The Impact of Renewable Energy Technology Innovation on Industrial Green Transformation and Upgrading: Beggar Thy Neighbor or Benefiting Thy Neighbor

Abstract

Renewable energy technology innovation (RETI) is a crucial way to improve energy poverty and combat climate change. However, few studies have examined the impact of RETI on industrial green transformation and upgrading (IGTU) from the perspective of spatial spillover and its regional boundary. Based on the theory of green growth and sustainable development, this paper expands the connotation boundary of IGTU and measures the IGTU levels of 30 provinces in China from 2006 to 2020 using an improved entropy weight method. Kernel density estimation and Moran’s I index are adopted to portray temporal and spatial patterns, the spatial Durbin model is employed to examine the influencing mechanism and spatial spillover effects of RETI on IGTU and its regional boundaries, and the differential impact of its spatial effects on time, region, resource endowment, and environmental regulation are explored further. The results show that (1) RETI and IGTU in China are steadily increasing, indicating a decreasing spatial differentiation pattern of “east–west”; (2) RETI significantly promotes local IGTU but inhibits neighboring IGTU, forming a “beggar-thy-neighbor” situation; (3) the spatial spillover effect of RETI on IGTU has significant regional boundaries—the spatial spillover effect significantly negative and persists up to 800 km, but it is significantly positive from 800 to 1400 km and shows a trend of increasing and then decreasing; and (4) the promoting effect of RETI on IGTU gradually increases over time, presenting spatial differences of promotion in the east and inhibition in the west. Furthermore, RETI has a more substantial promoting effect on IGTU in non-resource-based regions and strong environmental regulation regions. The findings for China provide concrete evidence for formulating targeted policies and seeking a path for IGTU for other developing countries.

1. Introduction

Nowadays, in order to address the dual pressures of increasingly severe climate change and slowing economic growth in the post-pandemic era, green transformation has become a new growth driver for governments to promote economic development [1,2]. As the growth engine of the national economy, the industrial transformation and upgrading mode increasingly influences the direction of national development [3,4]. Due to the rapid growth of traditional industrial civilization in the past few decades, fossil fuel consumption has aggravated environmental problems all over the world [5,6], and the extensive model of high energy consumption and high pollution has led to irreversible ecological environmental pollution and global warming [7]. The United Nations points out that if the global temperature exceeds 1.5 °C, 14% of species will face a high risk of extinction [8]. Phenomena such as flood disasters, crop harvest decline, water depletion, and wetland rainforest decline caused by increasing industrial activities will increase significantly in the next few decades, and the spread of disease will further accelerate. Faced with the resource and environmental crisis caused by traditional industrial civilization, all countries are actively exploring new paths for IGTU [9,10].

In recent years, major economies in the world have taken energy technology innovation as an essential breakthrough to promote the IGTU and actively formulated various policies and measures to seize the commanding heights of energy development [11,12]. For example, the United States has established “science and energy” as the first strategic theme, and it has recently released policies such as the Comprehensive Energy Strategy and the America First Energy Plan and deployed advanced energy technologies such as a new generation of nuclear energy, shale oil and gas, energy storage, and smart grids. The EU has taken the lead in putting forward the strategic goal of building a carbon-neutral economy in the European Green Agreement; launched the research, technology development, and demonstration framework program; and built a full-chain energy technology innovation ecosystem. Japan has issued strategic plans such as the Fifth Energy Basic Plan, the 2050 Energy and Environment Technology Innovation Strategy, and the Hydrogen Energy Basic Strategy, proposing to accelerate the development of renewable energy and comprehensively build a hydrogen-energy society. Apparently, energy technology innovation has attracted widespread attention all over the world. With the new round of technological revolution and industrial change developing deeply, RETI is gradually becoming a competitive focus to reshape the global energy landscape [13]. RETI is not only a key driver of renewable energy development, but also an engine to promote energy transformation and sustainable industrial development [7,14]. On the one hand, RETI significantly lowers the cost of energy production and consumption, promotes the development of new energy industries [15], improves the utilization of renewable energy, and builds a safe, autonomous, and low-carbon energy system [16]. On the other hand, RETI affects energy and environmental conditions more directly than traditional energy technologies [7], which has an essential impact on environmental governance. RETI is thus seen as a breakthrough to lead change and a solution to reduce energy–environment–economy conflicts [11,12].

The International Energy Agency (2021) pointed out that China will be the primary driver of installed renewable energy capacity in the coming years, followed by Europe, the United States, and India. In 2021, China’s renewable energy utilization ranked first in the world, and the installed capacity of renewable energy accounted for 31.9% globally. As the world’s largest producer and consumer of renewable energy, without exception, China is actively developing renewable energy technologies to cope with the extensive pollution caused by China’s rapid industrial development for more than 40 years [17,18]. In recent years, China’s total emissions of industrial waste gas and industrial wastewater have continued to increase. From 2000 to 2020, industrial waste gas emissions increased from 14 trillion m³ to 112 trillion m³, and industrial wastewater emissions increased from 40 billion to 92 billion metric tons, respectively [19]. The rapid industrial development has brought enormous pressure on environmental governance [20,21,22], and IGTU is thus imminent. In order to change the method of energy utilization in industrial production and manufacturing and increase the consumption of renewable energy, the Chinese government has actively taken a series of measures to promote renewable energy technological innovation. On the one hand, under the goal of a carbon peak and carbon neutrality, the Chinese government has actively introduced industrial policies to guide cleaner industrial production. For example, the 14th Five-Year Plan for Industrial Green Development and the Guiding Opinions on Accelerating the Establishment of a Sound Economic System for Green, Low-Carbon and Circular Development both emphasize the promotion of green and sustainable industrial development by optimizing and adjusting the industrial structure, improving resource-use efficiency, and strengthening industrial energy conservation and carbon and pollution reduction [23], thus providing an appropriate environment for developing RETI. On the other hand, fossil energy is still the primary source of energy consumption in China, of which coal consumption accounts for 56% and clean energy consumption accounts for only 25.5% [24]. To optimize the energy consumption structure and increase clean energy use, China actively promotes energy technology innovation [13]. The 14th Five-Year Plan for Scientific and Technological Innovation in the Energy Sector emphasizes that energy technology innovation should guide the industry to change the traditional energy consumption mode, accelerate the scale development of renewable energy, and promote IGTU. Renewable energy development has become an essential way for China to reduce the proportion of fossil energy consumption and promote IGTU [25].

Theoretically, the higher the level of RETI, the greater the opportunity for IGTU [7], but few studies have empirically verified this point. Moreover, as the new dual-cycle development pattern in China is constructed, industrial cooperation and exchange and the flow of production factors are increasing, resulting in knowledge spillover, which further accelerates industrial transfer and transformation and upgrading [26,27]. However, exploring the effects from the perspective of spatial spillover has been neglected. To fill this research gap, the following questions are explored in this paper: First, how can the connotation of IGTU be defined scientifically, and furthermore, how can a complete and reasonable index system be built to measure regional IGTU effectively? Second, is there a spatial spillover effect of the effect of RETI on IGTU? Furthermore, does it beggar thy neighbor or benefit thy neighbor? Third, is there a regional boundary of the spatial spillover of RETI on IGTU? Research on the above issues could promote high-quality industrial development and also provide experiences and insights to promote the global energy transition and climate governance.

Therefore, based on the theory of green growth and sustainable development and the data of 30 provinces in China from 2006 to 2020, this paper first defines the theoretical connotation of IGTU and employs the improved entropy weight method to measure it. Secondly, kernel density estimation and the Moran’s I index are adopted to portray the temporal and spatial differential pattern. Finally, the spatial panel Durbin model is adopted to investigate the influencing mechanism and the spatial spillover effect of RETI on IGTU and its regional boundary, and further examine the differential spatial effect under the constraints of time, region, resource endowment, and environmental regulation. Different from the previous studies, the contributions of this paper are focused on three aspects: (1) Based on the theory of green growth and sustainable development, the theoretical logic and conceptual boundary of IGTU are systematically combed and expanded from five dimensions of structural optimization, efficiency improvement, quality enhancement, environmental friendliness, and resource sustainability, and the comprehensive evaluation index system of IGTU is further constructed scientifically and rationally, which provides a research paradigm for effective measurement of IGTU. (2) The influence mechanism of RETI on IGTU is explored from the perspective of spatial spillover, which helps to fill the relevant research gaps and provides a new perspective for energy technology path selection in IGTU. (3) Using the threshold distance spatial weight matrix, the spatial spillover and its regional boundaries of RETI on IGTU are further examined, which provides empirical evidence for promoting IGTU within the effective range.

The rest of the paper is as follows: Section 2 reviews the relevant literature, Section 3 details the methodology and data, Section 4 reports and discusses the empirical results, and Section 5 presents the conclusions, discussion, and recommendations.

2. Literature Review

2.1. Measurement of IGTU

As we all know, the IGTU have been attracting academic attention, but there is no consistent conclusion on its concept definition and measurement. Existing studies on IGTU are mainly based on structural and efficiency perspectives [28,29,30,31]. From a structural standpoint, industrial transformation and upgrading is the optimization and adjustment of industrial structures, which can be improved by promoting rationalization and upgrading of industrial structures [7,32]. Specifically, industrial structure upgrading is a process in which the factor endowment of a country or region is transferred from the industrial sector with low production efficiency to one with high productivity, which can be measured by the proportion of the output value of the tertiary industry in GDP [18]. Some scholars also regard the ratio of the added value of the tertiary sector to that of the secondary sector as a proxy variable for industrial structure upgrading. Meanwhile, they believe that the tertiary industry has relatively clean and green production characteristics because of its lesser environmental pollution and resource consumption than the secondary industry [33,34,35].

Some scholars who agree with the efficiency perspective believe industrial transformation and upgrading reflect the transformation of industrial growth, which is the improvement and enhancement of efficiency [28,29]. Industrial green transformation focuses on industrial pollution control, emphasizing the realization of green and low-carbon industrial development [29]. Studies by scholars [28,36] believe that industrial green transformation requires considering energy and environmental factors and including energy inputs, pollutant emissions, and carbon emissions in the measurement framework to calculate total green factor productivity and adopt it as a proxy indicator. Gong et al. [37] argue that the improvement of desirable output and production technology and the reduction of undesired output should be comprehensively considered in manufacturing green transformation and upgrading, and total green factor productivity is the appropriate indicator to reflect green transformation and upgrading in the manufacturing industries. In addition, some scholars take the industrial industry as an example to construct a comprehensive evaluation index system to measure industrial green transformation based on energy and resource-intensive utilization, pollution reduction, industrial structure upgrading, productivity improvement, and sustainable development [31,38].

However, with the continuous integration of the technological revolution and industrial revolution, the new round of technology is increasingly leading to industrial changes, and IGTU is therefore endowed with new characteristics and connotations. IGTU not only includes traditional industrial structure adjustment and efficiency improvement, but also presents the new characteristics of digital, service-oriented, and sustainable development under the constraints of resources and environment in the intersection era of new technologies and energy revolutions [39]. Therefore, it is necessary to further define the connotation and measurement standards of IGTU in the new development context.

2.2. RETI and IGTU

Technological innovation is the basis of industrial transformation [40]. At present, research on RETI and IGTU mainly focuses on the influencing factors of RETI, and the impact of traditional technological innovation on industrial transformation or industrial structure upgrading. In terms of the factors that influence RETI, the relevant studies mainly focus on environmental supervision, R&D investment, financial development, power consumption, R&D intensity, and installed capacity of renewable energy [41,42,43,44,45]. In addition, some scholars have also explored the impact of RETI on the economy, society, and environment, such as technology diffusion, carbon emissions, economic gains, inclusive growth, and industry dynamics [46,47,48,49,50,51,52,53,54]. Existing studies on the exploration between RETI and IGTU mainly include three viewpoints. The first viewpoints hold that technological innovation has a driving effect on industrial upgrading. Endogenous growth theory holds that technological progress is an important reason for the sustained growth of enterprises [55,56,57,58]. Most studies support the view that technological innovation is essential in improving productivity [59,60], and strengthening technological innovation may accelerate industrial structural change [61,62]. Technological innovation mainly promotes industrial structure upgrading by improving production efficiency and reducing production costs [63,64]. The promoting effect has long-term stability, with the synergy between innovation and industry increasing over time [65,66]. With the increasing constraints of resources and the environment, technological innovation tends to develop in a green way [35]. Some scholars have studied the impact of green technology innovation. They found that advanced green innovation technologies improve enterprise productivity and strengthen competitiveness [67,68,69,70,71]. Energy-efficient technologies help to create a more sustainable industrial structure [72], and environmentally friendly technologies contribute to achieving environmental sustainability [73,74]. Meanwhile, with the transformation of China’s economic development model from investment-driven to innovation-driven, China would rely more on independent innovation to promote green technology innovation and thus promote industrial structure upgrading [75].

Contrary to the innovation-driven effect, few studies believe there is a negative impact between technological innovation and industrial development. With the continuous iteration and updating of technology, the improvement of technological innovation leads to innovative products endowed with more sophisticated technology. Compared with traditional technologies, green technology innovation would increase the operating cost and reduce initiatives to upgrade the industrial structure and green transformation [23].

Different from the above two viewpoints, the third one holds that there is an uncertain relationship between technological innovation and industrial transformation and upgrading. Humphrey and Schmitz [40] found that technological innovation indirectly promotes industrial transfer through the complex value chain and thus improves the core competitiveness of enterprises. Some studies have found that technological progress can reduce carbon intensity by optimizing industrial structures [76,77], which helps promote low-carbon and green industrial transformation. The impact of technological innovation on industrial upgrading has been strengthening over time, but the direction of the effect appears to be alternating positive and negative [27]. Such nonlinear influence is due to the unreasonable structure of innovation input, which may hinder industrial structure upgrading [78,79]. Other studies believe that green technology innovation also has a nonlinear impact. Genc and De Giovanni [80] consider that green technological innovation may have a possible compliance cost effect on green production and an innovation offsetting effect. Therefore, the link between green technology innovation and IGTU may be uncertain. When production costs increase in green technology innovation, manufacturing firms give up on IGTU due to lower profits; when green technology innovation can reduce production costs, manufacturing firms have more disposable profits to enhance innovation, which in turn promotes IGTU [23]. Bi et al. [81] further explored the impacts under both scenarios and found that technological innovations may increase or decrease production costs. Moreover, green energy technologies also affect economic globalization, reducing environmental pollution [82,83] and indirectly increasing the uncertainty of IGTU.

In addition, some of the previous related research studies have confirmed that with the development of information network technology and spatial geography, technological innovation has overcome geographical space constraints and achieved cross-regional innovation division and association. The impact of technological innovation on industrial structure upgrading is not limited to the local area, but also affects the industrial structure transformation in the surrounding areas through the spillover effect, indicating that the diffusion of technological innovation also promotes the transformation and upgrading of industrial structures [27,84]. As an essential component of technological innovation, RETI has received increasing academic attention. However, relevant studies have focused on the analysis of the impact of renewable energy consumption on technological innovation and the environment [85,86,87,88], and few studies have directly investigated the relationship between RETI and IGTU. RETI is considered an essential clean technology to reduce carbon emissions and pollution, which has the dual characteristics of cleanliness and technology diffusion [89,90,91]. Since accelerating industrial green development requires increasing renewable energy utilization, RETI has thus become a key force in leading industrial green transformation [7,92]. Strengthening RETI could change energy consumption patterns, reduce the utilization of traditional fossil energy by enterprises, and increase the proportion of clean energy utilization [11], contributing to reducing pollution emissions. Furthermore, enhancing RETI could promote efficiency improvement and increase enterprise productivity, which provides the foundation and support for industrial transformation and upgrading.

In summary, although previous studies have richly explored the impact of technological innovation on industrial structure upgrading or industrial green transformation, few studies have empirically examined the relationship between RETI and IGTU. Still, the following is still lacking in the existing research. First, with the promotion of emerging technologies and the increasing constraint of resources and environment, the connotation of IGTU has changed, but there is a lack of scientific definition. Second, the impact of traditional technological innovation on industrial structural transformation and upgrading has been widely explored. However, with the depth of the new round of the energy revolution and the advancement of carbon targets, RETI has gradually become an essential means to promote energy change and industrial green transformation [11]. However, the impact of RETI on IGTU is ignored, and the influencing mechanism is rarely discussed systematically. Third, the effect of technological innovation is restricted by geographical distance and administrative boundaries, and its spatial spillover effect is not continuous [93], but most studies lack this consideration. Therefore, when exploring the relationship between RETI and IGTU, the spatial effect and distance attenuation characteristics cannot be ignored. This study contributes to exploring the impact of RETI on IGTU from the perspective of spatial spillover and its attenuation boundaries, which provides policy implications for the government to formulate specific and differentiated energy technology innovation policies and promote energy transformation and industrial green and low-carbon development.

3. Methodology and Data

3.1. Methodology

3.1.1. Construction and Evaluation of the IGTU

A scientific and objective evaluation of IGTU can enhance IGTU under the goal of a carbon peak and carbon neutrality, which has important theoretical and practical significance for energy transformation and green and low-carbon development. Existing studies have not formed a unified definition of the theoretical connotation of IGTU. According to the relevant studies on industrial green transformation and industrial structure upgrading [30,31] and the current background of China’s economic transformation and development, this paper combines the realistic basis of the Guidance on Accelerating the Establishment of a Sound Green Low-Carbon Circular Development Economic System and the 14th Five-year Industrial Green Development Plan and refers to the study by Deng and Yang [38] to redefine IGTU. IGTU takes industrial structure adjustment as the core, with resource-saving, pollution reduction, and emission reduction as the critical direction, and finally realizes industrial structure optimization, efficiency improvement, quality enhancement, environmental friendliness, and resource sustainability by forming an organic combination of high-end industrial structure, low-carbon energy consumption, resource utilization recycling, clean production processes, green product supply, and production digitization.

Based on the theory of green growth and sustainable development, this paper further defines the theoretical connotation of IGTU and constructs an analytical framework from five dimensions: structural optimization, efficiency improvement, quality improvement, environmental friendliness, and resource sustainability. Simultaneously, fully considering the scientificity, comprehensiveness, representativeness, and availability of the index system construction, 39 indicators are finally selected to characterize IGTU comprehensively (

Table 1. Indicator system of IGTU in China.

First-Level Indicators	Second-Level Indicators	Third-Level Indicators	Unit	Attribute
Structural optimization	Rationalization	Thiel’s index	/	-
		Dual comparative coefficient	/	+
	Advancedization	Industrial upgrading index	/	+
		The proportion of the third industry to the second industry	%	+
		The proportion of the primary business income of high-tech industry in GDP	%	+
	Cleanliness	The proportion of the output value of the clean industry in industrial output value	%	+
		The proportion of the output value of high-energy-consuming industries in industrial output value	%	-
Efficiency improvement	Efficient	Capital productivity	%	+
		Labor productivity	CNY/person	+
		Total factor productivity	/	+
	Intensification	Labor-intensive industry agglomeration	/	-
		Capital-intensive industry agglomeration	/	+
		Technology-intensive industry agglomeration	/	+
Quality improvement	Digitalization	Computer communication manufacturing industry revenue as a proportion of industrial primary business income	%	+
		The proportion of computer communication manufacturing personnel in total employment	%	+
		Number of internet broadband access ports per capita	/person	+
		Length of fiber optic cable		+
		Digital financial inclusion index	/	+
	Servitization	Software industry service industry and industrial development coupling index	/	+
		The proportion of productive service industry employment in total employment	%	+
		Service sector overrun coefficient	/	+
		The proportion of the number of people in the service sector to that in the industrial sector	%	+
	High-endization	The competitive advantage of high technology products trade	/	+
		The complexity of export technology	/	+
		The proportion of new product sales revenue in industrial primary business revenue	%	+
Environmentally friendly	Energy-saving and carbon reduction	Energy consumption intensity	Tons/CNY	-
		Carbon emission intensity	Tons/CNY	-
	Pollution emission reduction	Industrial SO2 emission intensity	Tons/CNY	-
		COD emission intensity	Tons/CNY	-
		Industrial solid waste emission intensity	Tons/CNY	-
		Industrial smoke (dust) emissions intensity	Tons/CNY	-
Resource sustainability	Resource conservation	Industrial water resources utilization per industrial value added	m3/CNY	-
		Agricultural water resources utilization per agricultural output value	m3/CNY	-
		Industrial construction land scale per industrial value added	Km2/104 CNY	-
		Electricity consumption per GDP	kwh/CNY	-
	Recycling	The comprehensive utilization rate of industrial solid waste	%	+
		Harmless treatment rate of domestic waste	%	+
		Centralized treatment rate of sewage treatment plants	%	+
		The industrial SO2 removal rate	%	+

).

3.1.2. The Improved Entropy Weight Method

As an objective weighting method, the entropy weight method uses the attributes of each index to calculate the weight, which avoids the bias caused by human factors. According to the study by Dong et al. [94], the improved entropy method is used to calculate the IGTU index. If there are θ years, m provinces, and n indexes, then X_tij denotes index j of province i in year t. The specific calculation steps are as follows.

Firstly, the indicators are standardized (Equations (1) and (2)):

$$X_{tij}^{\prime }=\frac{X_{tij}-\min \left(X_{tj}\right)}{\max \left(X_{tj}\right)-\min \left(X_{tj}\right)}$$

(1)

$$X_{tij}^{\prime }=\frac{\max \left(X_{tj}\right)-X_{tij}}{\max \left(X_{tj}\right)-\min \left(X_{tj}\right)}$$

(2)

Secondly, the indicators are homogenized and the weights are calculated (Equation (3)):

$$p_{tij}=X_{tij}^{\prime }/{\displaystyle \sum _{t=1}^{\theta }{\displaystyle \sum _{i=1}^m{X}_{tij}^{\prime }}}$$

(3)

Thirdly, the information entropy is calculated (Equation (4)):

$$e_j=-\frac{1}{\ln \left(m\right)}{\displaystyle \sum }_{i=1}^m{p}_{tij}\ln (p_{tij})$$

(4)

Fourthly, the coefficient of variation is calculated (Equation (5)):

$$a_j=1-e_j$$

(5)

Fifthly, the weights of each indicator are calculated (Equation (6)):

$$W_j=a_j/{\displaystyle \sum }_{j=1}^n{a}_j$$

(6)

Finally, the composite score is calculated (Equation (7)):

$${\mathrm{S}}_{ti}={\displaystyle \sum }_{j=1}^n{\mathrm{W}}_j\times X_{tij}^{\prime }$$

(7)

3.1.3. Kernel Density Estimation

Kernel density estimation is a nonparametric method to estimate the probability density of a random variable, which examines the characteristics of the variable such as distribution, morphology, ductility, and polarization phenomena. The calculations are shown in Equations (8) and (9).

$$f\left(x\right)=\frac{1}{Nh}{\displaystyle \sum }_{i=1}^{N}K\ln \left(\frac{X_i-x}{h}\right)$$

(8)

$$K\left(x\right)=\frac{1}{\sqrt{2\pi }}\exp \left(-\frac{x^2}{2}\right)$$

(9)

where K(·) is the kernel function, X_i is an independent and identically distributed variable, x is the mean, N is the number of variables, and h is the smoothing parameter, called the bandwidth. The larger the bandwidth is, the smoother the estimated density curve is, and the larger the deviation is, the lower the estimation accuracy is.

3.1.4. Spatial Correlation Test

Before analyzing the spatial effect, the Moran’s I index is adopted to investigate the spatial correlation between IGTU and RETI. The calculation formula is as follows:

$$I_G=\frac{{\displaystyle \sum {}_{i=1}^n{\displaystyle \sum {}_{i\ne j}^n{W}_{ij}\left(x_i-\stackrel{\_}{x}\right)\left(x_j-\stackrel{\_}{x}\right)}}}{{\displaystyle \sum {}_{i=1}^n{\left(x_i-\stackrel{\_}{x}\right)}^2}}$$

(10)

$$I_L=x_i{\displaystyle \sum _{j\ne 1}^n{W}_{ij}{x}_j}$$

(11)

where I_G and I_L are the global spatial autocorrelation and local spatial autocorrelation indexes, respectively; n is the number of provinces; x_i and x_j are the observed values; and W_ij is the spatial weight matrix. According to the study by Liu et al. [26], matrices such as the adjacency matrix (W₁), geographical distance matrix (W₂), economic distance matrix (W₃), and economic-geographical distance matrix (W₄) are constructed for the empirical test.

3.1.5. Spatial Panel Durbin Model (SPDM)

IGTU and RETI in different provinces may affect each other spatially, and spatial factors need to be considered for spatial effects testing. The SPDM considers the spatial effects of both explanatory and explanatory variables [95]. Therefore, this paper adopts the SPDM to investigate the spatial effect of RETI on IGTU. The calculation formula is as follows:

$$\begin{array}{cc}IGT{U}_{it}=\hfill & {\alpha }_0+\rho {\displaystyle \sum _{i=1}^n{W}_{ij}}IGT{U}_{it}+{\alpha }_{1}RET{I}_{it}+{\alpha }_2{\displaystyle \sum _{i=1}^n{W}_{ij}}RET{I}_{it}+{\alpha }_3{X}_{it}+{\alpha }_4{\displaystyle \sum _{i=1}^n{W}_{ij}}{X}_{it}\hfill \\ & +{\mu }_i+{\nu }_t+{\epsilon }_{it}\hfill \end{array}$$

(12)

where ρ represents the effect of local IGTU on neighboring IGTU, α₂ is the spatial lag term coefficient of RETI, X_it denotes a series of control variables, and W_ij is the spatial weight matrix.

Furthermore, referring to the study of LeSage and Pace [96], the direct and indirect effects are decomposed using partial differentiation. The formulas for the converted model are as follows:

$$Y_{it}={(I_{NT}-\rho W_{ij})}^{-1}{X}_{it}(I_{NT}\beta +W_{ij}\delta )+{\left[I_{NT}-\rho W_{ij}\right]}^{-1}{\epsilon }_{it}$$

(13)

According to Formula (13), the derivative of the kth explanatory variable is derived as an independent variable, and the partial differential equations are as follows:

$$MEM=\frac{\partial Y_{it}}{\partial X_{it}}={(I_{NT}-\rho W_{ij})}^{-1}(I_{NT}\beta +W_{ij}\delta )$$

(14)

$$\begin{array}{l}M{E}_{Total}=\frac{1}{NT}\iota (MEM),\iota =ones(NT,1)\\ M{E}_{Direct}=\frac{1}{NT}Trace(MEM)\\ M{E}_{Indirect}=M{E}_{Total}-M{E}_{Direct}\end{array}$$

(15)

3.2. Data

3.2.1. Explained Variable: IGTU

The improved entropy weight method is used to measure the IGTU index, and the specific indexes are shown in Section 3.1.1.

3.2.2. Explanatory Variable: RETI

Existing studies mainly adopt the number of renewable energy patents and R&D investment in renewable energy technology to analyze RETI [15,85,97,98,99]. However, considering the technological depreciation and diffusion of RETI, this paper refers to the knowledge stock method of Lin and Zhu [89] to construct the RETI index to measure RETI reasonably. The specific calculation formula is as follows:

$$RET{I}_{it}={\displaystyle \sum _{s=0}^{t}RPA{T}_{it}\exp \left[-{\beta }_1(t-s)\right]}\left\{1-\exp \left[-{\beta }_2(t-s)\right]\right\}$$

(16)

where RPAT_it is the number of renewable energy patents. According to the study by Popp [100], the depreciation rate β₁ is 0.36, and the diffusion rate β₂ is 0.3.

3.2.3. Control Variables

(1) Urbanization level (UL): Urbanization development is a catalyst for structural transformation, which can accelerate the production factors’ flow and promote the industrial labor concentration, laying the foundation for industrial transformation and upgrading [101,102]. Therefore, the proportion of the urban population in the total population is used to measure the urbanization level.

(2) Government intervention (GI): Government intervention, especially fiscal spending, has an essential impact on the regional industrial structure transformation [103]. Therefore, the share of the financial expenditure budget in GDP is used as a proxy variable for government intervention.

(3) Energy consumption structure (ES): Coal consumption accounts for a third of China’s carbon emissions [104], which has an essential impact on the green transformation. The share of coal consumption in total energy consumption is thus used as a proxy variable for energy consumption structures.

(4) Financial development (FD): Financial products improve the efficiency of resource allocation [105] and help IGTU. Financial development is thus represented by the ratio of institutional deposit and loan balance in GDP.

3.2.4. Data Source

This paper takes panel data of 30 provinces in China from 2006 to 2020. Given the data availability, samples from Tibet, Hong Kong, Taiwan, and Macau are not included. The number of renewable energy patents was obtained from the Patent Search and Analysis System (http://www.pss-system.gov.cn (accessed on 20 October 2021). The social and economic data are from the China Statistical Yearbook, China Industrial Statistical Yearbook, China Tertiary Industry Statistical Yearbook, China Environmental Statistical Yearbook, China Energy Statistical Yearbook, China High Technology Industry Statistical Yearbook, China Foreign Trade Statistical Yearbook, Research Center for Digital Inclusive Finance (Peking University), and China Carbon Accounting Database. Descriptive statistics of variables are shown in

Table 2. Statistical description.

Variables	Obs	Mean	Max	Min	S.D
IGTU	450	0.309	0.722	0.122	0.137
RETI	450	25.73	291.7	0.013	43.87
ES	450	0.416	0.748	0.008	0.156
UL	450	0.553	0.941	0.275	0.137
GI	450	0.243	0.758	0.095	0.11
FD	450	3.134	7.578	1.454	1.082

.

4. Empirical Results

4.1. Temporal and Spatial Evolution Analysis

4.1.1. Temporal Evolution Characteristics

To portray the dynamic evolution trends of RETI and IGTU, the kernel density estimation was employed to depict the temporal trend of RETI and IGTU in China from 2006 to 2020 (Figure 1). Overall, the kernel density curve of RETI was characterized by a multi-peaked distribution and the center of the curve shifted to the right. The moving range during the 11th and 12th Five-year Plan (2006–2016) was significantly more extensive than during the 13th Five-year Plan (2016–2020), indicating that compared with the 13th Five-year Plan, the level of RETI increased faster during the 11th and 12th Five-year Plan. The peak height of the curve from 2006 to 2016 gradually increased and the width gradually decreased, indicating that the gap in RETI among Chinese provinces gradually decreased during the 11th and 12th Five-year Plan periods. Still, the width increased slightly and the RETI gap slightly widened during the 13th Five-year Plan period. This may be because environmental constraints have gradually increased over time and because of the differences in economic development, resource endowments, and the pressure of energy conservation and emission reduction among provinces. Local governments have adopted differentiated strategies to formulate strategic planning for renewable energy development and implemented policy choices with varying intensity, resulting in large differences in RETI among regions.

The kernel density curve of IGTU showed the characteristics of the single-peaked distribution. The center of the curve gradually shifted to the right, with the most significant movement from 2006 to 2016, indicating the fastest growth of IGTU during the 11th and 12th Five-year Plan periods and the steady improvement of IGTU during the 13th Five-year Plan periods. The curve trailed obviously to the right, and the extension showed broadening trends, indicating that the number of regions with higher IGTU increased, the areas with lower IGTU gradually shifted to those with higher IGTU, and the gap between areas gradually expanded. Possible reasons, on the one hand, are because, with the strengthening of resource and environmental constraints, the increase in environmental government costs has slowed down the growth rate of industrial green transformation and upgrading. On the other hand, with the enhancement of regional economic cooperation and the expansion of industrial transformation and transfer, the priority flow of high-quality mobile factors has moved to areas with good industrial foundation, leading to an expanding regional gap between stages.

4.1.2. Spatial Characteristics Analysis

To further clarify the spatial distribution of RETI and IGTU, this paper used ArcGIS10.3 to conduct a spatial visualization analysis of RETI and IGTU in China from 2006 to 2020. It can be seen in Figure 2 that RETI formed a pattern of low in the northwest and high in the southeast. The results are similar to the conclusions of Xin et al. [13]. Specifically, the high-RETI area mainly concentrated in the southeast coastal region and had an expanding trend in the surrounding areas. The low-RETI area was primarily distributed in the northwest and southwest regions, and its spatial extent was reduced with the continuous advancement of the energy-transition process. The southeast coast has abundant resources such as hydropower and tidal energy, which support technological innovation and help local renewable development. However, although the northwest regions have abundant resources of solar and wind power, it is difficult to support the development of RETI with relatively poor infrastructure, resulting in a development gap in RETI between the southeast and the northwest [106].

IGTU in China presented a step-down spatial differentiation pattern of east–central–west. Regions with high IGTU were sporadically distributed, mainly concentrated in southeast coastal areas, which have a unique location and economic advantages and can promote IGTU through technological innovation and knowledge spillover. The number of regions with higher IGTU increased from nine to 17, and the double-core distribution formed by Sichuan–Shan–Chongqing–Hubei and Liaoning–Shandong–Zhejiang–Fujian evolved to a contiguous distribution centered on Hebei–Henan–Hubei. Such areas have undertaken many industrial transfers, with close interaction of talent, technology, and capital, which contributes to enhancing resource utilization efficiency and promoting IGTU. The regions with lower IGTU evolved from a contiguous strip distribution to multi-core distribution, with the number decreasing from 12 to seven and a decrease of 41.67%, indicating that IGTU in most areas gradually improved.

The spatial range of the low IGTU region gradually narrowed, and the number decreased from four to one. The dual-core distribution formed by Gansu–Qinghai–Ningxia and Guizhou evolved to a single-core distribution dominated by Inner Mongolia. The industrial development of such regions is more dependent on resources, and the consumption mode based on traditional energy increases carbon emissions and aggravates environmental pollution, which is unfavorable to IGTU. The above results indicate apparent spatial heterogeneity between RETI and IGTU in the regions, and the spatial factors cannot be ignored when exploring the relationship between them.

4.2. Spatial Correlation Analysis

To test spatial correlation, the Moran’s I index was adopted to examine the spatial correlation of RETI and IGTU. Considering possible estimate bias, the four spatial weight matrices, including the adjacency matrix (W₁), geographical distance matrix (W₂), economic distance matrix (W₃), and economic–geographical distance matrix (W₄), were adopted to calculate the global Moran’s I index (

Table 3. Spatial autocorrelation test.

Year	RETI				IGTU
	W1	W2	W3	W4	W1	W2	W3	W4
2006	0.176 **	0.039 **	0.370 ***	0.004 **	0.359 ***	0.086 ***	0.117 **	0.155 ***
2007	0.285 ***	0.030 **	0.479 ***	0.019 **	0.312 ***	0.079 ***	0.123 **	0.127 **
2008	0.282 ***	0.035 **	0.450 ***	0.034 **	0.330 ***	0.073 ***	0.134 **	0.120 **
2009	0.262 ***	0.042 **	0.383 ***	0.057 **	0.320 ***	0.074 ***	0.158 **	0.120 **
2010	0.261 ***	0.040 **	0.366 ***	0.073 **	0.359 ***	0.083 ***	0.172 **	0.137 **
2011	0.281 ***	0.049 ***	0.361 ***	0.096 **	0.352 ***	0.076 ***	0.170 **	0.119 **
2012	0.231 **	0.030 **	0.311 ***	0.067 **	0.351 ***	0.076 ***	0.167 **	0.121 **
2013	0.258 ***	0.046 **	0.316 ***	0.124 **	0.349 ***	0.085 ***	0.182 **	0.137 **
2014	0.266 ***	0.040 **	0.344 ***	0.100 **	0.321 ***	0.074 ***	0.170 **	0.118 **
2015	0.257 ***	0.053 ***	0.343 ***	0.112 **	0.285 ***	0.059 ***	0.168 **	0.073 **
2016	0.296 ***	0.068 ***	0.286 ***	0.155 ***	0.316 ***	0.072 ***	0.182 ***	0.104 **
2017	0.227 **	0.047 ***	0.351 ***	0.078 **	0.344 ***	0.078 ***	0.184 ***	0.125 **
2018	0.210 **	0.061 ***	0.278 ***	0.125 **	0.327 ***	0.067 ***	0.165 **	0.110 **
2019	0.263 ***	0.087 ***	0.288 ***	0.166 ***	0.379 ***	0.081 ***	0.167 **	0.143 **
2020	0.326 ***	0.100 ***	0.297 ***	0.223 ***	0.433 ***	0.087 ***	0.145 **	0.164 ***

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

). The results in Table 3 show that the Moran’s I index of RETI and IGTU were significantly positive under the four spatial weight matrices from 2006 to 2020, demonstrating the existence of a spatial correlation between RETI and IGTU.

The Moran’s I index of RETI and IGTU showed a fluctuating upward trend during the examination period, demonstrating that the spatial aggregation of RETI and IGTU gradually strengthened, so it is necessary to take into account spatial effect when investigating the influencing mechanism of RETI on IGTU. Therefore, this paper adopts a spatial panel econometric model to examine the impact of RETI on IGTU. W₁ was used for empirical analysis in the next step, and W₂, W₃, and W₄ were adopted in the robustness test.

4.3. Estimation Result Analysis

4.3.1. Result of Model Selection

The previous section verified that RETI and IGTU were significantly spatially correlated, so selecting an appropriate spatial econometric model was necessary. As seen in

Table 4. Selection of spatial econometric models.

Tests	Statistics	Tests	Statistics
LM (lag) test	2.950 ***	Wald_spatial_lag	33.740 ***
Robust LM (lag) test	72.198 ***	LR_spatial_lag	27.880 ***
LM (error) test	109.405 ***	Wald_spatial_error	39.340 ***
Robust LM (error) test	178.653 ***	LR_spatial_error	38.690 ***
Hausman test	1.070 ***

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

, the LM, LR, Wald, and Hausman test results indicate that the SPDM with spatial-time fixed effects needed to be employed to investigate the spatial spillover effect.

4.3.2. Results of the SPDM

As seen from the results in

Table 5. Estimation results of the SPDM.

Variables	OLS	FE	SPDM
lnRETI	0.011 ***	0.067 ***	0.071 ***
	(2.11)	(5.74)	(10.60)
lnES	−0.187 ***	−0.138 **	−0.086 ***
	(−3.10)	(−2.42)	(−2.81)
lnUL	1.131 ***	1.140 ***	0.735 ***
	(9.69)	(10.25)	(11.02)
lnGI	−2.235 ***	−1.735 ***	−0.508 ***
	(−21.92)	(−12.22)	(−5.71)
lnFD	0.081 ***	0.077 ***	0.060 ***
	(5.75)	(5.59)	(8.07)
W×lnRETI			−0.057 ***
			(−3.95)
W×lnES			−0.211 ***
			(−3.21)
W×lnUL			−0.555 ***
			(−4.92)
W×lnGI			−0.231
			(−1.47)
W×lnFD			−0.014
			(−0.98)
Spatial rho			0.396 ***
			(8.18)
Variance sigma2_e			0.009 ***
			(14.80)
R2	0.773	0.808	0.855
N	450	450	450

Note:**, and *** indicate significance at the 5%, and 1% levels, respectively. The t-values are in parentheses.

, compared with the ordinary panel model, the R² of SPDM was significantly higher. By comparing the RETI coefficient, it was found that if the spatial effect were not considered, the influence of RETI on IGTU would be underestimated. Therefore, the SPDM results were selected for further analysis.

The value of spatial rho was 0.396 at a 1% significance level, demonstrating that there was a significant spatial spillover effect of IGTU, and regions with high IGTU would have a positive driving impact on IGTU in neighboring areas. This may be because geographically adjacent provinces have similar IGTU levels, and the process of industrial green transformation and upgrading is accompanied by industrial transfer, which accelerates the flow of factors such as talent, capital, and technology, contributing to the positive spillover of IGTU. RETI had a significant promoting effect on IGTU: when RETI increased by 1%, IGTU significantly improved by 0.071%. This may be because the improvement of RETI helps to improve the utilization efficiency of clean energy, promote the optimization of energy structure, weaken industries’ dependence on fossil energy, and reduce resource consumption and pollution emissions [12], thus promoting IGTU. From the control variable results, the relationship between ES and IGTU showed a negative change and passed the 1% significance test. The energy structure dominated by coal consumption would aggravate environmental pollution, leading to an inhibiting effect on IGTU. The relationship between UL and IGTU was positive with a coefficient of 0.735, passing the 1% significance level, which reflects that UL promoted IGTU significantly. With the improvement in the urbanization level, urban infrastructure and supporting services are constantly improved, which contributes to accelerating the element agglomeration and external economy in industrial development and thus promoting IGTU. When GI changed by 1%, IGTU significantly changed by 0.0508% in the opposite direction, demonstrating that GI hindered the level of IGTU. This is because government intervention may weaken the market resource allocation capacity, and the primary role of enterprises cannot be carried out entirely. Therefore, coordination between an efficient market and a competent government is needed to promote IGTU. The improvement of each unit of FD significantly increased the value of IGTU by 0.06, illustrating that FD significantly promoted IGTU. The increasingly mature capital market is conducive to investment in scientific and technological innovation [107]. Developing financial markets provides financial support for green technological innovation, contributing to improving industrial total factor productivity, thus driving IGTU.

Regarding spatial effects, the spatial weight coefficient results of W×lnRETI, W×lnES, and W×lnUL were −0.057, −0.211, and −0.555, respectively, which all passed the 1% significance level, illustrating that they had significant adverse spatial effects on IGTU. To avoid the result biases caused by the point estimate regression, this paper decomposed the spatial spillover effects and drew on the study of Lesage and Pace [96].

4.3.3. Spatial Effect Decomposition Analysis of SPDM

The results of spatial effect decomposition are reported in

Table 6. Decomposition of direct and indirect effects.

Variables	Direct	Indirect	Total
lnRETI	0.068 ***	−0.043 **	0.025 **
	(9.49)	(−2.02)	(2.01)
lnES	−0.115 ***	−0.380 ***	−0.495 ***
	(−3.68)	(−3.74)	(−4.24)
lnUL	0.710 ***	−0.414 ***	0.297 *
	(11.06)	(−2.65)	(1.68)
lnGI	−0.552 ***	−0.650 ***	−1.202 ***
	(−6.44)	(−2.86)	(−4.89)
lnFD	0.061 ***	0.014	0.075 ***
	(8.28)	(0.66)	(3.04)

Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. The t-values are in parentheses.

. The direct effect of RETI on IGTU was significantly positive, with a coefficient of 0.068, and the spatial spillover effect of RETI on IGTU was significantly negative, with a coefficient of −0.043, reflecting that RETI improved local IGTU but hindered neighboring IGTU, which formed a “beggar-thy-neighbor” situation. This may be because the improvement of RETI helps improve the industrial energy structure and promotes the transformation and upgrading of energy-intensive industries to high-end, diversified, and low-carbon industries and gradually forms green and low-carbon industrial clusters, which would in turn enhance local IGTU. However, the clustering of low-carbon industries produces a siphon effect, leading to the influx of capital, technology, and talent elements from the surrounding areas into the central area, which puts tremendous pressure on the sustainable development of the surrounding industries and is detrimental to the neighboring IGTU.

For the control variables, the coefficient of the direct effect and spatial spillover effect of ES on IGTU were negative, with coefficients of −0.115 and −0.38 and significance at the 1% level, illustrating that the energy structure inhibited the green transformation and upgrading of local and neighboring industries. It may be because not only does a coal-based energy consumption structure cause severe environmental pollution, but excessively relying on traditional energy would also lead to insufficient motivation for enterprises’ green innovation and backward production technology, which is unfavorable to IGTU. The results of the UL coefficient show that when UL increased by 1%, local IGTU improved by 0.71% and neighboring IGTU significantly decreased by 0. 41%, demonstrating that enhancement of UL promoted local IGTU but inhibited neighboring IGTU. On the one hand, urbanization development is conducive to improving local regional productivity and providing labor supply for industries; on the other hand, the local urbanization development expands the market demand and attracts advantageous industries and high-end talents from the neighboring areas, creating a siphoning effect and resulting in a “beggar-thy-neighbor” situation. When GI increased by 1%, local and neighboring IGTU significantly decreased by 0.55% and 0.65%, respectively, which shows that government intervention hindered local and neighboring IGTU. this suggests that government intervention may lower the marketization level and that it is difficult for enterprises to fully motivate the initiative to transform and upgrade, which is not conducive to promoting IGTU. The direct-effect coefficient of FD was a significant 0.061, and the spatial spillover effect coefficient of FD was an insignificant 0.014, indicating that the promotion of financial development on IGTU was positive in local areas but insignificant in neighboring regions. This is because financial development provided more financing opportunities for local enterprises, stimulates their innovative vitality, and provided financial security for local IGTU. However, due to the restrictions on the trans-regional operation of local financial institutions and the increased cost of trans-regional capital flow caused by the local administrative division, trans-regional financial cooperation was hindered, which was not conducive to promoting IGTU in surrounding areas.

4.3.4. Result of the Regional Boundary of RETI on IGTU

To further investigate whether the spatial spillover effect of RETI on IGTU has a regional boundary, drawing on the study of Liu et al. [26], this paper constructed threshold distance matrix W_dij to examine the spatial spillover effect with different geographical distance constraints. The calculation method is as follows:

$$W_{dij}=\{\begin{array}{ll}\frac{1}{d_{ij}}& d_{ij}\ge d\\ 0& d_{ij}<d\end{array}$$

(17)

In the threshold distance spatial weight matrix, the initial threshold distance d_ij was set as 200 km, and the incremental distance was set as 200 km and kept increasing to 2400 km. Furthermore, the trends of spatial spillover effect coefficients at the different distances were visualized (Figure 3).

There are mainly three characteristics of the effect of RETI on IGTU in Figure 3: (1) The effect of RETI on neighboring IGTU was significantly negative within 800 km, which declined to half at 600 km. Generally, 800 km was basically at the border of the province. The improvement of RETI gathered local high-quality resource elements. In addition, it attracted the inflow of high-quality capital, technology, and talent from neighboring areas, producing a siphon effect, which negatively affected the neighboring IGTU. However, with increasing geographical distance, the cross-regional cost of technical cooperation and information interchange gradually increased, which impeded the positive externality of RETI. (2) In the range of 800 to1400 km, RETI had a significant positive spatial spillover effect on IGTU, with a peak of 0.225 at 1000 km. China has eight integrated economic zones, the center of which is about 1000 km, and the increasingly close intra-regional technical cooperation and financial exchanges would drive neighboring IGTU through imitation and demonstration effects. (3) After 1600 km, the spatial spillover coefficient fluctuated randomly and insignificantly, with the spatial element decreasing sharply, showing that the influence of RETI on neighboring IGTU was limited by the regional boundary. In general, the above results provide strong evidence for the spatial distance decay hypothesis and verify that there are apparent regional boundaries for the spatial spillover effect of RETI on IGTU.

4.4. Heterogeneity Analysis

4.4.1. Temporal Heterogeneity

Given the economic policies and development strategies put forward by the Chinese government every five years, the samples were divided into three groups according to the 11th Five-year Plan (2006–2010), 12th Five-year Plan (2011–2015), and 13th Five-year Plan (2016–2020) to further examine the impact of RETI on IGTU. The results from models (1) to (3) in

Table 7. Heterogeneity results in time and regions.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
	11th Five-year Plan	12th Five-year Plan	13th Five-year Plan	Eastern	Central	Western
lnRETI	0.067 ***	0.068 ***	0.054 ***	0.107 ***	−0.004	−0.027 ***
	(7.29)	(5.48)	(4.11)	(12.04)	(−0.71)	(−5.46)
lnES	0.026	−0.007	−0.258 ***	−0.052	−0.054 **	0.029
	(0.52)	(−0.11)	(−5.37)	(−1.26)	(−2.03)	(1.26)
lnUL	0.942 ***	0.774 ***	0.274 **	0.659 ***	0.027	0.076
	(9.89)	(6.01)	(2.35)	(7.35)	(0.47)	(1.49)
lnGI	−0.593 ***	−0.473 ***	−0.866 ***	−0.076	0.120	−0.496 ***
	(−3.82)	(−3.13)	(−5.12)	(−0.63)	(1.58)	(−7.44)
lnFD	0.046 ***	0.050 ***	0.100 ***	0.076 ***	−0.034***	0.012 **
	(3.50)	(3.55)	(8.68)	(7.47)	(−4.97)	(2.15)
W×lnRETI	−0.070 ***	−0.038	−0.064 **	−0.082 ***	−0.007	0.013
	(−3.56)	(−1.34)	(−2.28)	(−4.14)	(−0.60)	(1.22)
W×lnES	−0.487 ***	−0.347 *	−0.151	−0.233 ***	0.075	0.007
	(−4.72)	(−1.90)	(−1.23)	(−2.63)	(1.36)	(0.15)
W×lnUL	−0.462 ***	−0.793 ***	−0.783 ***	−0.322 **	0.477 ***	−0.769 ***
	(−2.68)	(−3.65)	(−3.60)	(−2.09)	(4.87)	(−8.60)
W×lnGI	−0.224	−0.302	−0.387	−0.535 ***	0.002	0.326 ***
	(−0.90)	(−1.05)	(−1.13)	(−2.58)	(0.01)	(2.77)
W×lnFD	0.027	0.011	−0.009	0.003	−0.069 ***	0.038 ***
	(1.16)	(0.37)	(−0.33)	(0.14)	(−5.42)	(3.52)
Spatial rho	0.409 ***	0.350 ***	0.371 ***	0.516 ***	0.331 ***	0.419 ***
	(5.11)	(4.21)	(3.70)	(10.58)	(4.54)	(7.06)
Variance sigma2_e	0.006 ***	0.010 ***	0.006 ***	0.016 ***	0.006 ***	0.005 ***
	(8.55)	(8.58)	(8.52)	(14.59)	(14.72)	(14.65)
R2	0.880	0.728	0.826	0.791	0.345	0.552
N	150	150	150	165	120	165

Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. The t-values are in parentheses.

show that the driving effect of RETI on IGTU was more substantial during the 12th Five-year Plan period, but it decreased slightly during the 13th Five-year Plan period. This is because the green development direction of building a resource-saving and environment-friendly society was clearly defined in the 12th Five-year Plan period, and the 12th Five-year Plan for Renewable Energy Development issued by the National Energy Administration deployed a series of green development measures to promote energy-cleaning utilization and transformation. To achieve the green development goals, the industries have actively applied renewable energy technologies and integrated resource conservation and environmental protection into production, distribution, consumption, and construction, which promotes IGTU through industrial structure optimization, efficiency improvement, carbon reduction, and pollution reduction. During the 13th Five-year Plan period, although China’s installed hydropower, wind power, and solar power generation were at the forefront of the world, the sudden outbreak of COVID-19 at the end of the 13th Five-year Plan caused severe turbulence to the development of renewable energy. The reduction of new installed capacity for renewable energy power generation caused wind power supply tension, resulting in most wind-power projects not being online on time. As a result, the proportion of renewable energy consumption decreased and that of traditional fossil energy consumption increased, and industrial transformation and upgrading still relied on a large amount of traditional fossil energy consumption. Therefore, during this period, the promotion effect of RETI on local IGTU declined slightly, and the siphoning effect on neighboring IGTU weakened slightly.

In terms of the control variables, the inhibitory effects of ES and GI on IGTU were most substantial during the 13th Five-year Plan period, whereas UL and FD had significant boosting effects on IGTU in different periods. Among them, the promoting effect of UL on local IGTU decreased over time, mainly because China’s urbanization experienced a high growth rate of 3.2% during the 11th Five-year Plan period and slowed down to 2.0% during the 13th Five-year Plan period, which contracted market demand and slightly reduced the promoting effect on IGTU. In the process of UL promoting IGTU, local enterprises have attracted the inflow of high-quality capital and technology from surrounding areas, resulting in the outflow of neighboring high-quality elements, which weakens the IGTU development. The driving effect of FD on IGTU has increased over time, probably because a series of measures to accelerate the opening development in the financial sector have been implemented since 2018. Even during the COVID-19 period, the implementation of financial opening measures also continued to advance, and the pace of financial development sped up significantly, which contributed to providing financial resources and opportunities for green industry projects and accelerating IGTU. The inhibitory effect of ES on IGTU was most substantial during the 13th Five-year Plan period, probably because the pandemic led to the reduction of renewable energy electricity supply and an increase in the consumption of traditional energy sources such as coal and oil for industrial transformation and upgrading compared with the previous Five-year Plan period, which is not conducive to the green development of local industries. However, from spatial spillover effects, ES weakened neighboring IGTU gradually, which may be because of the higher coal consumption by local industries, making the proportion of traditional fossil energy in neighboring industries decrease, showing a trend of a decreasing inhibitory effect.

4.4.2. Regional Heterogeneity

The previous section confirmed that RETI and IGTU have significant spatial heterogeneity. To further examine the regional heterogeneity of RETI on IGTU, this paper investigated the impacts in the eastern, central, and western regions. It can be seen in columns (4)–(6) of Table 7 that RETI significantly promoted IGTU in the east and suppressed IGTU in the west, and RETI significantly inhibited neighboring IGTU in the eastern region. This may have correlations with regional factor endowment and industrial foundations. The eastern region has a better industrial foundation and is willing to take the initiative in reform, which can effectively play to the advantage of talents and technology concentration to improve RETI, which provides an opportunity for the eastern region to increase the clean energy utilization, optimize and adjust the energy structure, enhance the efficiency of resource utilization, and accelerate IGTU. At the same time, the process of technology spillover from RETI also produced a siphon effect. The promoting process of RETI on local IGTU attracted the inflow of superior elements that prefer IGTU in surrounding areas, which hindered the neighboring IGTU. However, the western region lags in economic development, with a weak industrial foundation and greater resource dependence. To catch up with central and eastern economic development, on the one hand, focus must be placed on expanding the industrial scale and choosing the rough development mode, ignoring industrial green development. On the other hand, the western region has more technical bottlenecks and a lack of management personnel; strengthening RETI is a challenge that will play a driving role in technological innovation in the short term, which is not conducive to promoting IGTU in the western region.

Based on the control variables, ES significantly inhibited central IGTU. this may be because the central region has taken over some backward industries in the eastern region and enhanced the industrial agglomeration. The development mode dominated by coal consumption is not conducive to IGTU. UL had a significant promotion effect on local IGTU in the eastern region but had a more substantial inhibitory impact on neighboring IGTU. This may be because the improving urbanization in the east attracts a large amount of labor force and expands the market demand, which promotes the high-quality mobile factors in the neighboring areas to gather in the east, resulting in UL favorably promoting the local IGTU but adversely affecting the neighboring IGTU in the east. GI had a significant inhibitory effect on local IGTU in the western region but significantly boosted neighboring IGTU in the west. This may be because of the weak industrial foundation in the west, and administrative interventions tend to disrupt the production and manufacturing in industrial development, causing enterprises to bear uncertain losses and leading to the dilemma in local IGTU. Moreover, with the increasing relocation and transfer of local industries, there are new opportunities for neighboring IGTU. FD significantly promoted local IGTU in the east and west, and significantly inhibited local IGTU in the central region. Regarding the spillover effect, FD significantly suppressed central IGTU and promoted western IGTU. This may be because RETI is mainly concentrated in the east and the financial development helps the development and application of RETI, which has a positive impact on local IGTU. Meanwhile, with industrial agglomeration being increasingly prominent in the east, the knowledge and technology effects exerted by RETI are easily spread to the industrial agglomeration areas and thus promote the neighboring IGTU in the east. In addition, the western region is abundant in renewable energy such as wind and solar power, and financial development provides financial support for renewable energy supply; thus, FD helps the western region take the lead in applying renewable energy, which in turn promotes the western local IGTU. With the expansion of renewable energy construction in the west and the integration of financial development, financial development is conducive to exerting positive externalities of RETI and promoting neighboring IGTU in western regions.

4.4.3. Resource Endowment Heterogeneity

Resource-based regions are essential carriers of China’s modern industry, with a disproportionately high share of non-renewable energy-extraction industries and a homogeneous industrial structure, which puts enormous pressure on the sustainable development of the economy, society, and environment [108]. To further examine whether the differences in resource endowments have heterogeneous effects on the above findings, regarding the study by Fan and Zhang [109], the sample was divided into resource-based and non-resource-based regions (

Table 8. Heterogeneous results of environmental constraints and resource endowment regions.

Variables	(1)	(2)	(3)	(4)
	Resource-Based Areas	Non-Resource-Based Areas	Weak Environmental Regulation	Strong Environmental Regulation
lnRETI	−0.008 *	0.074 ***	−0.078 ***	0.143 ***
	(−1.90)	(8.97)	(−9.37)	(11.74)
lnES	−0.021	−0.046	−0.040	−0.101 *
	(−1.08)	(−1.20)	(−1.04)	(−1.91)
lnUL	0.091 **	0.646 ***	1.119 ***	−0.562 ***
	(2.19)	(7.72)	(13.27)	(−4.86)
lnGI	0.145 ***	−0.664 ***	−0.494 ***	−0.228
	(2.65)	(−5.97)	(−4.56)	(−1.49)
lnFD	−0.011 **	0.072 ***	−0.047 ***	0.120 ***
	(−2.45)	(7.66)	(−5.08)	(9.28)
W×lnRETI	−0.025 ***	−0.032 *	0.104 ***	−0.136 ***
	(−2.99)	(−1.83)	(5.90)	(−5.19)
W×lnES	0.209 ***	−0.411 ***	−0.208 ***	−0.190 *
	(5.28)	(−5.00)	(−2.66)	(−1.68)
W×lnUL	−0.078	−0.419 ***	−1.257 ***	0.995 ***
	(−1.10)	(−2.97)	(−8.82)	(5.08)
W×lnGI	0.028	−0.152	0.036	−0.742 ***
	(0.29)	(−0.76)	(0.19)	(−2.76)
W×lnFD	−0.004	−0.020	0.022	−0.021
	(−0.40)	(−1.09)	(1.25)	(−0.82)
Spatial rho	0.356 ***	0.376 ***	0.423 ***	0.082 *
	(7.79)	(7.59)	(5.95)	(1.79)
Variance sigma2_e	0.003 ***	0.014 ***	0.013 ***	0.026 ***
	(14.99)	(14.82)	(14.72)	(14.95)
R2	0.337	0.816	0.612	0.668
N	135	315	330	120

Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. The t-values are in parentheses.

). The results show that RETI significantly inhibited local IGTU in resource-based regions but significantly promoted local IGTU in non-resource-based regions. This may be because resource-rich regions would increase the extraction and utilization due to lower-cost resources and reach a consensus on developing resource-based industries, resulting in the development of resource-based regions being prone to path dependence. In the short term, enhancing RETI makes it challenging to break the bottleneck of the “resource curse,” which still leads to high energy consumption, high carbon emissions, and increased pollution. On the contrary, non-resource-based regions are mostly leading regions in economic development, with more abundant human capital and technical support, which provide good development conditions for the R&D and application of RETI and promote IGTU in non-resource-based regions.

In terms of control variables, compared with resource-based regions, UL and FD promoted local IGTU in non-resource-based areas. This may be because non-resource-based areas have better economic development, and urbanization and financial development can strengthen the support of high-quality mobile factors, providing demand drivers and financial support for IGTU. However, the results of the spillover effects show that UL had a significant negative impact on neighboring IGTU in non-resource-based areas. This may be because there are fewer labor-intensive industries in non-resource-based areas, and urbanization development increases the labor transfer to labor-intensive industries in resource-based regions, which enhances the outflow of production factors and weakens the ability to improve new products and expand the industrial scale in neighboring regions, leading to IGTU development being hindered in neighboring areas. ES had a significant positive spatial spillover effect on IGTU in resource-based regions, but it significantly inhibited neighboring IGTU in non-resource-based regions. This may be because of the significant difference in resource endowment between resource-based and non-resource-based areas and the complementary resources within the area and neighboring regions, leading to a decrease in the IGTU level in areas with a high proportion of fossil-energy consumption, such as coal and oil.

4.4.4. Environmental Regulation Heterogeneity

The Porter Hypothesis believes that environmental regulation could induce firms to engage in technological innovation and improve their competitiveness [110], leading to the technological innovation level varying with the environmental regulation intensity. To examine whether different environmental regulation intensity has heterogeneous effects on the above results, referring to the study by Zhong et al. [111], the samples were divided into strong and weak environmental regulation areas (Table 8). The results demonstrate that in strong environmental regulation areas, RETI significantly promoted local IGTU and suppressed neighboring IGTU. In weak environmental regulation areas, RETI significantly inhibited local IGTU and promoted neighboring IGTU. It may be that stronger environmental constraints force enterprises to carry out RETI to optimize energy structures and improve the efficiency of clean energy utilization, which is conducive to promoting local IGTU. Conversely, in regions with weaker environmental constraints, enterprises have lower costs of pollution emissions and may only need end-of-pipe treatment to meet emission standards. Therefore, enterprises have less incentive to improve energy consumption structures through technological innovation, which significantly hinders the development of RETI and is detrimental to IGTU.

The results of the control variables show that UL significantly promoted local IGTU in areas with weak environmental regulations and suppressed local IGTU in regions with strong environmental regulations. This is probably because compared with areas with strong environmental regulations, resource-based industries are gathering in weak environmental regulation areas, which promotes the transfer of urban labor to weak environmental regulation areas. UL significantly suppressed neighboring IGTU in weakly regulated areas and promoted neighboring IGTU in strongly regulated regions, indicating that urbanization in the regions with weak environmental regulations makes it easy to attract and gather labor force from surrounding areas, resulting in the Matthew effect in neighboring labor force. FD significantly inhibited local IGTU in weakly environmentally regulated regions and promoted local IGTU in strongly environmentally regulated areas. This may be because the difference in RETI quality is caused by the different regional environmental regulations. The financial development gives priority to strong financial support for high-quality technologies, which enhances RETI quality and innovation effect of RETI in strong environmental regulation regions and is conducive to promoting IGTU in strong environmental regulation regions.

4.5. Robustness Test

4.5.1. Endogenous Test

Instrumental variables are the primary means to solving potential endogeneity. Referring to the study by Yuan et al. [112], this paper incorporated the instrumental variables of the explained variable lag of one period into the system-generalized moment-estimation model (SYS-GMM) and estimated it again to deal with the potential endogenous problems. According to the results in

Table 9. Robustness test.

Variables	(1)	(2)	(3)	(4)	(5)
	SYS-GMM	W2	W3	W4	Adding Control Variables
L.IGTU	0.933 ***
	(45.92)
lnRETI	0.069 **	0.068 ***	0.036 ***	0.064 ***	0.068 ***
	(2.06)	(11.16)	(4.18)	(9.45)	(12.24)
W×lnRETI		−0.070 *	−0.123 ***	−0.060 **	−0.007 **
		(−1.71)	(−5.37)	(−2.02)	(−1.82)
Control variables	YES
Spatial rho		0.008 **	0.105 *	0.300 ***	0.248 ***
		(2.05)	(1.78)	(3.77)	(3.79)
Variance sigma2_e		0.008 ***	0.010 ***	0.010 ***	0.005 ***
		(14.99)	(15.13)	(14.14)	(14.91)
AR(1)	−7.46
	(0.000)
AR(2)	1.37
	(0.172)
Sargan	293.13
	(1.00)
N	420	450

Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. The t-values are in parentheses.

, the time-lag term of IGTU was 0.933 at the 1% significance level, indicating the existence of path dependence of IGTU in the time dimension. The estimation results of other variables are generally consistent with the above, which proves the robustness of the main findings.

4.5.2. Changing the Space Weight Matrix

To help to eliminate the estimation bias caused by the selection of the spatial weight matrix [113], the model was estimated again by replacing the adjacency matrix (W₁) with the geographical distance matrix (W₂), economic distance matrix (W₃), and economic–geographical distance matrix (W₄). It can be seen from models (2) to (4) in Table 9 that under different spatial weight matrices, RETI significantly promoted local IGTU but inhibited neighboring IGTU, verifying the robustness of the results.

4.5.3. Adding Control Variables

Climate conditions are essential factors affecting pollution and emissions [13]. Given the deviation caused by missing variables, climate factors were included in the model for re-estimation. The results of column (5) in Table 9 show that despite the inclusion in the model of indexes such as mean annual sunshine duration, mean annual precipitation, and mean annual temperature, the relationship between RETI and IGTU remained unchanged, confirming the reliability of the previous results.

5. Conclusions and Discussion

5.1. Conclusions

Based on green growth and sustainable development theories and the data of 30 provinces from 2006 to 2020 in China, this paper firstly defines the theoretical connotation of IGTU and constructs an analytical framework in five dimensions: structural optimization, efficiency improvement, quality enhancement, environmental friendliness, and resource sustainability, and employs the improved entropy weight method to measure IGTU. Secondly, kernel density estimation and the Moran’s I index are used to portray spatial and temporal differentiation patterns. Finally, the SPDM is adopted to examine the influencing mechanism and spatial spillover effect of RETI on IGTU and its regional boundary, and further explore the spatial heterogeneity in time, region, resource endowment, and environmental regulation.

We conclude that (1) RETI increases steadily and rapidly, presenting a spatial pattern of low in the northwest and high in the southeast. IGTU fluctuates and shows a stepwise decreasing spatial differentiation pattern of east–central–west. The spatial dependence of IGTU and RETI gradually increases, and the spatial aggregation becomes prominent. (2) RETI promotes local IGTU significantly, but inhibits neighboring IGTU, presenting a “beggar-thy-neighbor” situation. (3) The effect of RETI on neighboring IGTU presents significant spatial distance attenuation characteristics. The negative spillover effect of RETI persists up to 800 km and halves at 600 km. However, from 800 to 1400 km, the spatial spillover effect is significantly positive and shows a trend of increasing and then decreasing. (4) RETI has the most substantial positive impact on the local IGTUs during the 12th Five-year Plan period. By region, the promotion effect of RETI on IGTU is most notable in the east, and the inhibition effect of that is significant in the west. (5) RETI improves local IGTU in non-resource-based areas and strong environmental regulation areas but has a significant inhibiting effect on local IGTU in resource-based areas and weak environmental regulation areas. From the spatial spillover effect, RETI promotes IGTU in regions with weak environmental regulation.

5.2. Policy Implications

Policy implications based on the above findings are put forward. First, the promotion of RETI on local IGTU is significant, meaning that IGTU needs to give full rein to the leading role of technological innovation and accelerate energy restructuring and transformation. RETI is the essential means to promoting energy revolution and prompting the industry to occupy a new round of competitive advantage. Therefore, the government should increase R&D investment; actively build a platform for cooperation between industry, academia, and research institutes; form several distinctive RETI enterprise clusters; accelerate the transformation of innovation achievements; improve energy utilization efficiency; optimize energy consumption structures; and create a green production system to boost IGTU.

Second, it is found that the negative spatial spillover effect of RETI on IGTU demonstrates an attenuating trend. This implies that the government should improve regional coordination mechanisms and promote regional integration. Mainly, it should break down factor barriers in local protection and market segmentation, strengthen the flow of talent and information and innovation achievement sharing, promote benign polarization and the trickle-down effect, and drive the synergistic development of regional IGTU.

Third, given that the impact of RETI on IGTU has significant heterogeneity in time, region, resource endowment, and environmental regulation, the government should plan the development scale and speed of RETI scientifically and reasonably according to the characteristics of the different development stages, which aim to avoid blind expansion and strive to maximize efficiency. The government should scientifically assess the regional resource endowment and formulate differentiated policies in light of local conditions. For eastern regions, due to their concentration of high-tech talent and the advantage of technological innovation, it is necessary to increase R&D investment in renewable energy technologies, actively promote the utilization of clean energy, play an exemplary leading role in RETI, and strive to gain competitive advantage in the new round of industrial revolution. The central and western regions have the advantage in wind and solar energy resources but lack the technology and a complete industrial chain. The integration of new energy infrastructure construction should be created, new energy manufacturing industry clusters should be created, the new energy industry chain and value chain should be promoted to medium–high levels, and the goal of IGTU should be achieved. In addition, resource-based regions should break through the limitations of a single industrial structure and guide the organic agglomeration of diversified industries. Emerging and alternative industries should be actively introduced, the industrial chain should be continuously extended, the transformation of old and new driving forces should be promoted, and new growth vitality should be injected into the regional IGTU. It is worth noting that stronger environmental regulations are conducive to stimulating the positive externalities of RETI to IGTU. Therefore, the government should adhere to the concept of green development, strengthen environmental regulation, implement diversified environmental policies, and accelerate the modernization of environmental governance and capacity to boost high-quality industrial transformation and upgrading.

5.3. Discussion

This study expands the conceptual boundary of IGTU based on green growth and sustainable development theories and reveals a “beggar-thy-neighbor” development pattern using SPDM among geographically adjacent provinces, which provides new evidence for optimizing the cooperation layout of RETI and promoting IGTU within effective spatial–regional boundaries. More importantly, how to achieve IGTU is not only a significant issue for China to achieve carbon neutrality, but also an increasing concern in the post-COVID-19 era for the whole world when dealing with energy transition and global climate. The conclusions of this study can provide empirical support for industrial green development in other developing countries, especially those dependent on oil resources and striving to improve industrial core competitiveness. Still, in terms of research breadth, this study could expand it further. First, due to the data availability, the analysis of RETI on IGTU at the provincial level is explored in-depth, and subsequent discussions can be further expanded using prefecture-level cities or enterprise data. Second, the evaluation of IGTU mainly selects quantitative indicators, ignoring the use of qualitative indicators. The studies can further incorporate qualitative factors such as policies and systems into the analysis framework for a more comprehensive evaluation of IGTU. Finally, future research could try to compare and explore the differences among countries worldwide, which is conducive to seeking a path for strengthening RETI cooperation and formulating targeted policies for IGTU based on different national conditions.