New Energy Technology Innovation and Industry Carbon Emission Reduction Based on the Perspective of Unbalanced Regional Economic Development

Abstract

Innovation in new energy technologies is a key driver in China’s efforts to achieve its environmental goals. However, the ability of different regions to develop and utilize new energy technologies may depend on their level of economic development. Based on a two-way fixed-effects panel data model, this paper empirically analyses the industry carbon emission reduction effect of new energy technology innovation and its heterogeneous performance at different stages of economic development, using data from 30 provinces and cities in China from 2000 to 2019. The results show that new energy technology innovation generally promotes CO₂ emissions in China. The specific effects are closely related to the characteristics of the industry and the stage of economic development. At the same time, the implementation of environmental regulations will inhibit this positive effect, while the adjustment of the industrial structure may promote this positive effect. This paper discovers the mechanism of heterogeneity in new energy technology innovation among different provinces with different levels of economic development. This finding helps to fully assess the carbon emission reduction capacity and potential of different provinces and facilitates the rational disaggregation and formulation of climate policy goals among regions.

1. Introduction

Since the Industrial Revolution, the rapid growth of the industrial economy has been the main driving force behind the sustained growth of the world economy. However, industry belongs to the high-energy-consuming and high-emission sector. If we rely on the industrial economy to promote economic growth for a long time, we are bound to consume a large amount of fossil energy, which leads to a continuous increase in the total amount of carbon dioxide emissions [1]. The accumulation of carbon dioxide emissions creates a series of environmental problems, such as global warming, sea level rise, and frequent outbreaks of extreme weather.

As a responsible power, the Chinese Government has realized the serious environmental problems arising from the large amount of carbon dioxide emissions. Since the 21st century, as the largest developing country, China’s rapid economic development has been accompanied by the problem of the “energy triangle”, i.e., economic development, energy security, and environmental protection. Energy technology innovation is an important way to realize energy savings and emission reduction [2]. The energy system is the basis for socio-economic development and an important way to influence carbon emissions [3]. China needs to further accelerate its energy transition in order to realize low-carbon green development. The driving factors of energy transition are broadly categorized into four main groups: economic factors, social factors, institutional factors, and technological factors, while the most basic and core driving factor is technological innovation [4]. Wu (2017) [5] in the green transformation of China’s industrial economy Laspeyres decomposition of the conclusion that energy technology reform is the main factor to promote China to achieve green economic growth. Meanwhile, the adjustment of the energy structure and new energy environmental effects play a weaker role than new energy technology innovation.

Energy technology innovation is an important way to realize energy conservation and emission reduction. Both the low-carbon and high-efficiency utilization of traditional fossil energy sources and the large-scale utilization of new energy sources at lower costs must rely heavily on technological innovation. For traditional fossil energy, technological innovation can improve energy efficiency and thus reduce energy consumption and CO₂ emissions in the production process, realizing energy savings and emission reduction. New energy sources are recognized as future energy sources due to their lack of carbon dioxide emissions [6]. Therefore, the mass use of new energy sources can improve energy security and mitigate climate change [7]. And the technological innovation of new energy enables the country to produce new energy output at a lower cost. In fact, the level of new energy technology innovation can effectively increase the capacity of new energy supply to meet energy demand and change the energy structure [8]. In order to address the tension between energy, the economy and the environment, the Chinese government has frequently used policy tools such as new energy tariff subsidies, technology funds and tax breaks to encourage the development of new energy technologies [9]. Accordingly, China’s new energy technology innovation level has been rising rapidly, and the number of patent applications has surged in the past three decades, which has become the core guarantee for China to fulfil its commitment as a great power on the global climate change issue [10].

Over the past three decades, research on the economics of climate change and energy has focused on identifying the key drivers of carbon emission growth and examining potential mechanisms for carbon emission reduction [11]. Currently, research on the environmental performance of specific green technologies, such as the development of new energy technologies, is on the rise [12]. Rosenbloom et al. [13] analyze that the driving force of the future energy transition will be carbon emission reduction. Hamamoto (2006) [14] and Zailani et al. [15] argued that unlike ordinary technological innovations, new energy technology innovations can simultaneously realize the economic and environmental benefits of enterprises. Nonetheless, two realities in recent years have forced us to re-examine the theoretical links between new energy technology innovation and carbon emissions. First, the environmental performance of new energy technology innovation may not be the same between regions with different income levels. China is a vast country, and with a GDP per capita of more than $20,000 in 2020, Beijing and Shanghai have ample funds to develop and utilize new energy technologies. However, four provinces, Gansu, Heilongjiang, Guangxi and Guizhou, have a GDP per capita of less than $7000, making economic development and improving people’s livelihoods their immediate priorities. Regions at different stages of economic development differ in their ability to develop and utilize new energy technology innovations, which leads to differentiated environmental performance impacts [16]. Second, the mechanisms by which new energy technology innovation affects environmental performance may also vary widely between regions with different income levels. Under policy intervention, the way in which rising carbon productivity drives environmental performance improvement in less developed regions may be insignificant or even counterproductive [17], and the mechanism of action may only be effective in some developed regions. Therefore, it is of theoretical significance and policy value to observe the impact of new energy technology innovation on carbon emissions from the perspective of regional development imbalance.

Inspired by existing studies, this paper focuses on analyzing the impact of new energy technology innovation on carbon emissions and the specific mechanisms under different income levels, using a sample of 30 provinces in China from 2000–2019. The marginal contribution of this paper to the existing literature is reflected in the following two aspects. First, it provides new evidence for understanding the heterogeneous impact of new energy technology innovation on carbon emissions. This paper uses a new empirical analysis strategy using a two-way fixed-effects model while constructing an interaction term for the level of economic development, which allows us to test the key hypothesis of whether the impact of new energy technology innovation on carbon emissions depends on the level of economic development. Earlier studies usually assumed that new energy technology innovation has a significant or insignificant homogeneous impact in all samples or in all samples within the same interval according to the threshold [18]. In contrast, this paper finds that new energy technology innovation in a given province has a significant impact on CO₂ emissions only in a given year, and the size of the marginal impact varies. Second, based on the perspective of regional development imbalance, this paper further analyses the specific impact mechanism of new energy technology innovation on carbon emissions at different stages of development. If new energy technology innovation has a differentiated impact on carbon emissions at different stages of development, it must stem from a differentiated impact mechanism. Based on the background of previous empirical studies, we creatively adopt the factor of regional development imbalance as a moderating variable to reveal through what specific channels new energy technology innovation affects carbon emissions under specific development conditions. The analytical strategy of this paper not only helps to understand the specific principles of the role of new energy technology innovation in mitigating climate change, but also provides targeted empirical conclusions for the formulation of regional climate and environmental policies.

The rest of this paper is organized as follows: a detailed literature review about climate change is provided in Section 2. Furthermore, a theoretical analysis which reveals the internal mechanism is presented in Section 3. The model specifications and data description are presented in Section 4. The empirical results are provided in Section 5. Then, we systematically discussed the innovation and limitations in our research compared with other studies in Section 6. Finally, the last section concludes this paper with some relevant policy implications.

2. Literature Review

With the increasing environmental climate, more scholars have shifted their attention to energy conservation and emission reduction. Studies on the relationship between technological innovation and carbon emissions have been gradually enriched. Research related to the economics of climate change and energy focuses on identifying key drivers of carbon emission growth and examining potential mechanisms for carbon emission reductions [1]. Since it is not possible to directly measure technological progress at the macro level, some literatures have tried to use the decomposition approach to study the impact of technological progress on carbon emissions. Referring to Chen [19], the impact of economic development on environmental pollutant emissions can be subdivided into three main effects: the scale effect, structural effect, and technological effect. In terms of the drivers of carbon emission growth, most studies based on the environmental Kuznets curve assumption, the STIPART modeling framework, the structural decomposition framework, and the factor decomposition method have affirmed the important driving role of economic scale expansion and population growth [20,21]. Boyce and He (2023) [22] quantified regional government policies and socio-economic development indicators in terms of social structure, emphasizing the importance of integrating local energy-specific policies into national carbon emission strategies. In terms of the drivers of carbon emission reduction, much literature based on scenario forecasting method and decomposition analysis affirm the future carbon emission reduction potential of technological progress [23,24]. Sun and Chen (2023) [25] argue that the direct impact of industrial structure upgrading and technological progress on regional CO₂ emissions is not clear, but the indirect impact by promoting energy structure transformation is both positive and obvious. But empirical studies based on economic theories have failed to obtain consistent research conclusions [26].

In new energy technology innovation, as an emerging branch in the field of technology innovation research, scholars have not yet reached a consensus on its categories, which are currently widely categorized as green technology (environmental technology, clean technology) and low-carbon technology [27,28]. However, how to go about defining the differences between the two has not yet been clearly delineated. The new energy technology innovation studied in this paper is the main component of carbon emission reduction technology in green technology innovation, also known as low-carbon technology breakthrough innovation, the specific scope of which includes solar energy, wind energy, ocean energy, biomass energy, nuclear energy, hydrogen energy, water energy, geothermal energy, chemical energy, and so on. Scholars take the new energy industry as the research object to explore the patent quality [29], development status [30], and performance evaluation [31] of new energy technology innovation. Currently, environmental performance research on the development of specific kinds of green technologies such as new energy technologies is emerging [32].

Reviewing the existing literature, most scholars affirmed the energy saving and emission reduction effects of new energy sources. Jordan et al. (2017) [33] emphasized the contribution of significant new energy technological innovations in the energy transition process. Through the research on the impact and mechanism of technological innovation, it is found that the improvement of the level of technological innovation realizes carbon emission reduction by improving the efficiency of energy use, adjusting the structure of energy consumption, and promoting the upgrading of local industries [34,35,36]. Wang et al. (2012) [37], based on data at the provincial level in China, found that innovations in new energy technologies can significantly reduce carbon emissions, and that this inhibitory effect gradually increases as the proportion of new energy use increases. For China, some studies have pointed out that the improvement of independent R&D capability and the introduction of foreign advanced technology can improve China’s carbon emission performance [38].

Some scholars have affirmed the environmental benefits of technological innovation. Conversely, there are a number of studies that question the environmental performance achieved by technological progress since the Industrial Revolution. It is argued that technological progress, which historically ignored environmental conservation and resource scarcity, has made a very limited contribution to carbon emission reduction targets [39]. Zhao et al. (2017) [40] pointed out that key technologies with carbon reduction potential must be vigorously developed in order to make the emission reduction effect of technological progress greater than the emission increase effect. Ou and Wang (2018) [41] point out the process of technology from R&D to application and promotion is affected by many external factors such as economy, policy, and market environment. For new energy technology innovation with a high cost, long R&D cycle, and high investment risk, it relies to a large extent on the joint guidance of the local economy, market and policy. A more liquid financial market [42], a standardized technology sharing and transfer environment [43], relatively strict environmental regulations, and complementary industrial support policies [44] can better support new energy technology innovations and their utilization, and in turn, achieve carbon emission reduction. This set of external conditions is often difficult to realize in economically underdeveloped regions. Yan et al. [45] pointed out that in terms of renewable energy technology innovation, the income of the region is an important factor that causes differences in the regional emission reduction effect, and that only when the income reaches a set level, renewable energy technology innovation can realize carbon emission reduction effect.

These studies have shown that there is significant regional and industry heterogeneity in the impact of technological innovation on carbon emission reduction [46]. At the same time, this heterogeneity is influenced by factors such as income, markets, environmental regulations, industrial structure, energy consumption structure, etc., leading to changes in the effect of technological emission reductions over time [47]. Upon reviewing existing literature, it becomes evident that there remains significant potential for the study of carbon reduction effects from innovations in renewable energy technologies. On one hand, most current studies have focused on industrial technological innovations, low-carbon technology innovations, and green technology innovations as their core explanatory variables to investigate their emission reduction performance. These variables encompass subsets of technologies with distinct characteristics and emission-reducing mechanisms. Consequently, the findings from these studies cannot be readily transferred to the carbon-reducing effects of renewable energy technology innovations, and they struggle to provide detailed and precise support for practical application. On the other hand, literature that does focus on renewable energy technology innovations primarily treats regions as their units of study, examining the mechanisms of their carbon-reducing effects and regional heterogeneities. These studies consider different industries as constituents of regional economies and analyze them holistically. Their conclusions cannot be simply applied to industries with distinct characteristics and fail to address the relationship between renewable energy technology innovations and carbon emissions from an industry heterogeneity perspective.

3. Theoretical Analysis

According to the theory of technology microhabitat [48], a technology will face many challenges in the economic, policy, and market environments from the process of research to the realization of marketable applications [49]. Most of the low-carbon technologies, mainly new energy technology innovations, exhibit characteristics such as high cost, high risk, long cycle, and low return. It is easier to fall into the “valley of death” of technological innovation in the process of gradually realizing the market-oriented conversion of technology and forming new industries, thus inhibiting the scale economic and environmental benefits of new energy technologies [50]. Based on the interpretation of the environmental Kuznets (EKC) curve and the energy rebound effect, the emission reduction effect brought about by technological advancement through energy efficiency improvement cannot fully offset the carbon emissions brought about by its promotion of economic growth, which ultimately leads to an increase rather than a decrease in carbon emissions.

A multitude of empirical studies indicate that the energy consumption effects of technological advancement possess a dual nature, encompassing both “energy-saving” and “energy rebound” effects. Ma and Stern (2008) [51] contend that technological advancements contribute to a decline in energy intensity and that the influence of technological progress on energy intensity varies across industries. Liu and Liu (2008) [52] believe that while technological progress can lead to an energy rebound effect, this rebound tends to show a diminishing trend over time. Conversely, the “Jevons Paradox” stands as a foundational concept for the energy rebound effect. Gardner and Joutz (1996) [53], based on data from the United States, discovered that technological advancements that lead to declines in energy prices result in a 1.1% increase in U.S. energy consumption due to the price effect. Zhou and Lin (2007) [54], utilizing post-reform China’s energy consumption data as their sample, found that the rebound effect for China’s energy consumption lies between 30% and 80%.

Existing domestic and international research on the relationship between new energy technology innovation and environmental performance mainly focuses on carbon dioxide emissions and green economic development. In terms of carbon dioxide emissions, the role of new energy technology innovation in reducing emissions has been recognized by most scholars. For example, Altintas and Kassouri [55] based on the sample data of European countries from 1985–2016 which contains energy R&D and carbon footprint. The results of linear and non-linear econometric modelling were analyzed in comparison with each other, and it was obtained that the level of technological innovation in European countries has a significant abatement effect on the carbon footprint. However, after dividing the types of energy technologies and geographic regions, the abatement effect of energy technologies has subtle changes. For example, Wang et al. (2012) [37] constructed a dynamic panel of energy technology patents and CO₂ emissions using inter-provincial panel data in China. Their study obtained that patents on fossil energy technologies have no significant effect on CO₂ emissions, while patents on carbon-free energy technologies have a significant inhibitory effect on CO₂ emissions, which is more significant in the eastern region. Berkhout (2000) [56] proposes a new definition from a microeconomics and macroeconomics perspective. When technological progress leads to an increase in energy efficiency, the energy consumption for producing the same unit of product will decrease. However, it cannot be ignored that producers will use more energy to substitute other means of production and therefore consume more energy.

In economically backward regions, the lack of investment in new energy technology research and development makes it very easy to fall into the “valley of death” where the capital chain breaks [57]. Moreover, due to the imperfections in the infrastructure of energy supply and consumption, the “Matthew effect” arises between the backwardness of economic development and the lack of innovation in new energy technologies. Ineffective investment in new energy technologies crowds out the original technological innovation activities of enterprises and seriously reduces the efficiency of enterprise resource allocation. This transformation not only fails to play the expected emission reduction benefits, but also harms the economic benefits of enterprises, which is not conducive to the realization of environmental benefits.

In regions lacking government policies for new energy development, the public sector is unable to support new energy technologies from development to market due to insufficient demand stimulation and insufficient funding. Enterprise new energy technology innovation is likely to be in a weak state, making it difficult to achieve commercialization and promotion [58]. With the continuous optimization of the policy environment, new energy technology innovation driven by demand and funding will realize the transformation from basic research to product application, and then gradually move to a new industrial ecology focusing on clean energy, thus achieving the dual goals of regional economic development and environmental improvement [59].

At the same time, the theory of biased technological progress suggests that the dynamic allocation of scarce R&D resources between different production sectors is a fundamental driver of structural change in the economy [60]. The rising level of innovation in new energy technology means that it is constantly “competing” for more scarce resources from other competitive fields. The combination of direct technology effects, indirect price effects and market scale effects gradually shapes the advantages of new energy technologies over traditional energy technologies. This will raise the relative output level of the new energy sector compared to the traditional high-carbon polluting sector, implying the promotion of a low-carbon and cleaner transformation of the industrial structure [61]. Bi Kexin et al. (2017) [62] found that breakthrough innovations in low-carbon technologies have a significant role in promoting industrial upgrading and help to realize the low-carbon transformation of the economy.

The rapid development of the theory of evolutionary economics has led many scholars to include institutional factors in the study of economic development (Aldieri et al., 2021) [63]. The rise of the institutional economics school, represented by Nelson, combines the theory of “innovation” with the theory of “institutions”. Currently, many scholars have integrated the theory of evolutionary economics into the energy field and believe that environmental regulatory policies are a major influence on the energy transition. Liu et al. (2018) [64] show that environmental regulation is beneficial in alleviating energy pressure. Pan et al. (2019) [65] found that environmental regulation can promote technological innovation, which in turn reduces energy intensity and improves energy efficiency. In contrast, Zhou et al. (2020) [66] show that this effect is characterized by an inverted “U” shape. Technological innovations stimulated by reasonable environmental regulations can accelerate the energy transition, but once this limit is exceeded, it can have a negative effect.

4. Model Construction

4.1. Variables and Data

4.1.1. Explained Variables

The accuracy of the evaluation of carbon emission intensity, as a measure of the effectiveness of carbon emission reduction, is crucial for empirical research. Carbon emission intensity is the ratio of carbon emissions to real GDP, representing the carbon emissions corresponding to each unit of economic output. Compared with the total carbon emissions or average carbon emissions, the carbon emission intensity indicator is more comparable to samples of different economic sizes, and it is logarithmic as an explanatory variable in this paper. This study accounts for China’s provincial carbon emissions with reference to the carbon emission measurement methodology provided by the IPCC.

$$C=\sum E_n·{\beta }_n·{\alpha }_n·\frac{12}{44}$$

(1)

In Equation (1), $C$ is the carbon emission. $E_n$ denotes the consumption of the nth energy source. ${\beta }_n$ is the carbon dioxide emission factor for the nth energy source, which can be obtained from the “IPCC Guidelines for National Greenhouse Gas Inventories 2006”. ${\alpha }_n$ is the standard coal conversion factor for the energy source. 12/44 is a carbon multiplication factor that indicates the molecular mass ratio of carbon to carbon dioxide.

4.1.2. Explanatory Variables

New energy technology innovation is mainly manifested in technological innovation in the development and utilization of non-fossil energy (such as wind energy, ocean energy, biomass energy, etc.) [67,68]. Therefore, this paper draws on Ye et al. (2018) [69] and Fan (2020) [20] to measure the level of new energy technology innovation through the number of patent applications for “new energy (clean energy and new energy)”, which includes solar energy, wind energy, ocean energy, biomass energy, nuclear energy, hydrogen energy, and other types of energy.

The data of new energy technology innovation comes from the public patent database under the search of “Shanghai Intellectual Property (Patent) Public Service Platform”. In this study, the scope of the search was positioned in “new energy” and “non-fossil energy” technologies, and the abstracts and keywords were set as “solar energy OR wind energy OR ocean energy OR biomass energy OR nuclear energy OR hydrogen energy OR hydroelectric energy OR geothermal energy OR chemical energy OR new energy OR new energy”, respectively. At the same time, the types of patents to be searched for are invention patents and utility model patents after excluding design patents.

4.1.3. Control Variables

Referring to Du et al. (2012) [70] and Lin and Zhu (2019) [10], the following control variables are considered in this paper. The level of urbanization (denoted as $urbanization$), which is measured by using the ratio of the regional year-end urban population to the resident population. Economic structure (denoted as $industrial$), which is measured by the output share of the secondary industry. Trade openness (denoted as $open$), measured using the ratio of total trade to GDP. Energy structure ($enestru$), measured by the share of regional electricity consumption in total national electricity consumption. The regional level of R&D ($research$), measured by the share of the regional internal expenditure on R&D funding in relation to GDP. The level of informatization ($information$), which is expressed by the share of total postal and telecommunication business to GDP. The mechanism variables used in the mechanism analysis are as follows: Industrial structure upgrading (denoted as $ind$) is measured by the output ratio of low-carbon industries to high-carbon industries. Among them, the division of high- and low-carbon industries is based on the median carbon emission intensity of each provincial subsector during the sample period. Environmental regulation ($enviro$), expressed as the share of completed investment in industrial pollution control in industrial value added. $urbanization$, $industrial$, $open$, $research$, $information$, and $enviro$ are obtained from the CSMAR database, and the original data for the variable $ind$ is obtained from Xu and Lin (2018) [71] and the China Energy Statistics Yearbook. The CO₂ emission data for each region are estimated with reference to Du et al. (2012) [70].

Table 1. Description of variables.

	Variables	Definition	Description
Explained variable	lnCarbon	Logarithm of carbon emission levels	Natural logarithm of carbon emission levels by province based on IPCC methodology
Explanatory variables	lnNewenergy	New Energy Technology Innovation	Natural logarithm of new energy patent applications
Control variables	urbanization	urbanization level (of a city)	Measurement of the ratio of urban population to resident population at the end of the year in the region
	industrial	industrialization level	Measuring the share of output in the secondary sector
	open	Trade openness	Measured by the ratio of total trade to GDP
	enestru	energy structure	the share of regional electricity consumption in total national electricity consumption
	research	R&D level	Regional R&D internal expenditures as a share of GDP.
	information	informatization level	Total post and telecommunications business as a percentage of GDP
Mediating variables	ind	industrial structure	Output ratio of low-carbon to high-carbon industries
	enviro	environmental regulation	Completed investment in industrial pollution control as a percentage of industrial added value

reports the specific characteristics of the variables and

Table 2. Descriptive statistics.

Variable	Obs	Mean	Std. Dev.	Min	Max
lnCarbon	600	5.230	0.988	−0.205	7.438
lnNewenergy	599	5.007	1.868	0.000	9.241
urbanization	600	0.503	0.157	0.139	0.896
industrial	600	0.347	0.085	0.113	0.559
open	600	0.304	0.374	0.013	1.721
enestru	600	0.033	0.023	0.003	0.110
information	600	0.058	0.032	0.014	0.236
ind	600	1.014	0.541	0.494	5.169
enviro	600	0.005	0.004	0.000	0.031

reports the descriptive statistics of the variables.

4.2. Benchmark Regression

In order to analyze the impact of new energy technology innovation on carbon dioxide emissions, this paper first considers the following benchmark model:

$$ln{Carbon}_{it}=\alpha +\beta ln{Newen}_{it}+{\gamma }^{\prime }{Z}_{it}+{\mu }_i+{\epsilon }_{it}$$

(2)

where α is a constant term. The subscripts $i$ and $t$ denote region i and year t, respectively. ${Carbon}_{it}$ is the carbon dioxide emissions of each province. ${lnNewen}_{it}$ is the key explanatory variable, which denotes the index of the progress of new energy technology. $Z_{it}$ is a vector set of control variables, which changes according to the characteristics of different industries. ${\mu }_i$ is the province fixed effect, and ${\epsilon }_{it}$ is the random error. The coefficient $\beta$ is used to describe the average (homogeneous) response of carbon emissions to changes in new energy technology innovation.

In order to test whether the impact of new energy technology innovation on carbon emissions depends on the level of economic development, this paper first adopts the following strategy:

$$ln{Carbon}_{it}=\alpha +\beta ln{Newen}_{it}+{\theta \left[ln{Newen}_{i,t-1}\times R_{it}\right]+\gamma }^{\prime }{Z}_{it}+{\mu }_i+{\epsilon }_{it}$$

(3)

where $R_{it}$ reflects the gap between the income level of province $i$ and the highest income level of all sample provinces. $R_{it}=\left[Y_{it}-\min \left(Y_{it}\right)\right]/[\max \left(Y_{it}\right)-\mathrm{m}\mathrm{i}\mathrm{n}(Y_{it}\left)\right]$, and $Y_{it}$ represents the per capita GDP. Model (3) adds an interaction term $ln{Newen}_{i,t-1}\times R_{it}$ to the baseline model, which assumes that the carbon abatement effect of new energy technology innovations is a linear function of the income gap $R_{it}$, i.e., $\frac{\partial ln{Carbon}_{it}}{\partial ln{Newen}_{i,t-1}}=\beta +\theta R_{it}$. If $\theta$ is significantly negative at a given level, it implies that the carbon reduction effect of new energy technology innovation rises with income growth.

$$ln{Carbon}_{it}=\alpha +\beta ln{Newen}_{it}+{\theta \left[ln{Newen}_{i,t-1}\times D_{it}\right]+\gamma }^{\prime }{Z}_{it}+{\mu }_i+{\epsilon }_{it}$$

(4)

where $D_{it}$ is a dummy variable, which is assigned 1 when GDP per capita is larger than the sample mean, and vice versa, it is assigned 0. Model III adds the interaction term of $D_{it}$ and $ln{Newen}_{i,t-1}$ on the basis of the baseline model. In Model (3), the marginal effect of $ln{Newen}_{i,t-1}$ can be expressed as $\beta$ and $\beta +\theta$ for low-income group and high-income group, respectively.

4.3. Mechanism Effect

Theoretically, new energy technology innovation may affect the CO₂ emission reduction situation by upgrading the industrial structure with environmental regulation. The new energy technology innovation system not only aims to promote the progress of energy technology, but also focuses its development on improving the affordability of new energy [72]. With the cost of new energy falling sharply, new energy technology innovation will promote new energy to traditional energy substitution, thus promoting the upgrading of the traditional industrial structure. And the government’s environmental regulatory policies (low-carbon technology subsidies, carbon emission standards, etc.) may stimulate enterprises to apply new energy low-carbon technologies and discourage the introduction of high-carbon technologies, which in turn will help promote the structural upgrading of the introduced technologies and drive them in the direction of decarbonization [73].

In summary, new energy technology innovation can influence its application in carbon emission through industrial structure decarbonization and government system. This paper tests the impact of new energy technology innovation on the two mechanism variables according to Models (5) and (6).

$${ind}_{it}=\alpha +{\lambda }_{1}ln{Newen}_{it}+{\gamma }^{\prime }{Z}_{it}+{\mu }_i+{\epsilon }_{it}$$

(5)

$${environ}_{it}=\alpha +{\lambda }_{2}ln{Newen}_{it}+{\gamma }^{\prime }{Z}_{it}+{\mu }_i+{\epsilon }_{it}$$

(6)

Among them, ${ind}_{it}$ t represents the change of industrial structure; ${environ}_{it}$ represents the environmental regulation level. ${\lambda }_1$ and ${\lambda }_2$ represent the effects of new energy technology innovation on industrial structure and the level of environmental regulation, respectively.

5. Results

5.1. Benchmark Regression Results

This paper uses panel data to estimate the impact of new energy innovation on carbon emissions. The results are shown in

Table 3. Baseline regression.

	(1)	(2)	(3)	(4)
	Pooled Regression	Model (2)	Model (3)	Model (4)
lnNewenergy	0.227 ***	0.254 ***	0.292 ***	0.263 ***
	[0.021]	[0.036]	[0.036]	[0.036]
RlnNewenergy			−0.177 ***
			[0.060]
DlnNewenergy				−0.060 ***
				[0.023]
urbanization	1.060 ***	0.607	0.923	0.749
	[0.312]	[0.681]	[0.690]	[0.672]
industrial	2.688 ***	0.871	0.851	1.152 **
	[0.368]	[0.547]	[0.513]	[0.485]
open	−1.071 ***	0.119	−0.079	−0.256 *
	[0.099]	[0.210]	[0.208]	[0.152]
enestru	16.969 ***	25.040 ***	27.262 ***	22.910 ***
	[1.451]	[7.592]	[7.072]	[5.638]
information	−0.775	−0.058	−0.428	−0.105
	[0.706]	[0.448]	[0.460]	[0.457]
_cons	2.439 ***	2.490 ***	2.420 ***	2.582 ***
	[0.209]	[0.437]	[0.417]	[0.407]
N	599	599	599	599
r2_o		0.491	0.609	0.622
r2	0.664	0.589	0.606

Standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

. Column (1) shows the estimation results of the mixed effects model, and Column (2)–Column (4) show the estimation results of Model (2) to Model (4) using the panel fixed effects model, respectively. From Columns (1) and (2), the coefficient of $ln{Newen}_{it}$ is positive, and the coefficient is still positive after adding the two-way fixed effect, which implies that the new energy technology innovation in general has a significant opposite effect on China’s carbon dioxide emission reduction. Compared with Model (2), Models (3) and (4) assume that new energy technology innovation affects carbon emissions in a more complex form, and the corresponding estimation results are shown in Columns (3) and (4). According to Model (2), an interaction term $ln{Newen}_{i,t-1}\times R_{it}$ is introduced, and the coefficient of this variable is estimated to be −0.13 and is significant at the 10% level, suggesting that a region’s level of economic development can inhibit the impact of innovations in new energy technologies on carbon emissions.

According to our findings, at the present time, new energy technology innovation promotes carbon dioxide emissions in China. This seems to be contrary to common sense and the reasons for developing new energy. This opposite effect may be related to the stage of economic development that China is currently in. The emission reduction effect brought about by technological progress through energy efficiency improvement cannot fully offset the carbon emissions brought about by its drive for economic growth, which ultimately leads to an increase rather than a decrease in carbon emissions. According to the results of the moderating effect, the level of economic development of a region will inhibit the promotion of new energy technology innovation on carbon emissions. This means that in a region with a high level of economic development, new energy technology innovation will further compensate for the carbon dioxide emissions brought about by economic growth.

In order to test the robustness of the initial scenario reflected in Column (3), this paper takes the interaction term $ln{Newen}_{i,t-1}\times D_{it}$ into account in Model (3). In Column (4), the estimated coefficient of $ln{Newen}_{i,t-1}\times D_{it}$ it is −0.04 and is significant at the 10% level, indicating that the new energy technology innovation can play a significant role in contributing to CO₂ emissions in the high-income group provinces. This result further supports the main conclusion of Column (2), which further supports the main conclusion of Column (3) and proves that the impact of new energy technology innovations on the level of carbon emissions is indeed related to the income level of the region.

5.2. Robustness Tests

5.2.1. Quantile Regression

There are large differences in China’s regional new energy innovation level, industrial structure, economic environment, and carbon emission level of various industries. At the same time, a large number of studies confirm the heterogeneity of regional and industry emission reduction effects. Therefore, this paper adopts the quantile regression model for robustness analysis to observe the emission reduction effect of new energy technology innovation under different carbon emission scales. The advantage of this method is that the regression relationship between independent variables and dependent variables at different quantile points can be studied separately. Meanwhile, quantile regression has no restriction on the overall distribution of the dependent variable, error distribution, and robust statistical properties for outliers.

The quantile regression results of the impact of new energy technology innovation on carbon emissions in heavy industry are shown in

Table 4. Quantile regression.

	Baseline Regression			Economic Factors
	25%	50%	75%	25%	50%	75%
lnNewenergy	0.264 ***	0.252 ***	0.252 ***	0.310 ***	0.289 ***	0.274 ***
	[0.028]	[0.016]	[0.016]	[0.030]	[0.018]	[0.019]
RlnNewenergy				−0.194 ***	−0.174 ***	−0.161 ***
				[0.043]	[0.025]	[0.027]
urbanization	0.809 **	0.569 **	0.569 **	1.053 ***	0.898 ***	0.791 ***
	[0.382]	[0.222]	[0.222]	[0.341]	[0.200]	[0.219]
industrial	1.291 **	0.792 **	0.792 **	1.307 **	0.767 **	0.391
	[0.615]	[0.358]	[0.358]	[0.565]	[0.333]	[0.362]
open	0.129	0.117	0.117	−0.163	−0.064	0.005
	[0.165]	[0.096]	[0.096]	[0.146]	[0.086]	[0.094]
enestru	27.075 ***	24.658 ***	24.658 ***	28.866 ***	26.965 ***	25.643 ***
	[4.638]	[2.697]	[2.697]	[4.215]	[2.473]	[2.708]
information	0.110	−0.090	−0.090	−0.277	−0.456	−0.581
	[0.894]	[0.519]	[0.519]	[0.854]	[0.501]	[0.549]
N	599	599	599	599	599	599

Standard errors in brackets. ** p < 0.05, *** p < 0.01.

, and there is no evidence that new energy technology innovation can reduce carbon emissions. After adding the consideration of economic factors, the effect of new energy technology innovation on promoting carbon emissions is more obvious. Overall, the regression results in Table 4 are roughly the same as those in Table 3, which can prove the robustness of the benchmark regression results.

5.2.2. Replacement of Explained Variables

To further test the robustness of the results, this paper retests Model (2)–Model (4) by replacing the explanatory variables. At present, the measurement of the level of carbon emissions can be expressed either in terms of total carbon emissions or in terms of per capita carbon dioxide emissions. Ma et al. (2015) [74] have used per capita carbon dioxide emissions as an energy savings and emission reduction indicator in their study. In this study, the natural logarithm of carbon dioxide emissions is replaced by per capita carbon dioxide emissions, and the regression results are shown in

Table 5. Replacement of explanatory variables.

	(1)	(2)	(3)	(4)
	Pooled Regression	Model (2)	Model (3)	Model (4)
lnNewenergy	−0.037 *	0.052 *	0.075 *	0.055 *
	[0.020]	[0.030]	[0.040]	[0.031]
RlnNewenergy			−0.105 *
			[0.055]
DlnNewenergy				−0.019
				[0.017]
urbanization	1.437 ***	−0.343	−0.157	−0.276
	[0.282]	[0.400]	[0.326]	[0.389]
industrial	1.128 ***	−1.858	−1.869	−1.816
	[0.429]	[1.406]	[1.343]	[1.361]
open	−0.434 ***	0.349	0.233	0.284
	[0.087]	[0.306]	[0.257]	[0.278]
enestru	−3.393 ***	−4.154	−2.845	−3.174
	[1.249]	[8.303]	[7.690]	[7.271]
information	0.091	−0.095	−0.313	−0.118
	[1.025]	[0.275]	[0.304]	[0.279]
_cons	−0.464 ***	0.817	0.776	0.784
	[0.169]	[0.728]	[0.684]	[0.789]
N	599	599	599	599
r2_o		0.001	0.011	0.004
r2	0.117	0.209	0.234

Standard errors in brackets. * p < 0.1, *** p < 0.01.

. From Table 5, the estimated coefficient of $ln{Newen}_{it}$ is always positive and significant at 10% level, which can prove the robustness of the results in Table 3. Meanwhile, from Column (3) of Table 5, the results of replacing the explanatory variables are still robust after considering the economic factors.

5.2.3. Lagging Treatment

Considering the time lag of new energy innovation affecting carbon emissions, this paper changes the core explanatory variable to the level of new energy innovation with one period lag in Column (1) of

Table 6. Lagging treatment.

	(1)	(2)	(3)
lnNewenergy	0.207 ***	0.217 ***	0.213 ***
	[0.039]	[0.031]	[0.032]
l_lnNewenergy	0.043	0.082 ***	0.058 **
	[0.030]	[0.025]	[0.026]
urbanization	0.517
	[0.587]
industrial	0.890
	[0.604]
open	0.161
	[0.199]
enestru	22.195 ***
	[7.135]
information	−0.400
	[0.392]
Rl_lnNewenergy		−0.162 ***
		[0.055]
Dl_lnNewenergy			−0.044 **
			[0.019]
l_urbanization		0.488	0.329
		[0.514]	[0.509]
l_industrial		0.588	0.532
		[0.472]	[0.493]
l_open		−0.072	0.010
		[0.150]	[0.162]
l_enestru		23.790 ***	24.555 ***
		[6.504]	[7.456]
l_information		1.401 **	1.481 **
		[0.639]	[0.641]
_cons	2.679 ***	2.724 ***	2.746 ***
	[0.409]	[0.323]	[0.349]
N	568	568	568
r2_o	0.479	0.607	0.546
r2	0.611	0.624	0.616

Standard errors in brackets. ** p < 0.05, *** p < 0.01.

. The estimation results show that the new energy innovation level in the lagged period still has a significant positive impact on the carbon emission level. Specifically, for every 1% increase in the debt ratio in the lagged period, the level of carbon emissions increases by 0.209%. After considering the income level of the regional economy, the estimation results are shown in Columns (2) and (3), and the estimation results in the lagged period are closer to the benchmark regression, indicating that the latter is robust.

5.3. Heterogeneity Analysis

5.3.1. Heterogeneity Test by Industry

In this paper, industry carbon emissions are used as the explanatory variables, and based on the characteristics of industry carbon emissions and combined with the classification of national economic industries (GB/T4754-2017 [75]), they are combined and divided into seven industries, including agriculture, forestry, animal husbandry and fishery, energy production-related industries, mining, manufacturing, construction, transportation and service industries. The carbon emission data of each industry were obtained from the China Carbon Accounting Databases (CEADs).

Table 7. Sub-industry heterogeneity test.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Agriculture, Forestry, Animal Husbandry and Fishery	Mining	Manufacturing	Energy Production-Related	Construction	Transportation	Service
lnNewenergy	0.163 ***	0.171 ***	0.217 ***	0.220 ***	0.291 ***	0.295 ***	0.163 ***
	[0.038]	[0.040]	[0.047]	[0.024]	[0.049]	[0.024]	[0.038]
urbanization	−0.761	0.778	0.364	0.248	−0.482	0.400	−0.761
	[0.473]	[0.701]	[0.612]	[0.375]	[0.464]	[0.314]	[0.473]
industrial	0.902	2.139 **	2.139 ***	0.545	1.461	0.006	0.902
	[0.978]	[0.940]	[0.742]	[0.519]	[1.240]	[0.651]	[0.978]
open	0.345	1.049	0.315	0.140	0.049	0.392 **	0.345
	[0.386]	[0.827]	[0.321]	[0.193]	[0.342]	[0.180]	[0.386]
enestru	7.216	7.400	21.778 **	30.234 ***	16.863	9.243	7.216
	[7.646]	[6.029]	[8.493]	[7.812]	[12.494]	[6.811]	[7.646]
information	−0.634	−2.102 ***	0.151	0.912 **	0.378	0.062	−0.634
	[0.613]	[0.609]	[0.504]	[0.360]	[1.309]	[0.547]	[0.613]
_cons	0.117	−0.699	1.098 *	2.316 ***	−2.108 ***	0.400	0.117
	[0.410]	[0.466]	[0.579]	[0.325]	[0.623]	[0.328]	[0.410]
N	595	579	596	596	596	596	596
r2_o	0.202	0.004	0.406	0.543	0.214	0.686	0.202
r2	0.237	0.310	0.530	0.730	0.364	0.752	0.237

Standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

reports the regression results of the impact of new energy technology innovation on the carbon emissions of each industry. Overall, new energy technology innovation has contributed to carbon emissions in various industries, with the greatest impact on the transportation industry and the least on the agriculture, forestry, and animal husbandry and fishery industries. This means that the transportation industry has a major impact on carbon dioxide emissions. Among the control variables, it can be observed that the energy mix plays an important role in the energy production sector. A higher level of electricity consumption contributes to a region’s CO₂ emissions. As the process of technological change accelerates, increased funding for research and development for polluting industries will lead to the emergence of high-tech equipment that will reduce regional CO₂ levels.

5.3.2. Heterogeneity Test by Urban and Rural Areas

Due to the limitations of different development situations in urban and rural areas, the scales of the industrial development and the industrial structure of different regions are different. Among them, the influencing factors that determine the ecologicalization of carbon emissions also have differences [76]. This paper will further be divided by urban and rural areas to compare the effect of new energy technology innovation in different regions. As can be seen from

Table 8. Heterogeneity test.

	(1)	(2)	(3)	(4)	(5)
	Eastern	Central	Western	Urban	Rural
lnNewenergy	0.279 ***	0.318 ***	−0.210	0.126 ***	0.124 **
	[0.042]	[0.039]	[0.317]	[0.037]	[0.050]
urbanization	0.371	−0.590	10.026	0.202	−0.219
	[0.560]	[0.744]	[7.070]	[0.307]	[0.603]
industrial	2.004 **	−0.153	6.060 *	−1.801 *	−1.428
	[0.796]	[0.687]	[2.615]	[0.902]	[0.987]
open	0.187	−1.383	−3.256	−0.003	−0.052
	[0.158]	[1.091]	[3.078]	[0.200]	[0.430]
enestru	14.757 *	23.844	27.174	−0.306	4.889
	[7.422]	[16.535]	[19.068]	[5.590]	[11.023]
information	−0.006	−0.564	−1.842	−0.796 **	−1.264 **
	[0.572]	[0.624]	[1.828]	[0.381]	[0.467]
_cons	1.945 **	3.748 ***	−0.423	1.685 ***	1.468 **
	[0.864]	[0.255]	[1.879]	[0.386]	[0.546]
N	220	220	159	599	599
r2_o	0.671	0.516	0.193	0.127	0.076
r2	0.660	0.721	0.623	0.344	0.236

Standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

, the new energy technology innovation all has no significant effect on the carbon emission level of urban and rural areas, in which the effect of rural areas is greater. At the same time, the level of informatization in a region will have a significant negative effect on carbon emissions.

5.3.3. Sub-Regional Heterogeneity Test

In order to explore the impact of new energy technology innovation on industry carbon emissions under different external environments, this paper divides China’s three major economic zones based on existing research. The 30 provinces in mainland China (Tibet is not included in the statistics due to incomplete data) are divided into three groups, eastern, central, and western regions, and the empirical test is carried out on the whole and each region, respectively. At the same time, with economic development and technological progress, the internal and external conditions for new energy technology innovation have changed, which means that the level of its carbon emissions may change. As can be seen from Table 8, the western region is mostly dominated by secondary industries, with more energy-consuming and highly polluting industries. Therefore, new energy technology innovation will inhibit carbon emissions, but the results in the table are not significant. In the eastern region, the level of economic development is higher. The pollution-suppressing effect of technological innovation can hardly compensate for the energy consumption and environmental damage caused by economic development. Thus, a positive facilitating effect is generated.

5.4. Mechanism Analysis

Low-carbonization adjustment of industrial structure is a potential channel for carbon emission reduction in the future. The co-growth of low-carbon and high-carbon industries has prevented industrial restructuring from playing its role as a channel for carbon emission reduction. In the future, as the level of low-carbon innovation in each province increases, low-carbon industries will become another engine to drive economic growth. New energy technologies will continue to promote the low-carbon adjustment of the industrial structure, and the potential of new energy industries to play a role in carbon emission reduction will be expected for a long time [77]. As shown in

Table 9. Mechanism analysis.

	Industry Structure			Environmental Regulation
	lnCarbon	Ind	lnCarbon	lnCarbon	Enviro	lnCarbon
lnNewenergy	0.254 ***	0.051 *	0.271 ***	0.254 ***	−0.001 ***	0.246 ***
	[0.036]	[0.027]	[0.033]	[0.036]	[0.000]	[0.041]
ind			−0.325 ***
			[0.098]
enviro						−10.796
						[17.846]
urbanization	0.607	−0.015	0.602	0.607	0.002	0.632
	[0.681]	[0.341]	[0.619]	[0.681]	[0.002]	[0.687]
industrial	0.871	−3.346 ***	−0.217	0.871	−0.002	0.848
	[0.547]	[0.508]	[0.665]	[0.547]	[0.003]	[0.542]
open	0.119	−0.430 **	−0.021	0.119	0.001	0.131
	[0.210]	[0.203]	[0.223]	[0.210]	[0.001]	[0.207]
enestru	25.040 ***	−7.479	22.610 ***	25.040 ***	0.092 **	26.035 ***
	[7.592]	[5.349]	[7.207]	[7.592]	[0.034]	[8.274]
information	−0.058	2.114 ***	0.629	−0.058	−0.014 **	−0.209
	[0.448]	[0.365]	[0.501]	[0.448]	[0.006]	[0.381]
_cons	2.490 ***	2.187 ***	3.201 ***	2.490 ***	0.005 ***	2.549 ***
	[0.437]	[0.344]	[0.429]	[0.437]	[0.002]	[0.398]
N	599	599	599	599	599	599
r2_o	0.491	0.158	0.560	0.491	0.002	0.482
r2	0.589	0.569	0.602	0.589	0.154	0.592

Standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

, new energy technology innovation will promote the development of low-carbon industries. However, due to the current insufficient development of low-carbon industries, the economies of scale effect of many industries has not yet been formed. This leads to an increase in the intensity of energy consumption brought about by the expansion of economic scale while also increasing the intensity of carbon emissions [25].

Whether environmental regulation can have a positive effect on global warming. To address this question, current research has centered on the “Green Paradox”, a hypothesis put forward by Sinn (2008) [78], which argues that environmental policies that reduce the demand for fossil fuels are ineffective in curbing climate warming. In this paper, we use the amount of investment completed in industrial pollution control as a proxy variable for environmental regulation. As can be seen from Table 9, the results of technological innovation of new energy will inhibit the amount of completed investment in industrial pollution control in a region, which represents the negative correlation between the results of technological innovation and industrial pollution. However, at the same time, environmental regulation presents a negative correlation to carbon emissions, but the results are not significant. The increase in the level of environmental regulation will make it face huge pollution control costs, crowding out the investment of enterprises in transforming production processes or green projects, which is not conducive to the completion of the transformation of enterprises. At the same time, some enterprises will choose to move away from regions with higher levels of environmental regulation and move to regions with lower environmental regulation. This reverse migration will lead to the deterioration of the industrial structure in the place of relocation, resulting in false upgrading in the place of relocation. The deterioration of the actual industrial structure will lead to an increase in energy consumption and the emission of pollutants, and an increase in the level of carbon emissions.

6. Discussion

In light of China’s advancements in new energy and its regional economic trajectory, this study employs a bi-directional fixed-effects methodology to discern the technological implications. By analyzing the ecological consequences of new energy technology innovation at the inter-provincial level, it has been ascertained that the progression of such innovations does not currently result in a concomitant reduction of carbon emissions. These observations endorse the theory of the carbon rebound effect, an extrapolation of the energy rebound effect. The outcomes of our research resonate with the findings presented by Ou and Wang (2018) [41]. This suggests that, at this juncture, leveraging new energy technology innovations and their application as a means to curtail carbon emissions remains an arduous challenge for China. Although the methodological basis of this study mirrors that of Wang et al. (2012) [37], the derived conclusions diverge substantially. The aforementioned scholars posited that technological innovations in new energy can lead to notable decrements in carbon emissions, a dampening effect that amplifies with an increasing share of new energy consumption.

Building upon empirical data, this research corroborates the environmental Kuznets curve (EKC) hypothesis. As delineated by Yu et al. (2019) [16], the variances in regional economic maturation influence the capabilities to cultivate and harness innovations in new energy technology. Regions in the nascent phase of economic ascension invariably witness a surge in natural resource consumption and an escalation in pollutant emissions due to the momentum of rapid growth, further exacerbating environmental strain [79]. Grounded in the findings of this research, it is posited that the degree of economic development exerts a pronounced moderating influence. Specifically, in regions marked by advanced economic maturation, innovations in new energy technologies serve as a counterbalancing mechanism, offsetting the CO₂ emissions engendered by economic proliferation.

Predominantly, extant scholarly works concerning new energy technology innovation deploy regions as their primary analytical unit. Notwithstanding, beyond purely economic considerations, the ramifications of new energy technology innovation may manifest divergently across regions, suggesting a region-specific heterogeneity [80]. Concurrently, Wang et al. (2012) [37] and Lin et al. (2019) [10] discerned that the influence of new energy technology innovation on carbon emissions is contingent upon the proportion of new energy utilization. Through a stratified analysis that distinguishes between sectors, urban versus rural demarcations, and geographical locales, it becomes evident that the transportation sector is the most significantly affected by new energy technology innovation. This observation underscores that China is currently navigating a pivotal transitional juncture in its transportation sector’s evolution. Moreover, the capability of new energy to counterbalance the pollution engendered by transportation remains suboptimal. Pertinently, the western precincts of China, predominantly characterized by secondary industries—notably energy-intensive and high-polluting sectors—suggest that new energy technology innovation could potentially exert a mitigating effect on carbon emissions.

Within our mechanistic analysis, the decarbonization of the industrial structure and environmental regulation are posited as intervening variables. A burgeoning body of academic literature contends that technological R&D can catalyze an industrial overhaul. Such a transformation is anticipated to diminish the prominence of sectors typified by intensive pollution and energy consumption, thereby augmenting investments in pristine machinery, advancing the R&D of green technologies, and ultimately accruing environmental dividends [81,82]. The empirical evidence garnered in this research reinforces this assertion. An ascendant trajectory of innovation in the realm of new energy expedites industry’s decarbonization. This industrial decarbonization invariably curtails a region’s carbon emissions.

Invoking the “Green Paradox” hypothesis as propounded by [77], there exists a presumption that environmental policy frameworks might be inept at dampening the appetite for fossil fuels as a stratagem against global warming. Our empirical assessment discerns a negative affinity between environmental edicts and carbon emissions, albeit the statistical significance remains tenuous. Paradoxically, innovation in new energy technology exhibits a persistent positive affiliation with carbon emissions. Our insights offer a more granulated elucidation of the perspectives presented by [83]. These scholars posit that policies designed to attenuate emissions inadvertently bolster energy provision and aggregate emissions. Our research amplifies this discourse, offering an enriched vantage point. Navigating the contemporary zeitgeist of the new energy renaissance, our study demystifies the intricate interplay of environmental regulations, positioning them within the continuum spanning technological innovation and carbon emissions.

7. Conclusions

7.1. Summary

As a growing body of literature focuses on the environmental performance of new energy technology innovations, a theoretically intuitive research topic gradually stands out: whether new energy technology innovations contribute to carbon dioxide (CO₂) emission reductions non-differently for regions with different income levels. Based on panel data from 30 provinces in China from 2000–2019, this paper combines the traditional linear approach and the two-way fixed-effects analytical framework to uncover new evidence on the regionally heterogeneous impact of new energy technology innovation on carbon emissions. The study in this paper enriches the existing research perspectives and is useful in clarifying the mechanism and effect of new energy technology innovation on carbon dioxide emissions reduction.

This paper finds that the impact of new energy technology innovation on carbon emissions depends on the income level of a region both in terms of significance and marginal scale. At present, the level of new energy technology innovation still struggles to cover the increase in energy consumption and pollution brought about by economic development and shows a positive promotional effect on carbon emissions. The higher the level of economic development of a region, the more obvious this promotional effect will be. At the same time, the size of the marginal impact of new energy technology innovation on different channels varies across regions. It is found that new energy technology innovation mainly affects carbon emission results through two channels, one is the level of decarbonization of the industrial structure of a region, and the other is the level of environmental regulation in a region. The research in this paper has learned that new energy technology innovation will promote the development of low-carbon industry. At present, the development level of low-carbon industry in China is low, and it is difficult to form a reduction effect on carbon emissions. At the same time, new energy technology innovation will inhibit the level of environmental regulation in the region. When the new energy continues to develop, the increase in the level of environmental regulation will make the local market participants face huge pollution control costs. At the initial stage, the region will appear to reject the environmental regulation, which will bring about an increase in the level of carbon emissions.

7.2. Policy Recommendations

The findings of this paper have the following policy implications. First, the carbon emission reduction capacities and potentials of different provinces should be fully assessed and taken into account when formulating and disaggregating provincial climate policy objectives. On the one hand, carbon emission reduction responsibilities should be determined based on historical carbon emission levels and existing carbon emission reduction technological capabilities (e.g., the level of development of new energy technologies). On the other hand, it should also take into account the trend of the carbon emission reduction potential of a particular province under economic development and set dynamic targets for carbon emission reduction responsibilities. Second, when formulating policies, different provinces should fully understand the stage of local innovation in new energy technologies, as well as the carbon emission reduction mechanisms that are significantly present at the current stage. Considering the scarcity of resources that will inevitably be faced during policy formulation and implementation, Chinese provincial governments should set their own policy priorities. Specifically, for provinces that do not have significant environmental regulation mechanisms, it is necessary to prioritize the promotion and adoption of new energy technologies and the replacement of traditional energy sources with new ones at this stage. For provinces with significant performance in both mechanisms, the one with the greater marginal effect should be chosen as the policy priority. Third, in order to realize the long-term goal of promoting carbon emission reduction through innovation in new energy technologies, the channels through which different provinces have the potential for sustained carbon emission reductions in the future should be assessed during the policy formulation process.

This study has several limitations. First, due to data availability, the scope of data in this paper is up to 2019. As the level of development of new energy technologies in China continues to increase and the importance of the environment grows, new developments beyond 2019 could be further assessed in the future, subject to data availability, to obtain new evidence of more policy-guiding value. In addition, the impact of new energy technology innovation on carbon emissions may have regional spillover effects. This means that new energy technology innovation in one region may affect the environmental performance of other regions through technological spillovers and industrial chain linkages, which is an important direction for further expansion of this study. Another important question derived from the conclusions of this paper is whether the incentive policies for new energy technology innovation will also depend on the level of regional economic development, since the carbon emission reduction effect of new energy technology innovation is constrained by the level of regional development. The study of this question will help to further explore how to optimize the policies related to new energy technology innovation, which is a work worth exploring in depth.

7.3. Limitations

This study has several limitations. Firstly, the present investigation primarily centers on the relationship between new energy technology innovation and carbon emissions in the context of regional developmental stages. Nevertheless, the dynamic interplay of this association in the long run, especially as economies progress, remains an avenue meriting further scholarly attention in subsequent studies. Secondly, our findings intimate that the prevailing innovative strides in the new energy domain seem inadequate in substantially curtailing carbon emissions or counteracting the environmental repercussions of economic advancement. In forthcoming research endeavors, Penguin aims to delineate strategies to bridge this ecological disparity, particularly during the transitional junctures of economic evolution. However, a salient challenge is the multifaceted nature of indicators associated with new energy development, marked by their heterogeneity and the inherent complexities in data acquisition. We are committed to addressing and surmounting these challenges in our future research pursuits.