businesses [32]. A developed transportation network provides the necessary conditions for the development of county-level e-commerce, thereby stimulating entrepreneurial activities in the region. Digital governance enhances government administrative efficiency and public service capabilities, contributing to the equalization of urban and rural public services. This equalization can significantly increase the availability of public transportation and road mileage per capita in rural areas. Additionally, the promotion of government digital governance will drive the implementation of new infrastructure projects, particularly the construction of digital government at the county and township levels. This will enhance rural infrastructure development, accelerate the widespread application of broadband internet and 5G technology in counties, and directly improve the informatization level in rural areas, creating favorable conditions for rural entrepreneurship.

Based on the above analysis, the following hypotheses are proposed:

H1: The National Pilot Policy of Information Accessibility, as a form of government digital governance, can promote county-level digital entrepreneurial vitality.

H2: The National Pilot Policy of Information Accessibility mainly enhances county-level digital entrepreneurial vitality through three pathways: promoting regional innovation, improving employment structure, and strengthening infrastructure.

3.3. Construction of a Theoretical Model

Drawing on the preceding literature review and proposed hypotheses, a corresponding theoretical framework is developed. The resulting model is presented in Figure 1.

4. Research Design

4.1. Model Construction

This study focuses on exploring whether digital governance can effectively enhance digital entrepreneurial vitality in county areas. To this end, this study uses the “National Pilot Policy of Information Accessibility” launched in 2014 as a quasi-natural experiment case and employs the difference-in-differences method to empirically test this issue. The analysis is based on differences between counties and further explores the impact of time changes. The specific model construction is as follows:

Entre_it = α + βTreat_it + γX_it + μ_i + λ_t + ϵ_it

(1)

where the dependent variable Entre_it is the digital entrepreneurial vitality of county i in year t, measured by the natural logarithm of the number of newly registered enterprises in the information transmission, software, and information technology service industries in that year. Treat_it represents the dummy variable for the Information Accessibility Pilot Project, with β being the coefficient of interest, indicating the average treatment effect of the policy on entrepreneurial vitality. X_it represents a series of control variables that may affect county digital entrepreneurial vitality, including economic development level, industrial structure, fiscal decentralization, foreign investment level, financial development level, internet penetration rate, and scientific and technological level. μ_i represents county fixed effects, controlling for all region-specific characteristics that do not change over time. λ_t represents year fixed effects, controlling for macro shocks nationwide. ϵ_it is the error term.

4.2. Data Sources

According to the “Digital Economy and Its Core Industry Statistical Classification (2021)” issued by the National Bureau of Statistics of China in May 2021, the digital economy includes digital industrialization and industrial digitization, which can be subdivided into 5 major categories, 32 medium categories, and 156 small categories [33]. This study uses data on newly registered enterprises in the information transmission, software, and information technology service industries from 2010 to 2020 across various counties in China, sourced from the Tianyancha enterprise database. In addition, control variables such as economic development level, industrial structure, and fiscal decentralization, among others, are derived from the “China County Statistical Yearbook” published during the same period. To maximize the construction of balanced panel data, this study examines data from 1206 counties in 22 provinces and autonomous regions and one municipality (Chongqing) from 2010 to 2020. It should be noted that during the data preprocessing stage, this study excluded certain county-level samples with severe missing data or statistical irregularities, primarily including counties under the jurisdiction of the Tibet Autonomous Region and several counties in Hebei Province. In Tibet, due to its unique geographic environment, weak statistical foundation, and distinct policy system, key indicators related to entrepreneurship and public finance are missing for multiple years, making it unsuitable for constructing a balanced panel. Additionally, the region’s digital infrastructure is significantly underdeveloped, and its governance structure differs substantially from that of eastern and central regions, which may compromise the consistency of the identification framework if included. In Hebei, some counties were excluded due to discontinuities in key variables resulting from changes in statistical definitions across years. Based on the principles of data completeness and institutional comparability, these samples were excluded to ensure the stability and credibility of model estimation. The final dataset consists of 10,578 observations. The integrated application of these data provides a comprehensive and robust empirical foundation for the study, ensuring the accuracy and reliability of the findings. It is important to emphasize that the external validity of the results primarily applies to county-level regions in eastern, central, and western China where administrative institutions and digital governance structures are largely comparable. Future research, upon access to higher-quality data, may conduct differentiated and extended analyses focusing on ethnic minority regions or other special governance units.

4.3. Variable Explanation and Descriptive Statistical Analysis

Dependent Variable: Digital Entrepreneurial Vitality (Entre). As a key indicator of regional entrepreneurial activity, entrepreneurial vitality has been widely used in macroeconomic research [34,35,36,37]. This study analyzed the business registration data to obtain the number of newly registered digital enterprises in each county every year. Compared with relying on the answers of respondents in micro-surveys about their self-employment status, using China’s business enterprise registration data to statistically calculate the number of newly established enterprises each year can more accurately capture the situation of formal entrepreneurial activities. This study takes the natural logarithm of the sum of the number of newly registered enterprises in the information transmission, software and information technology service industry in each county that year as a measure of digital entrepreneurial activity. At the same time, considering that the modernization level, social and economic development level of counties are different from those of districts, and that the social services, infrastructure level, educational and medical resources in districts are generally more abundant than those in counties, in order to more accurately study the impact of information-based public welfare pilot policy on the entrepreneurial activity of counties, this study excluded the data sample of districts. It is worth noting that this study uses the number of newly registered businesses in the “Information Transmission, Software, and Information Technology Services” sector as a proxy indicator for digital entrepreneurial vitality. This choice is primarily based on the fact that this sector reflects key characteristics of digital entrepreneurship, including technology orientation, platform-based models, and knowledge intensity. However, we also acknowledge that this definition has certain scope limitations. With the widespread application of the “Internet Plus” strategy, digital entrepreneurship has increasingly penetrated the peripheries of traditional industries such as retail, agriculture, e-commerce, and logistics. This is particularly evident at the county level, where various digitalized business models—such as agricultural livestreaming, e-commerce logistics, and online agricultural management systems—are not fully captured by the industry classification used. As a result, the indicator adopted in this study may underestimate the vitality of digital entrepreneurship in a broader sense, and primarily reflects entrepreneurial activity centered on IT and software technologies. This variable selection represents a trade-off made in light of data availability and consistency in industry classification. Future research, with access to richer data dimensions, may expand the measurement framework to encompass a broader definition of digital entrepreneurship and thereby provide a more comprehensive understanding of the digital entrepreneurial ecosystem.

Core Explanatory Variable: Implementation of the Information Accessibility Pilot Policy (Treat). This study uses the Information Accessibility Pilot Policy as a quasi-natural experiment. A dummy variable for the type of city where the county is located and the interaction term of the policy implementation time dummy variable (Group X Post) represent the treatment effect of the Information Accessibility Pilot Policy (Treat). Specifically, this study sets the Information Accessibility Pilot City Group as 1 (experimental group) and non-pilot cities as 0 (control group); the time dummy variable for policy implementation before and after 2014 is set to 0 and 1, respectively. If a county is in a pilot city after 2014, the interaction term takes the value of 1; if the county is in a pilot city before 2014 or if the city has never implemented the Information Accessibility Pilot Project, the interaction term takes the value of 0.

Control Variables: To comprehensively and accurately explore the impact of digital governance on urban entrepreneurial vitality, it is necessary to consider and control for other variables that may affect entrepreneurial vitality. Following the work of Zhao Tao et al. (2020) [37] and Zhan Yong and Li Shan (2022) [38], this study introduces a series of control variables in the analysis model: ① Economic development level (lnAGDP). Regions with higher economic development levels may have higher talent agglomeration and economic capacity, thereby possibly influencing the benefits of entrepreneurial activities and affecting residents’ entrepreneurial intentions. To reduce the impact of extreme values and heteroscedasticity, this study uses the natural logarithm of the actual per capita GDP, deflated by prices, to measure the economic development level of cities. ② Industrial structure (Indust_Stru). The differences in industrial structure of cities objectively reflect differences in resource endowments, input factors, and other conditions at the city level. These differences may lead to varying degrees of difficulty and space for entrepreneurship in different cities, thereby affecting urban entrepreneurial vitality. This study employs the ratio of the added value of the tertiary industry to the regional GDP as an indicator of industrial structure. ③ Fiscal decentralization (fiscal). This is quantified by the ratio of general fiscal revenue to general fiscal expenditure of the government. Fiscal decentralization may influence entrepreneurial vitality by shaping the fiscal incentive structures, resource allocation capacity, and policy-making flexibility of local governments, potentially exerting either positive or complex effects. Including it as a control variable in empirical analysis helps to isolate the impact of local fiscal structures on the effectiveness of entrepreneurship policies, thereby enhancing the robustness of the estimation results. ④ Financial development level (finance). A well-developed financial system can provide funding support for entrepreneurs of different types, scales, and natures, enrich sources of entrepreneurial capital, alleviate financing constraints, and provide incentive mechanisms and risk-sharing mechanisms for enterprises, thereby enhancing entrepreneurs’ confidence. This study uses the ratio of the balance of loans from financial institutions at the end of the year to GDP to indicate financial support in counties. ⑤ Population density (pop_den). Population density may influence entrepreneurial vitality by affecting market size, labor supply, information flow, and the agglomeration of resources, thereby promoting the emergence and diffusion of entrepreneurial opportunities. However, if excessively high, it may also lead to resource congestion and rising costs, which could have a suppressive effect on entrepreneurship. This is measured by the total household registration population of the county divided by the administrative area. ⑥ ICT infrastructure level (ict). The level of ICT infrastructure may enhance entrepreneurial vitality by improving information accessibility, reducing startup costs, expanding market channels, and promoting technological innovation. These effects collectively help to create an environment conducive to the integrated development of digital and traditional entrepreneurship. This is measured by the logarithm of the number of basic telephone users. By doing so, we can effectively remove the interference caused by these variables, ensuring a more accurate assessment of the impact of digital governance on county-level digital entrepreneurial vitality.

5. Empirical Analysis

5.1. Descriptive Statistical Results

Table 1. Descriptive Statistics of Main Variables.

Variable	Obs	Mean	Std. Dev.	Min	Max
entre	10,578	2.5762	1.0203	0	8.7183
treat	10,578	0.0933	0.2909	0	1
lngdp	10,578	10.1306	0.6556	7.932	12.801
indust_stru	10,578	0.3756	0.1107	0.0702	0.822
fiscal	10,578	0.2516	0.176	0.0093	1.4837
finance	10,578	0.6339	0.361	0	3.053
pop_den	10,578	0.0257	0.0247	0	0.1929
ict	10,578	10.0832	1.0521	2.5649	13.6026

presents the descriptive statistics of the main variables. As shown in Table 1, Entre represents the natural logarithm of the number of newly registered enterprises in the information transmission, software, and information technology service industries in each county each year, plus one. Its mean value is 2.5762, with a standard deviation of 1.0203, and the minimum and maximum values are 0 and 8.7183, respectively, indicating significant differences in digital entrepreneurial vitality among counties. The mean value of treat is 0.0933, indicating that 9.33% of the sample are counties belonging to “Information Accessibility Pilot Cities”. The descriptive statistics of the other variables align with findings from existing studies.

5.2. Baseline Model

Table 2. Baseline Regression Results.

Variables	(1)	(2)	(3)	(4)
	Entre	Entre	Entre	Entre
treat	0.6844 ***	0.1658 ***	0.3334 ***	0.1586 ***
	(20.4602)	(3.2521)	(11.2789)	(3.1989)
lngdp			0.3234 ***	−0.1042 *
			(20.3483)	(−1.6802)
indust_stru			1.5202 ***	0.7283 ***
			(17.0230)	(3.8966)
fiscal			0.0037	0.0158
			(0.0577)	(0.1445)
finance			0.2690 ***	−0.0741
			(10.1303)	(−1.3812)
pop_den			11.0902 ***	20.4963 ***
			(28.0919)	(3.7130)
ict			0.2630 ***	−0.0035
			(26.9830)	(−0.1977)
Constant	2.5123 ***	2.1102 ***	−4.4100 ***	2.4199 ***
	(245.8634)	(100.3000)	(−22.4119)	(3.5928)
Observations	10,578	10,578	10,578	10,578
R-squared	0.0381	0.3446	0.3000	0.3513
country fe	no	yes	no	yes
year fe	no	yes	no	yes
Number of countries		1206		1206

t-statistics in parentheses. *** p < 0.01, * p < 0.1.

reports the regression results of the impact of the Information Accessibility Policy on digital entrepreneurial vitality. Column (1) presents the estimation results without accounting for control variables and fixed effects. Column (2) includes county and year fixed effects. Columns (3) and (4) incorporate control variables based on the specifications in columns (1) and (2), respectively. The results show that the coefficients of treat are significantly positive in all cases, indicating that the Information Accessibility Pilot Project significantly enhances county-level digital entrepreneurial vitality. Taking the results in column (4) as the benchmark, the estimated coefficient of the Information Accessibility Pilot Project is 0.1586, indicating that, holding other conditions constant, the policy increased county-level digital entrepreneurial vitality by approximately 17%. Specifically, this policy effect corresponds to an average increase of about 2 newly registered digital enterprises per county per year, accounting for roughly 16% of the county-level mean value of digital entrepreneurial vitality. This demonstrates that the entrepreneurial promotion effect of the Information Accessibility Pilot Project has substantial economic significance. As mentioned earlier, the implementation and promotion of the pilot policy not only provide good policy support for pilot cities but also boost the confidence of venture investors, attract innovative elements such as talent and technology, and create a favorable institutional environment for entrepreneurship, thereby promoting county-level digital entrepreneurial vitality.

5.3. Parallel Trend Test

Baseline regression results indicate that counties located in “National Pilot Policy of Information Accessibility” pilot cities are associated with higher levels of digital entrepreneurial activity compared to those in non-pilot cities. However, this raises a potential concern: could such differences have already existed prior to the policy implementation? In other words, might the government have been inclined to select cities with higher pre-existing entrepreneurial activity as pilot sites? These potential sources of bias warrant further examination. Therefore, it is essential to determine whether the sample satisfies the parallel trends assumption—that is, whether counties in pilot and non-pilot cities exhibited broadly similar trends or no significant systematic differences in digital entrepreneurial activity prior to the implementation of the pilot program. In 2014, the National Development and Reform Commission, along with 12 other departments, designated 80 cities, including Shenzhen, as National Pilot Cities for Information Accessibility. Consequently, this study uses 2014 as the base year for the implementation of this policy. Following the approach of Beck et al. (2010) [39], the event study method is employed to examine whether the treatment and control groups exhibit parallel trends and to investigate the dynamic patterns associated with the entrepreneurial effects of the “Information Accessibility Pilot Cities.” The model is set as follows:

$${\mathrm{Entre}}_{\mathrm{it}}=\mathsf{\alpha }+{\sum }_{k=-4}^6{\mathsf{\beta }}_k{\mathrm{T}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{t}}_{i,t_0+k}+\mathsf{\gamma }{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\mu }}_{\mathrm{i}}+{\mathsf{\lambda }}_{\mathrm{t}}+{\mathsf{ϵ}}_{\mathrm{it}}$$

(2)

As shown in Figure 2, before the policy implementation, the estimated coefficients of the pilot group are statistically insignificant, suggesting no systematic differences in digital entrepreneurial activity between treatment and control counties prior to the policy. This indicates that the parallel trends assumption is not violated. After the policy implementation, the promotion effect of the Information Accessibility Pilot Project on digital entrepreneurship shows a trend of first increasing and then decreasing, consistent with the policy promotion process. In June 2014, the National Development and Reform Commission announced that 80 cities, including Shenzhen, had been selected as National Pilot Cities for Information Accessibility, and the pilot city work plans were approved in principle. In April 2016, the General Office of the State Council issued an implementation plan for promoting “Internet + Government Services” and implementing the “Information Accessibility Pilot Project,” proposing to accelerate the implementation of “Internet + Government Services” and deeply implement the Information Accessibility Project, aiming to establish a convenient, fair, inclusive, and efficient government service system. The plan proposed a two-step implementation within two years, with 80 pilot cities achieving the goal of “one application, one window acceptance, one network handling” within two years, significantly optimizing service processes, diversifying service models, and significantly improving public satisfaction with government services. By 2016, pilot cities had largely achieved the integration of government service processes, enabling “one ID for application, one-window acceptance, and one-network processing” for administrative service items. As a result, the policy demonstrated a significantly positive impact in the first two years of implementation. However, due to initial instability in policy outcomes and the fact that by 2017, cross-regional, cross-level, and cross-departmental coordination of government services had been preliminarily achieved among pilot cities—with over 80% of basic public services accessible online and replicable practices being gradually promoted nationwide—the effect of this exogenous shock in 2017 showed a temporary decline. Nevertheless, as the years of construction under the “Information Benefits to the People” pilot program increased, the digital transformation of local government governance continued to deepen. Consequently, the observed association between the policy and digital entrepreneurship strengthened overall.

5.4. Robustness Test

5.4.1. Placebo Test

To prove that the treatment effect comes from the establishment of the National Pilot Cities for Information Accessibility, a placebo test is conducted. The specific method is to randomly select some counties as the pseudo-experimental group of “Information Accessibility Pilot Cities,” and the remaining counties as the pseudo-control group. At the same time, a random year is selected for the pseudo-experimental group as the policy implementation time, and the wrong regression coefficients are obtained for the pseudo-samples. This process is repeated 500 times, and the probability distribution is shown in Figure 2. As shown in Figure 3, the placebo test error coefficients are symmetrically distributed around zero and exhibit clear characteristics of a normal distribution, with most p-values exceeding 1. In contrast, the actual policy effect estimate is 0.159, which significantly differs from the placebo results. Therefore, the results are consistent with the expectations of the placebo test. The findings suggest that there are no significant unobserved factors influencing the regression outcomes, thereby confirming the robustness of the baseline regression results.

5.4.2. PSM-DID

Considering the potential selection bias towards economically developed regions in the selection process of “Information Accessibility Pilot Cities,” the Propensity Score Matching-Difference-in-Differences (PSM-DID) method is used for robustness testing, referring to the approach of Bai Junhong et al. (2022) [40]. The nearest neighbor matching method is employed, and the covariates used for matching are consistent with the control variables. To ensure the validity of the PSM-DID approach, this study first conducted a covariate balance test between the treatment and control groups. Figure 4 illustrates the changes in standardized bias of the covariates before and after matching. The results show that, prior to matching, the standardized biases of most covariates were relatively large, with some variables (such as fiscal, ict, and lngdp) exceeding 30%, indicating substantial differences between groups. After applying propensity score matching, all covariate biases dropped significantly and approached zero, meeting the commonly accepted 10% threshold for standardized bias. This indicates that the matched treatment and control groups achieved satisfactory balance in covariates, thereby fulfilling the preconditions for applying the PSM-DID identification strategy. As shown in column (1) of

Table 3. Robustness Test.

Variables	(1)	(2)	(3)	(4)	(5)
	PSM-DID	New Digital Enterprises	Excluding Other Policies 1	Excluding Other Policies 2	Excluding the Pandemic
treat	0.1588 ***	23.0276 ***	0.1587 ***	0.1264 **	0.1522 ***
	(3.2007)	(2.8853)	(3.2670)	(2.5641)	(3.1687)
lngdp	−0.1003	9.7655	−0.1223 **	−0.1034 *	0.0393
	(−1.6128)	(1.4643)	(−1.9820)	(−1.6669)	(0.5922)
indust_stru	0.7444 ***	18.2024	0.6753 ***	0.6841 ***	0.9018 ***
	(3.9540)	(1.4371)	(3.7475)	(3.6789)	(4.7121)
fiscal	0.0132	30.5955	−0.0118	−0.0116	−0.1015
	(0.1191)	(1.0750)	(−0.1126)	(−0.1061)	(−0.9724)
finance	−0.0695	26.2182	−0.0629	−0.0764	−0.0849
	(−1.2919)	(1.2711)	(−1.1807)	(−1.4276)	(−1.6091)
pop_den	20.3728 ***	915.9411 **	12.2846 ***	20.3006 ***	18.3547 ***
	(3.7083)	(2.2430)	(3.1441)	(3.7251)	(3.7245)
ict	−0.0058	−0.3385	0.0070	−0.0043	−0.0042
		(−0.2595)	(0.4214)	(−0.2471)	(−0.2262)
big_data			0.4325 ***
			(8.1563)
broadband				0.1189 ***
				(3.0324)
Constant	2.4072 ***	−127.2870	2.7068 ***	2.4489 ***	1.0832
	(3.5442)	(−1.4782)	(4.1302)	(3.6467)	(1.5347)
Observations	10,522	10,578	10,578	10,578	9943
R-squared	0.3506	0.0248	0.3649	0.3527	0.3609
Number of countries	1200	1206	1206	1206	1206
country fe	yes	yes	yes	yes	yes
year fe	yes	yes	yes	yes	yes

Robust t-statistics in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

, the coefficient of treat is significantly positive at the 1% level, indicating that the “Information Accessibility Pilot Cities” construction significantly enhances county-level digital entrepreneurial vitality, and the conclusion is robust.

5.4.3. Replacing the Measure of Digital Entrepreneurial Vitality

As mentioned earlier, entrepreneurial vitality is usually measured using a standardized base. To facilitate comparison, this study follows the method of Bai Junhong et al. (2022), directly using the unstandardized number of new digital enterprises to measure entrepreneurial vitality [40]. As shown in column (2) of Table 3, the coefficient of treat is significantly positive at the 1% level, indicating that the National Pilot Policy of Information Accessibility significantly increases the number of new digital enterprises in counties, positively impacting digital entrepreneurial activities, and the results are robust.

5.4.4. Excluding the Interference of Other Policies

During the study period and within the sample range of this paper, there may have been other pilot policies overlapping with the effects of the “National Pilot Policy of Information Accessibility”, potentially leading to an underestimation or overestimation of its impact on digital entrepreneurial vitality. A review of existing literature reveals two highly relevant policies: the Broadband China strategy introduced in 2013, and the establishment of national big data comprehensive pilot zones in successive batches after 2016. Therefore, this study controls for these two policies’ dummy variables in the baseline regression model to minimize their impact on the estimation results. As shown in columns (4) and (5) of Table 3, the coefficients of the Information Accessibility Pilot Cities dummy variable remain significantly positive after controlling for these two policies, indicating a significant entrepreneurial effect of the Information Accessibility Policy. The policy significantly enhances county-level digital entrepreneurial vitality by breaking information silos, stimulating digital entrepreneurial enthusiasm in counties, and the results remain robust.

5.4.5. Excluding the Impact of the Pandemic

The COVID-19 pandemic outbreak in 2020 had a significant impact on China’s economic and social development, affecting county-level digital entrepreneurial vitality. To exclude the pandemic’s interference, this study refers to the approach of He Xiaoyu and Chu Deyin (2023) and removes the 2020 sample [41], electing data from 1206 counties from 2010 to 2019 for regression analysis. As shown in column (5) of Table 3, the coefficient of treat remains significantly positive at the 1% level, indicating that the Information Accessibility Pilot Cities construction significantly enhances urban entrepreneurial vitality even after excluding the 2020 pandemic interference, confirming the robustness of the conclusion.

5.5. Mechanism Test

From the theoretical analysis, government digital governance mainly promotes county-level digital entrepreneurial vitality through enhancing regional innovation, improving local employment structure, and strengthening infrastructure. Based on existing studies, for mechanism testing, this study follows the method of Wen Zhonglin et al. (2004) [42]. However, recent research by Jiang Ting (2022) [43] finds that this method may have issues such as omitted variables, endogenous variables, and high correlation between explanatory and mechanism variables. To systematically identify the transmission pathways through which the “National Pilot Policy of Information Accessibility” affects county-level digital entrepreneurial vitality, this study adopts the mediation analysis framework proposed by Jiang Ting (2022) [43] in empirical research on causal inference. This approach is not only intuitive and well-suited to panel data structures, but also enables the identification of the direct impact of policy variables on mediating variables while controlling for fixed effects, thereby improving the precision of mechanism identification and the explanatory power of theoretical models. Specifically, the Jiang method offers three key advantages for identifying causal mechanisms: First, it clearly defines the transmission logic between the dependent and mediating variables, avoiding the omission of key mechanisms or the conflation of causal pathways. Second, its model specification is straightforward, thus circumventing the complexity and potential bias associated with interpreting multiple regressions in the traditional “three-step mediation” method. Third, its theoretical structure aligns well with the transmission logic of policy interventions, making it particularly suitable for analyzing mediation effects in the context of quasi-natural experiments. Drawing on this methodology, the present study conducts separate regressions for three mediating variables—regional innovation capacity, employment structure, and infrastructure development—to rigorously and scientifically reveal how government digital governance enhances county-level digital entrepreneurship through multiple causal pathways. Building upon Equation (1), we specify the models for the mechanism analysis as follows:

M_it = α + βTreat_it + γX_it + μ_i + λ_t + ϵ_it

where M is the intermediary variable, sequentially replaced with three variables representing regional innovation capability, marketization level, and infrastructure construction. Other variables have the same meaning as before. If the coefficient β is significant, the intermediary effect is established.

First, regarding regional innovation capability, this study uses the logarithm of the number of patent applications (ln_innov) as a measure, following the research of Dai Xinling (2022) [44]. This indicator emphasizes innovation output results and comprehensively reflects regional innovation capability. As shown in column (1) of Table 3, the coefficient of ln_innov is 0.4238 and is significantly positive at the 1% level. As a mediating variable, the significance of regional innovation capacity indicates that the Information Benefits to the People policy effectively promotes the aggregation and transformation of entrepreneurial factors by enhancing the integration and reinvention capacity of local technological resources. This result supports the “capability empowerment” pathway hypothesis, suggesting that the strengthening of technological supply and knowledge diffusion capacity is a crucial condition for boosting county-level digital entrepreneurial vitality.

Second, employment structure. Following the research of Gillian Stevens (2004), this study uses the ratio of employment in the tertiary industry to total employment to represent regional employment structure (employ_str) [45]. Given the potential for reverse causality between third-sector employment share (employ_str) and digital entrepreneurial vitality—whereby a service-oriented employment structure may facilitate entrepreneurship, but the rise in digital startups may also reshape the employment composition—we adopt a lagged specification of employ_str to mitigate simultaneity bias. Specifically, we use the one-period lag of employ_str (L1.employ_str) in our mechanism regressions to ensure the temporal precedence of structural employment conditions relative to entrepreneurial responses. This approach strengthens causal inference by reducing contemporaneous endogeneity, while maintaining other mechanism variables and fixed effects to preserve estimation robustness. Although the use of lagged variables cannot fully eliminate endogeneity, it remains a widely accepted and practical strategy under county-level data constraints. As shown in column (2) of

Table 4. Mechanism Test Results.

Variables	(1)	(2)	(3)
	Innov	L1.Employ_str	Instru
treat	0.4238 ***	0.0180 ***	9.7136 ***
	(4.6257)	(3.3657)	(3.0171)
lngdp	0.1593	−0.0314 ***	5.6385 *
	(1.3475)	(−4.2524)	(1.7003)
indust_stru	−0.5115	−0.0426 **	−35.9206 **
	(−1.3566)	(−2.0831)	(−2.1574)
fiscal	0.3232 **	−0.0453 ***	−4.6981
	(2.1014)	(−4.9548)	(−1.2054)
finance	−0.0373	−0.0060	−27.7186
	(−0.3485)	(−1.1006)	(−1.4838)
pop_den	11.6650 **	−0.3408	−156.8514
	(2.1129)	(−0.6909)	(−1.5695)
ict	−0.1075 ***	0.0011	1.4780
	(−2.9367)	(0.5257)	(0.6874)
Constant	7.5702 ***	0.9267 ***	−20.9287
	(5.7663)	(11.9992)	(−0.4436)
Observations	7968	7145	7945
R-squared	0.1548	0.2589	0.0129
Number of countries	924	825	915
country fe	yes	yes	yes
year fe	yes	yes	yes

Robust t-statistics in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

, the coefficient of employ_str is 0.0108 and significantly positive at the 5% level, indicating that the “Information Accessibility Pilot Cities” construction promotes digital development, changes the employment structure, and enhances digital entrepreneurial vitality. With the integration of digital technology and government promotion of digitalization, traditional industries begin to transform, and emerging industries create numerous new entrepreneurial opportunities and pathways. The employment structure optimization mechanism indicates that industrial upgrading and talent attraction effects driven by the policy have directed more high-quality labor toward the digital economy sector. This structural shift enhances the efficiency of human resource allocation and represents a “human capital pathway” for promoting increased entrepreneurial vitality. This mechanism confirms the practical effect of the policy in unlocking the potential of digital entrepreneurship through the release of talent dividends.

Lastly, infrastructure level. Infrastructure construction level is a crucial indicator of a region’s entrepreneurial environment and attractiveness. This study follows Zheng Wei (2023) and uses the per capita road freight volume at the city level to represent regional infrastructure level (instru) [46]. As shown in column (3) of Table 4, the coefficient of instru is 9.7136 and significant at the 1% level, indicating that the deepening construction of “Information Accessibility Pilot Cities” enhances government digital governance, improving urban infrastructure efficiency and quality, increasing adaptability and sustainability, which is crucial for enhancing county-level digital entrepreneurial vitality. As the “physical carrier” of entrepreneurial activity, infrastructure plays a critical role in lowering entry barriers, improving information accessibility, and expanding market reach. Its significance confirms that digital entrepreneurship is highly sensitive to infrastructural conditions, constituting an “environmental support pathway” through which the policy exerts its impact.

The above analysis shows that although regional innovation capability, employment structure, and infrastructure level all have significant positive impacts on county-level digital entrepreneurial vitality, infrastructure level and regional innovation capability contribute the most, while employment structure contributes the least. This suggests that in future development, in addition to fully leveraging the regional innovation effects of the Information Accessibility Policy, efforts should be made to enhance infrastructure construction and attract and incentivize talent to alleviate employment structure constraints of new enterprises, thereby better promoting county-level digital entrepreneurial vitality.

6. Further Analysis

China’s vast territory exhibits significant regional differences in government governance efficiency and marketing environment. Given the regional disparity in government service efficiency, service capability, and other aspects, the analysis based on the overall sample may obscure the potential differences in the impact of digital governance transformation on county-level digital entrepreneurial vitality in different types of counties. Therefore, this study divides the sample cities into eastern, central, and western sub-samples based on their regional distribution and conducts heterogeneity tests. The specific regression results are shown in

Table 5. Heterogeneity Analysis at the County Level.

Variables	Eastern Counties	Central Counties	Western Counties
	Entre	Entre	Entre
treat	0.0661	0.1034	0.2152 ***
	(0.6243)	(1.3115)	(3.1936)
lngdp	0.6009 ***	−0.0871	−0.0938
	(3.6088)	(−0.8887)	(−1.1027)
indust_stru	−0.1616	1.6797 ***	0.6398 ***
	(−0.3511)	(5.2070)	(2.6830)
fiscal	−0.3796 *	0.4843 ***	0.0636
	(−1.8499)	(2.7901)	(0.3855)
finance	0.4401 ***	−0.4928 ***	0.0741
	(3.0732)	(−4.1514)	(1.1585)
pop_den	6.7766	18.9282 ***	25.2001
	(1.3933)	(3.1765)	(1.6368)
ict	0.1454 **	0.0282	0.0145
	(2.5015)	(1.0409)	(0.5922)
Constant	−5.7974 ***	1.8228 *	1.9925 **
	(−3.0050)	(1.7457)	(2.1629)
Observations	2038	4132	4408
R-squared	0.5024	0.3987	0.2862
Number of countries	245	508	453
country fe	yes	yes	yes
year fe	yes	yes	yes

Robust t-statistics in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

.

The regression results indicate that the impact of the “Information Accessibility Pilot Cities” policy on county-level digital entrepreneurial vitality varies by region. In the developed eastern region, the policy’s promotion effect on county-level digital entrepreneurial vitality is not significant; in the relatively developed central region, the policy’s impact is marginally significant at the 10% level; in the underdeveloped western region, the coefficient is 0.2152 and significant at the 1% level. From an economic perspective, this indicates that, holding other conditions constant, the policy increased county-level digital entrepreneurial vitality in western regions by approximately 24%, suggesting that in less-developed areas the entrepreneurial promotion effect of the policy is more pronounced and carries substantial policy significance. This suggests that government digital governance transformation can effectively overcome the economic development lag, low marketization level, and governance inefficiency in the central and western regions, significantly stimulating market vitality. The regional heterogeneity may be due to the superior governance efficiency, business environment, service level, and service capability in the developed eastern region compared to other regions, as well as the more developed commercial intermediary organizations, providing potential county entrepreneurs with multiple channels to obtain information and government services. Therefore, digital government construction has limited impact on improving governance efficiency, helping entrepreneurs obtain information, and serving entrepreneurs in the eastern region, resulting in limited influence on county-level digital entrepreneurial vitality. In the underdeveloped western region, the government’s digital governance transformation significantly improves governance efficiency, business environment, service efficiency, and service capability, while the lack of developed commercial intermediary organizations makes government digital governance crucial for entrepreneurs to obtain information and services, thus significantly impacting county-level digital entrepreneurial vitality. The central region’s governance efficiency, business environment, and commercial intermediary organizations are intermediate between the eastern and western regions, leading to an intermediate impact of government digital governance transformation on county-level digital entrepreneurial vitality.

It is important to note that the heterogeneity analysis reveals a significant positive effect of the Information Benefits to the People policy on county-level digital entrepreneurial vitality in western regions, whereas no statistically significant effect is observed in the eastern and central regions. This result requires a nuanced interpretation that considers both statistical power and regional characteristics. First, the “non-significant” findings in the eastern and central regions should not be directly interpreted as evidence of policy ineffectiveness. Further examination of sample sizes and residual variance reveals that the eastern region, affected by a limited number of pilot cases, has a relatively small sample size, resulting in larger estimation variance and potentially insufficient statistical power. Similarly, while the central region has a more adequate sample, the relatively high baseline level of infrastructure may limit the marginal improvement space, causing the policy effect to appear weaker and not statistically significant at conventional levels. Therefore, the non-significance in these regions should be understood as “effect uncertainty” rather than “absence of effect.” Second, the significant finding in the western region suggests that the policy exerts stronger marginal incentive effects in less-developed areas. Digital infrastructure investment, policy support, and factor integration generate greater institutional leverage in these regions, reflecting differentiated policy impacts. In summary, we recommend interpreting the non-significant results in the eastern and central regions as outcomes of limited statistical power and structural regional differences, rather than as indications of policy ineffectiveness, to avoid misjudging the heterogeneity of policy effects.

7. Discussion

7.1. Comparison with Previous Studies

The empirical results confirm that the National Pilot Policy of Information Accessibility significantly enhances county-level digital entrepreneurial vitality. This is consistent with prior findings in China and other contexts, which emphasize the role of digital governance in reducing transaction costs, improving institutional efficiency, and stimulating entrepreneurial activity [13,26]. Notably, by focusing on the county level, this study enriches existing research that has largely concentrated on cities or provinces. The stronger effects in western counties further highlight that government-led digital governance is particularly effective in regions where commercial intermediaries and market forces are less developed, echoing the institutional leverage of policy in less-advantaged contexts.

7.2. Theoretical Implications

The findings contribute to theoretical discussions by confirming three major transmission pathways. First, digital governance enhances regional innovation capacity, consistent with the knowledge spillover theory [29], as technological resources are aggregated and transformed into entrepreneurial opportunities. Second, the optimization of employment structure demonstrates the “human capital pathway”, where industrial upgrading and talent attraction improve labor allocation efficiency and entrepreneurial potential. Third, the improvement of infrastructure reflects the “environmental support pathway”, providing the physical foundation for lowering entry barriers, improving information accessibility, and expanding market reach. Taken together, these mechanisms underscore that digital governance should be regarded not only as an administrative reform, but also as a systemic institutional innovation that reshapes entrepreneurial ecosystems in less-developed regions.

7.3. Boundary of Generalization

The heterogeneity analysis further illustrates that policy effects are not uniform across regions. Significant positive impacts are found in western counties, while the effects are weaker or statistically insignificant in eastern and central regions. This pattern does not indicate policy ineffectiveness, but rather reflects structural differences in baseline conditions: eastern counties already possess relatively advanced governance efficiency and intermediary organizations, leaving less marginal space for improvement, whereas western counties benefit more strongly from institutional interventions. These results emphasize the importance of considering regional characteristics when evaluating the transferability of digital governance effects.

7.4. Contribution to Policy and Research

Overall, this study highlights that the National Pilot Policy of Information Accessibility represents an important institutional innovation that strengthens innovation, optimizes employment structure, and improves infrastructure, thereby stimulating digital entrepreneurship at the grassroots level. By focusing on counties, the study extends the scope of digital governance research beyond urban and provincial contexts, and offers new theoretical and empirical evidence for understanding how government-led digital transformation can support inclusive economic development.

8. Conclusion and Policy Implications

Currently, the digital economy is rapidly developing worldwide, and China is accelerating the construction of a digital nation. The critical role of digital infrastructure and digital finance in stimulating economic vitality has been widely recognized by scholars and policymakers. However, research on the economic impact of digital governance remains insufficient. In the process of China’s modernization, the construction of a digital government aligns with the demand for modernizing national governance, and digital transformation of government governance is becoming an action direction for all levels of government to seize opportunities in the digital era and build modern governance capabilities. It plays an important role in optimizing public services, enhancing residents’ sense of gain, and stimulating enterprise vitality. Therefore, it is worth exploring how government digital governance can stimulate digital entrepreneurial vitality in the digital economy era.

8.1. Conclusions

This study uses the registration data of Chinese enterprises from 2010 to 2020 for 1206 counties in China and employs the difference-in-differences method to analyze whether government digital governance construction can enhance county-level digital entrepreneurial vitality using the “National Pilot Policy of Information Accessibility” as a quasi-natural experiment. The research conclusions are as follows: First, the National Pilot Policy of Information Accessibility significantly stimulates county-level digital entrepreneurial vitality. The conclusion is further supported by a series of robustness tests, including parallel trend tests, placebo tests, PSM-DID method, excluding the impact of the pandemic, and other policies. Second, government digital governance construction mainly enhances county-level digital entrepreneurial vitality through three pathways: improving regional innovation capability, optimizing employment structure, and strengthening infrastructure. Third, sub-sample regression shows that the impact of government digital governance construction on county-level digital entrepreneurial vitality varies by region, with a more significant incentive effect in central and western counties than in eastern cities.

8.2. Policy Implications

Based on this, the following policy implications are derived:

First, efforts should focus on narrowing the urban–rural digital divide and extending infrastructure coverage deeper into county-level regions. At present, county areas in China lag significantly behind urban centers in terms of information infrastructure and digital literacy, creating a pronounced digital gap. Addressing this requires sustained fiscal input in broadband, 5G, and training, which entails high costs and long-term commitment. The time horizon of such policies is long-term, as improvements in infrastructure and digital skills will only gradually translate into sustained entrepreneurial vitality.

Second, greater coordination is needed within county-level governance systems to dismantle the fragmented structure of “multiple authorities and segmented management”. Fragmented standards and barriers to data sharing not only reduce efficiency but also create risks of unequal access to services. Building integrated government service platforms and unified data standards requires governance capacity, and the success of such reforms depends on the ability of local governments to mobilize resources and overcome institutional inertia.

Third, it is necessary to develop a differentiated digital entrepreneurship ecosystem tailored to the diversity of county contexts. Counties vary greatly in terms of economic structure, labor mobility, and industrial foundations, and “one-size-fits-all” approaches are likely to fail. Policies should therefore grant local governments sufficient flexibility to combine digital governance with local comparative advantages, such as “agriculture + digital” or “tourism + digital”. The risks here lie in policy misalignment or resource misallocation, but with adequate governance capacity, such strategies can produce both short-term benefits (entrepreneurial entry) and long-term outcomes (enterprise survival and innovation).

Fourth, dynamic evaluation and incentive mechanisms should be strengthened to enhance local government proactiveness. Establishing multidimensional evaluation systems and linking them with fiscal transfers can ensure that resources are allocated based on performance. However, this also carries risks of formalism or distorted incentives. The success of such performance-based incentives depends on transparent governance and the careful design of indicators. Furthermore, evaluations should consider not only immediate outcomes but also longer-term entrepreneurial quality, thereby balancing short-term signals with sustainable development.

Taken together, these policy implications demonstrate that government-led digital governance is most effective when accompanied by sufficient fiscal support, strong governance capacity, and coordinated institutional arrangements. Under such conditions, digital governance can not only stimulate immediate entrepreneurial activity but also foster sustained entrepreneurial development and contribute to the long-term transformation of county-level economies.

8.3. Limitations

Although this study has deeply explored the impact of government digital governance on county-level digital entrepreneurial vitality, there are still some limitations that need to be addressed in future research. First, the study relies on county-level registration data from 2010 to 2020, which effectively captures entrepreneurial activity but does not reflect firm quality dimensions such as survival, growth, or innovation. Moreover, the outcome measure is narrowly focused on IT-related and information service enterprises, which limits the ability to generalize results to other sectors. Second, all counties within pilot cities are classified as treated, yet significant intra-city heterogeneity exists—particularly in top-tier cities with stronger infrastructure and talent pools—which may bias estimates and suggests the need for more granular designs. Third, the allocation of pilot policies was not random and may have been influenced by fiscal capacity, industrial structure, or institutional readiness, raising concerns of potential selection bias. Fourth, the results may be confounded by overlapping public policies, such as the Broadband China strategy and national big data pilot zones, making it difficult to fully isolate the effects of the Information Accessibility pilot. Fifth, the year 2020 presents a unique context due to the COVID-19 pandemic and related policy interventions, which may have distorted both enterprise registration and governance outcomes, rendering results particularly sensitive to this year. Finally, the associations identified in this study should be interpreted as correlational rather than strictly causal. Although robustness checks mitigate some endogeneity concerns, the evidence cannot definitively establish causal mechanisms. These limitations should be borne in mind when applying the conclusions to policy or theoretical generalizations.