Digital Economy, Technological Innovation and Green High-Quality Development of Industry: A Study Case of China

Abstract

This research project investigates the direct and indirect influences of the digital economy in promoting the green high-quality development of industry. We analyze the baseline regression, spatial spillover effect, intermediary effect, and threshold effect of the digital economy and its technological innovation processes on the green high-quality development of industry in 31 provinces and cities in China during the period between 2012–2020. The results reveal that the digital economy generates a positive impact and nonlinear incremental character on the green high-quality development of industry, and technological innovation is the mediating factor of both. Digital economy and technological innovation have a threshold effect. When they cross the threshold value, the force will be strengthened. The econometric analysis of spatial spillover shows the green high-quality development of intra-regional and inter-regional industries is promoted by the digital economy. The dimension of regional differences in the analysis shows that the effect of the digital economy is not evident in the western region of China, while the impact is a lot more prominent in the eastern than in the middle area of the country. Study results provide measures of the green high-quality development of industry, such as the application of green scenarios, differential development, and collaborative development in the digital economy.

1. Introduction

With the adoption of massive reforms to support a market-driven economy, China has maintained high growth levels, and strategically established a global paradigm for economic development within the last 30 years. People’s living standards and quality of life have been greatly improved. Nevertheless, the economic growth has generated some unanticipated socioeconomic problems which China now shares with developed economies at global levels: maximization of energy efficiency, environmental pollution, and imbalance in its industrial infrastructure.

In the 18th National Congress of the Communist Party of China (CPC), China has incorporated and implemented the new development concept of coordination, greenness, innovation, sharing, and openness to stimulate creativity and innovation, encouraging openness towards global partners for the enhancement of options and opportunities focused on the resolution of green issues vis-a-vis sustainable economic development. The data released by the Global Environmental Performance Index 2020, jointly issued by Columbia University and Yale University, rank China 130th and 120th in terms of Global Environmental Performance Index and ecosystem vitality [1]. From the perspective of the global initiative to build a green ecological civilization, China urgently needs to reflect on the seriousness of its environmental and ecological concerns. Presently, the economy of China has embarked on a new phase of high-quality development, which gives the highest priority to green issues, which include attaining low pollution, low energy consumption, and sustainable development, with a clear focus on the global value chain in the context of industry’s green development.

The green high-quality development of industry is the great challenge faced by China with the beginning of the new development period. The popular and mass use of the new generation of IT with the ascendancy of AI, the blockchain, and big data, provides a strategic support system for green high-quality development of industry. Since 2017, “Digital economy” has been the centerpiece of government reports. In the 19th National Congress Report, the CPC advocated to “promote the deep integration of the Internet, big data, AI and the real economy” [2], within the programs of the green high-quality development of industry and digital economy. Can the green high-quality development of industry be promoted through the digital economy with the help of technological innovation, and how does it work? These are significant for the study of the green high-quality development of industry. This paper starts by examining the mechanism and processes of the digital economy applicable to green development and analyzing and measuring its direct and indirect effects. By formulating a system of measuring the impacts of the processes of the digital economy and industrial green development and using the level of technological innovation as the intermediary variable, this study empirically tests the mediating effect, spatial spillover effect, and threshold influence of the digital economy on green high-quality development of industry.

2. Review of Related Literature

2.1. Green Development of Industry

The contemporary discourse on high-quality industrial development includes the five concepts of “innovation, coordination, green, openness, and sharing” of high-quality economic development in the modern economic age. This alters the existing industrial development paradigm, which is today superseded by cultivating and enhancing the new development of kinetic energy, advancing and rationalizing the industrial structure, and producing high-quality, value-added, complex, and highly personalized products to fulfill the needs of people’s requirements for an improved quality of life [3]. Numerous studies have examined the paradigm of high-quality industry-focused development [4] and the construction of an industrial high-quality evaluation index system [5]. The empirical evidence is largely industry-specific for comprehensive measurement and evaluation of manufacturing [6], industry [7], and the high-technology industry [8,9], providing good references for accurately determining the driving force and processes of industrial high-quality development.

The concept of green development originates from discussions on the green and ecological economy, a new model for sustainable development focused on the protection of the natural environment in the contexts of the ecological environment and resource carrying capacity [10]. Industrial green development is a response to green development at the industrial level [11]. Both domestic and foreign studies on industrial green development usually focus on agriculture, industry, energy-intensive industries, and modern service industries, emphasizing the construction of green development models, measurement of industrial green development level and affecting factors, and green development processes [12,13,14].

In this context, the green high-quality development of industry attains a high level, which is an ideal state [15]. From the viewpoint of the development process, green high-quality development of industry transforms, only enhancing the economic benefits into the quality of green development. It shapes and enriches the sustainable development capacity of the industry based on sound ecological practices and fosters the evolution of a circular economy consisting of the interplay of ecology and the green and development goals focused on delivering economic, social, and ecological benefits [16]. Indeed, resource constraints and increasingly serious environmental pollution challenge industry towards the integration of industrial economic benefits, ecological environment, and social harmony for the realization of green and high-quality development.

2.2. Digital Economy

The concept of “digital economy” was primarily put forward by Prof. Don Tapscott and discussed from micro-business paradigms and the latest technologies [17]. Subsequently, foreign and domestic studies focused their discussions on the connotation, evolutionary laws, characteristics, measurement methods, value creation, and the risk management aspects of the digital economy [18,19]. Studies claimed that the digital economy represents a novel economic form based on digital technology, with digital data and digital platforms as its two basic features. Digital data make enterprise business traceable, controllable, and traceable. Once data are transformed into digital intelligence and commercialized, it can create value [20,21]. A digital platform provides an online interaction mechanism for market participants, optimizes resource allocation by reducing information asymmetry and becomes a strategic driving force for industrial structure up-gradation and economic high-quality development [22,23]. Most studies on digitization in high-quality industrial development focused on the two driving paths of digital industrialization and industrial digitization. In digital industrialization, studies mainly analyzed the development trend and prospect of using digital technologies, including mobile Internet, new-generation semiconductors, IoT, and AI, which seek to expand new models, cultivate new ones, and create industries [24]. Among them, digital infrastructure is the carrier for the development and integration of the digital economy. It is also essential to fortify the framing and support of IoT, industrial internet, and cloud computing [25]. Taking advantage of digital industry clusters to create high-quality digital industry chains increases the level of intelligent manufacturing, research, and development. An industrial Internet ecosystem integrates government, industry, academia, research, and application and builds a strong ecological base for the advancement of the industrial digital economy [26]. Digital industrialization strengthens industrial innovation efficiency and expands industrial collaboration levels with the help of the diffusion effect, the spillover effect of digital technology, and the platform effect [27,28,29]. Studies show that the digital transformation of the industry at national and regional levels prioritizes manufacturing, energy, and agriculture. Lee et al., argued that information access is a catalyst for upgrading and transformation of industrial structure and it is based on the industrial linkage standpoint. Supported and led by new-generation digital technology, the amalgamation of change, traditional technology, and digital technology-driven innovations accelerates the growth, differentiation, and reorganization of industries by transforming traditional industrial production methods, production processes, and industrial organization models [30]. This trend not only promotes the technological upgrading of traditional industries and enhances production efficiency [31], but also forms new products, services, business models, and new modes, thereby extending the industrial chain, reconstructing the industrial ecosystem, and promoting high-quality industrial development [32,33].

2.3. Digital Economy vis-à-vis Green Development of Industry

The digital economy, driven by digital technology, not only provides new production factors and changes in production methods towards the realization of green industrial development but also improves the efficiency of energy utilization and limits and eliminates environmental pollution [34]. Studies examined its role and impact on green development in industry, manufacturing, agriculture, and construction from theoretical and empirical perspectives. From the theoretical perspective, Liu and Zheng analyzed the digital economy and focused on the prospect of transforming traditional industries, arguing that the utilization of new IT setups to develop innovative green technologies [35,36], such as emission reduction, flexible manufacturing, energy conservation, and biological manufacturing. These infuse existing production technologies, processes, and circulation processes with high pollution and high energy consumption into more resource-sensitive and energy-efficient processes with a diminished ratio of wastage in all links of the industrial chain and reduce carbon emissions. Analyzing the digital transformation of green industries, Qian argued that digital technology promotes the life-cycle management of the whole industrial chain of emerging green and low-carbon industries and decreases energy consumption in the industrial development system [37]. Qian [38] and Wu and Zhu [39] suggested that digital technology helps to acquire ecological and environmental protection information quickly, authentically, and effectively; they proposed real-time supervision of traditional industries and green industries in R&D and design, manufacturing, transportation management, and recycling, and increase the level and efficiency of accurate monitoring and management of energy and carbon emissions. Fan and Xu [40] emphasized the contribution of residents’ digital literacy. The improvement of residents’ digital level can effectively and sustainably replace low-energy-consumption segments and strengthen the degree of substitution of factors for energy consumption. From the empirical perspective, Papalia R B. and Bertarelli S [41], Cheng et al. [42] constructed industrial green total factor productivity and reported that the digital economy generates a statistically significant positive promotion impact on industrial green production efficiency, enabling industrial greenness. Yin and Tian [43] used the Tobit model to determine that the digital economy exerts a substantial isotropic driving effect on the innovation efficiency of high-technology industries, and innovation supports the green development of these industries. Moreover, the green economy can contribute to the digital economy. Furthermore, Verhoef E.T. and Nijkamp P [44] and Jiang [45] claimed that the green economy should participate in the discourse of the digital economy, and comprehensively implement the concept of green development to accelerate the process of green digital transformation.

Based on the studies previously cited, many valuable studies have been accumulated on the digital economy, industrial high-quality development and green industrial development. Nevertheless, there are still some deficiencies: First, the current literature only analyzes the meaning and path of green high-quality development of industry, and the index evaluation system is incomplete by green total factor productivity. Secondly, the general literature mainly analyzes the impact of the digital economy on industrial green development or industrial high-quality development. There are few studies on the impact of green and high-quality development of industry from theoretical and empirical perspectives. Based on the extent of research studies, the innovation of this study to the existing literature is as follows: (1) the segregation of the impact mechanism of the digital economy on green development of the industry into direct and indirect effects, and systematically assess the intrinsic dynamics of the digital economy acting on green development of the industry; (2) the construction of the system of green development of the industry from three aspects: economic benefits, social benefits, and ecological benefits; (3) the strengthening of the regional heterogeneity of the digital economy on green development of industry and the articulation of the main development ideas of green development of industry in different regions.

3. Theoretical Research and Hypothesis

The digital technology in the digital economy is the driving force that provides new production factors, changes in production methods, energy utilization efficiency, and environmental pollution improvement for the green high-quality development of industry. It will not only generate direct impact through its advanced and network characteristics but also realize indirect impact with nonlinear characteristics through scale and spillover effects.

3.1. Direct Impact of the Digital Economy on Green High-Quality Development of Industry

Under the dual carbon target, technological innovation, as a production factor, lies at the core of this initiative to promote green high-quality development of industry. It also serves as the embodiment of the digital economy. (1) The digital economy achieves its green high-quality development of industry through digital technologies, such as AI, industrial Internet, and supercomputing centers, by upgrading production methods, processes, equipment use, and precise management efficiency of industrial green development, which effectively shortens the development cycle of technologies and products, enhancing the market fit of new technologies or new products, reducing the carbon emissions and the ratio of resource and energy consumption in each link of the industrial chain [46]. (2) The digital economy provides an effective information-exchange platform for green high-quality development of industry, expands access to information channels and market scale, eliminates mismatches between supply and demand, facilitates resource search and matching for making industrial green development more efficient, realizes the integration and sharing of various resources among different industries, enterprises, and regions, and systematizes resource allocation. (3) By establishing the management mode and regulatory mechanism of industrial green development, the market is guided to unceasingly expand the demand for industrial clean technology, green technology equipment, and green products, to force the industry toward green development and stimulate the industry to attain transformation in strategies of green development, green products and services, and industrial organization system. Hence, the below hypothesis is postulated in this study:

Hypothesis 1.

The digital economy creates a positive impact on green high-quality development of industry, and the digital economy can achieve the green high-quality development of industry goals through technological innovation.

3.2. Indirect Influence of the Digital Economy on Green High-Quality Development of Industry

(1) The threshold effect of the digital economy on green high-quality development of the industry. The digital economy industry informs the operations of various industries, creates new industries and business models, and reduces the space-time boundary of industrial development through the massive application of 5G, IoT, Cloud computing, and blockchain. Improving the speed of knowledge and information transmission through Internet platforms and digital platforms, the industry expands intelligent interconnection and data-sharing. Producers, retailers, traders, and enterprises can obtain production materials quickly and efficiently with near-zero marginal cost with synergistic efficiency. in the industrial chain. The precise allocation of resources that limits the use of resources improves. Additionally, it can accelerate the integration of digital technology and traditional technology, empower every link of industrial development, and develop sophistication in networking initiatives, such as joint innovation between enterprises and enterprises, joint innovation between enterprises and university research institutions, and business linkage. With these, consumers can access and use with precision and speed green consumer products developed with the help of digital technology [36]. As the number of participants increases, the benefits of promoting green industrial development would increase geometrically with the expansion of the digital economy. As a result, this study proposes the following hypothesis:

Hypothesis 2.

Different stages of the digital economy development have different degrees of effect on the green high-quality development of industry, with nonlinear characteristics.

(2) The spatial spillover influence of the digital economy on the green high-quality development of industry. The rapid development of the digital industry not only drives the comprehensive gathering of technology, capital, talents, and other elements but also accumulates scientific research institutions, related enterprises, government departments, and intermediaries, creating a perfect digital economy network. The network markedly decreases the space-time distance between regions and expands the connectivity of economic activities between regions, generating spatial spillover effects. In contrast, the digital economy’s development makes the market more transparent and fairer, forcing enterprises to seek long-term sustainable development in the escalating market through green innovation. This directs the innovation subjects in the industrial chain towards green transformation through the market signal generated by the forward-backward correlation effect [36], thus forming regional industrial agglomeration. Industrial regional agglomeration infrastructure can be shared and built together to centralize environmental pollution management facilities, improve resource utilization efficiency, and reduce pollution emissions. The digital industry agglomeration fosters the diffusion and dissemination of green ideas, green technologies, and knowledge-creating extensive spatial spillover characteristics through effective communication, learning, and coordination among various players. Thus, the below hypothesis is presented in this study:

Hypothesis 3.

The digital economy promotes the green high-quality development of industry of inter-regional industries through spatial spillover effects.

4. Research Method and Variable

4.1. Research Method

To validate the previous research hypothesis, the first phase is the construction of a benchmark regression model to analyze the direct influence of the digital economy on the green high-quality development of the industry.

$$Gd{i}_{it}={\alpha }_0+{\alpha }_{1}di{g}_{it}+{\alpha }_2{X}_{it}+{\mu }_i+{\sigma }_t+{\epsilon }_{it}$$

(1)

$Gd{i}_{it}$ denotes the index of green high-quality development of industry; $di{g}_{it}$ denotes the digital economy’s development level; $X_{it}$ denotes a series of control variables; $i$ is province; $t$ is period; ${\alpha }_0$ and the below variables ${\beta }_0 ,{\gamma }_0, {\phi }_0, {\eta }_0$ denote constant terms that are not explained by independent variables, the long-term stable part of the existence; ${\alpha }_1,{\alpha }_2$ denote the regression coefficients; ${\mu }_i$ denotes individual fixed effect; ${\sigma }_t$ denotes time fixed effect; and ${\epsilon }_{it}$ is a random error term.

To test the conceivable indirect influences of the digital economy on the green high-quality development of industry, it is essential to determine whether technological innovation serves as a mediating variable between the two. The methods are as follows: first, we test the coefficients of the digital economy’s development level in the model (1) based on passing the significance test. The following were formulated to test the hypothesis: a constructed regression model of the digital economy ($di{g}_{it}$) on the possible mediating variable technological progress ($T{I}_{it}$) and the regression model of the digital economy ($di{g}_{it}$) and the mediating variable on green industrial development ($Gd{i}_{it}$). The significance of the regression coefficients in models (2) and (3) can be tested by determining whether technological innovation has a mediating effect.

$$T{I}_{it}={\beta }_0+{\beta }_{1}di{g}_{it}+{\beta }_2{X}_{it}+{\mu }_i+{\sigma }_t+{\epsilon }_{it}$$

(2)

$$Gd{i}_{it}={\gamma }_0+{\gamma }_{1}di{g}_{it}+{\gamma }_{2}T{I}_{it}+{\gamma }_3{X}_{it}+{\mu }_i+{\sigma }_t+{\epsilon }_{it}$$

(3)

The indirect effect, besides the intermediary effect analysis, must also consider the network effect and aggregation effect of the digital economy on the green high-quality development of industry: the nonlinear dynamic effect of the digital economy and technological innovation on the green high-quality development of industry. This indirect effect can be tested by setting the threshold effect model (4) per the benchmark model (1).

$$Gd{i}_{it}={\phi }_0+{\phi }_{1}di{g}_{it}\times I(Ad{j}_{it}\le \theta )+{\phi }_{2}di{g}_{it}\times I(Ad{j}_{it}>\theta )+{\phi }_3{X}_{it}+{\mu }_i+{\epsilon }_{it}$$

(4)

Among them, the digital economy and technological innovation are threshold variables and $I(·)$ are indicator functions. If the conditions in parentheses are satisfied, $I(·)=1$. Else, $I(·)=0$.

Finally, the Spatial Durbin Model ($\mathrm{SDM}$) was constructed by adding the spatial interaction terms of both control variables to test the spatial spillover effect (1).

$$Gd{i}_{it}={\eta }_0+\rho WGd{i}_{it}+{\varphi }_{1}Wdi{g}_{it}+{\eta }_{1}di{g}_{it}+{\varphi }_{2}W{X}_{it}+{\eta }_2{X}_{it}+{\mu }_i+{\sigma }_t+{\epsilon }_{it}$$

(5)

Model (5), $WGd{i}_{it}$ represents the spatial lag term of green high-quality development of industry; ($Wdi{g}_{it}$) denotes the spatial lag term of the digital economy; ($W{X}_{it}$) stands for the spatial lag term of control variables; $W$ denotes the spatial weight matrix; $\rho$ denotes the spatial auto-regressive coefficient; The above variables ${\beta }_1$, ${\beta }_2$, ${\gamma }_1$, ${\gamma }_2$, ${\gamma }_3$, ${\phi }_1$, ${\phi }_2$, ${\phi }_3$, ${\eta }_1, {\eta }_2$ denote the influence coefficient of independent variable on dependent variable; ${\varphi }_1$ and ${\varphi }_2$ denote the spatial interaction term of the digital economy and control variables. In this study, we used the adjacency matrix and geographic distance matrix for regression.

4.2. Variables and Data

4.2.1. Variable

The system of green high-quality development of industry ($Gdi$). According to the results of the theoretical analysis, the system of green high-quality development of industry can be constructed from 3 primary indicators, 18 measurement indicators of industrial green economic benefits, ecological and environmental benefits, and social benefits (

Table 1. Measurement Results of the Index System for Green Development of Industry.

Primary Indicator	Weights	Secondary Indicator	Tertiary Indicator	Indicator Measurements	Weights	Attributes
Industry Economy Benefits	1.125	Effect of Structural Optimization and Upgrading	Rationalization of Industrial Structure	New Theil index	−0.107	−
			Advanced Industrial Structure	The proportion of output value of secondary and tertiary industries in GDP (%)	0.193	+
		Spatial Clustering Effect	Industry Aggregation	Location entropy	0.221	+
		Resource Allocation Effect	Labor Productivity	Regional production output/employment (yuan/person)	0.207	+
		Science and Technology Innovation Effect	Technology Allocation Efficiency	R&D expenditure in GDP (%)	0.206	+
				Number of patent applications (pieces)	0.168	+
		International Competitiveness Effect	International Market Share of Products	The proportion of industrial imports and exports in regional output (%)	0.112	+
Industry Ecology Environment Benefits	−0.966	Green Development	Energy Consumption	Regional energy consumption elasticity	0.240	−
				Regional electricity consumption elasticity	0.202	−
			Pollution Emission Rate	Wastewater emissions from 10,000 yuan of gross regional product	0.155	−
				Emissions of exhaust gases from 10,000 yuan of gross regional product	0.163	−
				Solid waste emissions from 10,000 yuan of gross regional product	0.192	−
		Environmental Construction	Environmental Pollution Control Intensity	The proportion of industrial pollution control investment in industrial value added (%)	0.181	+
			Environmental Pollution Control Efficiency	Comprehensive utilization rate of solid waste (%)	−0.133	+
Industry Social Benefits	0.841	Social Harmony	Employment Benefits	Urban registered unemployment rate (%)	−0.375	−
			Income Benefits	Disposable income per inhabitant	0.407	+
			Consumer Spending	Per capita consumption expenditure	0.498	+
			Social Security	Social security fund expenditures/GDP	0.470	+

Notes: GDP = gross domestic product, R&D = Research and development. Attributes + devotes that the variable is a positive indicator, Attributes—devotes that the variable is a negative indicator.

).

Industrial economic benefits can be expressed by market performance. Per the industrial organization theory, market performance is the final economic results in terms of cost, price, output, technology, and profit formed by certain market behaviors of enterprises. It is typically measured by industrial structure optimization and upgrading effect, spatial agglomeration effect, resource allocation effect, science and technology innovation effect, and international market competition effect [47]. Optimization and upgrading of industrial structure comprise rationalization of industrial infrastructure. The rationalization phenomenon reflects the degree of effective utilization of industrial resources and the degree of coordinated development between industries. Advanced industrial structure denotes the process of the continuous evolution of industrial structure toward high level from low level through technological-driven innovation. It is measured by adopting the new Thayer index of Gan et al. [48]. The calculation formula used is as follows: $NTL={\displaystyle \sum _m^n(\frac{Y_m}{Y}})Ln\left(\frac{Y_m}{L_m}/\frac{Y}{L}\right)$, where $NTL$ is smaller, that is, the degree of coupling between industrial and employment structure is better, and the industrial structure is more reasonable; $n$ denotes the number of industries; $Y_m$ denotes the industrial output value; $Y$ is $GDP$, L_m is the industrial employment, and $L$ is the total number of employment. The advanced industrial structure was estimated by the ratio of the output value of secondary and tertiary industries in GDP. The larger the value, the higher the industrial structure level. The industrial spatial agglomeration effect reflects the coordinated degree of industrial resources in regional distribution, which can be measured by locational entropy. The calculation formula is as follows: $L{Q}_i=\frac{e_i/f_i}{E_i/E}$, where $e_i$ denotes secondary employment of province $i$; $f_i$ denotes the total employment of province $i$; $E_i$ denotes national secondary employment; $E$ denotes the total national employment. The higher the location entropy index, the higher the level of high technology industry concentration in the region. The effect of science and technology innovation primarily examines the configuration effect owing to management, technological innovation, and business model innovation in the process of industrial development. It can be computed by the proportion of R&D expenditure and the number of patent applications. The impact of international competition reflects the competitive ability of the industry in the international market and is calculated by the proportion of regional industrial import and export value to a regional output value.

The ecological and environmental benefits of the industry primarily comprise the improvement of the environmental construction and ecological environment. The green development of the ecological environment is reflected by energy consumption efficiency and pollution emission efficiency. The energy consumption efficiency of the industry reflects the energy utilization. The lower the energy consumption, the higher the quality of industrial green development. It is measured by energy consumption elasticity and electricity consumption elasticity. Ecological pressure largely comes from the impact of industrial exhaust gas, wastewater, and emissions, and the efficiency of industrial pollution emissions can be measured by wastewater-, carbon gas-, and solid-waste emissions of 10,000 Yuan GDP. The industrial environmental construction efficiency was measured by both efficiency and investment intensity of environmental pollution control. The investment intensity of environmental pollution control was measured by the proportion of industrial pollution-control investment in industrial value added, and the efficiency of environmental pollution control was measured by the comprehensive utilization rate of the solid waste.

The social benefits of the corporations emphasize people-oriented facilities, fulfilling people’s high sense of achievement, access, and happiness in the shared development. Employment, income, and consumption are the most direct manifestation of people’s quality of life. The disposable per-capita income of residents, urban registered unemployment rate, consumption expenditure of residents (per-capita), and the utilization of the social security fund measure the social benefits of industrial green development.

The construction of a digital economy system. Currently, many domestic and international research institutions have published digital economy measurement methods, for example, the European Union, the International Telecommunication Union, the Digital Economy Advisory Committee of the U.S. Department of Commerce, China Center for Information Industry Development (CCID), and Tencent Research Institute [49]; however, these measurement methods have different standards, differences in measurement focus, and some data are hard to obtain. Based on the principles of science, accessibility, comparability, and comprehensiveness, this study referred to the digital economy measurement method of CCID and constructed a digital economy development system from seven secondary indicators in three dimensions: digital infrastructure level, digital application level, and digital technology innovation (

Table 2. Measurement Results of the Index System for Digital Economy.

	Primary Indicator	Weights	Secondary Indicator	Weights
$\mathrm{The}\mathrm{Level}\mathrm{of}\mathrm{Digital}\mathrm{Economy}(dig$)	Digital Infrastructure Level	0.346	Number of Internet broadband port access (million households)	0.227
			Internet penetration rate (%)	0.390
			Phone penetration rate (units per 100 people)	0.383
	Digital Technology Innovation	0.313	Share of ICT employment in total regional employment (%)	0.313
	Digital Application Level	0.341	Number of domain names (million)	0.355
			Number of websites per 100 enterprises (number)	0.261
			E-commerce sales	0.384

).

Technological innovation ($TI$). “Proportion of R&D expenditure” was selected as one of the indicators to measure the allocation effect of technological innovation in green high-quality industrial development. To evade the endogenous problem, the proportion of technology market turnover in GDP was considered a proxy variable of technological innovation that might produce an intermediary effect.

Control variables. To comprehensively examine the effects of the digital economy on high-quality, green industrial development, control variables are needed that could affect high-quality, green development of the industry. Five indicators were selected in this study as control variables, for example, economic development ($GDP$ per capita), foreign direct investment ($FDI$), urbanization level ($Urban$), local fiscal expenditure ($LFXE$), and financial development level ($Fin$). The original indicators used were: $GDP$ per capita, the proportion of regional foreign investment in regional GDP, the proportion of the nonagricultural population in total population, the proportion of total fiscal expenditure in the regional GDP at year-end, and the proportion of total institutional loans and deposits in a gross regional product at the end of the year. To decrease heteroscedasticity, we took the natural log of GDP per capita.

4.2.2. Data

A panel of data from 31 provinces and cities in China for a time period ranging from 2012 to 2020 has been selected in this study. The original data are obtained from the China Science and Technology Statistical Yearbook, China Statistical Yearbook, China Population and Employment Statistical Yearbook, China Insurance Statistical Yearbook, some provincial and municipal statistical bulletins, and iiMedia Data Center.

Table 3. Statistical Characteristics of Main Variables.

	Variable	Number of Observations	Mean Value	Standard Deviation	Maximum Value	Minimum Value
Explained variable	$Gdi$	279	0.108	0.329	1.352	−0.408
Explanatory variable	$dig$	279	0.253	0.132	0.828	0.077
Intermediary variable	$TI$	279	1.427	2.612	16.101	0.0005
Control variables	$Ln(AGDP)$	279	10.930	10.184	12.008	9.889
	$Urban$	279	0.572	0.329	0.866	0.228
	$FDI$	279	0.369	0.354	1.726	0.047
	$LFXE$	279	0.295	0.218	1.379	0.120
	$Fin$	279	3.384	1.146	7.552	1.568

shows the statistical characteristics of the major variables. The results showed that the mean value of $Gdi$ was 0.108, the maximum value was 1.352, and the minimum value was −0.408, indicating that there were significant differences among different regions. There are also obvious differences in digital economy level (dig), technological innovation (TI) and control variables.

4.3. Measuring Method of the Composite Index

We used SPSS20.0 to conduct a principal component synthesis analysis on the green development of the industry ($Gdi$) and digital economy ($dig$). The steps were as follows: (1) We used linear interpolation to interpolate the missing values of each variable and normalize them. (2) The KMO test showed that all indicators of KMO were >0.5, indicating a correlation between variables. The principal component analysis can be applied to compute the weight of each indicator. (3) According to the principle that the eigenvalue is >1, factor extraction was performed to obtain the eigenvalue of the factor, variance interpretation rate, and accumulation rate [50]. The factor score was calculated as follows: $F_i={\lambda }_{1i}{C}_1+{\lambda }_{2i}{C}_2+{\lambda }_{3i}{C}_3+\cdots +{\lambda }_{ni}{C}_n$, where $F_i$ denotes the first $i$ principal component; ${\lambda }_n$ denotes the component score coefficient, which is $C_n$ divided by the square root of its corresponding eigenvalue. (4) According to the product of the principal component weight and the component score after accumulation, the comprehensive evaluation value of each index was calculated according to the product of the principal component weight and component score. Principal component weight = variance interpretation cumulative rate. (5) First-level indicators were used to comprehensively evaluate the value (see Table 1 and Table 2, respectively), and then the principal component analysis was performed to acquire the comprehensive indexes of the green high-quality development of industry and the digital economy.

5. Empirical Test

5.1. Direct Effect Test

This study has used MATLAB to test the direct, indirect, and spatial effect.

Table 4. Results of Baseline Regression.

Variable	(1)	(2)
$dig$	2.192 *** (0.077)	0.648 *** (0.046)
$LnAGDP$		0.344 *** (0.057)
$Urban$		−0.003 * (0.001)
$FDI$		0.079 * (0.044)
$LFXE$		0.014 * (0.018)
$Fin$		0.185 ** (0.042)
constant	−0.447 *** (0.022)	−3.934 *** (0.557)
Fixed Province	NO	YES
Fixed time	NO	YES
N	279	279
$R^2$	0.245	0.311

Note: ***, **, * in the table are passing the significance test at 1%, 5%, 10% confidence level, respectively, and the same in the table below.

demonstrates the results of the benchmark regression of the impact of the digital economy on the green development of industry: the direct effect test. The regression results (1) and (2) revealed that the computed coefficients of the digital economy were significant, and both were positive. When there was a 1% increase in the development of the digital economy, the level of green high-quality development of industry increased by 0.648%. This suggests that the green development of the industry is promoted by the digital economy. The results (2) of adding control variables showed that the regression coefficients of local fiscal expenditure, economic development level, foreign investment, and financial development level were positive and statistically significant. This indicates that the introduction of foreign investment, improvement of the economic development at the regional level, government’s macro-control, and the financial market provides strong support for high-quality, green high-quality development of industry. The regression coefficient $Urban$ was statistically significant at –0.003, suggesting that the promotion of urbanization exerted a small inhibiting effect on green high-quality development of industry, which could be because some towns blindly pursued urbanization rate in the early stage of new urbanization, resulting in the disconnection between industries and cities and hollowing out of industries, thereby hindering green high-quality development of industry.

Table 5. Mediation Effect Test.

Variable	$\mathit{G}\mathit{d}\mathit{i}\left(1\right)$	$\mathit{T}\mathit{I}\left(2\right)$	$\mathit{G}\mathit{d}\mathit{i}\left(3\right)$
$dig$	0.648 *** (0.046)	7.257 *** (1.357)	0.353 * (0.256)
$TI$			0.041 *** (0.012)
Control variables	YES	YES	YES
Fixed province	YES	YES	YES
Fixed time	YES	YES	YES
Number of provinces	31	31	31
N	279	279	279
Number of periods	9	9	9
$R^2$	0.319	0.330	0.241

Note: ***, * in the table are passing the significance test at 1%, 10% confidence level, respectively, and the same in the table below.

populates the results of mediating effects of the digital economy affecting the green high-quality development of industry. Model (1) is the result in Table 4, which verifies that there is a positive effect of the digital economy on the green development of the industry. Model (2) validates that the digital economy can enhance the mediating variable—technological innovation. Model (3) demonstrated that the estimated coefficient of the dependent variable decreases from 0.648 to 0.353 after adding the mediating variable, indicating that a mediating role is played by the enhancement of technological innovation in the digital economy’s influence on the green high-quality development of industry. Hence, Hypothesis 1 is verified.

5.2. Indirect Effect Test

Based on the cumulative effect and network effect of the digital economy, this research study has used the threshold model to determine whether the digital economy exerts an indirect effect on the green high-quality development of industry. First, we used the bootstrap method to report whether there is a threshold effect. By repeating the sampling 500 times, the results showed that the digital economy significantly passed the single threshold, and the technological innovation passed the double threshold effect test. Then, the threshold regression coefficients were estimated according to model (4).

Table 6. Results of Threshold Regression.

Variables	Threshold Variables
	$\mathit{d}\mathit{i}\mathit{g}$	$\mathit{T}\mathit{I}$
$\mathrm{Threshold}\mathrm{value}{\pi }_1$	0.4387	−0.1295
$\mathrm{Threshold}\mathrm{value}{\pi }_2$		−0.0023
$dig\times I(Adj\le {\pi }_1)$	0.212 *** (0.249)	−0.331 (0.205)
$dig\times I({\pi }_1<Adj\le {\pi }_2)$	0.593 ** (0.240)	0.252 * (0.201)
$dig\times I(Adj>{\pi }_2)$		0.773 *** (0.190)
Control variables	YES	YES
Fixed province	YES	YES
N	279	279
$R^2$	0.235	0.273

Note: ***, **, * in the table are passing the significance test at 1%, 5%, 10% confidence level, respectively, and the same in the table below.

illustrates that the threshold value of the digital economy is 0.4387, and there is a significantly positive impact of the digital economy on the green development of industry in the threshold interval, and its positive influence increases when the digital economy crosses the threshold value. This finding indicates that at the early stage of the digital economy, digitalization of traditional and green industries has a long construction period, high cost and large scale, lacks senior talent in the digital economy, the system of institutional construction and supervision is imperfect, the energy consumption of infrastructure is high on the development of digital economy, and the upgrading of traditional industries depends on it. With the deepening integration of the digital economy and the real economy, relevant green regulations and preferential policies ensure a stable market and sustainable development environment, prompting fundamental changes in green production mode and residents’ green lifestyle. Thus, it brings strong development momentum for the green development of the industry. When technological innovation crosses the two thresholds, the impact on green high-quality development of industry demonstrates nonlinear incremental characteristics. Furthermore, a benign interaction is formed between technological innovation, digital economy, and green high-quality development of industry. Hence, Hypothesis 2 is supported.

5.3. Spatial Spillover Effect

The spatial spillover effects of the impact of the digital economy on the green high-quality development of industry are illustrated through the use of a spatial econometric model. First, the two indicators were tested for spatial autocorrelation using the Moran index method (Moran’s I).

Table 7. Spatial Global Autocorrelation Characteristics.

Year	$\mathit{d}\mathit{i}\mathit{g}$		$\mathit{G}\mathit{d}\mathit{i}$
	Moran’s I	Z-Value	Moran’s I	Z-Value
2012	0.206 ***	2.483	0.255 ***	2.859
2013	0.133 ***	1.756	0.358 ***	3.992
2014	0.100 ***	1.422	0.378 ***	4.182
2015	0.096 ***	1.371	0.339 ***	3.821
2016	0.098 ***	1.386	0.363 ***	4.035
2017	0.084 ***	1.245	0.360 ***	3.982
2018	0.074 ***	1.135	0.308 ***	3.445
2019	0.077 ***	1.158	0.310 ***	3.523

Note: *** in the table is passing the significance test at 1% confidence level, and the same in the table below.

presents the results, showing that Moran’s indexes of both green developments of the industry and digital economy during 2012–2019 were statistically significant at the 1% level. This indicates that the digital economy and green high-quality development of industry of 31 provinces and cities in China have spatial autocorrelation and spatial agglomeration state during this period. Then, we used SDM for the regression test (

Table 8. Test Results of Spatial Effects of Digitalization on Green Development of Industry.

Model	SAR Model		SDM Model
Weight Matrix	Geographical Distance	Adjacency	Geographical Distance	Adjacency
Variables	(1)	(2)	(3)	(4)
$\rho$	0.529 *** (0.070)	0.452 (0.064)	0.635 *** (0.067)	0.420 *** (0.072)
$dig$	0.425 * (0.221)	0.225 (0.224)	0.481 ** (0.236)	0.414 * (0.247)
$W\times dig$			0.056 (0.109)	0.082 * (0.025)
Direct effect	0.453 * (0.240)	0.234 * (0.227)	0.406 * (0.034)	0.256 * (0.135)
Indirect effect	0.450 * (0.255)	0.163 * (0.168)	0.563 * (0.104)	0.185 * (0.127)
Control variables	YES	YES	YES	YES
Log-L	315.380	311.808	303.549	300.663
$R^2$	0.160	0.159	0.106	0.126

Note: ***, **, * in the table are passing the significance test at 1%, 5%, 10% confidence level, respectively, and the same in the table below.

), which confirmed that the spatial autoregressive coefficients in the Spatial lag model (SAR) and Spatial Dubin Model (SDM) were positive at a 1% level of significance, and the spatial interaction coefficients of the digital economy were positive at 10% level of significance. This indicated an exogenous spatial interaction effect of the digital economy and an endogenous interaction effect of green high-quality development of industry in the sample. Finally, we used spatial partial differentiation to decompose the spatial effects into indirect and direct effects. The direct effect of the digital economy was 0.406, indicating that every 1% increase in the digital economy caused a 0.406% increase in the green development of industry in the region. In addition, the indirect effect was 0.563, suggesting that there was a spatial spillover effect of the digital economy in the zone. Its economic implications were that the digital economy in the region could directly influence the green industry of the neighboring region, or it could affect the green industry of the neighboring region by influencing the digital economy of the neighboring region. Hence, Hypothesis 3 is validated.

5.4. Analysis of Regional Heterogeneity

The spatial distribution of the digital economy and green high-quality development of industry presented obvious heterogeneity owing to the large difference in resource endowment among provinces and cities. Thus, it is of certain reference importance to analyze the influence of the digital economy on the green development of the industry from regional differences. In this study, the region was divided into three zones, namely the west, east, and central region.

Table 9. Heterogeneity Results.

Variables	Eastern Region	Middle Region	Western Region
$dig$	1.315 * (0.662)	0.277 * (0.297)	0.007 (0.504)
constant	−1.473 (1.353)	−2.916 *** (0.784)	0.333 (1.142)
Province fixed	YES	YES	YES
Time fixed	YES	YES	YES
N	72	99	108
$R^2$	0.209	0.262	0.132

Note: ***, * in the table are passing the significance test at 1%, 10% confidence level, respectively, and the same in the table below.

shows that both the east and central regions exerted a significant influence of the digital economy on the green high-quality development of industry, while the western region did not demonstrate a significant impact. The advantage of the eastern region was significantly ahead of the central region, which could be because the eastern zone has relatively high-technology talents, sufficient capital, a good market development environment, and a relatively sufficient release of the driving force of the digital economy. The western zone is rich in energy resources, which guarantees China’s energy supply. However, the proportion of high pollution, high energy consumption, and high emission industrial units is larger, which hinders green industrial development. Although the digital economy is developing rapidly in some regions in western China, for example, Chengdu–Chongqing region has become one of China’s five digital economy clusters with the development advantages of core industries, such as integrated circuits and intelligent connected vehicles, but the promotion of green high-quality development of industry is not obvious. Under the goal of carbon peak and carbon neutrality, enterprises should take the road of digitalization and green integration to jointly promote the green high-quality development of industry.

5.5. Robustness Tests

Considering that the results of the above-mentioned empirical tests might be affected by the selection of indicators and other factors, we adopted two methods to conduct robustness tests. (i) Setting the province fixed effect. Provinces with a higher economic development have a first-mover advantage in the digital economy, which might affect the change in the macro-environment. In this regard, this paper sets the province fixed effect and the interaction effect of province and year to mitigate the changes in macro environment.

Table 10. Robustness Test Results.

Variables	Excluding Macro Environment Impact		Changing the Explanatory Variable
	(1)	(2)	(3)	(4)
$dig$	2.046 *** (0.035)	0.657 *** (1.357)	0.356 * (0.070)	0.524 (0.094)
Control variables	YES	YES	NO	YES
Province fixed	YES	YES	YES	YES
Time fixed	YES	YES	YES	NO
Province × Time	NO	YES	NO	NO
N	279	279	279	279
Period number	9	9	9	9
$R^2$	0.424	0.283	0.306	0.341

Note: ***, * in the table are passing the significance test at 1%, 10% confidence level, respectively, and the same in the table below.

(Columns 1 and 2) indicates that the empirical test findings mentioned above remain robust considering macro-environmental factors. (ii) The replacement of explanatory variables. The digital economy is a crucial carrier of modern computer information networks. This study replaced the digital economy with the output value of the computer electronic-communication-equipment manufacturing industry. Columns (3) and (4) in Table 10 indicate that the impact of the digital economy on the green high-quality development of industry is still valid after considering the endogenous problem.

6. Conclusions and Policy Implications

6.1. Conclusions

Based on the spatial panel data of 31 provinces and cities of China for the period ranging from 2012 to 2020, an evaluation system of the digital economy and green high-quality development of industry was formulated. Fixed effects, intermediary effects, threshold effects, and spatial econometric models were used to empirically test both direct and indirect impacts of the digital economy on the green development of industry and the course of action adopted from multiple perspectives. The main findings of this research study can be summed up as: (1) The digital economy generates an overall positive influence on the green development of industry and is a major driving force for green high-quality development of industry in China. A focus on regional differences shows the significant measurable effect of the digital economy in the eastern zone in contrast to the development in the central and western zones of China. (2) The indirect effect of the digital economy on the green development of industry has nonlinear dynamic characteristics. The digital economy crosses the threshold value, and the effect force is enhanced. Technological innovation plays a mediating role in determining the effect of the digital economy on the green development of the industry. (3) The digital economy expands a spatial spillover effect that promotes green high-quality development of industry within the region and among regions.

6.2. Policy Implications

It can be seen from the research conclusion that there is a need to pay closer attention to the application of digital technology innovation in the green transformation of industry and develop the differentiation strategy, and parallel to this, coordinate the development of the digital economy.

First, promoting industrial green transformation and upgrading with digital technology innovation. (1) we will develop green and clean technologies and resource recycling technologies for traditional industries through digital technologies and the application of innovative operation processes for traditional industries such as R&D, manufacturing, marketing, logistics, and services. The remanufacturing and reusing of waste products will improve the efficiency of resource utilization at the source, process, and end, and reduce the intensity of the carbon emissions of environmental pollution. (2) The integration of the digital economy and environmental and ecological protection industries, and the promotion of the use of green technologies, green design, and smart use of new infrastructure will reduce energy consumption and improve resource utilization. The building of green data centers and green information networks, developing green storage and building a digital industrial system with green recycling and low-carbon development. (3) The use of digital technology to raise the level of efficiency of carbon accounting of investment and financing projects and enterprises by financial institutions through digital technology, establishing the carbon footprint disclosure system of financial institutions, to provide preferential investment and financing conditions for green and low-carbon enterprises and projects and force enterprises to green transformation will be useful in a forward-looking envisioning of green development.

Second, gradually narrowing the regional digital divide and the gap in green industrial development with differentiated strategies. (1) Based on the local advantageous industries and development high points, each region implements regional differentiation strategies. Each region creates digital economy industries with regional characteristics, creates digital economy industry clusters with strong radiation, and invests additional resources in the digital integration development of industries. (2) Focusing on the regional iconic green industry chains, the central and western regions cultivate and support leading enterprises to build complete chains of the industrial Internet platform. Actively implementing digitalization and intelligent technology transformation and enhancing the drive of green industries accelerates the green development of the industry.

Finally, the regional digital economy might develop in collaboration with contiguous units to expand the benefits of the spatial spillover effect on high-quality, green industrial development. (1) This will strengthen the construction of digital infrastructure and data centers and platforms that connect regional digital economy industries. The building of national hubs for regional and national integrated big data centers. (2) Establishing and promoting green and smart parks based on digital technology to serve clustered enterprises and industries. Collecting and supervising the environmental pollution emission data of enterprises and parks by using digital technologies, such as industrial Internet, big data, adjusting the energy demand, product production, and service mode of enterprises following the green and low-carbon objectives. This can not only promote the green development of the industry but also produce spatial spillover effects on relevant enterprises and industries inside and outside the region.