Next Article in Journal
Setting and Testing of the Out-of-Step Protection at Mongolian Transmission System
Next Article in Special Issue
Carbon Emission Efficiency and Low-Carbon Optimization in Shanxi Province under “Dual Carbon” Background
Previous Article in Journal
Flexible Films as Anode Materials Based on rGO and TiO2/MnO2 in Li-Ion Batteries Free of Non-Active Agents
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

The Influencing Effects of Industrial Eco-Efficiency on Carbon Emissions in the Yangtze River Delta

1
Research Institute of Central Jiangsu Development, Yangzhou University, Yangzhou 225009, China
2
Key Laboratory of Regional Sustainable Development Modeling, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
*
Author to whom correspondence should be addressed.
Submission received: 13 November 2021 / Revised: 26 November 2021 / Accepted: 30 November 2021 / Published: 6 December 2021

Abstract

:
A low-carbon economy is the most important requirement to realize high-quality integrated development of the Yangtze River Delta. Utilizing the following models: a super-efficiency slacks-based measure model, a spatio-temporal correlation model, a bivariate LISA model, a spatial econometric model, and a geographically weighted random forest model, this study measured urban industrial eco-efficiency (IEE) and then analyzed its influencing effects on carbon emission in the Yangtze River Delta from 2000 to 2017. The influencing factors included spatio-temporal correlation intensity, spatio-temporal association type, direct and indirect impacts, and local importance impacts. Findings showed that: (1) The temporal correlation intensity between IEE and scale efficiency (SE) and carbon emissions exhibited an inverted V-shaped variation trend, while the temporal correlation intensity between pure technical efficiency (PTE) and carbon emissions exhibited a W-shaped fluctuation trend. The negative spatial correlation between IEE and carbon emissions was mainly distributed in the developed cities of the delta, while the positive correlation was mainly distributed in central Anhui Province and Yancheng and Taizhou cities. The spatial correlation between PTE and carbon emissions exhibited a spatial pattern of being higher in the central part of the delta and lower in the northern and southern parts. The negative spatial correlation between SE and carbon emissions was mainly clustered in Zhejiang Province and scattered in Jiangsu and Anhui provinces, with the cities with positive correlations being concentrated around two locations: the junction of Anhui and Jiangsu provinces, and within central Jiangsu Province. (2) The direct and indirect effects of IEE on carbon emissions were significantly negative, indicating that IEE contributed to reducing carbon emissions. The direct impact of PTE on carbon emissions was also significantly negative, while its indirect effect was insignificant. Both the direct and indirect effects of SE on carbon emissions were significantly negative. (3) It was found that the positive effect of IEE was more likely to alleviate the increase in carbon emissions in northern Anhui City. Further, PTE was more conducive to reducing the increase in carbon emissions in northwestern Anhui City, southern Zhejiang City, and in other cities including Changzhou and Wuxi. Finally, it was found that SE played a relatively important role in reducing the increase in carbon emissions only in four cities: Changzhou, Suqian, Lu’an, and Wenzhou.

1. Introduction

The rapid economic development of China, which has been characterized by industrialization-based urbanization, consumes large amounts of fossil energy, making China the main global CO2 emitter [1,2,3]; thus, the country suffers severe pressure from the international community to reduce carbon emissions. Since China’s rapid industrialization is still developing, its high proportion of secondary industries, characterized by high pollution, high emissions, and high energy consumption, contributes the majority of carbon emissions [4,5]. Meanwhile, China’s traditional industrial structure has been identified as the main reason for the overall increase in carbon emissions [6]. In the context of Chinese efforts to construct an ecological civilization and China’s green commitment to building a cleaner and more beautiful world, green, circular, and low-carbon development measures have become inherently important strategies to promote high-quality and sustainable development. These are considered effective, practical measures to achieve peak carbon emission goals by 2030 and carbon neutrality by 2060. As such, the continuous adjustment, optimization, and transformation of industrial structure should be oriented towards green, ecological, and environment-friendly approaches to reduce industrial pollutant emissions and improve industrial eco-efficiency (IEE) [7,8]. This provides an adequate rationale for considering the impacts of IEE.
As the region with the fastest growing and most developed urbanization and industry in China, heavy industry and related economic structures predominate in some areas of the Yangtze River Delta, resulting in a continuous increase in energy consumption and large scale carbon emissions. With the aim of improving green and high-quality integrated development of the Yangtze River Delta, the gradual decoupling of carbon emissions from economic growth has become an important issue. Hence, this study analyzed the spatio-temporal correlation characteristics between carbon emissions and IEE, as well as the effects of IEE on carbon emissions, for cities located in the Yangtze River Delta.
The paper is structured as follows: In Section 2, the literature related to the subject of our research is introduced. Section 3 introduces the methods and data sources, including the multidimensional evaluation index of IEE and the spatial econometric methods used in analyzing the effect of IEE on carbon emissions. Section 4 presents the results, including panel gray correlation characteristics, spatio-temporal association patterns, spatial effects, and local importance analysis of the effect of IEE. Finally, Section 5 provides the conclusions and recommendations of the study.

2. Literature Review

General eco-efficiency is defined as the effectiveness of producing maximum economic outputs together with minimizing natural resource and environmental degradation [9,10]. Due to the flexibility and generality of the concept, eco-efficiency has become a popular quantitative indicator of sustainable development and has been widely applied to various areas of environmental and ecological impact assessment, such as products [11,12], sectors [13,14], and industries [15,16,17]. Following the core idea of eco-efficiency [18,19,20], IEE has the dual goals of industrial economic growth and industrial carbon emissions reduction, effectively uniting the material exchange and energy conversion of industrial ecosystems. Further, IEE also couples the subsystems of industrial economy, energy consumption, and environment impact, and accommodates industrial development and resource and environmental protection [21,22,23]. For these reasons, IEE has become an effective indicator to measure the green and ecological levels of industrial development and has been used as a representative index to evaluate the coordination between industrial development and protection of the environment. Therefore, improving IEE may be a more feasible approach for reduction in carbon emissions compared with other carbon reduction tactics.
The measurement methods associated with eco-efficiency have gradually developed from a single ratio model into complex simulation models, including the ecological footprint model, life cycle accounting, stochastic frontier analysis, the environmentally sustainable value model, and the environmentally extended input-output model [13,24,25,26]. Notably, due to its advantages of requiring neither a specific functional form nor explicit weights to aggregate the indicators [10,27], the nonparametric data envelopment analysis (DEA) technique has become the most commonly used method to handle the multiple inputs and outputs of efficiency measurement, both flexibly and effectively [28]. To better adapt to the complexity and multi-dimensionality of regional ecosystems, the original DEA model has been further improved to produce the super-efficiency slacks-based measure (Super-SBM) model, which deals with the undesirable outputs and indistinguishable efficiency ranks of decision-making units (DMUs) [29,30]. By incorporating undesirable outputs, the Super-SBM model is more applicable to evaluating IEE that considers not only the environmental performance of industrial production, but also embodies industrial productivity under environmental constraints [19,24,31,32].
Existing studies have theoretically and empirically analyzed the impacts of multiple socio-economic factors including economic growth [33,34], foreign direct investment [35,36], and trade [37,38], as well as policy factors, including policies on local environmental expenditures [39], environmental regulation [40,41], and carbon tax [42], etc. Other targets of analysis have included population factors such as size [43,44] and density [45,46], as well as urbanization [47,48], and technical factors, such as research and development investment [5,49], and green technology innovation [50,51,52] applied to carbon performance in China. Meanwhile, in terms of the effects of industry, several factors have been identified, such as the rise in the proportion of secondary industries, significantly promoting carbon intensity [53], industrial agglomeration contributing to reducing industrial carbon intensity through economy of scale, and technology spillover [54], and the upgrading and optimization of industrial structure to promote energy conservation and carbon reduction [55,56,57]. It has also been shown that industrial green transformation and adjustment has significantly reduced carbon intensity [58]. This wide basis of research provides multi-mechanistic and multi-dimensional indicator selection criteria with which to analyze the influencing factors of carbon intensity. Based on the STIRPAT (stochastic impacts by regression on population, affluence, and technology) model and the availability of data, six control variables were selected to detect the direction and degree of their impact on carbon emissions, as described below.
There is a bidirectional relationship between economic growth and carbon emissions. Carbon emissions are considered to be the consequence of production processes and economic activity [3], in that increase (or decrease) of economic growth will lead to increase (or decrease) in carbon emissions. However, the magnitude of the effect of economic growth on carbon emissions may be differentiated at different levels of the economy. In addition, the negative externalities generated by carbon emissions will lower economic growth performance [59]. Therefore, per capita gross domestic product (PGDP) was selected as an indicator to characterize economic development level.
Considering another aspect of the economy, the local fiscal expenditure on environmental protection, education, science, and technology can play a role in improving environmental quality, enhancing residents’ awareness of environmental protection, and promoting the progress of energy-saving technology, thereby reducing carbon emissions. However, it can be anticipated that differences in social and economic conditions in different regions will impair the carbon emissions reduction effect of fiscal expenditure across the regions [39]. Therefore, the proportion of local fiscal expenditure in GDP was selected as an indicator to reflect the impact of fiscal expenditure on carbon emissions (denoted as FIN).
In terms of foreign economic factors, the impact of foreign investment on regional environmental conditions can be generally divided into two categories. The first is based on the “pollution heaven” hypothesis, which holds that developed countries transfer heavy polluting industries and high carbon industries to developing countries under a low environmental regulation threshold [3]. The second is the proposed “pollution halo” hypothesis, which holds that when foreign-funded enterprises in developed countries move to host countries, they also bring with them greener and cleaner production technologies that improve local production and environmental protection levels, thus helping to reduce carbon emissions [3]. Therefore, the amount of completed foreign direct investment in each city (FDI) was used as an indicator to represent the status of foreign direct investment, while the average annual price of the renminbi (RMB) exchange rate over a period of years was converted to RMB as a unit, and the GDP index was adjusted to offset the effect of price changes.
It is known that as urbanization accelerates, many economic activities tend to aggregate, while the construction of large-scale infrastructure emerges and energy consumption increases, which improves urban carbon intensity. Meanwhile, urbanization brings economies of scale and for guiding green energy consumption and building ecological towns, urbanization effectively contributes to carbon emissions reduction [60,61]. Therefore, urbanization (URB) was selected as an important indicator and was derived using the proportion of urban population to permanent population.
In terms of the effect of environmental regulation on carbon emissions, there are two opposite views. On the one hand, the government promotes the effect of “forced emission reduction” through a series of command-and-control environmental regulations, which shuts down enterprises with high pollution and high energy consumption or forces them to reduce production scale or to use low-carbon technology. Forced emission reduction is also enacted through a series of market-motivated environmental regulations, which encourage enterprises to use environmental protection technologies and raises energy costs of enterprises by imposing emission taxes, environmental protection taxes, pollution treatment subsidies, etc., thus reducing energy consumption intensity and carbon emissions to a certain extent [62,63]. On the other hand, according to the “green paradox” theory, proposed by Sinn [64], measures aiming to reduce carbon emissions that tax fossil fuel consumption to develop renewable energy have been found to reduce global demand for fossil fuels and undermine the wealth maximization of resource owners. When it is anticipated that such measures could hurt the future price of their resources, resource owners accelerate the extraction of their resources, which results in higher rates of fossil fuel extraction in the short term and increases carbon emissions in the short term. Therefore, this study used the ratio of total environmental governance investment to regional GDP to characterize environmental regulation (denoted as ENV).
In terms of technology, technological progress has positive and negative effects on carbon emissions through the development of production and emission reduction technologies. Production technology progress improves the productivity of factors and energy efficiency, while emission reduction technology progress promotes the reduction in emission intensity [65,66]. However, due to the energy rebound effect and the effects of an extensive economic growth mode, technological progress has been found to stimulate economic activity and increase energy consumption, in turn offsetting energy savings achieved by improved efficiency [67,68,69]. Therefore, the number of patent applications granted was selected to characterize technological progress (denoted as PAT).
As the first law of geography states that everything is related to everything else, and near things are more related than distant things [70], it can be said that spatial relationships are highly consequential in socio-economic and resource environment analyses [71,72]. It has been confirmed that spatial dependence effects exist in terms of considering regional eco-efficiency [73] and carbon emissions performance [74]. Accordingly, spatial autocorrelation phenomena can be judged by using exploratory spatial data analysis, while specific causal relationships can be explained using the spatial econometric model [75,76]. Given the obvious core-periphery characteristics of its socio-economic development, the socio-economic phenomena of the Yangtze River Delta urban agglomeration show not only temporal correlation but also spatial correlation to some extent. Hence, it is reasonable to use the bivariate LISA [77,78] to reveal the spatial heterogeneous relationships between local carbon emissions and nearby IEE, as well as to use the spatial econometric model to estimate the effects of IEE on carbon emissions.
In summary, it can be stated that many related aspects in the measurement of carbon emissions, the factors influencing carbon emissions, and the connotation and measurement of IEE have been studied. However, deficiencies in our understanding of the topics remain. First, although related studies have analyzed multiple effects of industry on carbon emissions, almost no attention has been paid to the impacts of IEE, which not only integrates green processes and environmental protections but also embodies the synthesis effect of industrial scale and technological progress. As such, it is unknown whether the improvement in IEE and its decomposition terms promote or impede carbon emissions. Second, few studies have explored the spatio-temporal association effects between IEE and carbon emissions among different cities, such as the spatial spillover effect of IEE on carbon emissions from adjacent cities, or the spatio-temporal heterogeneity effect of IEE on carbon emissions among different cities. These questions deserve close attention, since the answers are not only important for helping China to reduce carbon emissions, but also for providing a reference for industrializing regions.

3. Materials and Methods

3.1. Super-SBM Model

The Super-SBM model has the advantages of avoiding radial and oriented deviation, incorporating undesirable outputs into the efficiency evaluation model, and discriminating between the efficient DMUs equal to 1 [79,80,81,82]. Hence, the Super-SBM model, with undesirable outputs, better represents the nature of regional IEE evaluation.
Now we consider the industrial production process with n DUMs which uses m input factors to produce s1 desirable outputs and s2 undesirable outputs. The three vectors are respectively expressed as: x R m ,         y d R s 1 ,         y u R s 2 among which x , y d , and y u d represents the inputs, desirable outputs, and undesirable outputs. The matrices of Y d > 0 , Y u d > 0 are defined as X = [ x 1 , , x n ] R m × n , Y d = [ y 1 d , , y n d ] R s 1 × n , and Y u d = [ y 1 u , , y n u ] R s 2 × n . Assume that the DMUk ( x k , y k d , y k u d   ) is slacks-based-measure-efficient. Then, the Super-SBM with undesirable outputs can be defined as:
β S E = min 1 + 1 m i = 1 m s i x i k 1 1 s 1 + s 2 ( r = 1 s 1 s r d y r k d + t = 1 s 2 s t u d y t k u d )
s . t .     x i k j = 1 , j k n x i j λ j s i y r k d j = 1 , j k n y r j d λ j + y r d y t k u d j = 1 , j k n y t j u d λ j + y t u d 1 1 s 1 + s 2 ( r = 1 s 1 y r d y r k d + t = 1 s 2 y t u d y t k u d ) > 0 λ ,     s ,     s + 0   i = 1 ,     2 , m ;     j = 1 ,     2 , n ( j k ) ;       r = 1 ,     2 , s 1 ;     t = 1 ,     2 , s 2
where β S E is the objective function whose value can be more than 1, λ j is a weighting factor, the subscript k means a DMU whose efficiency is being estimated, and the vectors s , s d , and s u d denote the slacks of inputs, desirable outputs, and undesirable outputs, respectively.
The above Super-SBM model is based on the assumption of constant returns-to-scale, thus by adding the restrictions j = 1 , k n w j = 1 , it can convert a variable return to scale. Further, the overall IEE factor can be decomposed into pure technical efficiency (PTE) and scale efficiency (SE) [83]. The factor PTE refers to the efficiency of resource allocation, utilization, pollution control, management, and production technology. The factor SE refers to the efficiency of the scale of resources, in terms of measuring whether the DMUs are the optimal production scale [84]. Using MaxDEA Ultra software, the IEE, PTE, and SE of 41 cities in the Yangtze River Delta urban agglomeration from 2000 to 2017 were obtained.

3.2. Panel Gray Correlation Model

The gray correlation model is a mature and effective tool to deal with uncertainty systems that have only “partial information known” and which are influenced by multiple factors [85,86]. However, being restricted by either the single object dimension or time dimension, the traditional model cannot overcome the defects of inconsistent gray relational order caused by the change of object ranking in the panel data. To better consider the differences of object and time dimensions of the panel data, Dang et al. [87] put forward the panel gray correlation model, which introduces incremental differences to investigate the development characteristics of different indicators in each period from the time dimension. The new model also utilizes deviation differences to measure the distribution characteristics of different indicators in each object from the object dimension and extracts the relative differences of development degrees and the direction of different indicators in both object and time dimensions in order to judge the extent of positive and negative correlation. Due to the complexity of the derivation process needed to calculate the incidence coefficient, the specific formula expression is omitted from this paper and can be referred to in Dang et al. [87].

3.3. Bivariate LISA Model

The bivariate LISA statistic is defined as [72,78]:
B L I S A = N i = 1 N j i N w i j z i ce z j i e e / i = 1 N j i N w i j
where N is the total number of cities; w i j is spatial weight matrix based on queen contiguity; z j c e is the z-score standardization of carbon emissions in city i ; and z j i e e is the z-score standardization of industrial eco-efficiency in city j . The heterogeneous clusters from the local bivariate LISA can be divided into four types, namely, HiHj, LiLj, LiHj, and HiLj. If the city is identified as Highcarbon emission-HighIee, it means that its carbon emissions are significantly higher than the average value of all cities, and the industrial eco-efficiency of adjacent cities is significantly higher than the average value of all cities. This means that the city has a higher level of carbon emissions and is surrounded by cities with higher industrial eco-efficiency.

3.4. Spatial Econometric Model

The general form of the spatial econometric model is known as the spatial Durbin model (SDM), which is hereafter used in the text, and can be defined as follows [72,75]:
y i t = λ j = 1 n w i j y j t + β X j t + j = 1 n w i j X i j t γ + c i + a t + ε i t ,               ε i t ~ i . i . d ( 0 , δ 2 )
where y i t is the carbon emission for city i at year t ; n is the total number of cities; w i j y j t is the spatial lag of carbon emission; X i t is the independent variables; β is the corresponding coefficients for X i t ; λ is the spatial spillover of adjacent regional carbon emission; is the corresponding spatial lag coefficient of the adjacent regional independent variables; is an independently and identically distributed error term for i and t with zero mean and a variance of σ 2 ; and c i and α t denote the spatial and temporal effects, respectively. Before the model is estimated, the hypotheses H0: γ = 0 and H0: γ + λβ = 0 should be performed to test whether SDM can be simplified to the spatial lag model (SLM) and the spatial error model (SEM), respectively. These tests can adopt the form of a Wald or LR test.
For estimating the outcome of the SDM, note that the estimated coefficient of the explanatory variables has no significance; however, the direct and indirect effects of the explanatory variables should be fully understood as follows: The direct effect refers to the influence of the change of independent variables in adjacent regions on the local area, while the indirect effect refers to the influence of the change of independent variables in a certain region on its neighboring areas [74].

3.5. Data Sources

Different from the conventional econometric model, which uses remote sensing light to estimate carbon emissions, Chen et al. [88] adopted variant coefficient models and the normalized difference index to estimate the carbon emissions of 334 prefecture-level cities in China from 1992 to 2017. Their method was based on more accurate and easier-to-understand techniques of inter-calibration of the DMSP/OLS (Defense Meteorological Program Operational Line-Scan System) and NPP/VIIRS (National Polar-orbiting Partnership Visible Infrared Imaging Radiometer Suite) datasets, and was found to be effective for describing long-term, full-coverage CO2 data at the scale of prefecture-level cities.
In this study, based on a related IEE evaluation system and data availability, four indicators were selected to characterize industrial land, labor, electricity, and water consumption. Industrial added value was selected as the expected output indicator, while industrial waste-water discharge, exhaust emissions, and smoke and dust emissions were the undesirable output indicators. The evaluation system components for input-output and variable description are listed in Table 1.
All datasets used in determining the indicators of IEE and related independent variables were obtained from the Jiangu Statistical Yearbook (2001–2018), Anhui Statistical Yearbook (2001–2018), Zhejiang Statistical Yearbook (2001–2018), China Statistical Yearbook for Regional Economy (2001–2014), the Statistical Bulletins of National Economic and Social Development for each city, and the statistical websites of certain provinces and cities compiled by the government during 2001–2018. According to the administrative division in 2020, the data of Chaohu City before 2011 were incorporated into Hefei, Ma’anshan, and Wuhu cities. The final determined number of cities for this study was 41. In order to account for inflation, relevant economic data such as the industrial added value, GDP, and financial expenditure were converted based on the year 2000.

4. Results

4.1. Panel Gray Correlation between IEE, PTE, SE, and Carbon Emissions

Before calculating the correlation coefficients between IEE, PTE, SE, and carbon emissions, the logarithmic operator was used to preprocess indicators of each city, then both the spatial and temporal correlation coefficients were obtained, and are visualized in Figure 1 and Figure 2, respectively. As can be seen from Figure 1, the temporal correlation intensity between IEE and carbon emissions exhibited an inverted V-shaped fluctuation trend, and the correlation coefficient increased continuously from 0.008 in 2000 to 0.190 in 2008, then continuously dropped, reaching 0.026 in 2017, with an annual average value of 0.043. This implies that IEE exerted first an enhanced, and then weakened, positive impact on carbon emissions, but that the impact was relatively low. The temporal correlation intensity between PTE and carbon emissions was generally below zero with an annual average value of −0.051 and exhibited a W-shaped fluctuation trend. This illustrates that the improvement of PTE could effectively promote the reduction in carbon emissions. The temporal correlation coefficient between SE and carbon emissions was generally similar to the trend between IEE and carbon emissions and exhibited an inverted V-shaped fluctuation trend with an annual average value of 0.088, implying that the improvement in IEE was dependent on SE; meanwhile, SE was positively correlated with increase in carbon emissions.
The spatial correlation coefficient was divided into four intervals: below 0, and three equal intervals from 0 to the maximum value of 0.503, 0.630, and 0.496, respectively. It can be seen from Figure 2 that the average spatial correlation intensity between IEE and carbon emissions was 0.104. The negative correlation cities accounted for 31.71% and were mainly clustered in southern Zhejiang Province and southern Jiangsu Province. These are developed cities that followed green industrial production and paid more attention to ecological environmental protection, which contributed to reducing carbon emissions. While the positive correlation cities accounted for 68.29%, with the high correlation intensity cities mainly being distributed in central Anhui Province, and Yancheng and Taizhou cities. In order to speed up industrial development, these cities developed large numbers of labor intensive and polluting low-end industries, which brought environmental damage and increased carbon emissions. The average spatial correlation intensity between PTE and carbon emissions was 0.132, and the spatial pattern was higher in the central part of the delta and lower in the northern and southern parts. The negative correlation cities accounted for 24.39%, while the positive correlation cities accounted for 68.29%, implying that most of the urban industrial enterprises were lagging behind in the research, development, and use of green production and pollution control technologies, resulting in an increase in carbon emissions. The average spatial correlation intensity between SE and carbon emissions was −0.024. The negative correlation cities accounted for 48.78% and were mainly clustered in Zhejiang Province and scattered in Jiangsu and Anhui provinces. The positive correlation cities accounted for 51.22% and were mainly concentrated at the junction of Anhui and Jiangsu provinces and within central Jiangsu Province, implying that the improvement of SE effectively promoted carbon emissions reduction.

4.2. Spatio-Temporal Association Pattern between IEE and Carbon Emissions

Although the panel gray correlation revealed the spatio-temporal correlation between IEE and carbon emissions, it ignored the underlying spatio-temporal association type between IEE and carbon emissions. Hence, the bivariate LISA spatio-temporal clustering was used to identify the association type, in order to investigate their local features more clearly, both significant and non-significant association types in 2000, 2008, and 2017 were visualized, as shown in Figure 3.
In 2000, the Lcarbon emissionLIEE type was mainly clustered in the northern cities of Anhui, and scattered in Nanjing, Suzhou, Suqian, Nantong and Jiaxing city, which had less industry-intensive economies with long-term economic development being characterized by low energy consumption and low emissions. In 2008, the spatial scope of Lcarbon emissionLIEE-type cities further expanded with Wuxi, Lianyungang, Wenzhou, Jinhua, Hefei, and Tongling city joining the Lcarbon emissionLIEE type. In 2017, the spatial pattern of Lcarbon emissionLIEE-type cities was almost identical to that of 2008, with some minor adjustments. For instance, Lianyungang, Suzhou, Nantong, and Huaibei city retreated from the Lcarbon emissionLIEE type, and Wuhu city became an Lcarbon emissionLIEE type.
In 2000, there were six cities belonging to the Hcarbon emissionHIEE type, including Shanghai, Xuzhou, Huai’an, Zhoushan, Yangzhou, and Lishui city, with the high carbon emissions surrounded by high IEE cities, which was due to rapid urbanization and economic growth that promoted carbon emissions. In 2009, except for removal of Yangzhou from the Hcarbon emissionHIEE type, other cites remained as Hcarbon emissionHIEE type, and Zhenjiang and Taizhou became the Hcarbon emissionHIEE type. In 2017, the spatial pattern of the Hcarbon emissionHIEE type was generally similar to that of 2008, but Lishui exited from the Hcarbon emissionHIEE type, and Yangzhou once again became the Hcarbon emissionHIEE type.
The number of cities belonging to the Hcarbon emissionLIEE type was relatively small, and mainly centered around the Hcarbon emissionHIEE type, with Changzhou, Yancheng, and Huainan remaining as the Hcarbon emissionLIEE type in 2000, 2008, and 2017.
In 2000, the Lcarbon emissionHIEE type cities were mainly distributed in Zhejiang and Southern Anhui, which benefited from the improvement in low-carbon transportation and infrastructure construction, the implementation of low-carbon life, and the promotion and development of forest carbon sequestration, thus, effectively improving the low-carbon production living ecological space. In 2008, the Lcarbon emissionHIEE type cities tended to contract to the northern Zhejiang and southern Anhui areas. In 2017, the spatial pattern of Lcarbon emissionHIEE type cities was generally similar to that of 2008, with Lianyungang, Nantong, Huaibei, and Lishui city evolving into the Lcarbon emissionHIEE type.

4.3. The Spatial Effect of IEE, PTE, and SE on Carbon Emissions

Using Stata15.0 software, first, the pooled ordinary least squares (OLS) regression was estimated to diagnose the spatial autocorrelation of OLS residuals. The Moran’s I test of OLS residue was 5.377, significant at the 1% statistic level, illustrating a strong spatial effect in the OLS residue. Second, the Lagrange Multiplier (LM) and Robust LM tests were used to evaluate whether the SLM model or the SEM model was more applicable than the non-spatial effect model. The test results are shown in Table 2. The LM (lag), LM (error), and the Robust LM (error) tests were significant at the 1% level, and the Robust LM (lag) was significant at the 10% level. In comparison, the value of the LM (error) and Robust LM (error) tests were higher than those of LM (lag) and Robust LM (lag) tests, implying that the SEM model was more suitable for capturing the spatial effects of the model. Since both the SLM and SEM models are non-nested spatial models, Wald and likelihood ratio (LR) tests were then needed to test whether the SDM model could be simplified to the SLM or SEM model. The statistical results determined that the SDM model could be used instead of either the SLM or SEM models. Moreover, the Hausman test showed that the fixed effect should be considered. Hence, the SDM model including the fixed effect was selected to explain the impact of IEE and its decompression terms on carbon emissions.
The estimated results of the impacts of IEE, PTE, and SE on carbon emissions were obtained (Table 3). The results showed that the spatial autocorrelation coefficients of carbon emissions were significantly positive at the 1% significance level, implying that the increase in carbon emissions in neighboring cities exerted a positive spatial spillover effect on local carbon emissions, further revealing the importance of joint control for regional carbon emissions.
For the SDM model, the estimated coefficients of explanatory variables could not directly reflect their marginal effect; moreover, it was difficult to accurately measure the direct influence of independent variables on the dependent variable. Therefore, a partial differential equation was used to calculate the direct effects, indirect effects, and total effects of each independent variable, as shown in Table 4.
According to Table 4, the direct effect of IEE on carbon emissions was significantly negative at the 1% level, implying that the improvement of IEE contributed to reducing local carbon emissions. The indirect effect of IEE on carbon emissions of neighboring cities was significantly negative at the 10% level, indicating that the improvement in local IEE emerged as the determinant for the reduction in local carbon emissions, and brought a reduction in carbon emissions in surrounding cities. The overall effect of IEE was significantly negative at the 5% level, indicating that improvement in IEE could reduce carbon emissions. This is because the Yangtze River Delta region has striven to accelerate the upgrade of the industrial economic growth mode and has made significant efforts to phase out industrial enterprises with higher energy consumption and which create serious environmental pollution. Moreover, the region has also actively implemented ecological environmental governance policies and measures to improve the ecological environment. Considering this, in the context of regional collaborative governance, this study shows that the impact of IEE on carbon emissions exhibited a marked local effect and a regional spillover effect.
The direct impact of PTE on carbon emissions was significantly negative at the 1% level, implying that the improvement of PTE contributed to reducing local carbon emissions. In terms of the indirect effects and total effect, PTE did not play a significant role in reducing the carbon emissions of adjacent cities, implying that the effect of PTE was confined to the local city. This was probably because the improvement in PTE was influenced by management, institutions, and technology absorption capacity. Accordingly, if the function, scale, and development concept of industrial structure in the surrounding cities were greatly different from the local cities, the requirements for low-carbon technology application and innovation would not be met. Moreover, there existed a certain gap between the current PTE and the optimal technological frontier, which jointly restricted the positive spillover effect of PTE.
Both the direct effect and indirect effect of SE on carbon emissions were significantly negative at the 5% level, indicating that the improvement in SE would not only alleviate local carbon emissions, but also reduce carbon emissions in the surrounding cities through the spatial spillover effect. The overall effect of SE was significantly negative, implying the optimal SE and the positive environmental externality of industrial agglomeration formed by the optimal configuration of industrial production factors were conducive to mitigating carbon emissions.
With respect to the impact of control variables on carbon emissions, the significance and sign of each of the control variables was generally similar under the regression models of SDM-IEE, SDM-PTE, and SDM-SE. Specifically, the direct effect of PGDP was significantly positive, but its indirect effect was negative and insignificant, illustrating that economic development likely increased local carbon emissions but had little impact on adjacent cities. This is probably because the high carbon-producing features of the economy and energy structure were very prominent in the delta region. Although the economic growth of the Yangtze River Delta region had entered the middle and late stages of rapid industrialization development, secondary industry still dominated the economic structure over the long term, resulting in rapid growth demand for coal, oil, natural gas, and other fossil energy.
In terms of financial considerations, the increase in FIN not only exerted a significantly positive impact on local carbon emissions, but also significantly increased carbon emissions in adjacent cities. On the one hand, this was because, local governments assumed the responsibility for financial expenditure for economic development, education, medical care, social security, and environmental protection with limited tax resources. Under the premise of limited financial resources, governments competed to attract more opportunities to increase fiscal revenues. However, environmental protection investment was usually large-scale and slow to take effect, and furthermore, was not used as a performance evaluation indicator in this study. Hence, local governments were more inclined to lowering the threshold of environmental protection and shifting to expenditure for capital construction. On the other hand, environmental pollution had a positive externality, in that, in the case of local governments lacking cooperation, each government would take on free-riding to reduce its own environmental protection investment and reduce its responsibility for protecting the environment, thereby aggravating the local and adjacent cities’ carbon emissions.
The direct and indirect impact of FDI on carbon emissions was negative, but not significant, indicating that the “pollution refuge effect” was not validated in the Yangtze River Delta region. Instead, the inflow of FDI through the pollution halo effect and the scale effect brought an advanced production technology and management mode, which exerted a restraining effect on carbon emissions. However, the promotion effect of foreign investment on economic growth stimulated all cities to lower local thresholds of environmental regulation to attract foreign investment, thereby introducing pollution-intensive industries, which aggravated energy consumption and carbon emissions. Hence, the negative environmental effect offset the positive environmental effect of FDI, resulting in an insignificant impact of FDI.
The direct impact of URB on carbon emissions was positive, but not significant. This was owing to energy consumption being the most important driving force of urbanization. With the improvements of urbanization, increasingly more energy and resources were consumed, which inevitably increased carbon emissions. Yet, with the construction of resource-saving and environment-friendly urban societies, the urbanization of the delta region gradually evolved into a new type, characterized by green production, green living and consumption patterns, and green economic transformation. This counteracted the positive carbon effects of urbanization, resulting in an insignificant impact. Meanwhile, the indirect effect of URB on carbon emissions was significantly negative. This was because the urban economy development required large amounts of funding and infrastructure construction. However, due to urban competition, the more resources a city gathered, the less resources its surrounding cities possessed. This effect, therefore, inhibited the development of urbanization in neighboring regions and thus contributed to reducing carbon emissions.
The direct impact of PAT on carbon emissions was significantly positive, while the indirect impact of PAT on carbon emissions was insignificant. These effects were because the related technological progress was not biased towards energy conservation or emissions reduction, but rather toward product innovation and productivity improvement. This promoted the expansion of enterprise-scale progress and led to an increase in carbon emissions intensity. In addition, the introduction and application of low-carbon technology increased short-term enterprise production costs, which was not conducive to enterprises participating in market competition.
The direct effect of ENV on carbon emissions was negative, but not significant, while the indirect impact of ENV on carbon emissions was positive and insignificant. These effects were because there remained prominent contradictions between regional economic development, environmental governance, and energy consumption in the short term. Although effective environmental regulation could force enterprises to invest in measures to reduce carbon emissions, such as pollution control, production process improvement, and environmental technology innovation, the inhibition emission effect was ultimately offset by the green paradox effect. That paradox effect was created from the expansion of reproduction to compensate for the increased cost arising from short-term environmental regulations [64], which was thereby not conducive to long-term low-carbon development.

4.4. Analysis of the Local Importance of IEE, PTE, and SE on Carbon Emissions

The SDM model revealed the overall impacts of IEE, PTE, and SE, but without considering their local effects. Therefore, to further explore the importance of IEE, PTE, and SE at the local scale, the geographically weighted random forest regression (GRF) [89] was applied to determine the local importance effects of IEE, PTE, and SE to the increase in carbon emissions. The values were expressed in the form of percentage increase in mean square error (%IncMSE); the higher the value of %IncMSE, the greater the importance of a given variable. It can be seen from Figure 4 that the importance of IEE, PTE, and SE varied dramatically in space. The positive effect of IEE had obvious spatial heterogeneity on the increase in carbon emissions, in that its importance increased from the east to west. Compared with other cities in Jiangsu and Zhejiang provinces, the improvement of IEE was more likely to alleviate the increase in carbon emissions in cities clustered in northern Anhui Province. Generally, the importance of PTE exhibited a similar spatial pattern compared with IEE, in that cities with a higher importance effect were mainly clustered in northwestern Anhui Province, in south Zhejiang Province, and were scattered in a few cities, such as Changzhou and Wuxi. While SE was not a relatively important factor affecting the increase in carbon emissions, its importance distribution was more dispersed in space, with relatively high importance in Changzhou, Suqian, Lu’an, and Wenzhou cities.

5. Conclusions and Recommendations

The spatial-temporal evolution and formation mechanisms of IEE and carbon emissions have been extensively discussed, and the impact of industrial production on carbon emissions has been investigated from single dimensions of industrial production indicators. However, the relationship between IEE and carbon emissions has not received the attention it deserves, and the exact link between IEE and carbon emissions has not been systematically examined. Hence, we used the Super-SBM model to measure IEE and its decomposition terms for 41 cities in the Yangtze River Delta from 2000 to 2017. Based on these measurements, the impacts of IEE and its decomposition terms on carbon emissions were further analyzed from spatio-temporal correlation and heterogeneous perspectives. The main conclusions are as follows:
(1)
The temporal correlation intensity between IEE and SE and carbon emissions exhibited an inverted V-shaped variation trend, while the temporal correlation intensity between PTE and carbon emissions exhibited a W-shaped fluctuation trend. In space, the negative correlation between IEE and carbon emissions was mainly distributed in the developed cities of the Yangtze River Delta, while the positive correlation cities were mainly distributed in central Anhui Province, and Yancheng and Taizhou cities. The spatial correlation intensity between PTE and carbon emissions exhibited a spatial pattern of being higher in the central part of the delta and lower in the northern and southern parts. The negative spatial correlation between SE and carbon emissions was mainly clustered in Zhejiang Province and scattered in Jiangsu and Anhui provinces, with the positive correlation cities being concentrated along the junction of Anhui and Jiangsu provinces and within central Jiangsu Province.
(2)
The direct and indirect effects of IEE on carbon emissions were significantly negative, implying that the improvement in IEE contributed to reducing carbon emissions. The direct impact of PTE on carbon emissions was significantly negative, while its indirect effect was insignificant. Both the direct and indirect effects of SE on carbon emissions were significantly negative.
(3)
The importance impacts of IEE, PTE, and SE on carbon emissions increase exhibited significant spatial heterogeneity. The positive effect of IEE was more likely to alleviate the increase in carbon emissions in northern Anhui City. In terms of technology, PTE was more conducive to reducing the increase in carbon emissions in northwestern Anhui City, southern Zhejiang City, and other cities including Changzhou and Wuxi. Finally, it was found that SE played a relatively important role in reducing the increase in carbon emissions in only four cities, Changzhou, Suqian, Lu’an, and Wenzhou.
This study on the impact of IEE on carbon emissions provides importance guidance for planning the construction of ecological civilizations and high-quality integrated development in the Yangtze River Delta. In order to further apply the potential of IEE in promoting carbon emission reduction, relevant policy suggestions have been proposed as follows.
First, ecological industry should be the fundamental direction of industrial development in the future. Cities in the Yangtze River Delta should continue to speed up green industrial transformation and upgrading, making full use of the current technological levels to give full scope to advantageous industries, but also strengthening the ability of green technological innovation and renewal, to improve IEE and realize the unification of the industrial economy and ecological construction. Moreover, it is necessary to continuously improve technical management levels and production scale efficiency, as well as to speed up the progress of technical efficiency. Such efforts would aid in coordinating the development of industrial management levels, technology applications, and scale expansion.
Second, cities in the Yangtze River Delta should promote the co-governance of carbon emissions due to the spatial spillover effect of carbon emissions identified in this study. Efforts should be made to establish a “five-synergy” operating mechanism featuring “coordinated case-handling, prevention, construction, improvement, and restoration,” thus creating a social ecological environment pattern of multiple co-governance and sharing. Such efforts would prevent and control carbon emissions at the source to reduce emissions increases. Moreover, in order to exert the positive spillover effects of IEE, PTE, and SE on carbon emissions reduction, cities should strengthen the collaborative research and development of green production technology and emissions reduction technology. To this end, cities should also strengthen the exchange of advanced organization and management modes between industrial enterprises.
Third, to achieve the decoupling of carbon emissions from economic growth, low-carbon circular economies should be vigorously developed to prevent economic growth being sacrificed. To this end, the government should promote resource-conserving and environmentally friendly urbanization, as well as raise public awareness of low-carbon living and green consumption, while at the same time formulating a long-term regulation mechanism to reduce carbon emissions considering local conditions. Moreover, local governments and enterprises should increase their investment in green science and technology innovation, as well as increase the introduction of low-carbon equipment and enhance regional cooperation. In addition, when introducing foreign investment, local governments should raise the environmental access threshold and refuse to bring in low-efficiency, high-consumption, or high-polluting foreign enterprises. Moreover, foreign direct investment should be guided into energy conservation and environmental protection and fully facilitate the pollution halo effect.
We acknowledge that there are some limitations to this study. First, academics have not reached a consensus on the evaluation system of IEE, and due to the influence of the availability of data, the measurement of IEE does not fully consider social policies and other closely related factors. Second, although it has been confirmed that improvement in IEE contributes to reducing carbon emissions, its underlying influencing mechanisms require further consideration, that is, whether IEE can indirectly affect carbon emissions through other intermediate variables, which can be assessed with the help of a mediation model. Third, we need to further explore the heterogeneous influence of IEE on the carbon emissions of cities of different sizes and the differential influence of IEE on carbon emissions in different periods.

Author Contributions

Conceptualization, Z.L. and D.S.; methodology, Z.L. and X.Z.; formal analysis, Z.L. and D.S. writing—original draft preparation, Z.L. and X.Z.; writing—review and editing, Z.L. and D.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Humanities and Social Sciences Youth Foundation, Ministry of Education of the People’s Republic of China, grant number 20YJCZH080, Social Science Foundation of Jiangsu Province, grant number 20SHD009, the Yangzhou University Qing Lan Project, and the Yangzhou Lv Yang Jinfeng Project in 2020.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Data available from the authors upon request.

Acknowledgments

We acknowledge the support from the Humanities and Social Sciences Youth Foundation, Ministry of Education of the People’s Republic of China (20YJCZH080), Social Science Foundation of Jiangsu Province (20SHD009), the Yangzhou University Qing Lan Project, and the Yangzhou Lv Yang Jinfeng Project in 2020. We also acknowledge the precious advice of the referees.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Wu, Q.Y.; Zhang, X.L.; Shang, Z.Y.; Li, Z.J. Political-economy based institutional industry complex and sustainable development: The case of the salt-chemical industry in Huai’an, China. Energy Policy 2015, 87, 39–47. [Google Scholar] [CrossRef]
  2. Chen, J.D.; Cheng, S.L.; Nikic, V.; Song, M.L. Quo vadis? Major players in global coal consumption and emissions reduction. Transform. Bus. Econ. 2018, 17, 112–132. [Google Scholar]
  3. Lv, Y.L.; Liu, J.N.; Cheng, J.Q.; Andreoni, V. The persistent and transient total factor carbon emission performance and its economic determinants: Evidence from China’s province-level panel data. J. Clean. Prod. 2021, 316, 128198. [Google Scholar] [CrossRef]
  4. Yang, Y.; Cai, W.J.; Wang, C. Industrial CO2 intensity, indigenous innovation and R&D spillovers in China’s provinces. Appl. Energy 2014, 131, 117–127. [Google Scholar]
  5. Luan, B.J.; Huang, J.B.; Zou, H. Domestic R&D, technology acquisition, technology assimilation and China’s industrial carbon intensity: Evidence from a dynamic panel threshold model. Sci. Total Environ. 2019, 63, 133436. [Google Scholar]
  6. Zhang, K.K.; Xu, D.Y.; Li, S.R. The impact of environmental regulation on environmental pollution in China: An empirical study based on the synergistic effect of industrial agglomeration. Environ. Sci. Pollut. Res. 2019, 26, 25775–25788. [Google Scholar] [CrossRef] [PubMed]
  7. Zhang, N.; Wei, X. Dynamic total factor carbon emissions performance changes in the Chinese transportation industry. Appl. Energy 2015, 146, 409–420. [Google Scholar] [CrossRef]
  8. Liu, Z.; Zhang, H.; Zhang, Y.J.; Zhu, T.T. How does industrial policy affect the eco-efficiency of industrial sector? Evidence from China. Appl. Energy 2020, 272, 115206. [Google Scholar] [CrossRef]
  9. Kuosmanen, T. Measurement and analysis of eco-efficiency: An economist’s perspective. J. Ind. Ecol. 2005, 9, 15–18. [Google Scholar] [CrossRef]
  10. Huang, J.H.; Yang, X.G.; Cheng, G.; Wang, S.Y. A comprehensive eco-efficiency model and dynamics of regional eco-efficiency in China. J. Clean. Prod. 2014, 67, 228–238. [Google Scholar] [CrossRef]
  11. Dahlstrom, K.; Ekins, P. Eco-efficiency trends in the UK steel and aluminum industries: Differences between resource efficiency and resource productivity. J. Ind. Ecol. 2005, 9, 171–188. [Google Scholar] [CrossRef]
  12. Zhao, R.; Liu, Y.Y.; Zhang, Z.Y.; Guo, S.D.; Tseng, M.L.; Wu, K.J. Enhancing eco-efficiency of agro-products’ closed-loop supply chain under the Belt and Road initiatives: A system dynamics approach. Sustainability 2018, 10, 668. [Google Scholar] [CrossRef] [Green Version]
  13. Shao, L.G.; Yu, X.; Feng, C. Evaluating the eco-efficiency of China’s industrial sectors: A two-stage network data envelopment analysis. J. Environ. Manag. 2019, 247, 551–560. [Google Scholar] [CrossRef] [PubMed]
  14. Tenente, M.; Henriques, C.; da Silva, P.P. Eco-efficiency assessment of the electricity sector: Evidence from 28 European Union countries. Econ. Anal. Policy 2020, 66, 293–314. [Google Scholar] [CrossRef]
  15. Li, Z.J.; Sarwar, S.; Jin, T. Spatiotemporal evolution and improvement potential of agricultural eco-efficiency in Jiangsu province. Front. Energy Res. 2021, 9, 746405. [Google Scholar] [CrossRef]
  16. Liu, C.H.; Gao, M.D.; Zhu, G.; Zhang, C.X.; Zhang, P.; Chen, J.Q.; Cai, W. Data driven eco-efficiency evaluation and optimization in industrial production. Energy 2021, 224, 120170. [Google Scholar] [CrossRef]
  17. Sun, Y.Y.; Hou, G.L. Analysis on the spatial-temporal evolution characteristics and spatial network structure of tourism eco-efficiency in the Yangtze River Delta urban agglomeration. Int. J. Environ. Res. Public Health 2021, 18, 2577. [Google Scholar] [CrossRef] [PubMed]
  18. Caiado, R.G.; de Dias, R.F.; Mattos, L.V.; Quelhas, O.L.G.; Filho, W.L. Towards sustainable development through the perspective of eco-efficiency-A systematic literature review. J. Clean. Prod. 2017, 165, 890–904. [Google Scholar] [CrossRef] [Green Version]
  19. Shah, I.H.; Dong, L.; Park, H.S. Tracking urban sustainability transition: An eco-efficiency analysis on eco-industrial development in Ulsan, Korea. J. Clean. Prod. 2020, 262, 121286. [Google Scholar] [CrossRef]
  20. de Araújo, R.V.; Espejo, R.A.; Constantino, M.; de Martin, P.M.; Taveira, J.C.; Lira, F.S.; Herrera, G.P.; Costa, R. Eco-efficiency measurement as an approach to improve the sustainable development of municipalities: A case study in the Midwest of Brazil. Environ. Dev. 2021, 39, 100652. [Google Scholar] [CrossRef]
  21. Kluczek, A.; Olszewski, P. Energy audits in industrial processes. J. Clean. Prod. 2017, 142, 3437–3453. [Google Scholar] [CrossRef]
  22. Guan, Y.R.; Huang, G.H.; Liu, L.R.; Huang, C.Z.; Zhai, M.Y. Ecological network analysis for an industrial solid waste metabolism system. Environ. Pollut. 2019, 244, 279–287. [Google Scholar] [CrossRef] [PubMed]
  23. Guan, Y.R.; Huang, G.H.; Liu, L.R.; Zhai, M.Y.; Zheng, B. Dynamic analysis of industrial solid waste metabolism at aggregated and disaggregated levels. J. Clean. Prod. 2019, 221, 817–827. [Google Scholar] [CrossRef]
  24. Yu, Y.T.; Huang, J.H.; Zhang, N. Industrial eco-efficiency, regional disparity, and spatial convergence of China’s regions. J. Clean. Prod. 2018, 204, 872–887. [Google Scholar] [CrossRef]
  25. Yang, L.; Yang, Y.T. Evaluation of eco-efficiency in China from 1978 to 2016: Based on a modified ecological footprint model. Sci. Total. Environ. 2019, 662, 581–590. [Google Scholar] [CrossRef] [PubMed]
  26. Gao, P.; Yue, S.J.; Chen, H.T. Carbon emission efficiency of China’s industry sectors: From the perspective of embodied carbon emissions. J. Clean. Prod. 2020, 283, 124655. [Google Scholar] [CrossRef]
  27. Huang, J.H.; Xia, J.J.; Yu, Y.T.; Zhang, N. Composite eco-efficiency indicators for China based on data envelopment analysis. Ecol. Indic. 2018, 85, 674–697. [Google Scholar] [CrossRef]
  28. Toma, P.; Miglietta, P.P.; Zurlini, G.; Valente, D.; Petrosillo, I. A non-parametric bootstrap-data envelopment analysis approach for environmental policy planning and management of agricultural efficiency in EU countries. Ecol. Indic. 2017, 83, 132–143. [Google Scholar] [CrossRef]
  29. Lee, T.; Yeo, G.T.; Thai, V.V. Environmental efficiency analysis of port cities: Slacks-based measure data envelopment analysis approach. Transp. Policy 2014, 33, 82–88. [Google Scholar] [CrossRef]
  30. Zhou, C.S.; Shi, C.Y.; Wang, S.J.; Zhang, G.J. Estimation of eco-efficiency and its influencing factors in Guangdong province based on Super-SBM and panel regression models. Ecol. Indic. 2017, 86, 67–80. [Google Scholar] [CrossRef]
  31. Wang, X.M.; Ding, H.; Liu, L. Eco-efficiency measurement of industrial sectors in China: A hybrid super-efficiency DEA analysis. J. Clean. Prod. 2019, 229, 53–64. [Google Scholar] [CrossRef]
  32. Zhou, Y.; Kong, Y.; Sha, J.; Wang, H.K. The role of industrial structure upgrades in eco-efficiency evolution: Spatial correlation and spillover effects. Sci. Total. Environ. 2019, 687, 1327–1336. [Google Scholar] [CrossRef] [PubMed]
  33. Azam, M.; Khan, A.Q. Testing the Environmental Kuznets Curve hypothesis: A comparative empirical study for 28 low, lower middle, upper middle and high income countries. Renew. Sustain. Energy Rev. 2016, 63, 556–567. [Google Scholar] [CrossRef]
  34. Wang, Q.; Su, M. Drivers of decoupling economic growth from carbon emission-an empirical analysis of 192 countries using decoupling model and decomposition method. Environ. Impact Assess. Rev. 2020, 81, 106356. [Google Scholar] [CrossRef]
  35. Zhang, C.G.; Zhou, X.X. Does foreign direct investment lead to lower CO2 emissions? Evidence from a regional analysis in China. Renew. Sustain. Energy Rev. 2016, 58, 943–951. [Google Scholar] [CrossRef]
  36. Malik, M.Y.; Latif, K.; Khan, Z.; Butt, H.D.; Hussain, M.; Nadeem, M.A. Symmetric and asymmetric impact of oil price, FDI and economic growth on carbon emission in Pakistan: Evidence from ARDL and non-linear ARDL approach. Sci. Total Environ. 2020, 726, 138421. [Google Scholar] [CrossRef] [PubMed]
  37. Ertugrul, H.M.; Cetin, M.; Seker, F.; Dogan, E. The impact of trade openness on global carbon dioxide emissions: 32 Evidence from the top ten emitters among developing countries. Ecol. Indic. 2016, 67, 543–555. [Google Scholar] [CrossRef] [Green Version]
  38. Safi, A.; Chen, Y.Y.; Wahab, S.; Ali, S.; Yi, X.R.; Imran, M. Financial instability and consumption-based carbon emission in E-7 countries: The Role of Trade and Economic Growth. Sustain. Prod. Consum. 2021, 27, 383–391. [Google Scholar] [CrossRef]
  39. Cheng, S.L.; Chen, Y.T.; Meng, F.X.; Chen, J.D.; Liu, G.Y.; Song, M.L. Impacts of local public expenditure on CO2 emissions in Chinese cities: A spatial cluster decomposition analysis. Resour. Conserv. Recycl. 2021, 164, 105217. [Google Scholar] [CrossRef]
  40. Chen, X.; Chen, Y.E.; Chang, C.P. The effects of environmental regulation and industrial structure on carbon dioxide emission: A non-linear investigation. Environ. Sci. Pollut. Res. 2019, 26, 30252–30267. [Google Scholar] [CrossRef]
  41. Pei, Y.; Zhu, Y.M.; Liu, S.X.; Wang, X.C.; Cao, J.J. Environmental regulation and carbon emission: The mediation effect of technical efficiency. J. Clean. Prod. 2019, 236, 117599. [Google Scholar] [CrossRef]
  42. Ding, S.T.; Zhang, M.; Song, Y. Exploring China’s carbon emissions peak for different carbon tax scenarios. Energy Policy 2019, 129, 1245–1252. [Google Scholar] [CrossRef]
  43. Jeong, K.; Kim, S. LMDI decomposition analysis of greenhouse gas emissions in the Korean manufacturing sector. Energy Policy 2013, 62, 1245–1253. [Google Scholar] [CrossRef]
  44. Quan, C.G.; Chen, X.J.; Yu, S.S.; Ye, X. Analysis on the influencing factors of carbon emission in China’s logistics industry based on LMDI method. Sci. Total Environ. 2020, 734, 138473. [Google Scholar] [CrossRef]
  45. Wu, Y.; Shen, L.Y.; Zhang, Y.; Shuai, C.Y.; Yan, H.; Lou, Y.L.; Ye, G. A new panel for analyzing the impact factors on carbon emission: A regional perspective in China. Ecol. Indic. 2019, 97, 260–268. [Google Scholar] [CrossRef]
  46. Ali, U.M.; Gong, Z.M.; Ali, M.U.; Asmi, F.; Muhammad, R. CO2 emission, economic development, fossil fuel consumption and population density in India, Pakistan and Bangladesh: A panel investigation. Int. J. Finance Econ. 2020, 1–14. [Google Scholar] [CrossRef]
  47. Zang, X.L.; Zhao, T.; Wang, J.; Guo, F. The effects of urbanization and household-related factors on residential direct CO2 emissions in Shanxi, China from 1995 to 2014: A decomposition analysis. Atmos. Pollut. Res. 2017, 8, 297–309. [Google Scholar] [CrossRef]
  48. Sun, W.; Huang, C.C. How does urbanization affect carbon emission efficiency? Evidence from China. J. Clean. Prod. 2020, 272, 122828. [Google Scholar] [CrossRef]
  49. Wang, Q.; Zhang, F.Y. Does increasing investment in research and development promote economic growth decoupling from carbon emission growth? An empirical analysis of BRICS countries. J. Clean. Prod. 2020, 52, 119853. [Google Scholar] [CrossRef]
  50. Lee, K.H.; Min, B. Green R&D for eco-innovation and its impact on carbon emissions and firm performance. J. Clean. Prod. 2015, 108, 534–542. [Google Scholar]
  51. Du, K.R.; Li, P.Z.; Yan, Z.M. Do green technology innovations contribute to carbon dioxide emission reduction? Empirical evidence from patent data. Technol. Forecast. Soc. Chang. 2019, 146, 297–303. [Google Scholar] [CrossRef]
  52. Wen, Q.; Chen, Y.; Hong, J.K.; Chen, Y.; Ni, D.F.; Shen, Q.P. Spillover effect of technological innovation on CO2 emissions in China’s construction industry. Build. Environ. 2020, 171, 106653. [Google Scholar] [CrossRef]
  53. Wang, Q.; Wu, S.D.; Zeng, Y.E.; Wu, B.W. Exploring the relationship between urbanization, energy consumption, and CO2 emissions in different provinces of China. Renew. Sustain. Energy Rev. 2016, 54, 1563–1579. [Google Scholar] [CrossRef]
  54. Chen, D.K.; Chen, S.Y.; Jin, H. Industrial agglomeration and CO2 emissions: Evidence from 187 Chinese prefecture-level cities over 2005–2013. J. Clean. Prod. 2018, 172, 993–1003. [Google Scholar] [CrossRef]
  55. Li, K.; Lin, B.Q. The nonlinear impacts of industrial structure on China’s energy intensity. Energy 2014, 69, 258–265. [Google Scholar] [CrossRef]
  56. Mi, Z.F.; Pan, S.Y.; Yu, H.; Wei, Y.M. Potential impacts of industrial structure on energy consumption and CO2 emission: A case study of Beijing. J. Clean. Prod. 2015, 103, 455–462. [Google Scholar] [CrossRef] [Green Version]
  57. Cheng, Z.H.; Li, L.S.; Liu, J. Industrial structure, technical progress and carbon intensity in China’s provinces. Renew. Sustain. Energy Rev. 2018, 81, 2935–2946. [Google Scholar] [CrossRef]
  58. Hou, J.; Teo, T.S.H.; Zhou, F.L.; Lim, M.K.; Chen, H. Does industrial green transformation successfully facilitate a decrease in carbon intensity in China? An environmental regulation perspective. J. Clean. Prod. 2018, 184, 1060–1071. [Google Scholar] [CrossRef]
  59. Waheed, R.; Sarwar, S.; Wei, C. The survey of economic growth, energy consumption and carbon emission. Energy Rep. 2019, 5, 1103–1115. [Google Scholar] [CrossRef]
  60. Wang, Z.H.; Yin, F.C.; Zhang, Y.X.; Zhang, X. An empirical research on the influencing factors of regional CO2 emissions: Evidence from Beijing city, China. Appl. Energy 2012, 100, 277–284. [Google Scholar] [CrossRef]
  61. Wang, Z.H.; Zhang, B.; Liu, T.F. Empirical analysis on the factors influencing national and regional carbon intensity in China. Renew. Sustain. Energy Rev. 2016, 55, 34–42. [Google Scholar] [CrossRef]
  62. Chen, H.Y.; Hao, Y.; Li, J.W.; Song, X.J. The impact of environmental regulation, shadow economy, and corruption on environmental quality: Theory and empirical evidence from China. J. Clean. Prod. 2018, 195, 200–214. [Google Scholar] [CrossRef]
  63. Zhang, S.Y.; Li, H.X.; Zhang, Q.; Tian, X.; Shi, F. Uncovering the impacts of industrial transformation on low-carbon development in the Yangtze River Delta. Resour. Conserv. Recycl. 2019, 150, 104442. [Google Scholar] [CrossRef]
  64. Sinn, H.W. Public policies against global warming: A supply side approach. Int. Tax Public Finance. 2008, 15, 360–394. [Google Scholar] [CrossRef]
  65. Dinda, S. Production technology and carbon emission: Long-run relation with short-run dynamics. J. Appl. Econ. 2018, 21, 106–121. [Google Scholar] [CrossRef]
  66. Zhang, F.; Deng, X.Z.; Phillips, F.; Fang, C.L.; Wang, C. Impacts of industrial structure and technical progress on carbon emission intensity: Evidence from 281 cities in China. Technol. Forecast. Soc. Chang. 2020, 154, 119949. [Google Scholar] [CrossRef]
  67. Yang, L.S.; Li, Z. Technology advance and the carbon dioxide emission in China-Empirical research based on the rebound effect. Energy Policy 2017, 101, 150–161. [Google Scholar] [CrossRef]
  68. Wu, L.P.; Chen, Y.; Feylizadeh, M.R.; Liu, W.J. Estimation of China’s macro-carbon rebound effect: Method of integrating Data Envelopment Analysis production model and sequential Malmquist-Luenberger index. J. Clean. Prod. 2018, 198, 1431–1442. [Google Scholar] [CrossRef]
  69. Wang, Z.H.; Sun, Y.F.; Wang, B. How does the new-type urbanisation affect CO2 emissions in China? An empirical analysis from the perspective of technological progress. Energy Econ. 2019, 80, 917–927. [Google Scholar] [CrossRef]
  70. Tobler, W.R. A computer movie simulating urban growth in the Detroit region. Econ. Geogr. 1970, 46, 234–240. [Google Scholar] [CrossRef]
  71. Anselin, L. Spatial effects in econometric practice in environmental and resource economics. Am. J. Agric. Econ. 2001, 83, 705–710. [Google Scholar] [CrossRef]
  72. Hu, M.J.; Pang, Y.M.; Jin, T.; Li, Z.J. Spatio-Temporal Evolution of EIWB and Influencing Factors: An Empirical Study from the Yangtze River Delta. Appl. Spat. Anal. Policy 2021, 14, 999–1024. [Google Scholar] [CrossRef]
  73. Zhou, Y.; Kong, Y.; Zhang, T. The spatial and temporal evolution of provincial eco-efficiency in China based on SBM modified three-stage data envelopment analysis. Environ. Sci. Pollut. Res. 2020, 27, 8557–8569. [Google Scholar] [CrossRef] [PubMed]
  74. Wang, Y.D.; Zheng, Y.M. Spatial effects of carbon emission intensity and regional development in China. Environ. Sci. Pollut. Res. 2021, 28, 14131–14143. [Google Scholar] [CrossRef]
  75. Elhorst, J.P. Matlab software for spatial panels. Int. Reg. Sci. Rev. 2014, 37, 389–405. [Google Scholar] [CrossRef] [Green Version]
  76. Bi, C.; Zeng, J.J. Nonlinear and spatial effects of tourism on carbon emissions in China: A spatial econometric approach. Int. J. Environ. Res. Public Health 2019, 16, 3353. [Google Scholar] [CrossRef] [Green Version]
  77. Cheng, Z.H. The spatial correlation and interaction between manufacturing agglomeration and environmental pollution. Ecol. Indic. 2016, 61, 1024–1032. [Google Scholar] [CrossRef]
  78. Song, W.Z.; Wang, C.; Chen, W.Q.; Zhang, X.L.; Li, H.R.; Li, J. Unlocking the spatial heterogeneous relationship between per capita GDP and nearby air quality using bivariate local indicator of spatial association. Resour. Conserv. Recycl. 2020, 160, 104880. [Google Scholar] [CrossRef]
  79. Tone, K. A slacks-based measure of efficiency in data envelopment analysis. Eur. J. Oper. Res. 2001, 130, 498–509. [Google Scholar] [CrossRef] [Green Version]
  80. Tone, K. A slacks-based measure of super-efficiency in data envelopment analysis. Eur. J. Oper. Res. Sci. 2002, 143, 32–41. [Google Scholar] [CrossRef] [Green Version]
  81. Tone, K. Dealing with Undesirable Outputs in DEA: A Slacks-Based Measure (SBM) Approach; North American Productivity Workshop: Toronto, ON, Canada, 2004; pp. 44–45. [Google Scholar]
  82. Zhang, J.R.; Zeng, W.H.; Wang, J.N.; Yang, F.L.; Jiang, H.Q. Regional low-carbon economy efficiency in China: Analysis based on the Super-SBM model with CO2 emissions. J. Clean. Prod. 2017, 163, 202–211. [Google Scholar] [CrossRef]
  83. Cheng, G. Data Envelopment Analysis: Methods and MaxDEA Software; Intellectual Property Publishing House Co. Ltd.: Beijing, China, 2014. [Google Scholar]
  84. Peng, H.S.; Zhang, J.H.; Lu, L.; Tang, G.R.; Yan, B.J.; Xiao, X.; Han, Y. Eco-efficiency and its determinants at a tourism destination: A case study of Huangshan National Park, China. Tour. Manag. 2017, 60, 201–211. [Google Scholar] [CrossRef]
  85. Škrinjarić, T. Dynamic portfolio optimization based on grey relational analysis approach. Expert Syst. Appl. 2020, 147, 113207. [Google Scholar] [CrossRef]
  86. Fang, G.C.; Wang, Q.L.; Tian, L.X. Green development of Yangtze River Delta in China under Population-Resources-Environment-Development-Satisfaction perspective. Sci. Total. Environ. 2020, 727, 138710. [Google Scholar] [CrossRef] [PubMed]
  87. Dang, Y.G.; Shang, Z.J.; Wang, J.J.; Feng, Y. Grey incidence model for relationship between indicators of panel data and its application. Control. Decis. 2019, 34, 1077–1084. [Google Scholar]
  88. Chen, J.D.; Gao, M.; Cheng, S.L.; Liu, X.; Hou, W.X.; Song, M.L.; Li, D.; Fan, W. China’s city-level carbon emissions during 1992–2017 based on the inter-calibration of nighttime light data. Sci. Rep. 2021, 11, 3323. [Google Scholar] [CrossRef] [PubMed]
  89. Georganos, S.; Grippa, T.; Gadiaga, A.N.; Linard, C.; Lennert, M.; Vanhuysse, S.; Mboga, N.; Wolff, E.; Kalogirou, S. Geographical random forests: A spatial extension of the random forest algorithm to address spatial heterogeneity in remote sensing and population podelling. Geocarto Int. 2021, 36, 121–136. [Google Scholar] [CrossRef] [Green Version]
Figure 1. Temporal correlation intensity between IEE, PTE, SE, and carbon emissions.
Figure 1. Temporal correlation intensity between IEE, PTE, SE, and carbon emissions.
Energies 14 08169 g001
Figure 2. Spatial correlation intensity between (a) IEE and carbon emissions, (b) PTE and carbon emissions, (c) SE and carbon emissions.
Figure 2. Spatial correlation intensity between (a) IEE and carbon emissions, (b) PTE and carbon emissions, (c) SE and carbon emissions.
Energies 14 08169 g002
Figure 3. Local spatial association between IEE and carbon emissions in (a) 2000, (b) 2008 and (c) 2017.
Figure 3. Local spatial association between IEE and carbon emissions in (a) 2000, (b) 2008 and (c) 2017.
Energies 14 08169 g003
Figure 4. Spatial distribution of the importance of (a) IEE, (b) PTE and (c) SE.
Figure 4. Spatial distribution of the importance of (a) IEE, (b) PTE and (c) SE.
Energies 14 08169 g004
Table 1. The evaluation indicators of industrial eco-efficiency (IEE).
Table 1. The evaluation indicators of industrial eco-efficiency (IEE).
Primary IndicesSecondary IndicesUnitMaxMinMeanStdev
Input indicatorsIndustrial land areaTen thousand tons736.81.550.6191.82
Industrial employeeTen thousand people62911.4112.8391.29
Industrial electricitykW·h1202.041.63128.83168.12
Industrial water consumptionTen thousand tons320,40097825,567.9847,975.59
Expected output indicatorIndustrial added valueTen thousand yuan7606.4515.1826.891030.34
Undesirable output indicatorIndustrial waste water dischargeTen thousand tons85,735548.912,930.3515,649.08
Industrial exhaust emissionsTen thousand tons17,041141880.612483.51
Industrial smoke and dust emissionsTen thousand tons31593.4525.26573.73
Table 2. Test results of spatial model.
Table 2. Test results of spatial model.
TestStatisticsTestStatistics
LM (lag) test71.181 ***LR test spatial lag93.89 ***
Robust LM (lag) test2.978 *LR test spatial error133.23 ***
LM (error) test549.296 ***Wald test spatial lag68.92 ***
Robust LM (error) test481.094 ***Wald test spatial error92.87 ***
Hausman test−115.18
Note: *** Significant at the 0.01 level; ** Significant at the 0.05 level; * Significant at the 0.1 level.
Table 3. Regression results of spatial Durbin models.
Table 3. Regression results of spatial Durbin models.
SDM-IEESDM-PTESDM-SE
IEE−0.035 (−2.64) ***
PTE −0.036 (−1.83) *
SE −0.033 (−1.89) *
PGDP0.159 (6.90) ***0.154 (6.70) ***0.148 (6.62) ***
FIN0.079 (7.69) ***0.081 (7.86) ***0.081 (7.83) ***
FDI−0.003 (−0.60)−0.004 (−0.80)−0.004 (−0.72)
URB0.064 (1.59)0.067 (1.67) *0.065 (1.62)
PAT0.025 (3.57) ***0.025 (3.59) ***0.027 (3.95) ***
ENV−0.008 (−1.35)−0.008 (−1.40)−0.008 (−1.44)
W×IEE−0.021 (−0.87)
W×PTE 0.001 (0.03)
W×SE −0.042 (−1.29)
W×PGDP−0.112 (−3.48) ***−0.112 (−3.46) ***−0.111 (−3.53) ***
W×FIN0.012 (0.83)0.013 (0.90)0.011 (0.78)
W×FDI−0.004 (−0.45)−0.006 (−0.68)−0.005 (−0.61)
W×URB−0.236 (−5.34) ***−0.235 (−5.35) ***−0.243(−5.48) ***
W×PAT−0.015 (−1.71) *−0.017 (−1.85) *−0.013 (−1.38)
W×ENV0.023 (1.76) *0.022 (1.69) *0.022 (1.70) *
rho0.671 (24.48) ***0.677 (25.26) ***0.667 (23.99) ***
R20.9580.9570.958
Sigma20.0040.0040.004
Log likelihood936.483933.938934.999
Note: *** Significant at the 0.01 level; ** Significant at the 0.05 level; * Significant at the 0.1 level.
Table 4. Direct, indirect, and total effects of spatial Durbin models.
Table 4. Direct, indirect, and total effects of spatial Durbin models.
Direct EffectIndirect EffectTotal Effect
SDM-IEESDM-PTESDM-SESDM-IEESDM-PTESDM-SESDM-IEESDM-PTESDM-SE
IEE−0.046
(−2.90) ***
−0.126
(−1.95) *
−0.171
(−2.32) **
PTE −0.041
(−1.78) *
−0.069
(−0.61)
−0.110
(−0.88)
SE −0.048
(−2.27) **
−0.178
(−2.07) **
−0.226
(−2.27) **
PGDP0.157
(7.00) ***
0.151
(6.72) ***
0.145
(6.69) ***
−0.009
(−0.12)
−0.016
(−0.22)
−0.029
(−0.41)
0.148
(1.92) *
0.135
(1.68) *
0.116
(1.55)
FIN0.097
(9.78) ***
0.099
(10.04) ***
0.098
(9.93) ***
0.182
(6.19) ***
0.193
(6.61) ***
0.18
(6.14) ***
0.278
(8.67) ***
0.292
(9.14) ***
0.278
(8.70) ***
FDI−0.005
(−0.90)
−0.006
(−1.22)
−0.006
(−1.11)
−0.018
(−0.78)
−0.026
(−1.11)
−0.023
(−1.00)
−0.023
(−0.88)
−0.032
(−1.24)
−0.028
(−1.12)
URB0.017
(0.47)
0.02
(0.55)
0.018
(0.48)
−0.539
(−7.67) ***
−0.541
(−7.57) ***
−0.55
(−7.9) ***
−0.522
(−7.23) ***
−0.521
(−7.05) ***
−0.533
(−7.46) ***
PAT0.025
(3.61) ***
0.025
(3.50) ***
0.029
(4.19) ***
0.001
(0.03)
−0.004
(−0.18)
0.013
(0.67)
0.025
(1.21)
0.021
(0.91)
0.042
(1.95) *
ENV−0.003
(−0.41)
−0.004
(−0.47)
−0.004
(−0.52)
0.052
(1.30)
0.049
(1.21)
0.048
(1.22)
0.049
(1.06)
0.046
(0.98)
0.044
(0.97)
Note: *** Significant at the 0.01 level; ** Significant at the 0.05 level; * Significant at the 0.1 level.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Li, Z.; Zheng, X.; Sun, D. The Influencing Effects of Industrial Eco-Efficiency on Carbon Emissions in the Yangtze River Delta. Energies 2021, 14, 8169. https://0-doi-org.brum.beds.ac.uk/10.3390/en14238169

AMA Style

Li Z, Zheng X, Sun D. The Influencing Effects of Industrial Eco-Efficiency on Carbon Emissions in the Yangtze River Delta. Energies. 2021; 14(23):8169. https://0-doi-org.brum.beds.ac.uk/10.3390/en14238169

Chicago/Turabian Style

Li, Zaijun, Xiang Zheng, and Dongqi Sun. 2021. "The Influencing Effects of Industrial Eco-Efficiency on Carbon Emissions in the Yangtze River Delta" Energies 14, no. 23: 8169. https://0-doi-org.brum.beds.ac.uk/10.3390/en14238169

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop