Digital Economy, Technological Innovation, and Environmental Quality Improvement

Abstract

Based on China’s provincial panel data from 2013 to 2020, this paper makes an empirical study on the influence of digital economy development on regional environmental quality by using the fixed effect model and mediating effect model. The results show that the development of a digital economy can effectively reduce environmental pollution and improve environmental quality. Secondly, a mechanism analysis shows that a digital economy can promote technological innovation and improve environmental quality; that is, there is a mediating effect when the total number of patents and utility model patents is a proxy variable of technological innovation, and there is no mediating effect when invention patents and design patents are proxy variables. Finally, through further analysis of regional differences, it is found that the development of the digital economy has obvious regional heterogeneity characteristics to the improvement of environmental quality, and the central and western regions are higher than the eastern regions. These conclusions provide new policy directions and inspiration for improving the regional environment, improving people’s quality of life, and promoting high-quality economic development.

1. Introduction

China’s economic development has made great strides since reform, but extensive development has led to the deterioration of the country’s environmental quality. While the Ministry of Ecology and Environment released a national status report on ecological environment quality, the country’s ecological environment quality improved in 2021, with the total emissions of major pollutants and CO₂ emissions per unit of gross domestic product falling. However, there are problems with air and water pollution and garbage disposal, and new measures to improve the quality of the environment are urgently needed. At the same time, China’s economy has grown manifold, with China’s Gross Domestic Product (GDP) continuing to rank second in the world after 2012 and increasing its share of the global economy year after year. The period of 2012–2021 will see China’s economy grow at an average annual rate of 6.68%, higher than the world’s average growth rate over the same period. By 2021, China’s GDP will be US $17.73 trillion, increasing by 8.1% year-on-year. From 2012–2021, the digital economy plays an important role in China’s economic development, with an average growth rate of 15.9% between 2012 and 2021, increasing its contribution to the economy from 21.6% to 39.8%. By 2021, with a scale of US $7.1 trillion and ranking second in the world, the digital economy has become an important force driving China’s economic development. The digital economy played a key role in fighting the New Crown epidemic and restoring production and life, helping China’s economy to sustain its growth momentum. In the future, the digital economy will continue to play an important role in driving China’s sustained economic development and provide new opportunities for high-quality economic development Currently, the digital economy is integrated into many fields of economic development, and it has played an important role in promoting regional development and accelerating industrial transformation and upgrading [1]. In this context, studying whether the digital economy plays a role in improving environmental quality is of great theoretical and practical significance for further evaluating the social and economic benefits achieved as well as expanding knowledge of the factors improving environmental quality.

This can not only improve quality of life, but it is also the key to promoting high-quality economic development. In recent years, related issues have attracted much scholarly attention. Many studies have measured and identified the influencing factors of environmental quality. For example, regarding environmental quality measurement, the current research can be divided into two types: one measures environmental quality based on a single pollutant index [2]; the second builds a quality index system [3]. Regarding influencing factors of environmental quality, economic studies have identified direct factors, including economic scale growth [4], open innovation [5], industrial structure adjustment [6], technological innovation level [7], and corporate social responsibility [8]. Indirect factors include environmental regulations [9], foreign trade [10], and foreign direct investment [11,12].

Since Tapscott first proposed the concept of digital economy in 1996, domestic and foreign scholars have increasingly enriched their research on it. In terms of theoretical research, foreign scholars have made mature progress on the scientific concept [13] and basic characteristics [14] of the digital economy in the early stage. With the development of the digital economy, scholars can pay more attention to its evolution. The existing research is mainly conducted from three levels, namely, industrial digitalization, digital industrialization, and digital governance. In the research related to industrial digitalization, the existing achievements focus on the research for improving the traditional industrial production technology, and production process level, through digital empowerment [15]. In the research related to digital industrialization, the existing achievements include the formation of digital industrialization [16] and the promotion of digital industrialization on high-quality economic development [17]. In the research related to digital governance, more attention is paid to the process of promoting the healthy development of the digital economy, including the government’s promotion of digital economy development [18] and the government enterprise-coordinated governance of the digital economy [19]. In terms of empirical research, early relevant research is relatively scarce; The primary reason is the lack of a unified digital economy measurement index. With the digital economy’s development and the deepening of related research, measuring the digital economy has attracted the attention of government departments and academia domestically and abroad. Current efforts to measure the digital economy exist in two categories: The first is the direct method; that is, under a defined scope, the scale and volume of the digital economy in a certain area are counted or estimated [20]. The second is the comparison method, that is, based on multiple-dimensional indicators, the development situation between different regions is compared to judge the relative development of the digital economy or specific fields [21].

In academia, comparative methods are commonly used. Earlier, scholars such as Wunnava and Leiter [22] used this method to measure the level of digital economic development. Recently, with rapid digital development in China, empirical research has yielded rich results and focused more on investigating the digital economy’s impact, such as analyzing how it affects industrial transformation and upgrading, [1] employment quality [23], and regional innovation [24]. However, fewer results focus on digital economic development’s impact on environmental quality and the majority focus on the Internet’s impact on it. For example, progress in Internet technology can improve environmental quality through dynamic environmental monitoring, government information gathering, enhancing public participation, and increasing the intelligence of environmental protection industries [25]. It can also affect environmental quality by promoting the transformation and upgrading of the industrial structure and technology [26]; or through Internet platforms that promote residents’ positive attitudes towards environmental protection using informal Internet educational platforms to enhance residents’ environmental literacy, thereby improving environmental quality [27]. However, compared with the Internet’s development, the digital economy is a broader conceptual category, and the logical mechanism of how it affects environmental quality requires further exploration. In addition, digital economic development is likely to produce a degree of technological innovation; for example, it has the characteristics of openness and integration, which can promote knowledge sharing, enable intelligent research and development, and increase the probability of technological innovation [28,29]. This can improve environmental quality by enhancing the energy utilization rate of polluting enterprises and inventing new energy sources [30]. To this end, exploring the key ways in which the digital economy affects environmental quality requires conducting a systematic investigation in the context of technological innovation.

In summary, the existing literature provides rich results regarding the related research fields of the digital economy and environmental quality, but less research focuses on the relationship between the two. Its impact on environmental quality is only analyzed from specific perspectives regarding the Internet or the digital economy, and the results are largely one-sided. Therefore, the digital economy still lacks theoretical and empirical analysis regarding the improvement of environmental quality; research on its path mechanism from the perspective of technological innovation is even more lacking. In view of this, this study combines the digital economy, technological innovation, and environmental quality improvement in the same research framework. First, we theoretically analyze the inherent mechanism of the three’s relationship; then, based on 2013–2020 China inter-provincial panel data, we test the digital economy’s impact on environmental quality, the intermediary role of technological innovation, and the differences in the intermediary role of different levels of technological innovation. We further analyze how regional heterogeneity in the digital economy affects environmental quality and provide new ideas for exploring the digital economy’s impact mechanism on environmental quality.

Based on previous research, this paper discusses the relationship between the digital economy, technological innovation, and environmental quality improvement by using a complete theoretical framework and quantitative analysis. There are at least four possible contributions to the existing research, including 1. The construction of an index system to evaluate and measure digital economy development from three dimensions: digital economy infrastructure index, digital economy application index and digital industry development index. 2. Integration of the digital economy into the framework of environmental quality improvement analysis theoretically expand the study of factors affecting environmental quality improvement in the environment and analyze the path affecting environmental quality from the angle of technological innovation. This provides a theoretical basis for promoting the development of the digital economy, improving technological innovation capabilities, and improving environmental quality policy. 3. Using provincial panel data from 2013–2020, the paper uses the fixed-effect model and mediating effect model to make an empirical test of the influence of digital economy development on the improvement of environmental quality and the mediating mechanism to complement relevant domestic research. 4. Analyzes the regional heterogeneity of the relationships to provide important practical inspiration for different regions to implement strategies and promote coordinated regional development.

2. Materials and Methods

The development of the digital economy is characterized by high efficiency, integration, and diffusion, which can reduce environmental pollution. Relevant research is still in its infancy, with more research focusing on mutually beneficial networks and environmental quality. By summarizing the current research and application of the digital economy in reducing environmental pollution, one may observe that the digital economy can improve environmental quality by reducing the emission level of pollutants from regional polluters at the enterprise level. Regional environmental pollution comes from both production and distribution. The development of the digital economy can reduce environmental pollution and improve environmental quality from both aspects by enabling the use of digital platforms to directly match energy supply and demand, improve energy utilization efficiency, reduce pollutant emissions, and improve environmental quality. It also can reduce the cost of information search, reduce information asymmetry, and promote accurate matching of supply and demand. From the point of view of distribution, large production enterprises or third-party logistics enterprises can use digital technology to realize intelligent sorting of goods and intelligent warehousing control, improve logistics efficiency, reduce pollutant emissions during vehicle transportation and improve environmental quality [31].

At the industry level, with the continuous deep penetration of digital technology and three industries, there are differences in the penetration between the three industries, which affect the development of the industry to a certain extent, and then the quality of the environment. In terms of return on capital, the return on capital in both primary and secondary industries shows a downward trend. In the trend of profit maximization, digital capital tends to shift to the tertiary industry [32]. Therefore, the fusion of the digital economy and traditional industry presents the characteristic of "reverse integration". It will integrate with the tertiary industry, improve the service level of the industry, and gradually integrate with the primary and secondary industries [33]. In the short term, the integration of the digital economy and the tertiary industry can increase the proportion of the tertiary industry in the total industry, transfer capital to the tertiary industry, and reduce investment in the primary and secondary industries. While the tertiary industry discharges relatively less, the environmental quality will be improved. In the long run, the digital economy gradually merges with industries one and two and can effectively improve production efficiency, reduce unnecessary output, reduce pollutant emissions, and improve environmental quality. Therefore, in the short and long term, this “reverse integration” approach can effectively reduce industrial pollution levels and improve environmental quality.

With the spread of the digital economy, the idea of environmental protection can be implemented quickly. At the government level, the development of the digital economy enables the government to use digital technology to promote new green consumption concepts, such as green waste recycling and green travel, which significantly improve the quality of the environment [34]. At the enterprise level, platform companies can use big data to gather information about environmental protection, gather people who are environmentally conscious but can’t put methods into practice, and use the platform to actively participate in environmental protection and improve the quality of the environment [35]. At the personal level, the use of a mobile Internet platform to publish various environmental pollution events, through the platform’s expansion effect, enhances the environmental awareness of the government and enterprises to improve the quality of the environment.

Based on the above analysis of the mechanism of action, hypothesis H1 is proposed:

Hypothesis 1 (H1):

Developing the digital economy can reduce environmental pollution and improve environmental quality.

The research on the impact of technological innovation on environmental quality improvement started early. The theory that technological innovation can reduce environmental pollution can be traced back to Ehrlich and Holdren’s research that proposed a comprehensive assessment model that combines environmental factors with population size, wealth per capita and skill level, noting that technological advances can reduce environmental pollution from population growth [36]. Grossman & Krueger put forward the theory of three effects of international trade and environmental pollution and pointed out that the influence of international trade on environmental quality is mainly manifested in scale effect, structure effect and technology effect [37]. Subsequently, this theory has been widely used to study the relationship between technological innovation and environmental pollution. Summarizing the application practice of existing research and technological innovation in reducing environmental pollution, it can be seen that the impact of technological innovation on environmental quality can be summarized as follows: Technological innovation can not only improve energy utilization efficiency but also promote the development of new energy sources, reduce environmental pollution and improve environmental quality [38]. Technological innovation, especially breakthroughs in energy-saving technologies, have enabled enterprises to increase the use of energy-saving equipment, greatly increase energy utilization efficiency, reduce the emission of major pollutants, and improve the quality of the environment. The improvement of technological innovation has created conditions for the development and utilization of new renewable energy sources, and effective utilization can effectively reduce environmental pollution and improve environmental quality. At the personal level, science, technology, and innovation can change residents’ lifestyles and reduce their pollution emissions. The innovation of Internet technology has led to the continuous development of e-commerce and intelligent logistics. In addition, online platforms have relatively low prices and a wide variety of products, so people choose to shop online more. It reduces energy consumption and pollution emissions because of space constraints that force residents to use transportation to designated locations [39]. Technological innovation has changed the way residents travel. For example, residents can reduce detours by navigating by car, or they can choose how to travel by using the navigation software’s smart congestion alerts. The improvement of residents’ travel efficiency can reduce the environmental pollution caused by necessary travel and improve the environmental quality. At the industrial level, technological innovation can promote transformation and improvements reducing environmental pollution and improving environmental quality. Of the three major industries, the secondary sector has the highest intensity of pollution emission intensity. By improving the efficiency of industrial production through technological innovation, pollution emissions from secondary industries can be significantly reduced. At the same time, due to the low emission intensity of the tertiary industry, the continuous breakthrough of digital technology and the integration and development of the tertiary industry accounts for a large proportion, promoting the upgrading of industrial transformation, reducing emission intensity, and improving environmental quality [40]. Technological breakthroughs and innovations will have a disruptive effect on the development of the industry, creating new industries. Under the general trend of high-quality development of the economy, new industries are inclined to lower pollution industry, which will inevitably reduce pollutant emissions and improve environmental quality.

Based on the above analysis of the mechanism of action, hypothesis H2 is proposed:

Hypothesis 2 (H2):

Increasing the level of technological innovation can reduce environmental pollution and improve environmental quality.

Technological innovation is an important channel to improve environmental quality. The level of technological innovation directly influences the improvement of environmental quality. The development of the digital economy can directly or indirectly enhance the level of technological innovation, thus improving the quality of the environment. The “direct effect” refers to the direct improvement of technological innovation. The agglomeration of knowledge in the field of the digital economy is highly innovative, and the development of the digital economy is bound to be accompanied by the improvement of technological innovation. The convergence of the digital economy will also lead to new technological innovations. The improvement of digital technology will improve the ability of the real economy to be served by the digital economy and promote the integration and innovation of technology in all fields [29]; the indirect effect is to increase R&D investment and thus improve the level of technological innovation. It is embodied in the digital economy, which improves productivity by reducing the cost of information search and communication, facilitating the matching of supply and demand, and achieving economies of scale. The embedding of digital technology enables enterprises to more accurately locate user needs in order to meet users’ diverse needs at a low cost. As average costs fall and the types of products and services increase, economies of scale will emerge. A diversified business expands the profit source of the enterprise, and the increase in profit can promote the increase of R&D investment, thus promoting the improvement of the technological level [41]. Through the "direct effect" of improving the level of technological innovation induced by the development of the digital economy, environmental quality can be improved. The continuous improvement of digital technology through technological innovation can replace some traditional technologies and improve environmental quality. Technological innovations arising from the development of the digital economy can improve energy efficiency through accurate budgeting and monitoring of digital technologies. The related digital fusion technology innovation induced by the digital economy can improve the environment quality by continuously improving internet shopping, intelligent navigation, etc. Finally, with the development of digital technology, the reverse integration of industrial upgrading with agricultural and industrial service transformation can effectively reduce pollution emissions and improve environmental quality. The technological innovation induced by the digital economy through the “indirect effect” mainly involves increasing R&D investment, increasing the probability of technological innovation, and strengthening the regulatory role of technological innovation in the relationship between the digital economy and environmental quality.

Based on the above analysis of the mechanism of action, hypothesis H3 is proposed:

Hypothesis 3 (H3):

The development of the digital economy has raised the level of science, technology and innovation and improved the quality of the environment.

To test the influence of the development of the digital economy on the quality of the environment in China, a simplified econometric model is set up as follows:

$$E{Q}_{it}={\beta }_0+{\beta }_{1}D{E}_{it}+{\beta }_2{X}_{it}+{\mu }_i+{\epsilon }_{it}$$

(1)

In the formula, I and t represent the region and time, respectively; $D{E}_{it}$ represents the digital economy; $E{Q}_{it}$ represents environmental quality as noted by the emission intensity of industrial sulfur dioxide (SO₂); $X_{it}$ represents a set of control variables; ${\mu }_i$ represents individual effects, reflecting unobservable regional heterogeneity; ${\epsilon }_{it}$ is a random disturbance term.

To examine the regulatory role of technological innovation in the impact of digital economic development on environmental quality, this paper proposes to establish a mediating effect model and formulate Formulas (2) and (3) based on Formula (1):

$$INN{O}_{it}={\alpha }_0+{\alpha }_{1}D{E}_{it}+{\alpha }_2{X}_{it}+{\mu }_i+{\epsilon }_{it}$$

(2)

$$E{Q}_{it}={\gamma }_0+{\gamma }_{1}INN{O}_{it}+{\gamma }_{2}D{E}_{it}+{\gamma }_3{X}_{it}+{\mu }_i+{\epsilon }_{it}$$

(3)

In Formulas (2) and (3), INNOit represents the level of technological innovation in region i at time t, which is represented by the number of patent applications for invention patents (INNO₁), utility model patents (INNO₂) and design patents (INNO₃).

Data and Variable Description

Reference is made mainly to the “China Statistical Yearbook”, “China Science and Technology Statistical Yearbook”, “China Environmental Statistical Yearbook”, and provincial statistical yearbooks over the years. Considering the available data, the survey period is from 2013 to 2020. The final sample is panel data for 30 provinces (districts, municipalities) (excluding Tibet, Hong Kong, Macao, and Taiwan) from 2013 to 2020. The missing data were supplemented by means of mean estimation and ARIMA filling.

Specific variables are measured as follows:

Environmental Quality (EQ): With reference to the practice of Dai Lihua et al. [7], out of consideration for data integrity and accuracy. The ratio of industrial sulfur dioxide (SO₂) to GDP is used to represent its emission intensity; it is used as a proxy variable of environmental pollution and processed logarithmically. The higher the data, the worse the pollution and the lower the environmental quality. In the robustness test, the emission intensity of industrial nitrogen oxides (NO), industrial wastewater (WATER) and industrial solid waste (SOLID) are presented as a pollution proxy variable for environmental pollution, and data processing is the same as industrial sulfur dioxide (SO₂).

Digital Economy (DE): Measurement of the digital economy involves a wide range of factors, making it difficult to substitute a single indicator for a variable. Based on the index system (ITU digital economy development put forward by the United Nations International Telecommunication Union, Shanghai Academy of Social Sciences and Caixin Think Tank, this paper refers to the research results of Han Lu et al. [42]. The digital economy infrastructure index, digital economy application index and digital industry development environment index are used as the primary indicators of the digital economy indicator system. The secondary indicators of the digital economy indicator system are selected according to the availability and timeliness of data, the provincial digital economy development level indicator system is established (see

Table 1. China’s provincial digital economy evaluation index system.

Name	First-Level Indicator	Secondary Indicators
Digital Economy Development Evaluation Index System	Digital Economy Infrastructure Index	Mobile phone users (10,000 households)
		Internet broadband access users (10,000 households)
		Software business revenue (100 million yuan)
	Digital Economy Application Index	Number of companies with e-commerce activities (number)
		E-commerce sales (100 million yuan)
		Digital Financial Inclusion Index (%)
	Digital Industry Development Index	Number of electronic and communication equipment manufacturing enterprises (person)
		Electronic and communication equipment manufacturing practitioners
		Average number of people (pieces)

), and the comprehensive indicators of the digital economy are obtained by linear weighting. For specific processing methods, please refer to the research of Liu Jun et al. [43]. Taking 2013 as the base period and ensuring the comparability of the provincial indices in different statistical years, the calculation formula is constructed as follows:

$$X_{i,t}=\frac{V_{i,t}-V_{min0}}{V_{max0}-V_{min0}}\ast 6+1$$

(4)

In the formula, t represents the index year, V_max0 and V_min0 represent the maximum and minimum values of the original data in the base year, the processed data are comparable between different years, and the index value of the base period ranges from 1 to 7. The non-base period indicator range can be less than 1 or greater than 7, which can reflect the development of the digital economy after the base period. In terms of weight processing, the method of the weight of each level of indicators = 1/number of indicators of this level is used so that the weight of each secondary indicator relative to the total index = the weight of the secondary indicator to which the indicator belongs * the weight of the primary indicator to which the indicator belongs. After the weight is determined, the linear weighting method is used to calculate the digital economy development level of each province.

Technological Innovation (INNO). Two indicators, innovation input and innovation output, are commonly used to measure the level of technological innovation in the region. Innovation input refers to the regional R&D investment, and innovation output refers to the regional patent application volume or authorization volume. Because innovation output reflects the actual innovation level of the region better than input, this paper chooses innovation output as the proxy variable of technological innovation. In innovation output, the number of patent applications and the number of patent grants is often used as proxy variables, and patent grants have a certain time lag, so the number of patent applications (INNO) is used as a proxy variable for technological innovation. Patent applications can be divided into invention patent applications (INNO₁), utility model patent applications (INNO₂) and design patent applications (INNO₃). In this paper, the total number of patent applications and the number of patent applications of three types in each province (district and city) are selected for logarithmic processing to represent the level of technological innovation and logarithmic processing.

Referring to the existing research, control variable selection: (1) Economic development level (GDP). The proportion of GDP of each province (autonomous region and city) in the national GDP is selected as a proxy indicator to measure the level of regional development. (2) The level of foreign direct investment (FDI). The proportion of total foreign direct investment in each province (autonomous region and municipality) to GDP indicates the level of foreign direct investment. (3) The level of industrialization (INDU). The ratio of the added value of the secondary industry to GDP in each province (autonomous region and city) is selected to represent the level of industrialization. (4) The level of import and export trade (IM). The proportion of the import and export trade volume of each province (autonomous region and city) to GDP represents the regional import and export trade level (5) Technology scale (TEC). Selecting the proportion of the technology market turnover in each province (autonomous region and city) to GDP represents the regional technology scale. The data description statistics of the above variables are shown in

Table 2. Descriptive statistics of each variable.

Variable	Variable Name	Observations	Mean	Standard Deviation	Min	Max
DE	Digital economy	240	1.124	0.644	0.333	4.357
SO2	Industrial SO2 Emission Intensity	240	2.373	1.481	−3.019	6.609
NO	Industrial NOx Emission Intensity	240	2.987	0.855	−1.552	6.566
SOLID	Industrial Solid Waste Emission Intensity	240	8.243	1.240	3.376	11.783
WATER	Industrial wastewater discharge intensity	240	11.353	0.435	7.005	13.357
INNO1	Invention patents	240	8.341	1.388	4.511	11.166
INNO2	Number of utility model patents	240	9.827	1.338	5.652	12.914
INNO3	Design patents	240	8.518	1.574	4.836	12.461
GDP	The level of economic development	240	0.052	0.274	0.003	4.254
FDI	Level of foreign direct investment	240	0.564	2.218	0.002	34.233
INDU	Level of industrialization	240	0.501	0.245	0.004	3.967
IM	Import and export trade level	240	23.835	24.714	0.716	127.055
TEC	Technology scale	240	0.016	0.029	0.000	0.175

.

3. Results

3.1. Benchmark Regression Results

Based on the results of the F and Hausman test, a fixed effect model was selected to estimate the impact of the digital economy on environmental quality and robust standard errors were used for each model. The results are shown in

Table 3. Impact of digital economy on environmental quality.

Variable	(1)	(2)	(3)	(4)	(5)
	lnSO2	lnSO2	lnNO	lnSOLID	lnWATER
DE	−2.870 ***	−2.965 ***	−1.400 ***	−0.374 ***	−0.723 ***
	(−7.353)	(−9.185)	(−9.107)	(−3.520)	(−8.944)
GDP		−1.007 ***	−1.011 ***	−1.040 ***	−1.022 ***
		(−40.884)	(−72.972)	(−124.061)	(−113.896)
FDI		−0.023 ***	−0.017 ***	0.007 ***	−0.003 ***
		(−9.016)	(−13.528)	(7.971)	(−4.970)
INDU		0.600 ***	0.644 ***	0.534 ***	0.575 ***
		(6.072)	(7.541)	(20.785)	(13.830)
IM		−0.021	−0.006	0.006	−0.005
		(−1.180)	(−0.858)	(1.265)	(−1.486)
TEC		−13.837 **	4.083	1.854	2.201
		(−2.354)	(0.978)	(0.750)	(1.039)
Fixed effects	Yes	Yes	Yes	Yes	Yes
Constant term	5.601 ***	6.189 ***	4.385 ***	8.265 ***	12.020 ***
	(12.760)	(12.931)	(22.532)	(58.227)	(79.574)
Observations	240	240	240	240	240
Adj-R2	0.682	0.787	0.762	0.777	0.840
number of individuals	30	30	30	30	30

Note: **, *** indicate significance at the 5%, and 1% levels, respectively.

.

Column (1) reflects regression results without other control variables, with a digital economy coefficient of −2.870, which passed the 1% salience test, indicating that the development of the digital economy has reduced environmental pollution and improved environmental quality.

Column (2) gives an estimate of the control variables to be added. Added to the control variable, the digital economy coefficient is −2.965, passing the 1% significance level test. The data shows that the impact of the digital economy on sulfur dioxide emissions is negative, with or without control variables, and was significant at a significant level of 1%. Therefore, Hypothesis 1 is validated.

When compared to three indicators of industrial nitrogen oxides, industrial solid waste, and industrial wastewater in the robustness test, the impact coefficients of the digital economy on these pollutants were −1.400, −0.374 and −0.723, respectively. Further, all passed the 1% significance level test, suggesting that developing a digital economy could improve environmental quality by reducing emissions of these three pollutants.

The order in which digital economic development affects emissions of these four pollutants is: industrial sulfur dioxide > industrial nitrogen oxides > industrial wastewater > industrial solid waste.

3.2. Robustness Test and Endogeneity Analysis

To improve the reliability of the empirical results, a stability test and an endogeneity test were carried out. First, we tested the robustness of the model by substituting explanatory variables, i.e., industrial nitrogen oxide emission intensity, wastewater emission intensity, and solid waste emission intensity, to represent environmental pollution in place of industrial sulfur dioxide emission intensity. The experimental results are presented in columns (3–5) of Table 3, where the direction and significance level of the impact of the core explanatory variable on environmental pollution are consistent with the regression results in column (2), indicating that the estimation results in this paper are stable. Secondly, this study uses the 2SLS instrumental variable regression method to test endogeneity, referring to Zheng Wanteng et al. [44] and selects the Internet penetration rate as an instrumental variable. The results are shown in

Table 4. Endogenous problem handling.

Variable	(1)	(2)
	DE	lnSO2
IV-INTERNET	3.253 ***
	(9.90)
DE		−2.873 ***
		(−9.64)
GDP	0.122	−0.729 **
	(1.20)	(−2.43)
FDI	−0.017	−0.110 ***
	(−1.35)	(−2.98)
INDU	−0.037	0.546
	(−0.32)	(1.58)
IM	0.013 ***	0.024 ***
	(10.79)	(4.25)
TEC	2.395 **	−13.897 ***
	(2.40)	(−4.46)
Fixed effects	Yes	Yes
Constant term	0.057	5.088 ***
	(0.61)	(17.91)
Observations	240	240
Adj-R2	0.576	0.280
C-D Wald F	-	242.98

Note: **, *** indicate significance at the 5%, and 1% levels, respectively.

. The instrumental variable (IV-INTERNET) is very positive for the digital economy regression result, indicating a strong correlation between the two. At the same time, the C-D WaldF value is 242.98, indicating no weak instrument variables. The regression coefficient of the digital economy is significantly negative, which shows that the conclusions of this study are still valid after endogenous factors are considered.

3.3. Test of Mediating Effect

To investigate the mediating effect of technological innovation in the relationship b tween digital economy and environmental quality, three models need to be established:

• Baseline models without mediating variables.
• Mediating variables are models that explain variables.
• Baseline models add models of mediating variables.
• The three models must satisfy three conditions:
• The influence of digital economy on environmental quality should be significant.
• The influence of digital economy on mediating variables should be significant.
• When considering mediating variables, the influence of digital economy on environmental quality should diminish or even disappear, which means that the influence of the digital economy on environmental quality comes from mediating variables in part or in whole.

As can be seen from column (1) of

Table 5. Mediating effect of technology effect between the digital economy and environmental quality.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	SO2	INNO	SO2	INNO1	SO2	INNO2	SO2	INNO3	SO2
DE	−2.965 ***	1.015 ***	−2.302 ***	0.733 ***	−2.778 ***	1.466 ***	−2.065 ***	0.357 *	−2.916 ***
	(−9.185)	(6.346)	(−5.137)	(5.065)	(−6.844)	(7.448)	(−4.480)	(1.838)	(−8.393)
INNO			−0.653 ***
			(−3.215)
INNO1					−0.255
					(−1.412)
INNO2							−0.614 ***
							(−3.835)
INNO3									−0.137
									(−0.701)
Control variable	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant term	6.189 ***	9.625 ***	12.477 ***	9.098 ***	8.508 ***	8.272 ***	11.266 ***	8.232 ***	7.319 ***
	(12.931)	(43.113)	(6.632)	(40.744)	(5.160)	(29.826)	(8.471)	(32.510)	(4.988)
Observations	240	240	240	240	240	240	240	240	240
Adj-R2	0.787	0.621	0.808	0.470	0.790	0.657	0.818	0.222	0.788
Number of individuals	30	30	30	30	30	30	30	30	30

Note:*, *** indicate significance at the 10%, and 1% levels, respectively.

, the impact coefficient of the digital economy on the quality of the environment was −2.965, which passed the significance level test of 1%, consistent with the results in column (2) of Table 3. As can be seen from column (3), the technical innovation had an impact factor of −0.653 on sulfur dioxide emissions and passed the 1% significance level test. The results show that using the total number of regional patent applications as a proxy variable for technological innovation can reduce environmental pollution and improve the quality of the environment. Therefore, Hypothesis 2 is validated.

In combination with columns (1–3), when the digital economy is added to the model along with the number of patent authorizations, the digital economy has a smaller impact on environmental quality than when interpreted separately, and the coefficients are significant, meeting all three conditions for mediating effect. The data show that the digital economy plays a partial intermediary role in reducing environmental pollution by increasing patent authorizations, upgrading regional technological innovation, and reducing the environmental pollution. Therefore, Hypothesis 3 is validated.

This paper further analyzes the existence and difference of technological innovation mediating effect represented by three kinds of patent applications: invention, patent, and design. When combined with columns (1, 4, 5) and (1, 8, 9), it can be seen that when both the digital economy and invention, digital economy and design patents are added to the model, the invention patents and design patents have little effect on sulfur dioxide emissions and the mediating effect does not satisfy three conditions. This shows that the digital economy cannot increase the level of technological innovation in the region by increasing the number of patents for inventions patents and designs, thus reducing the environmental pollution.

In conjunction with columns (1, 6, 7), we can see that the digital economy has a positive impact on utility model patents. When both the digital economy and utility model patents are added to the model, the influence of the digital economy on environmental quality is smaller than when it is explained separately, and the coefficient is significant, satisfying the three conditions of mediating effects. The data show that the digital economy has a certain intermediary effect on reducing environmental pollution by increasing the number of patents for utility model patents, raising the level of regional technological innovation, and reducing the environmental pollution. The reason for this may be that there are fewer patents for inventions and those related to energy consumption and consumption reduction that are directly or indirectly integrated into the development of the digital economy, and the effect on environmental quality is not obvious; but there are more patents for utility models related to energy conservation and energy consumption, which significantly improve environmental quality.

The technology innovation level of design patents is low and cannot play an effective intermediary role in the digital economy and environmental quality improvement. The mediating effect was further tested using the more rigorous Bootstrap (sampling times were set to 1000 times with a 99%confidence interval), as shown in Table 5. The results (see

Table 6. Results of the mediating effect test (Bootstrap test).

Mediating Variable	Influence Effect	Coef.	S.E.	Z	P	99% Confidence Interval
INNO	Indirect effect	−0.663	0.228	−2.91	0.004	[−1.251, −0.075]
	Direct effect	−2.302	0.306	−7.53	0.000	[−3.088, 1.515]
INNO2	Indirect effect	−0.900	0.287	−3.14	0.002	[−1.639, −0161]
	Direct effect	−2.065	0310	−6.66	0.000	[−2.863, −1.267]

) showed that a 99% confidence interval (CI) did not include 0 when total regional patent filings were used as a proxy variable for technological innovation, either directly (−3.088, −1.515) or indirectly (−1.251, −0.075), reconfirming the existence of mediating effect. When regional utility model patents were used as proxy variables for technological innovation, 99% of the confidence interval did not contain 0 for either direct effects (−2.863, −1.267) or indirect effects (−1.639, −0161), again verifying the existence of intermediate effects.

4. Discussion

China’s economic development attracts worldwide attention, but the problem of unbalanced regional development still exists. Differences in the stage of economic development, the level of institutional mechanisms, and resource endowments in different regions will lead to significant regional differences in the impact of the digital economy on environmental quality. From a macroscopic point of view, the level of system and industrial structure in the eastern region is higher than that in the central and western regions. Embedding digital technology in the central and western regions could greatly improve the level of system and industrial structure, thus improving the quality of the environment. Therefore, from the institutional and industrial structure level, the digital economy may play a marginal role in improving environmental quality in these areas than the eastern region. At the micro level, the development of the digital economy has gradually blurred the boundaries between regions, allowing production, circulation, and consumption entities in less-developed regions to access new services through digital technologies. As a result, the growth of the digital economy has benefited more “tail” users and played an inclusive role in the digital economy. As a result, the marginal effects of digital technologies in improving environmental quality by production, distribution and consumption in the central and western regions may be greater than in the east. Based on the above analysis, the influence of the digital economy on the quality of the environment has regional differences, and the influence of the central and western regions may be greater than that of the eastern regions. To test this hypothesis, we divide the samples into three subsamples: East, Central and West for further analysis.

From the regional heterogeneity results in

Table 7. Regional heterogeneity of the impact of the digital economy on environmental quality.

Variable	(1)	(2)	(3)	(4)	(5)
	National	National	Eastern	Central	Western
	lnSO2	lnSO2	lnSO2	lnSO2	lnSO2
DE	−2.870 ***	−2.965 ***	−1.650 ***	−3.735 ***	−3.850 ***
	(−7.353)	(−9.185)	(−3.854)	(−5.802)	(−6.724)
Control variable	No	Yes	Yes	Yes	Yes
Fixed effects	Yes	Yes	Yes	Yes	Yes
Constant term	5.601 ***	6.189 ***	8.533 ***	5.036 ***	5.415 ***
	(12.760)	(12.931)	(3.578)	(7.219)	(4.139)
Observations	240	240	88	64	88
Adj-R2	0.682	0.787	0.835	0.922	0.770
Number of individuals	30	30	11	8	11

Note: *** indicate significance at the 1% levels.

, the coefficients of the digital economy (DE) are −1.650 in the eastern region, respectively, and have passed the 1% significance level test. The central region is −3.735 and has passed the 1% significance level test; the western region is −3.850 and has passed the 1% significance level test. This shows that the development of the digital economy in eastern, central, and western regions has reduced environmental pollution and improved the quality of the environment. However, there is a certain heterogeneity in the degree of influence, which is manifested in “West” > “Middle” > “East”. The results validate this analysis. The reason for this may be the influence of industry transfer, the pollution in the central and western regions is more serious than in the eastern regions, and the marginal effect of the digital economy on productivity in the central and western regions is stronger. Moreover, the digital economy is inclusive, and its growth can benefit more “tail” users, serving more midwestern businesses to reduce pollution and improve environmental quality.

5. Conclusions

This paper analyzes the influence of the digital economy on the improvement of environmental quality and the possible regulatory role of technological innovation in the relationship between the two. According to the analysis, the influence of the digital economy on the improvement of environmental quality may vary regionally. The intensity of sulfur dioxide emission intensity was used as a proxy variable for environmental pollution in a sample of China’s provincial panel data from 2013−2020. The influence of digital economy development on regional environmental quality is experimentally tested by using the fixed effect model and mediating effect model. The results show at the national level, the development of a digital economy can effectively reduce sulfur dioxide emissions from industry and improve environmental quality. Compared to other environmental pollution indicators, the order of impact is industrial sulfur dioxide > industrial nitrogen oxides > industrial wastewater > industrial solid waste. This shows that the digital economy can effectively reduce environmental pollution and improve environmental quality.

The mediating effect shows that the digital economy can reduce environmental pollution and improve environmental quality by promoting technological innovation. Among them, when the number of patent applications and the number of utility model patents as proxy variables of technological innovation, there is a certain intermediary role. This intermediary effect does not exist when invention patents and designs are used as proxy variables for technological innovation. This shows that in the digital economy, there are differences in the mediating role of innovation at different levels of technology in improving environmental quality.

In the further subregional discussion on the impact of the development of the digital economy on the improvement of environmental quality, it is found that the improvement of environmental quality caused by the development of the digital economy has obvious regional heterogeneity. Compared with the eastern region, the development of the digital economy in the central and western regions has a greater impact on the quality of the environment.

The conclusion of this paper is of great practical significance to promote the development of the regional digital economy, enhance the economic effect of the digital economy, effectively improve the quality of the environment in our country, and then promote the development of a high-quality economy. Relevant policy recommendations include:

Increasing the construction of communications infrastructure and accelerating the digital transformation of the industry. Governments should pay attention to the role of the digital economy in improving environmental quality. Make full use of the new opportunities brought about by the development of the digital economy, push the economy from “extensive” to “intensive”, reduce environmental pollution and improve the quality of the environment.

Increase investment in R&D and raise the level of technological innovation. All regions should combine their development practices and increase their investment in R&D and manpower. With the help of digital and integrated innovation, the number of patents related to energy conservation and energy consumption has continuously increased, adding new impetus to the improvement of regional environmental quality.

To promote coordinated regional development and promote high-quality economic development. The development of the digital economy in different regions has different effects on the quality of the environment. The eastern region should continue to play the “icing on the cake” role of the digital economy in improving the quality of the environment and encourage the digital economy to develop to a higher level, while the central and western regions should seize the new opportunities brought about by the development of the digital economy.