Digital Government Development, Local Governments’ Attention Distribution and Enterprise Total Factor Productivity: Evidence from China

Abstract

Building digital government is an important means for the government to improve the public service ability and optimize the business environment, which directly affects the production and operation activities of micro-enterprises. Based on the panel data of listed enterprises and municipal government portal website performance in China, this paper empirically investigates the impact of digital government development on enterprise total factor productivity (TFP) and the moderating effect of the local government’s attention distribution. The research results showed that digital government development significantly improved the enterprise TFP, and this conclusion remained unchanged after a series of robustness tests using instrumental variables, one-stage lag of explained variables, and debiased machine learning models. We also found that the greater the pressure faced by local governments and the longer the chief officials’ tenure, the more attention local governments paid to building digital government, and the more obvious the role of digital government development in promoting enterprise TFP. Heterogeneity test results showed that the information disclosure, online service, and public participation all had a positive effect on enterprise TFP, while the user experience had no effect on it. Digital government development had a more obvious role in promoting enterprise TFP of central and western regions, non-SOEs, and technology-intensive enterprises. Moreover, reducing enterprise rent-seeking, attracting new enterprise entry, and increasing enterprise R&D investment are important mechanisms for digital government development to improve enterprise TFP.

1. Introduction

China is in a critical period of economic transformation and upgrading. Improving enterprise TFP by optimizing the business environment, and then promoting high-quality economic development is the top priority of the work of local governments at all levels. The new generation of digital technology represented by the Internet, big data, artificial intelligence (AI), and block chain brings human society into a new “digital era” and gives rise to a new digital government model which is the organic integration of digital technology and governance theory [1,2,3,4]. From the central government to government at the county level, digital technology has been used to optimize the government service process, and governance innovation practices represented by government portal websites, government Weibo, and open data platforms have continued to spring up. According to the “UN GOVERNMENT SURVEY 2020”, China’s e-government development index rose from 0.681 in 2018 to 0.795 in 2020, ranking 20 places higher than in 2018 and reaching a “very high” level of global e-government development. Building digital government has gradually become an important means for the local government to improve organizational operation efficiency [5] and the positive correlation between digital government development and the optimization of the business environment has been recognized by the academic community [4,6]. Thus, can digital government development improve enterprise TFP? The answer to this question will not only help to objectively evaluate the economic effect of digital government development, but also provide new insights for improving enterprise TFP.

With the wave of building digital government sweeping the world, the academic community has examined digital government’s connotation, performance, influencing factors, and governance logic, obtaining abundant achievements. Some empirical studies on the economic effect of digital government development have mainly focused on its positive influence t on curbing corruption [7], attracting foreign direct investment [8], promoting the development of digital economy [9] and other fields from a macro-perspective. Compared to those studies, micro-perspective research is much less, and only a few studies analyze the impact of digital government development on enterprise innovation [10,11]. However, as the micro-foundation of high-quality economic development, whether and how enterprise TFP is affected by the development of digital government has not attracted due care. Moreover, China’s special political system environment and the important role of local governments and chief officials in the process of building digital government have been ignored by existing studies when identifying the consequences of digital government development.

According to the technology application theory, the internal factors and external environment of an organization will determine the effect of technology application. Over the years, under the vigorous promotion of the central government, China has made great progress building digital government, but the development and the implementation effect of digital government varies significantly in different regions, and there are still many problems such as “zombie websites” and “information islands”. While the economic development level, public demand, cultural characteristics, and other factors can explain the differences in building regional digital government [12], the influence of digital technology practice scenarios cannot be ignored. In China, the political system, characterized by inter-governmental relations and personnel promotion, constitutes the institutional environment for the application of government digital technology [13,14]. As a typical “top managerial” project, the degree of emphasis and attention distribution of local governments and chief officials have a direct impact on the process of building digital government and its implementation effect. Fan et al. [5], Tan et al. [14], Li and Ma [15], and Xiao et al. [16], based on digital governance practices of China’s local governments, found that institutional factors, such as technology application ability, peer competition pressure, and attention distribution of chief officials, would affect the application of digital technology by local governments. Therefore, the identification of the economic effects of digital government development would be biased if the initiative of local governments and chief officials was not regarded.

In view of this, based on the panel data of listed enterprises and cities in China, this paper investigates the impact of digital government development on enterprise TFP and the moderating effect of local government attention. The contribution of this paper is reflected in three aspects: firstly, the building of digital government is growing vigorously and has a profound impact on the economy and society. However, research on the effect of digital government development is focused on practical instruction and experience summarization, and large-sample empirical research is relatively rare. Although Wang et al. [10], Qu and Wang [11] analyzed the effect of digital government development on enterprise innovation, the change of enterprise resource allocation efficiency in the context of digital government development has not been noted. In a study by Zhang et al. [17], the impact of building service-oriented government on enterprise TFP is examined by PAT (Public Affairs Television), mainly according to the logic of strengthening mass supervision-perfecting business environment-improving production efficiency. However, building digital government relies on formal systems to regulate local governments’ behavior and strengthen their public service functions. The influence of digital government development on enterprise business environments is more profound and comprehensive. This paper examines the impact of digital government development on enterprise TFP, supplying new empirical evidence for identifying the economic effects of digital government development and providing useful references for improving enterprise TFP.

Secondly, this paper investigates the mechanism of building digital government affects enterprise TFP from three aspects: enterprise rent-seeking, new enterprise entry, and enterprise innovation, which is helpful for opening the mechanism “black box” of the impact of digital government development on micro-enterprise. Meanwhile, we decompose the building of digital government into four subdivisions: disclosure, online services, public participation, and user experience, and analyze the heterogeneous effects conducive to revealing the micro-effects of building digital government and understanding digital government practices.

Finally, in China, local governments have different attitudes towards building digital government, and its process and implementation effect vary greatly in different regions; yet, the reasons behind this are ignored in existing studies [10,11]. Local governments’ behavior is the specific externalization of the system and China’s special political system is an important perspective to explain China’s growth miracle. It is also necessary to combine the political system and the application scenarios of digital technology to understand the incentives of local governments to strengthen the building of digital government and to examine the effect of digital government development. Within the context of technology application, combined with the characteristics of China’s political system and local governments’ behavior, this paper investigates the influence of the local governments’ attention distribution on the relationship between building digital government and enterprise TFP. This study can provide theoretical support for governments at all levels to comprehensively promote the process of building digital government.

2. Theoretical Analysis and Research Hypotheses

2.1. The Impact of Digital Government Development on Enterprise TFP

Digital government development can promote the integration of offline and online government services, consolidate public data resources, strengthen business collaboration, and optimize government services, which provide a fairer business environment for market entities, thus affecting enterprise production and resource allocation [2,6]. Firstly, digital government development can improve information transparency, regulate government behavior, reduce enterprise rent-seeking, and then improve enterprise TFP. According to Robert, the condition of corruption occurrence is equal to monopoly power, adds discretion power and subtracts responsibility [18]. Theoretically, the effective control of “monopoly power” and “discretionary power” and the improvement of supervision levels can minimize the occurrence of corruption. However, under the Sinicism decentralization system, “monopoly power” and “discretionary power” are highly combined, and there is a relatively serious information asymmetry between the public and the government, which greatly reduces the cost of government officials’ corruption behavior [19]. In addition, due to an imperfect market system, establishing a good relationship with government officials can help enterprises obtain more resources, such as government subsidies, credit sources, and business qualifications. Therefore, enterprises are also keen on rent-seeking behavior. This not only disturbs the normal market order, but also distorts the innovation and development strategy of enterprises, and seriously hinders enterprise technological progress and the promotion of TFP [20]. The governance reform driven by digital technology reduces the information asymmetry between the government and the public [10] and broadens the channels for the public to supervise the government’s behavior [21], which can regulate the use of local governments’ “monopoly power” and “discretionary power” and contribute to the “power operating transparently”. Additionally, the building of digital government establishes a new communication mechanism between the government and enterprises, and it can compress enterprise rent-seeking space, inhibit enterprise rent-seeking behavior, and encourage enterprises to seize competitive advantages by improving production efficiency.

Secondly, digital government development can attract new enterprise entry and improve enterprise production efficiency through the “learning effect” and “competition effect”. Building digital government is essential to re-engineer government public service processes and improve the efficiency of government services through the co-construction and sharing of government information, which is conducive to improving the relationship between the government and the market, optimizing the business environment, and thus attracting new enterprises [6,17]. The entry of new enterprises will promote the spread and application of external new technologies and ideas in the local market, accelerate knowledge spillover and knowledge diffusion, create favorable conditions for mutual cooperation and communication, and is conducive to improving the production efficiency of local enterprise. At the same time, with the entry of new enterprises, the market competition will be further intensified. In order to obtain or maintain competitive advantages, enterprises can only choose to continuously optimize resource allocation and improve TFP.

Finally, digital government development can stimulate enterprise innovation vitality and promote enterprise TFP. Digital government development compresses enterprise rent-seeking space, which not only reduces the possibility of entrepreneurs replacing “innovation strategy” with “rent-seeking strategy”, encouraging enterprises to focus on productive activities, but also reduces enterprise non-productive expenditure, so that enterprises have more funds for technological innovation activities. Meanwhile, with the optimization of the business environment, more resources will flow into high-efficiency areas. Faced with competitive pressure, enterprises have stronger demands for technological innovation, transformation and upgrading in the pursuit of profit maximization, and accelerate the pace of technology research and development. Technological innovation is a fundamental method to improve enterprise TFP, and the increase in innovation investment will directly promote enterprise TFP. Based on the above analysis, our first hypothesis is stated as follows:

H1. 

Digital government development can improve enterprise TFP.

H1a. 

Digital government development improves enterprise TFP by reducing rent-seeking.

H1b. 

Digital government development improves enterprise TFP by attracting new enterprise entry.

H1c. 

Digital government development improves enterprise TFP by stimulating enterprise innovation.

2.2. The Moderating Effect of the Local Governments’ Attention Distribution

The building of digital government is an innovative practice of the government to improve public service ability by using digital technology. According to the technology application theory, the institutional environment will affect an organization’s technology adoption and its implementation effectiveness, and the desired results can only be achieved when the new technology is compatible with its operational environment [22,23]. In China, the political system is characterized by inter-governmental relations and personnel promotion that constitute the institutional environment for the application of government digital technology [13,14]. Although the central government and local governments have issued relevant policies and made building of digital government a “top managerial project” to strengthen their unified leadership, the different incentives from inter-governmental relations and officials’ promotions still directly affect the attention distribution of local governments and chief officials to the building of digital government, and also restrict the level and implementation effect of regional digital government development.

In China, the relationship between governments can be summarized as the vertical top-down relationship and the horizontal learning and competition relationship [24,25], of which the top-down inter-governmental relationship is particularly important for shaping local governments’ behavior. Under the system of administrative centralization and fiscal decentralization, the superior government leads the subordinate government through financial resources allocation, personnel appointment and removal, and dependence on the superior government, which will affect the subordinate government’s cognition and implementation of the “superior policy”. Based on sample data from the United States and the United Kingdom, Menzel [26] and Walker [27] found that the attitude of the superior organization affected the subordinate organization’s adoption of new technologies. Fan et al. [14] used data from China’s government portal websites to find that, with the increase in pressure faced by the subordinate government, the performance of government portal websites was enhanced significantly. Along with deepening the process of building digital government, the central government and local governments at all levels are actively exploring new ways of building digital government, but the distinctive “top-down” characteristic is still imprinted in various places, and the pressure from the superior government directly affects the subordinate government’s building of digital government and implementation effect. Under the same conditions, with the increase in pressure from the superior government, subordinate governments will allocate more attention and resources on building digital government, and its impact on regional business environment and enterprise TFP may be more obvious. Therefore, our second hypothesis is proposed as follows:

H2. 

Superior pressure has a positive moderating effect on the relationship between digital government development and enterprise TFP; that is, the greater the pressure from the superior government, the more obvious the role of digital government development in promoting enterprise TFP.

Under the Sinicism decentralization system, the superior government completes the governance of subordinate government officials through a relative performance evaluation mechanism, and its content becomes a baton to control chief officials’ behavior, constituting another institutional constraint on local governments’ behavior [25,28]. After the Third Plenary Session of the 11th Central Committee of the CPC, the work focus of the Party and the government shifted to economic construction, and the economic growth speed became the core content of chief officials’ performance evaluation, which greatly aroused the enthusiasm of chief officials and local governments to develop the economy. With the transformation and upgrade of China’s economic development mode, scientific and technological innovation and green development have been gradually incorporated into chief officials’ performance evaluation system, strengthening chief officials’ attention to the quality of economic growth [29,30]. However, the energy of chief officials is limited. Under the multi-dimensional performance assessment system, chief officials’ attention distribution depends on the relative importance of different assessment contents. Empirical studies show that the economic growth rate still occupies an important position in chief officials’ performance evaluation, and chief officials tend to focus on economic development in the early period of their tenure with strong promotion incentives [31]. Moreover, due to the weak attraction in the explicit governance mechanism and the design of incentive mechanisms, activities such as building government portal websites which are unlikely to increase chief officials’ promotion advantages, are usually at a disadvantage in the chief officials’ attention distribution [14]. Therefore, chief officials who are in the early stages of their official tenure may pay relatively little attention to building digital government, which will restrict the effectiveness of the practice of building digital government. With the increase in the tenure of chief officials and the decline of promotion incentives, building digital government may gain more attention, and its impact on enterprise TFP will be further strengthened. Based on the above analysis, our third hypothesis is proposed as follows:

H3. 

The tenure of chief officials has a positive moderating effect on the relationship between digital government development and enterprise TFP; that is, with the increase in the chief officials’ tenure, the role of digital government development in promoting enterprise TFP is more obvious.

3. Research Design

3.1. Sample Selection and Data Sources

This paper selected Chinese A-share listed companies from 2010 to 2017 as the research sample (according to the published data of the Guomai e-government website, the sample period is from 2010 to 2017). Based on previous research, the samples were selected under the following rules: (1) financial companies were excluded; (2) ST and ST* companies during the sample period were removed; (3) the companies with missing core explanatory variables were deleted; (4) the companies with abnormal financial data, such as negative total assets, total assets less than current assets or intangible assets, and negative liabilities, were removed. Ultimately, 9937 observation samples were obtained. In order to avoid the impact of outliers, the continuous variables were Winsorized at the levels of 1% and 99%.

The enterprise data used in this paper were from CSMAR and Wind databases. The digital government development data was retrieved from the Guomai e-government website (http://www.echinagov.com/data/, accessed on 18 January 2023), the data on chief officials’ tenure from websites such as People’s Daily Online and Baidu Baike, and other urban data from the China City Statistical Yearbook and EPS data platform.

3.2. Model Construction

To examine the impact of digital government development on enterprise TFP, the following model was established:

$$TF{P}_{it}={\alpha }_0+{\alpha }_{1}EGover{n}_{ict}+\alpha X_{it}+{\mu }_d+{\lambda }_t+{\epsilon }_{it}$$

(1)

where the subscripts i, d, c, and t denote the enterprise, industry, city, and year, respectively. The explained variable is enterprise total factor productivity (TFP). The core explanatory variable is digital government development (EGovern). If the estimated coefficient ${\alpha }_1$ is significantly positive, it indicates that digital government development can promote enterprise TFP, which supports H1. The X is a set of enterprise-level control variables, ${\mu }_d$ represents industry fixed effect, ${\lambda }_t$ represents year fixed effect, and ${\epsilon }_{it}$ is random perturbation terms.

According to the above analysis, local governments’ attention distribution will affect the process and implementation effect of the city’s digital government building. In order to identify this influence, based on Equation (1), we added the interaction term of local governments’ attention distribution (Attention) and digital government development (EGovern), in the model shown in Equation (2):

$$TF{P}_{it}={\beta }_0+{\beta }_{1}EGover{n}_{ict}+{\beta }_{2}EGover{n}_{ict}\times Attentio{n}_{ict}+{\beta }_{3}Attentio{n}_{ict}+\beta X_{it}+{\mu }_d+{\lambda }_t+{\epsilon }_{it}$$

(2)

where Attention means local governments’ attention distribution. In this model, we mainly focused on the coefficient of the interaction term. If ${\beta }_2$ is significantly greater than 0, it indicates that the promotion effect of digital government development on enterprise TFP will be affected by the local government’s attention distribution, supporting H2 and H3. The meanings of other variables are the same as in Equation (1).

3.3. Variable Description and Descriptive Statistics

3.3.1. Enterprise TFP (TFP)

In the existing literature, the OP, LP, OLS, GMM, FE, and ACF methods are usually used to measure enterprise TFP. Since OLS and FE methods cannot solve the endogeneity problem and lead to the biased estimation of enterprise TFP, the OP, LP, GMM, and ACF methods are used more widely [32]. In this paper, OP and LP methods were used to measure enterprise TFP in benchmark regression, and GMM and ACF methods were used to measure enterprise TFP in the robustness test. Referring to Dai et al. [33], the selection and treatment of the index of added value and factor input are as follows: with the industrial producer ex-factory price index where the enterprise is located to exclude the impact of price fluctuations, the net fixed assets are adopted to measure capital input. Investment is measured by the capital expenditure, scilicet cash paid for fixed assets, intangible assets, and other long-term assets deduct the withdrawal of the disposal of fixed, intangible, and other long-term assets, and all are deflated by the fixed asset price index to eliminate the impact of price fluctuations. We used the number of employees to measure labor input, and the above indicators were expressed logarithmically.

3.3.2. Digital Government Development (EGovern)

Government portal websites are the main channels for transmitting government data resources, and the digital governance practices of China’s government also stem from the “Government Online Project”, so the performance of government portal websites is regarded as a key indicator reflecting the government’s digitalization level [10,15,34]. Among studies on digital government development, this paper chose to construct the variable of digital government development (EGovern) based on the performance scores of China’s government portal websites published by the Guomai e-government website. This was mainly based on two considerations: first, the Guomai e-government website comprehensively evaluates the performance of government portal websites in 31 provinces, municipalities directly under the central government, and prefecture-level cities in China from the aspects of open government, public participation, online services, and user experience, providing a wealth of information. Second, the Guomai e-government website started to publish this data in 2010, and up to 2017, there are a total of eight periods of panel data, which could better support the empirical analysis of this paper. The data is also often used in the research field of China’s digital government development [10,11]. In the benchmark regression, the comprehensive score of government portal performance divided by 100 was used to measure a city’s digital government development.

3.3.3. Attention Distribution (Attention)

Combined with China’s political system environment and local governments’ behavior characteristics, this paper used the financial dependence (Attention-Finance) and chief officials’ tenures (Attention-Tenure) to indirectly measure local governments’ attention distribution to building digital government mainly from the perspectives of superior pressure and chief officials’ promotion incentives. Referring to Zhang [35] and Fan et al. [14], the degree of local governments’ financial dependence(Attention-Finance) is measured by the difference between general public budget revenue and general public budget expenditure divided by general public budget revenue. Referring to Geng et al. [31], chief officials’ tenure(Attention-Tenure) is measured by the tenure of the secretary of the municipal party committee. If the secretary of the municipal party committee takes office between January and June, the current year is regarded as the first year of his tenure, and Attention-Tenure is equal to 1; if the secretary of the municipal party committee takes office between July and December, the next year is regarded as the first year of his tenure, and Attention-Tenure is equal to 0.

Referring to Bennett et al. [36], Qi and Yang [37], and Liu et al. [38], control variables include enterprise growth ability (TobinQ), return on assets (ROA), asset-liability ratio (Lev), ownership concentration (Top5), enterprise age (Age), and enterprise size (Size). The description of main variables and their descriptive statistics are shown in

Table 1. Definitions and descriptive statistics of main variables.

Variable Name	Calculation Method	Data Sources	Reference	N	Mean	Std. Dev.	Min	Max
Total factor productivity (TFP_OP, TFP_LP)	OP	CSMAR and Wind databases	Dai et al. (2021) [33]	9937	7.877	0.942	4.928	12.553
	LP			9937	8.801	1.050	5.568	13.533
Digital government development (EGovern)	Comprehensive score of government portal website performance/100	Guomai e-government website	Wang et al. (2022) [10]; Qu and Wang (2022) [11]	9937	0.700	0.133	0	0.895
Official tenure (Attention-Tenure)	The tenure of party secretary	People’s Daily Online and Baidu Baike	Geng et al. (2016) [31]	9865	3.299	1.844	0	11
Superior pressure (Attention-Finance)	The difference between general public budget revenue and general public budget expenditure divided by general public budget revenue	China City Statistical Yearbook and EPS data platform	Zhang (2015) [35]; Fan et al. [15] (2018)	9879	−0.583	0.972	−18.740	0.351
Enterprise growth ability (Tobin Q)	Tobin Q value	CSMAR and Wind databases	Bennett et al. (2020) [36]; Qi and Yang (2021) [37]; Liu et al. (2020) [38]	9937	2.768	1.884	0.930	11.440
Return on assets (ROA)	Net profit at end of period / total assets at end of period			9937	0.043	0.054	−0.148	0.210
Asset-liability ratio(Lev)	Ending liabilities / ending total assets			9937	0.407	0.205	0.0509	0.886
Ownership concentration (Top5)	The shareholding ratio of the top five shareholders			9937	0.523	0.148	0.191	0.845
Enterprise age(Age)	Years of establishment			9937	14.916	5.360	4	28
Enterprise scale (Size)	The logarithmic of total ending assets			9937	21.894	1.116	19.740	25.263

.

4. The Impact of Digital Government Development on Enterprise TFP

4.1. Benchmark Regression

Table 2. Benchmark regression results.

Variable Name	(1)	(2)	(3)	(4)
	TFP_OP	TFP_LP	TFP_OP	TFP_LP
EGovern	0.470 ***	0.503 **	0.516 ***	0.558 ***
	(0.165)	(0.197)	(0.112)	(0.105)
Tobin Q			−0.014 ***	−0.013 **
			(0.005)	(0.005)
ROA			3.119 ***	3.788 ***
			(0.214)	(0.192)
Lev			0.754 ***	0.864 ***
			(0.076)	(0.095)
Top5			0.209 ***	0.254 ***
			(0.073)	(0.088)
Age			−0.001	0.000
			(0.002)	(0.002)
Size			0.542 ***	0.656 ***
			(0.013)	(0.013)
Constant	7.548 ***	8.449 ***	−4.852 ***	−6.568 ***
	(0.116)	(0.135)	(0.302)	(0.292)
Industry fixed effect	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
N	9937	9937	9937	9937
R2 adjusted	0.275	0.231	0.736	0.764

Note: Dependent variable is enterprise TFP. *** and ** represent the significance levels at 1% and 5%, respectively. The standard error is clustered at the city level and reported in parentheses. The same is below. Source: Author’s calculation.

lists the regression results of the city’s digital government development on enterprise TFP. Column (1) and Column (2) are the results only controlling the industry effect and year effect, and Column (3) and Column (4) are the results of adding a series of control variables on the basis of Columns (1) and (2). Regardless of whether control variables were added or not, the estimated coefficient of digital government development (Egovern) was always significantly positive at the 1% level; that is, digital government development promoted the enterprise TFP. According to the estimation results in Column (4), when the level of the city’s digital government development increased by 0.1 units (the standard deviation of the variable of digital government development was 0.133), the TFP of enterprises in the jurisdiction increased by 0.058 units (0.1 × 0.558) on average. Theoretically, digital government development provides a more open, legalized, and convenient business environment for market entities. On the one hand, it can reduce the information asymmetry between the government and enterprises, compress the power of rent-setting and rent-seeking spaces, make enterprises focus on productive activities, and improve enterprise TFP. On the other hand, it can stimulate the competition vitality of market entities, inspire enterprise innovation, and then promote enterprise TFP. This paper further examines how digital government development improves enterprise TFP.

4.2. Robustness Test and Endogeneity Discussion

First, instrumental variables were used. In this paper, the city’s digital government development was exogenous compared to a single enterprise. Since government policies are formulated based on the overall performance of enterprises in the jurisdiction, the reverse causality may still exist in the model. Meanwhile, although a series of control variables are added to the model, there may be other factors that affect enterprise TFP and the city’s digital government development. If they are included in the random perturbation term, the estimation results will be biased. In order to alleviate the endogeneity problem, this paper took the government portal website performance of the province where the enterprise was located as the instrumental variable of the city’s digital government development and used the two-stage least squares (2SLS) method to re-estimate Equation (1). The Kleibergen–Paap rk LM test rejected the original hypothesis that the instrumental variables and endogenous explanatory variables were not correlated, and the Kleibergen–Paap rk Wald F value was significantly greater than the critical value of the weak instrumental variables test. Therefore, it can be considered that the instrumental variable selected in this paper was reasonable. The estimation results of the second stage are shown in Column (1) of

Table 3. Robustness test and endogeneity discussion.

Variable Name	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	TFP_LP	TFP_GMM	TFP_ACF	TFP_LP	TFP_LP	TFP_LP	TFP_LP
EGovern	0.902 ***	0.631 ***	0.514 ***	0.561 ***	0.581 ***	0.691 ***	0.475 ***
	(0.085)	(0.113)	(0.109)	(0.113)	(0.045)	(0.083)	(0.071)
Constant		−0.980 ***	−8.387 ***	−6.528 ***		−6.502 ***	−6.333 ***
		(0.375)	(0.280)	(0.297)		(0.144)	(0.230)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	9845	9937	9937	7860	9937	8791	2578
R2 adjusted	0.696	0.413	0.835	0.766		0.763	0.754
Kleibergen–Paap rk LM	1568.036 [0.000]
Kleibergen–Paap rk Wald F	3084.050 [0.000]
					165.110

Note: The value in brackets is the statistical P value. *** represent the significance levels at 1%, respectively. Source: Author’s calculation.

. The estimated coefficient of digital government development was still significantly positive at the 1% level, indicating that the above conclusion still holds after considering the endogeneity of the model.

Second, the measurement of enterprise TFP was replaced. We used GMM and ACF methods to re-measure enterprise TFP. The regression results are shown in Columns (2) and (3) of Table 3. The estimated coefficient of digital government development was significantly positive, consistent with the benchmark regression results.

Third, considering that the influence of digital government development on enterprise TFP may exist in time lags, we lagged the explanatory variables by one period. The regression results are shown in Column (4) of Table 3. The estimated coefficient of digital government development was significantly positive at the 1% level, indicating the previous regression results were robust and reliable.

Fourth, since the model misspecification would also affect the parameter estimation results, we relaxed the assumption of the linear model and used the debiased machine learning method proposed by Chernozhukov et al. to re-estimate the impact of digital government development on enterprise TFP [39]. The regression results are shown in Column (5). The estimated coefficient of digital government development did not change significantly and was positive at the 1% level, which once again verified that the conclusions in this paper are robust and valid.

Finally, the data was replaced. We used the data from the “Survey and Evaluation Report on Online Government Service Capacity of Provincial Governments and Key Cities” released by the E-Government Research Center of the Party School of the Central Committee of C.P.C (National Academy of Governance) in 2015, 2017, 2018, 2019, 2020 and 2021 to reconstruct the variable of regional digital government development. The digital government development of provinces was measured by the index of integrated government service capacity of provincial government divided by 100, and the regression results are shown in Column (6) in Table 3. The city digital government development was calculated using the index of integrated government service capacity of 32 key cities (most of them provincial capitals and cities above the sub-provincial level) divided by 100, and the regression results are shown in Column (7) in Table 3. The regression results unveiled that the regression coefficient of digital government development was still significantly positive at the 1% level. Thus, the previous findings of our study were robust and valid.

4.3. Heterogeneity Analysis

4.3.1. The Influence of Subdivisions of Digital Government Development

Building digital government is a systematic project, involving market supervision, social management, public services, open government affairs and other aspects. Using the government portal websites performance evaluation index system constructed by the Guomai e-government website, this paper investigated the impact of subdivisions of digital government development on enterprise TFP from four aspects: information disclosure, online services, public participation, and user experience. The regression results are shown in

Table 4. Test results of subdivisions of digital government development.

Variable Name	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	TFP_OP				TFP_LP
Information disclosure	1.314 ***				1.367 ***
	(0.277)				(0.287)
Online services		0.869 ***				1.011 ***
		(0.229)				(0.208)
Public participation			1.090 ***				0.933 ***
			(0.285)				(0.246)
User experience				0.305				0.544
				(0.416)				(0.366)
Constant	−4.790 ***	−4.684 ***	−4.628 ***	−4.592 ***	−6.488 ***	−6.405 ***	−6.289 ***	−6.301 ***
	(0.297)	(0.289)	(0.281)	(0.285)	(0.292)	(0.281)	(0.278)	(0.287)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	9937	9937	9937	9937	9937	9937	9937	9937
R2 adjusted	0.735	0.735	0.734	0.734	0.763	0.763	0.761	0.762

Notes: *** represent the significance levels at 1%, respectively. Source: Author’s calculation.

. The estimated coefficients of information disclosure, online services, and public participation were all positive and significant at the 1% level, while the estimated coefficient of user experience did not pass the significance test. The reasons are as follows: in the index system established by the Guomai e-government website, information disclosure, online services, and public participation mainly focused on the quality of public information in government portal websites, the convenience of public services, and timeliness of feedbacks to public inquiries. The higher the score, the better the effect of the government’s digital technology application, and the greater its effect on improving the business environment and promoting enterprise production and operation. The user experience dimension mainly evaluated the performance of government portal websites from the aspects of internationalization, personalized services, sensory experience, and so on. It had no direct impact on the improvement of government governance and the optimization of the regional business environment. Therefore, user experience did not have a significant effect on enterprise TFP.

4.3.2. Regional Heterogeneity

China has a vast territory, diverse natural conditions and human environments, and the economic developmental level of the eastern, central, and western regions are obviously different; thus, the digital government development in different regions is also unequal. We divided the sample into three sub-samples from the eastern, central, and western regions to explore the diverse nexus between digital government development and enterprise TFP, and the regression results are shown in

Table 5. Results of the regional heterogeneity test.

Variable Name	(1)	(2)	(3)	(4)	(5)	(6)
	TFP_OP			TFP_LP
	The Eastern Region	The Central Region	The Western Region	The Eastern Region	The Central Region	The Western Region
EGovern	0.330 *	0.524 ***	0.776 ***	0.418 ***	0.667 ***	0.874 ***
	(0.169)	(0.175)	(0.202)	(0.151)	(0.218)	(0.174)
Constant	−4.865 ***	−4.201 ***	−5.053 ***	−6.798 ***	−6.073 ***	−6.345 ***
	(0.391)	(0.732)	(0.973)	(0.372)	(0.780)	(0.958)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	6503	1995	1439	6503	1995	1439
R2 adjusted	0.753	0.735	0.731	0.778	0.766	0.752
SUEST test chi2		3.09 *	4.47 **		2.71 *	5.09 **

Notes: ***, **, and * represent the significance levels at 1%, 5%, and 10%, respectively. Source: Author’s calculation.

. In all regressions, the coefficient of digital government development was positive and significant at least at the 10% level, indicating that the positive effect of digital government development on enterprise TFP was established in different regions. The SUEST test results showed that digital government development had a more obvious impact on enterprise TFP in the central and western regions compared to the eastern region. Wan et al. found that there was a nonlinear relationship between the governance effect of digital government development and the level of economic development [40], and the positive impact of digital government development on the governance level was more significant in low-income countries. In China, compared to the central and western regions, the eastern region has a higher level of economic development and business environment, so that the role of digital government development in improving the government’s ability to perform its duties and optimizing the business environment is lower than that of the central and western regions. Furthermore, enterprises in the eastern region have a relatively higher TFP and are less affected by policy shocks, so the optimization of the business environment has a more significant effect on enterprises in the central and western regions with a relatively lower TFP [41].

4.3.3. Enterprise Heterogeneity

Differences in enterprise’ technology level, governance structure, and industry characteristics may lead to the heterogeneous impact of digital government development on enterprise TFP. Therefore, according to the enterprise ownership structure, we divided the sample into two sub-samples of state-owned enterprises (SOEs) and non-state-owned enterprises (non-SOEs), and the regression results are shown in

Table 6. Test results of the heterogeneity of enterprise ownership structure.

Variable Name	(1)	(2)	(3)	(4)
	TFP_OP		TFP_LP
	SOEs	Non-SOEs	SOEs	Non-SOEs
EGovern	0.308 *	0.483 ***	0.328 *	0.589 ***
	(0.174)	(0.123)	(0.183)	(0.126)
Constant	−4.750 ***	−4.728 ***	−6.188 ***	−6.541 ***
	(0.530)	(0.344)	(0.542)	(0.352)
Control variables	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
N	3139	6793	3139	6793
R2 adjusted	0.744	0.712	0.778	0.745
SUEST test chi2	2.99 *		2.78 *

Notes: *** and * represent the significance levels at 1% and 10%, respectively. Source: Author’s calculation.

. The coefficient of digital government development was significantly positive at least at the 10% level, indicating that digital government development had a positive impact on the TFP of both SOEs and non-SOEs. By comparing the regression coefficients of different sub-samples and SUEST test results, it was found that digital government development had a greater effect on the TFP of non-SOEs. In China, non-SOEs face greater financing constraints and institutional costs than SOEs, which restricts enterprise operation and development [17,42,43]. Digital government development has brought more efficient and open market environments for enterprises, which can reduce the cost of non-SOEs in dealing with the local government. As a result, the impact of digital government development on non-SOEs is more prominent.

Additionally, referring to Lu and Dang [44], we divided the sample into two sub-samples of technology-intensive and non-technology-intensive enterprises, and the regression results are shown in

Table 7. Results of the industry heterogeneity test.

Variable Name	(1)	(2)	(3)	(4)
	TFP_OP		TFP_LP
	Technology Intensive Industry	Non-Technology Intensive Industry	Technology Intensive Industry	Non-Technology Intensive Industry
EGovern	0.702 ***	0.307 ***	0.745 ***	0.391 ***
	(0.162)	(0.117)	(0.156)	(0.116)
Constant	−4.562 ***	−4.941 ***	−6.421 ***	−6.608 ***
	(0.477)	(0.446)	(0.598)	(0.428)
Control variables	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
N	3184	6753	3184	6753
R2 adjusted	0.710	0.740	0.749	0.767
SUEST test chi2	5.31 **		5.25 **

Notes: *** and ** represent the significance levels at 1% and 5%, respectively. Source: Author’s calculation.

. The coefficient of digital government development was still significantly positive and was greater in technology-intensive industries. This result showed that technology-intensive industries are more susceptible to the influence of the market environment and innovation environment, and benefit more from digital government development. This is consistent with the results of Zhang et al. [17], who demonstrated that service-oriented government construction had a more obvious effect on the productivity improvement of technology-intensive enterprises.

4.4. Mechanism Analysis

Furthermore, in order to investigate whether digital government development can affect enterprise TFP by reducing enterprise rent-seeking, attracting new enterprises entry, and promoting enterprise innovation, the following models were constructed:

$$Ren{t}_{it}/Newfir{m}_{ct}/R{D}_{it}={\delta }_0+{\delta }_{1}Egover{n}_{ict}+\delta Z_{it}+{\mu }_d+{\gamma }_c+{\lambda }_t+{\epsilon }_{it}$$

(3)

$$TF{P}_{it}={\theta }_0+{\theta }_{1}Ren{t}_{it}/Newfir{m}_{ct}/R{D}_{it}+\theta X_{it}+{\mu }_d+{\lambda }_t+{\epsilon }_{it}$$

(4)

where Rent indicates enterprise rent-seeking, which is measured by the excess management cost of enterprises according to Dechow et al. [45] and Chen et al. [46]. Newfirm indicates new enterprise entry, which is measured by the natural logarithm of the number of city’s new enterprise entry in the current year. The RD represents enterprise innovation, which is defined as enterprise R&D expenditure divided by enterprise operating income. The Z_it is a set of control variables, ℽ_c is the city fixed effect, and other variables have the same meaning as in Equation (1).

Columns (1) and (2) of

Table 8. Mechanism test results.

Variable Name	(1)	(2)	(3)	(4)	(5)	(6)
	Rent	TFP_LP	Newfirm	TFP_LP	RD	TFP_LP
EGovern	−0.031 **		0.232 ***		1.236 ***
	(0.014)		(0.023)		(0.259)
Rent		−0.733 ***
		(0.172)
Newfirm				0.082 ***
				(0.016)
RD						0.124 ***
						(0.011)
Constant	−0.023 ***	−6.200 ***	1.306 ***	−5.920 ***	−1.944 ***	−6.022 ***
	(0.005)	(0.283)	(0.311)	(0.348)	(0.741)	(0.273)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
City fixed effect	No	No	Yes	No	No	No
N	9818	9818	7650	7650	8004	8004
R2 adjusted	0.005	0.761	0.561	0.752	0.534	0.791

Note: Dependent variables of Columns (1), (3), and (5) are enterprise rent-seeking, new enterprise entry, and enterprise innovation, respectively. Dependent variables of Columns (2), (4), and (6) are the enterprise TFP calculated by the LP method. *** and ** represent the significance levels at 1% and 5%, respectively. Source: Author’s calculation.

show the results of enterprise rent-seeking as the mechanism variable. It can be seen that the influence coefficient of digital government development on enterprise rent-seeking was −0.031, which was significant at the 5% level, indicating that digital government development was helpful to build the “cleaning and close” relationship between government and enterprise, and reduce the occurrence of enterprise rent-seeking behavior. The results of Column (2) showed that the influence coefficient of enterprise rent-seeking on enterprise TFP was significantly negative, which was consistent with the conclusions of previous studies [41]. The improvement of production level and operational efficiency is the fundamental way for the long-term sustainable development of enterprises. However, when the market environment is not perfect, enterprises are keen to obtain competitive advantages through rent-seeking, which not only occupies enterprise limited resources, but also restricts the improvement of enterprise productivity. Thus, H1a was verified.

Columns (3) and (4) of Table 8 show the results of new firm entry as the mechanism variable. It can be seen that the influence coefficient of digital government development on new enterprise entry was 0.232, which was significant at the 1% level, indicating that digital government development can provide a fairer competition environment and attract new enterprise entry. When the level of the city’s digital government development increased by 0.1 units, the number of new enterprise entries increased by 2.32 percent. The results of Column (4) show that the coefficient of new enterprise entry on enterprise TFP was significantly positive, indicating that new enterprise entry can bring new technologies and new ideas to the local market, stimulate enterprise vitality, and improve the efficiency of enterprise resource allocation. Thus, H1b was verified.

Columns (5) and (6) of Table 8 show the results of enterprise innovation as the mechanism variable. According to the estimation results of Column (5), the influence coefficient of digital government development on enterprise R&D was 1.236, which was significant at the 1% level. Digital government development can create a good business environment and innovation environment for enterprises, mobilize enterprise innovation motivation, and improve enterprise R&D investment. When the level of the city’s digital government development increased by 0.1 units, the level of R&D investment of enterprises in the jurisdiction increased by 0.12 units. Regardless of the perspective of statistical significance or economic significance, the impact of digital government development on enterprise innovation could not be ignored, which was consistent with the research of Wang et al. [10]. Column (6) shows that R&D investment had a significantly positive impact on enterprise TFP, which has been widely confirmed by previous studies. Overall, digital government development can standardize government behavior, make the government affairs environment more efficient and convenient and the market environment more open and fairer, which enables enterprises to focus on productive activities and continuously optimize resource allocation. Thus, H1c was verified.

5. The Impact of Local Governments’ Attention Distribution

As mentioned above, under the Sinicism decentralization, both the superior pressure and officials’ tenure will affect local governments’ attention distribution and in return, affect the process of regional digital government development and its implementation effect. Based on Equation (2), we investigated the influence of local governments’ attention distribution on the relationship between digital government development and enterprise TFP.

Table 9. Moderating effect of superior pressure.

Variable Name	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Full Sample		The Eastern Region		The Central Region		The Western Region
	TFP_OP	TFP_LP	TFP_OP	TFP_LP	TFP_OP	TFP_LP	TFP_OP	TFP_LP
EGovern	0.558 ***	0.621 ***	0.391 **	0.373 **	0.640 ***	0.738 ***	0.210	0.109
	(0.135)	(0.054)	(0.174)	(0.149)	(0.231)	(0.171)	(0.192)	(0.133)
EGovernv×Superior pressure	0.092 **	0.066 *	0.238	0.127	0.241 **	0.227 **	0.145 *	0.144 **
	(0.041)	(0.037)	(0.175)	(0.112)	(0.106)	(0.099)	(0.083)	(0.062)
Superior pressure	−0.053	−0.044 **	−0.102	−0.098	−0.078	−0.121 **	−0.002	0.039
	(0.043)	(0.020)	(0.116)	(0.075)	(0.062)	(0.061)	(0.045)	(0.031)
Constant	−4.867 ***	−6.629 ***	−4.755 ***	−6.414 ***	−4.486 ***	−6.339 ***	−5.283 ***	−6.350 ***
	(0.316)	(0.159)	(0.318)	(0.304)	(0.766)	(0.379)	(0.754)	(0.434)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	9879	9879	6502	6502	1986	1986	1391	1391
R2 adjusted	0.736	0.764	0.737	0.764	0.760	0.786	0.756	0.767

Notes: ***, **, and * represent the significance levels at 1%, 5%, and 10%, respectively. Source: Author’s calculation.

shows the results of the mediating effect of the superior pressure faced by the local government. Columns (1) and (2) are the full-sample regression results. The coefficient of digital government development and the interaction term of digital government development (EGovern) and superior pressure faced by the local government (Superior pressure) were significantly positive, indicating that superior pressure will strengthen the positive impact of digital government development on enterprise TFP. Building digital government is an important means for the government to optimize the business environment and is also a typical “top managerial project”. The pressure from the superior government affects the subordinate government’s attention to building digital government. Under the same conditions, the greater the pressure faced by subordinate governments, the better the implementation effect of building digital government; namely, the greater the promotion effect on enterprise TFP, thus proving H2.

Furthermore, we divided the sample into three sub-samples: the eastern region, the central region and the western region, and the results are shown in Columns (3) to (8) of Table 9. The regression coefficients of the interaction term were positive at least at the 10% significance level only for the central and western region samples. In other words, in the relatively underdeveloped central and western regions, increasing the attention of local governments can play a positive role in building digital government, optimizing the business environment, and stimulating the vitality of market entities.

Table 10. Moderating effect of official tenure.

Variable Name	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Full Sample		The Eastern Region		The Central Region		The Western Region
	TFP_OP	TFP_LP	TFP_OP	TFP_LP	TFP_OP	TFP_LP	TFP_OP	TFP_LP
EGovern	0.391 **	0.373 **	0.284 *	0.369 ***	0.259 **	0.301 *	0.097	0.040
	(0.174)	(0.149)	(0.167)	(0.114)	(0.119)	(0.178)	(0.238)	(0.227)
EGovern ×Attention-Tenure	0.051 *	0.054 *	0.067	0.079	0.092 **	0.074 *	0.053 *	0.042 **
	(0.029)	(0.030)	(0.051)	(0.056)	(0.044)	(0.039)	(0.028)	(0.021)
Attention-Tenure	−0.030	−0.025	−0.072 **	−0.060 **	−0.042	−0.053	−0.016	−0.028
	(0.020)	(0.022)	(0.032)	(0.030)	(0.036)	(0.040)	(0.027)	(0.029)
Constant	−4.755 ***	−6.414 ***	−4.612 ***	−6.577 ***	−4.365 ***	−6.093 ***	−5.134 ***	−6.195 ***
	(0.318)	(0.304)	(0.435)	(0.388)	(0.730)	(0.835)	(0.761)	(0.740)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	9865	9865	6475	6475	1964	1964	1426	1426
R2 adjusted	0.737	0.764	0.754	0.778	0.752	0.778	0.744	0.755

Notes: ***, ** and * represent the significance levels at 1%, 5%, and 10%, respectively. Source: Author’s calculation.

shows the results of the moderating effect of official tenure. Among them, Columns (1) and (2) are the regression results of the full sample, and Columns (3) to (8) are the results of sub-samples in the eastern, central, and western regions. In the full sample, the estimated coefficient of the interaction term of digital government development (EGovern) and chief officials’ tenure (Tenure) was positive at the 10% significance level, reflecting that under the same conditions, the longer the tenure of chief officials, the more obvious the promotion effect of digital government development on enterprise TFP; thus, the H3 was confirmed. Under the current officials’ assessment system, although the weight of regional economic growth speed has decreased, it still occupies an important position. Therefore, chief officials usually focus on regional economic development in the early stages of their tenure, while the attention distribution to government portal websites is relatively limited [14]. With the decrease in promotion incentives, chief officials will pay more attention to affairs such as building of digital government, and the promotion effect of digital government development on enterprise TFP will also increase. From the regression results of sub-samples, the positive moderating effect of chief officials’ tenure was also more obvious in the central and western regions.

6. Conclusions and Implications

Based on the data of China’s listed companies and government portal websites from 2010 to 2017, this study examined the impact of digital government development on enterprise TFP and the moderating effect of local governments’ attention distribution. We found that cities’ digital government development can create a good market environment for enterprise, which helps to improve enterprise TFP. This conclusion still holds after a series of robustness tests using instrumental variables, one-stage lag of explained variables, and debiased machine learning models. By subdividing the content of building digital government, we also found that the three subdivided dimensions of information disclosure, online service, and public participation all had a positive effect on enterprise TFP, while the user experience dimension had no effect on enterprise TFP. The results of sub-sample regression showed that digital government development had a more significant promoting effect on enterprise TFP in central and western regions, non-SOEs, and technology-intensive enterprises. The mechanism test results showed that digital government development can regulate the government’s behavior and provide a more open and efficient environment for market entities, which will reduce enterprise rent-seeking, attract new enterprise entry, and increase enterprises’ R&D investment, thus improving enterprise TFP. Finally, as a typical “top managerial project”, local governments’ attention distribution will positively affect the relationship between digital government development and enterprise TFP. With the increase in the superior government’s pressure and the chief officials’ tenure, the local government will allocate more attention to building digital government, and digital government development will have a more obvious effect on enterprise TFP.

This paper provides empirical references for the government to improve the level of building digital government and promote the high-quality development of enterprises. Based on the above research conclusions, the following suggestions are put forward. First, to further improve the level of digital government development. At present, China’s building of digital government has entered a comprehensive acceleration period, and the idea of using digital reform to help the transformation of government functions has been deeply rooted in the hearts of the people. However, from the evaluation results of many institutions, there are great differences in the level of building digital government in different regions. To meet the urgent needs of economic and social development and create new prospects for building digital government, it is also necessary to strengthen reform thought, support building of digital government in terms of funds, qualified personnel, and technology, upgrade relevant infrastructure, and fully promote the digital and intelligent model of government public services.

Additionally, strengthening the unified leadership of the practice of building digital government and giving full play to system advantages will ensure the integrated development of building digital government. As a “top managerial project”, it is necessary to increase the attention distribution of local governments and mobilize the initiative of chief officials. On the one hand, it is necessary to adhere to the overall layout, perform a good job in the top-level design, improve the supporting system and working mechanism, strengthen the linkage between departments, form a joint force, and ensure the solid progress of the digital government building. On the other hand, we should strengthen the assessment, improve the digital literacy of the cadre team, incorporate digital governance into the officials’ evaluation system, and transform the institutional advantages into a strong driving force for building digital government.

Finally, using digital technology scientifically can maximize the digital empowerment effect. Digitalization and intelligence are not an end in themselves, the purpose of building digital government is to improve the government’s ability to perform its duties and meet people’s yearning for a better life. However, at present, some local governments are not scientific enough in the application of digital technology, and there are problems of technology worship and abandonment, which not only violates the original intention of building digital government, but also makes it difficult to exert the effect of digital empowerment. To optimize the business environment through building digital government, there is a need to start from the needs of enterprises, pay attention to the voice of enterprises, perfect the feedback mechanism, and provide efficient and convenient services for enterprises. While optimizing government services, it is also essential to focus on improving the government’s digital governance and regulatory capabilities, strengthening network and data security, and building a solid security defense line for digital government development.

Our study findings must be interpreted in light of two limitations. First, in view of the objective factors, our sample period was from 2010 to 2017; it is necessary to expand the time period of the sample data in future research so that the research can draw more diverse inferences. Second, this paper mainly explored the impact of the moderating effect of local governments’ attention distribution from superior pressure and chief officials’ tenure, which is consistent with the practice of China’s digital government development. However, the conclusions may not be applicable to other countries beyond mere reference values, and future research may look into other intervening variables targeted at different application environments.