Spatial and temporal pattern evolution and influencing factors of energy–environmental efficiency: A case study of Yangtze River urban agglomeration in China

Abstract

Improving energy–environmental efficiency is prerequisite for sustainable development. In order to explore ways to improve energy–environmental efficiency, this paper uses the undesired output slack-based model to measure the energy–environmental efficiency of the Yangtze River urban agglomeration based on the input and output index data from 2008 to 2017, and its spatial and temporal pattern evolution is analyzed by using kernel density estimation, Gini coefficient, and coefficient of variation. Moreover, the Tobit regression model is used to analyze the influencing factors of the energy–environmental efficiency of the Yangtze River urban agglomeration. The results indicate that the energy–environmental efficiency of each city is increased continuously, and the regional differences are gradually narrowed. The spatial pattern is changed from polar nucleus type to valley type, and finally the distribution characteristics of “overall high” are formed. Overall, the energy–environmental efficiency presents a spatial layout of “high in the east and low in the west.” The regression results show that the level of economic development and energy–environmental efficiency are “U-type” associated characteristics, and government regulation and population density have significant positive effects on it. Industrial structure and technological progress have negative effects on it, and the effect of opening degree is not significant.

Keywords

Energy–environmental efficiency,slack-based model,spatial and temporal patterns kernel density estimation Tobit regression model

Introduction

As a strategic resource, energy plays an important role in national and regional economic development,¹ but energy consumption also generates environmental challenges in countries throughout the world.² The global environmental problem has increased with increase in energy demand.^3,4 China is currently the largest energy producer and consumer in the world.⁵ Energy consumption and CO₂ emissions have been at top level in the world for many years.⁶ According to BP,⁷ China accounted for 23.2% of global energy consumption and 33.6% of global energy consumption growth in 2017, ranking first in global energy growth for 17 consecutive years. The energy and environmental issues are main hurdles for sustainable development.⁸ In the face of the increasingly prominent contradiction between the economy and the environment, how to achieve sustainable economic and social development and take into account the green and healthy development of the environment has become a topic of great concern to governments and all sectors of society. With the promulgation of national environmental protection policies and people’s high standards for environmental quality, improving the utilization efficiency of energy and environment is of great significance for China’s economic transformation and environmental sustainable development.⁹ It is not only an inevitable choice to achieve sustainable economic and social development,¹⁰ but also an important means to achieve energy conservation and emission reduction.¹¹ As a competitive region in China, the Yangtze River urban agglomeration has many problems,¹² such as unreasonable energy, environmental consumption structure, and serious environmental pollution.¹³ Therefore, this paper takes the Yangtze River urban agglomeration as an example to study the spatial and temporal differentiation characteristics of energy–environmental efficiency, and analyzes the influencing factors of energy–environmental efficiency, which has important sense for promoting the coordinated development of energy–environmental efficiency at the regional level.

Research on energy efficiency mainly has two perspectives: single factor and total factor energy efficiency. Since single factor energy efficiency does not consider the substitution between energy and other input factors, many studies have focused on energy efficiency under the framework of total factor. For instance, Hu and Wang¹⁴ first analyzed the energy efficiency of China by using the total factor index, and introduced DEA into China’s energy efficiency analysis. It has effectively overcome the shortcomings of traditional single factor energy efficiency research methods. Since then, DEA has been widely used to evaluate the environmental and energy performance of decision-making units from the perspective of global economic development.¹⁵ Similar studies using DEA to study total factor energy efficiency include Borozan,¹⁶ Ouyang et al.,¹⁷ and Zhou et al.¹⁸ However, the studies above did not take undesired outputs into account; Mandal¹⁹ pointed out that the DEA model that ignores undesired outputs will cause deviation of energy efficiency measurement. Wang et al.²⁰ confirmed the existence of deviation by comparing the energy efficiency considering undesired output and not considering undesired output. Li and Hu²¹ found that energy efficiency would be overestimated if environmental factors were not considered. Wu et al.’s²² research shows that ignoring undesired output will make the total factor energy efficiency measurement result higher. With the extensive attention of the academic community on environmental pollution in the process of energy consumption (Lin, 2012)²³, the slack-based model (SBM) based on undesired output is widely used in the evaluation of total factor energy efficiency.^24,25 The SBM model introduces the slack variable into the objective function, which solves the deviation problem caused by the neglect of the slack variable in traditional DEA model. Although the SBM model has been widely used in many fields, few scholars have combined energy and environment to study the energy–environmental efficiency by using the SBM model.

The research on the temporal and spatial evolution characteristics of energy efficiency is mainly based on national and provincial scales. Pan et al. (2012)²⁶, Yang and Liu,²⁷ and Guan and Xu²⁸ measured energy efficiency on China’s provinces by using panel data and found that energy efficiency presented the characteristics of regional differentiation; the polarization between the east and the west is serious. However, few scholars have been involved in the evolution of the spatial and temporal pattern of energy efficiency considering environmental constraints. Moreover, studies on the influencing factors of energy and environmental efficiency are even more scarce, most of the references are about influencing factors of energy efficiency. For example, Cui et al.²⁹ analyzed the main factors affecting energy efficiency and found that technology level and management indicators are important factors. Yao et al.³⁰ analyzed the influence of technology gap on China’s energy efficiency and carbon emission performance with DEA model considering group heterogeneity. In recent years, a few scholars have paid attention to this aspect, such as Chen et al.³¹ and Xiao et al.³² However, they studied the energy–environmental efficiency of the industrial sector at the national and provincial levels, respectively, without paying attention to the factors affecting the energy–environmental efficiency at the regional level. Therefore, the influencing factors of energy–environmental efficiency at the regional level need to be further explored.

Most scholars only consider the expected economic output and lack the analysis of the environmental factors of undesired output when studying the energy efficiency problem. Moreover, the researches on the spatial and temporal differentiation of energy efficiency are mostly based on national and provincial scales, and the literature of energy–environmental efficiency from the perspective of cities was ignored. Therefore, this paper takes Yangtze River urban agglomeration as the research object. Combined with the existing research³³ and taking into account the characteristics of the Yangtze River urban agglomeration, this paper defines energy–environmental efficiency as the energy efficiency that takes environmental constraints into account under the total factor framework. Through incorporating environmental factors into the measurement of energy efficiency, and analyzing its spatial and temporal pattern evolution characteristics and influencing factors, so as to provide reference for realizing the win-win situation of regional economic benefit and ecological environment benefit.

The rest of the paper is structured as follows: the second section is “Research methods and data processing”; the third section is the “Analysis and discussion of energy–environmental efficiency results”; the fourth section is the “Analysis of influencing factors of energy–environmental efficiency”; and the last section is the “Conclusions, findings, and recommendations.”

Research methods and data processing

SBM model

The SBM model was proposed and extended by Tone.³⁴ The efficiency problem under undesired output can be more effectively evaluated (Tu and Liu, 2011)³⁵ by putting the relaxation variable into the objective function. The undesired output SBM model can better solve the problem of environmental pollution factor not being included in the model and input–output relaxation, so it is an ideal research method. In this paper, environmental factors are incorporated into the SBM model. The non-radial SBM model formula with the same return to scale can be written as

\begin{array}{l} \min EEN = \frac{1 - \frac{1}{M} \sum_{m = 1}^{M} \frac{s_{m}^{x}}{x_{k^{′} m}^{t^{′}}}}{1 + \frac{1}{N + 1} (\sum_{n = 1}^{N} \frac{s_{n}^{y}}{y_{k^{′} n}^{t^{′}}} + \sum_{i = 1}^{I} \frac{s_{i}^{b}}{b_{k^{′} i}^{t^{′},}})} \\ s . t . \sum_{t = 1}^{T} \sum_{k = 1}^{K} z_{k}^{t} x_{k m}^{t} + s_{m}^{x} = x_{k^{′} m}^{t}, m = 1, 2, \dots \dots, M \\ \sum_{t = 1}^{T} \sum_{k = 1}^{K} z_{k}^{t} y_{k n}^{t} - s_{n}^{y} = y_{k^{'} n}^{t^{'}}, n = 1, 2, \dots \dots, N \\ \sum_{t = 1}^{T} \sum_{k = 1}^{K} z_{k}^{t} b_{k i}^{t} + s_{i}^{b} = b_{k^{'} i}^{t^{'}}, i = 1, 2, \dots \dots I \\ z_{k}^{t} \geq 0, s_{m}^{x} \geq 0, s_{n}^{y} \geq 0, s_{i}^{b} \geq 0, k = 1, 2, \dots \dots K \end{array}

(1)

where

M

is the total number of indicators,

N

is the number of expected output indicators,

I

is the number of undesired output indicators, and

z_{k}^{t}

is the weight of each city. The

s_{m}^{x}, s_{n}^{y}, s_{i}^{b}

represent the relaxation vector of input, expected output, and undesired output, respectively. The

x_{k^{'} m}^{t^{'}}, y_{k^{'} n}^{t^{'}}, b_{k^{'} i}^{t^{'}}

show the input and output value of the

k

th city in

t

years,

EEN

is the urban energy–environmental efficiency value, and

0 < EEN < 1

. The

EEN

= 1 means that the evaluation unit DEA is effective and is on the production front surface, and the urban energy–environmental efficiency is high; when

EEN

< 1, the DEA is invalid and indicates that the urban energy–environmental efficiency is not high and there is potential for improvement.

Kernel density estimation

Kernel density estimation is a non-parametric method for estimating the probability density function. The density distribution of different years can be calculated by the kernel density estimation, which is convenient for comparative analysis of the temporal variation of a particular economic activity. Suppose $x_{1}, x_{2}, \dots, x_{n}$ are n sample points of independent and identical distribution F, assume that the probability density function is f, and the kernel density can be written as

f_{n} (x) = \frac{1}{n} \sum_{i = 1}^{n} K_{h} (x - x_{i}) = \frac{1}{n h} \sum_{i = 1}^{n} K (\frac{x - x_{i}}{h})

(2)

where

K (\frac{x - x_{i}}{h})

is the kernel density function and h is the bandwidth. The bandwidth of this paper uses the system default value in EViews 8.0 software. Following Jiang et al.³⁶ the Gaussian kernel function can be estimated as

Gaussian = \frac{1}{\sqrt{2 π}} e^{- \frac{1}{2} t^{2}}

(3)

Gini coefficient and coefficient of variation

The Gini coefficient is an economic indicator that measures the overall level of income distribution differences by quantitative analysis and is widely used to study and analyze regional differences. This study uses the formula of the Locational Gini Coefficient proposed by Krugman. Its function form can be written as

G = \frac{1}{2 N^{2} \bar{c}} \sum_{i = 1}^{N} \sum_{j = 1}^{N} | c_{i} - c_{j} |

(4)

where G represents Gini coefficient and N represents the number of cities. In this paper, N = 8,

c_{i}

and

c_{j}

represent the energy–environmental efficiency of city i and city j, respectively.

\bar{c}

indicates the average energy–environmental efficiency of the city.

Further, the coefficient of variation is a statistic that measures the dispersion degree of observations in a set of data. It can be written as

C . V = \frac{σ}{| \bar{c} |}

(5)

where C.V. represents the coefficient of variation,

σ = \sqrt{\frac{\sum {(c_{i} - \bar{c})}^{2}}{N}}

, which represents the standard deviation of urban energy–environmental efficiency, and

\bar{c}

represents the average energy efficiency of each city, N is the number of samples (N = 8).

Tobit regression model

In order to avoid the deviation of the results produced by the calculation of the general panel regression model effectively, this paper uses the Tobit regression model to analyze the influencing factors of the energy–environmental efficiency. Tobit standard panel model can be written as

Y_{i}^{*} = α x_{i} + μ_{i}, μ_{i} \sim N (0, σ^{2}), i = 1, 2, \dots, n

(6)

where i denotes the ith decision unit,

Y_{i}^{*}

is the true sample value,

x_{i}

is the influencing factor of the ith decision-making unit, and

μ_{i}

is a random interference term obeying the normal distribution

Y_{i} = \{\begin{matrix} Y_{i}^{*}, Y_{i}^{*} > 0 \\ 0, Y_{i}^{*} \leq 0 \end{matrix}

(8)

Since the efficiency value obtained by the SBM model is between 0 and 1, the explained variables of the regression equation are limited to this interval. Based on the traditional Tobit regression model and selected indicator variables, the function form can be written as

Y_{i, t}^{*} = α + β_{1} X_{1} + β_{2} X_{1}^{2} + β_{3} X_{2} + β_{4} X_{3} + β_{5} X_{4} + β_{6} X_{5} + β_{7} X_{6} + ε_{i, t}

(8)

Y_{i, t} = \{\begin{array}{l} 0, Y_{i, t}^{*} \leq 0 \\ Y_{i, t}^{*}, {0 < Y}_{i, t}^{*} \leq 1 \\ 1, Y_{i, t}^{*} > 1 \end{array}

(9)

where

Y_{i, t}

is the energy–environmental efficiency value of the tth year of the ith city.

β_{1}, β_{2}, \dots, β_{7}

are the coefficients corresponding to each indicator, and

ε_{i, t}

is the random error options, assumed normally distributed with zero mean value and constant variance.^37,38

Indicator selection and data source

Evaluation index selection

The panel data from 2008 to 2017 were used, and input–output relationship of the total factor energy–environmental efficiency was estimated based on the Cobb–Douglas production function. Based on the previous studies, we selected appropriate input and output indicators combining with the unique characteristics of Yangtze River urban agglomeration.

The input indicators are described as follows: capital is measured by capital stock indicators; labor is represented by the number of employees in eight cities of Yangtze River urban agglomeration; energy is denoted by the total energy consumption of Yangtze River urban agglomeration.

The output variables are as follows: The desired output is represented with the gross regional product of the Yangtze River urban agglomeration. This output is measured by the real GDP during the period from 2008 to 2017, where 2008 is the base year. The undesired output is environmental pollution; in view of the availability of data, the synthetic variables of industrial SO₂ and industrial soot emissions are selected as the undesired output indicators. The description of each indicator is shown in Table 1.

2. Data sources

One-decade panel data (from 2008 to 2017) are used for this study, collected from different sources such as China Environmental Statistical Yearbook, the Jiangsu Statistical Yearbook, the China Energy Statistical Yearbook, and the statistical yearbooks and environmental status bulletins of each city. Taking the comparability of the results into account, the unified conversion is based on the 2008 constant price GDP. The GDP data are derived from the Jiangsu Statistical Yearbook from 2008 to 2017. The capital input index draws on the research of Shan³⁹ and uses the “perpetual inventory method” to calculate the capital stock. The specific formula is given in equation (10)

K_{i, t} = I_{i, t} + (1 - δ_{i, t}) K_{i, t - 1}

(10)

where

K_{i, t}

represents the capital stock of i city in year t,

I_{i, t}

represents the fixed asset investment amount of i city in year t, and

δ_{i, t}

indicates the depreciation rate of fixed assets of i city in year t. In order to derive the capital stock of each city from 2008 to 2017, we have determined the initial capital stock, that is the capital stock of each city in Yangtze River Urban Agglomeration in 2008. The formula for calculating the initial capital stock can be written as

K_{i, 0} = \frac{I_{i, 0}}{δ + r_{i}}

(11)

where δ represents the depreciation rate of urban fixed assets, and

r_{i}

represents the growth rate of urban fixed asset investment. Following Young⁴⁰ the

δ + r_{i}

is approximated by 10%. Due to the geographical proximity of each city, the economic diffusion effect is strong, and the depreciation rate of fixed assets is uniformly expressed by 9.6%.

Analysis and discussion of energy–environmental efficiency results

Energy–environmental efficiency measurement results

The energy–environmental efficiency value is calculated by using the SBM model (Table 2), and the histogram of the energy–environmental efficiency change is plotted with these data (Figure 1).

Table 1.

Evaluation index system of energy–environmental efficiency measurement.

Indicator type	Indicator selection	Indicator description	Document basis
Input indicator	Capital investment	Capital stock (converted at constant price in 2008)	Li et al.⁴⁶
	Labor input	Number of employees
	Energy input	Total energy consumption
Output indicator	Economic aggregate	GDP (converted at constant price in 2008)	Wu and Dong⁴⁷
Undesired output indicator	Environmental pollution	Industrial SO₂ emissions
Undesired output indicator	Environmental pollution	Industrial wastewater discharge

GDP: gross domestic product.

Table 2.

Calculation results of energy–environmental efficiency under the undesired output SBM model.

	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	Mean value
Changzhou	0.652	0.674	1.000	1.000	1.000	1.000	1.000	1.000	0.893	0.972	0.919
Nanjing	0.589	0.578	0.636	0.619	0.658	0.695	0.786	0.865	1.000	1.000	0.743
Nantong	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	0.799	1.000	0.980
Suzhou	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Taizhou	0.588	0.617	1.000	0.603	0.592	0.590	0.620	0.754	0.697	0.691	0.675
Wuxi	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Yangzhou	0.600	0.620	0.654	0.640	0.680	0.683	0.756	0.801	1.000	1.000	0.743
Zhenjiang	0.677	0.708	0.741	0.599	0.645	0.659	0.730	0.789	0.741	1.000	0.729
Mean value	0.763	0.775	0.879	0.808	0.822	0.828	0.862	0.901	0.891	0.958	0.849

Figure 1.

Distribution of energy–environmental efficiency in Yangtze River urban agglomeration from 2008 to 2017.

Table 3.

Explanation of influencing factors of energy–environmental efficiency.

	Variable name	Variable description	Document basis
Interpreted variable	Energy–environmental efficiency (Y)	Energy and environment efficiency value	Vaninsky⁴⁸
	Economic development level (X1)	Take the logarithm of GDP per capita	Qi and Luo⁴⁹Bai et al.⁵⁰
	Industrial structure (X2)	Secondary industry gross output value/GDP	Qi and Luo⁴⁹Bai et al.⁵⁰
Explanatory variables	Openness (X3)	Total import and export/GDP	Meng and Li⁵¹
	Technological progress (X4)	Energy consumption per unit of GDP	Li and Zhou⁵²
	Government regulation (X5)	Government environmental investment/GDP	Li and Shi⁵³
	Population density (X6)	Take the logarithm of the resident population per unit area	Zeng⁵⁴

GDP: gross domestic product.

Table 4.

Tobit regression results of influencing factors of energy–environmental efficiency.

Variable	Coefficient	Std. error	z-Statistic	Prob.
X1	−4.597**	2.129	−2.159	0.0309
X1²	0.217**	0.096	2.265	0.0235
X2	−1.588***	0.449	−3.536	0.0004
X3	−0.043	0.037	1.164	0.2445
X4	−0.038*	0.100	0.014	0.0479
X5	1.782***	0.328	−5.429	0.0000
X6	0.187**	0.078	2.389	0.0169
C	24.164	11.809	2.046	0.0407

Note: ***, **, and * indicate the regression coefficients are significant at 1, 5, and 10%, respectively.

Table 5.

Tobit regression results of influencing factors of energy–environmental efficiency.

Variable	Coefficient	Std. error	z-Statistic	Prob.
X1	−2.827*	1.703	−1.660	0.0568
X1²	0.142**	0.076	1.862	0.0226
X2	−1.548***	0.318	4.865	0.0000
X4	−0.160*	0.096	1.670	0.0648
X5	1.627***	0.302	−5.375	0.0003
X6	0.0546**	0.070	−0.781	0.0247
C	14.131	9.561	1.478	0.1394

Note: ***, **, and * indicate the regression coefficients are significant at 1, 5, and 10%, respectively.

It is found that the cities with the highest energy–environmental efficiency are Suzhou and Wuxi. From 2008 to 2017 period, the efficiency value is equal to 1. It indicates that the evaluation unit DEA is effective and at production frontier; Nantong and Changzhou are closely followed, most of the years are also on the front curve, and the annual average energy–environmental efficiency exceeds 0.9. The two cities with the lowest energy–environmental efficiency are Taizhou and Zhenjiang although there are certain gaps compared with other cities; however, it has reached a higher efficiency level around 0.7 and has a large space for energy saving. Overall, it can be seen that the energy–environmental efficiency of the Yangtze River urban agglomeration is in a good situation.

The histogram of the energy–environmental efficiency change is plotted in Figure 1. Since 2008, except Suzhou, Wuxi, and Nantong, the energy–environmental efficiency of the other five cities has shown a rising trend; the energy–environmental efficiency of the six cities has reached 1 by 2017. In terms of overall distribution, the average energy–environmental efficiency of the Yangtze River urban agglomeration reached 0.849 in the past 10 years, and the overall trend showed a gradual upward trend. The energy input–output structure was gradually optimized, and the energy-saving and efficiency-oriented approach was gradually advanced.

Spatial and temporal differences of energy–environmental efficiency

Drawing on the research results of Sun and Li,⁴¹ and taking the rationality of the data into account, the annual efficiency data of 2008, 2011, 2014, and 2017 are used to draw the kernel density distribution map of the energy–environmental efficiency of the Yangtze River urban agglomeration. The kernel density distribution of energy–environmental efficiency of Yangtze River urban agglomeration is given in Figure 2.

Figure 2.

Kernel density distribution of energy–environmental efficiency.

From the shape of the kernel density curve, the energy–environmental efficiency of the Yangtze River urban agglomeration shows a double-peak state from 2008 to 2017. The difference in kernel density values corresponding to the peaks is large; it shows that the spatial difference in energy–environmental efficiency is significant. Our results are in line with Pantaleo et al.⁴² At the same time, as the opening on both sides of the curve shrinks, it depicts that the gap between high energy efficiency and low energy efficiency is gradually narrowing, and the differentiation phenomenon tends to ease.

From the position point of view, the kernel density curve is shifted to the rightward as a whole, and the kernel density value corresponding to the peak gradually increased, indicating that the environmental efficiency is generally increasing. From the perspective of input–output, the energy consumption growth rate in the 10-year period has dropped from 6.1 to 5.4%. At the same time, the discharge of industrial waste water and waste gas in urban agglomeration has also declined. It indicates that by the development and utilization of new clean green energy, the energy structure is gradually adjusted and transformed, and environmental problems are gradually improved, so the energy–environmental efficiency presents a rising trend on the whole.

From the aspect of kurtosis, the nucleus of the kernel density curve tends to be steep on the right side, and the kernel density value of the right peak gradually increases, indicating that most of the urban energy–environmental efficiency gathers toward the high efficiency area gradually. The improvement of Nanjing and Yangzhou is large overall, indicating that the two cities are shifting toward a green and low-carbon development model gradually. The reduction of energy consumption has promoted the improvement of the ecological environment and provided a good space for economic development. In turn, the economic development provides financial support for dealing with energy and environmental issues, and the energy–environmental efficiency has improved.

Evolution of spatial pattern of energy–environmental efficiency

Based on the efficiency data of four equal intervals in 2008, 2011, 2014, and 2017, and the average energy–environmental efficiency from 2008 to 2017, the spatial distribution map of energy–environmental efficiency of the Yangtze River urban agglomeration is drawn. From the energy–environmental efficiency radar map of the Yangtze River urban agglomeration (Figure 3), it can be seen that the energy–environmental efficiency level in urban agglomeration is generally high, and the efficiency values are concentrated between 0.6 and 1.

Figure 3.

Radar diagram of energy–environmental efficiency from 2008 to 2017.

Based on the division method of efficiency value by Liu and Shen⁴³ and Cheng et al.,⁴⁴ the energy–environmental efficiency of the Yangtze River urban agglomeration is divided into four levels: inefficient area (0–0.6), medium efficiency area (0.6–0.75), higher efficiency area (0.75–0.9), and high efficiency area (0.9–1.0), and the energy–environmental efficiency hierarchical map is obtained (Figure 4).

Figure 4.

Layering diagram of energy–environmental efficiency in different years.

Figure 4(a) shows that there were three types of energy–environmental efficiency rating zones in 2008. The energy–environmental efficiency declined step by step from the southeast to the northwest, showing the polar nuclear spatial distribution with Wuxi, Suzhou, and Nantong as the core, forming a central–peripheral spatial layout. Figure 4(b) shows that the spatial structure of energy–environmental efficiency changed slightly in 2011, showing a spatial pattern that spread from the periphery to the middle. Figure 4(c) shows that the energy–environmental efficiency presented a valley-like spatial distribution pattern of “high east-west and middle low” in 2014. Figure 4(d) indicates that the energy–environmental efficiency in 2017 showed a spatial distribution pattern of “overall high.” During this period, except for Taizhou, the other seven cities have reached the level of high efficiency, the spatial polarization effect is weakened, and the diffusion effect is enhanced.

According to the classification of the average energy–environmental efficiency of the Yangtze River urban agglomeration in 2008–2017 (Figure 5), it can be seen that the energy–environmental efficiency shows a small polarization phenomenon in space, and generally forms the spatial distribution feature of energy–environmental efficiency “high in the east and low in the west.”

Figure 5.

Energy–environmental efficiency layering from 2008 to 2017.

Spatial differences in energy–environmental efficiency

Gini coefficient and coefficient of variation are used to compare the overall difference and dispersion of data, and analyze the spatial variation trend of energy–environmental efficiency of Yangtze River urban agglomeration. Figure 6 indicates that although the Gini coefficient fluctuated slightly from 2008 to 2017, it shows a slight downward trend as a whole. It means that the spatial difference in energy–environmental efficiency among cities decreased gradually. The coefficient of variation and the Gini coefficient show a similar fluctuation trend, reflecting a more obvious dispersion degree of energy–environmental efficiency among cities, and further indicate that the spatial difference of energy–environmental efficiency shrinks with time gradually. In general, the Gini coefficient and the coefficient of variation jointly indicate that the energy–environmental efficiency difference among cities shows a gradual narrowing trend from 2008 to 2017.

Figure 6.

Gini coefficient and variation coefficient of energy–environmental efficiency from 2008 to 2017.

Analysis of influencing factors of energy–environmental efficiency

Drawing on existing research⁴⁵ and combining the characteristics of the Yangtze River urban agglomeration, this study chooses six factors: economic development level, industrial structure, openness, technological progress, government regulation, and population density. The detailed description of these variables is presented in Table 3. The Tobit regression model is used to analyze the impact degree on the energy–environmental efficiency of the Yangtze River urban agglomeration.

Tobit regression results

The regression results are demonstrated in Table 4. The external opening factor (X3) does not pass the test at the 10% significance level, indicating that its effect on the energy–environmental efficiency is not significant, so it is removed from the model and the regression is performed again. The results are as follows (Table 5).

Discussion of results

The level of economic development is negatively correlated with the energy–environmental efficiency, while the square of economic development level is positively correlated with the energy–environmental efficiency, indicating that the level of economic development and energy–environmental efficiency show a “U-shaped” state. From 2008 to 2017, the energy–environmental efficiency first decreased and then increased. The reasons are as follows: First, in the early stage of economic development, high energy consumption, high emissions, and high pollution under the high-speed growth mode inevitably lead to high input and low output. Second, with the gradual promotion of the green and low-carbon economy, the improvement of production mode and the strengthening of urban ecological civilization construction, the energy input has become more rationalized, the consumption structure has been improved, and the energy–environmental efficiency has gradually been improved.

The industrial structure, energy consumption per unit of GDP, and energy–environmental efficiency are negatively correlated. This study compares the measured energy–environmental efficiency value with the corresponding industrial structure and finds that the proportion of secondary industry in each city had declined in the past 10 years. The proportion of the tertiary industry had gradually increased. Until 2016, the proportion of the secondary industry in most cities has been less than 50%, while the overall energy–environmental efficiency has also shown an upward trend. The energy consumption per unit of GDP is significantly negatively correlated with the energy–environmental efficiency of the Yangtze River urban agglomeration. The input of energy factors will lead to higher environmental output and environmental degradation, resulting in a decline in energy–environmental efficiency. The degree of openness is negatively correlated with the energy–environmental efficiency, but not significant.

Government regulation, population density, and energy–environmental efficiency are positively correlated. From a macro perspective, local governments can improve the energy consumption structure and environmental pollution of the region effectively by adjusting the energy consumption structure and introducing environmental protection policies. Population density has a significant positive correlation with energy–environmental efficiency, and the improvement of living standard, cultural level, and environmental awareness brought about by increase of population density has a greater promoting effect on energy–environmental efficiency than the negative effect caused by the increase of ecological environment pressure. The increase in population density causes changes in urban function and spatial layout.

Conclusions, findings, and recommendations

Conclusions, findings

To study the energy–environmental efficiency problem at region and city level, this paper uses the input and output index data of the Yangtze River urban agglomeration, and integrates the environmental factors into the SBM model to analyze the evolution characteristics of its energy–environmental efficiency, and uses the Tobit regression model to explore the influencing factors of energy–environmental efficiency, which has certain reference significance for improving regional energy–environmental efficiency. Based on study findings the following conclusion can be drawn:

By calculating the energy–environmental efficiency values of the Yangtze River urban agglomeration, it is found that except Suzhou, Wuxi, and Nantong, the energy–environmental efficiency of other cities shows the upward trend of fluctuation and converges into the high-efficiency area gradually. However, the difference in energy–environmental efficiency among cities still exists and it shrinks with time gradually. From the overall distribution point of view, the energy–environmental efficiency showed the trend of declined first and then rose in the past 10 years, indicating that the energy input and output structure of the Yangtze River urban agglomeration is optimized gradually, and it is gradually developed in the direction of energy saving and high efficiency.

It is found that the energy–environmental efficiency has been increasing continuously over the past 10 years, and the spatial difference has been fluctuating with the passage of time, but the difference still exists. The spatial pattern has evolved from the polar nucleus to the valley type gradually, and it forms a distribution pattern of “overall height” finally. The role of spatial polarization effect is gradually weakened, the diffusion effect is gradually enhanced, and the spatial pattern is gradually developed from a single center to an integrated direction. The spatial layout characteristics of energy–environmental efficiency “high in the east and low in the west” are formed as a whole.

The energy–environmental efficiency of the Yangtze River urban agglomeration is not only related to the changes in the input and output factors of each city, but also related to the overall external environment. Through the Tobit regression model analysis, it is found that the level of economic development and energy–environmental efficiency show “U-shaped” correlation characteristics. Government regulation is the key driving factor to improve the energy–environmental efficiency of Yangtze River urban agglomeration. The increase of population density also has a significant role in promoting energy–environmental efficiency. Industrial structure, technological progress will hinder the improvement of energy–environmental efficiency; the degree of openness to the energy–environmental efficiency is not significant.

Recommendations

There are differences in the development of energy and environmental efficiency among cities in the Yangtze River urban agglomeration. To improve energy–environmental efficiency it is necessary to abandon the extensive economic development model and build a new manufacturing industry cluster along the Yangtze River. For cities with low energy–environmental efficiency in urban agglomeration such as Yangzhou, Zhenjiang, and Taizhou, they should make full use of regional labor force advantages, resource advantages, and industrial advantages, and actively develop regional characteristic industries. For cities with high energy–environmental efficiency such as Suzhou and Wuxi, priority can be given to the development of strategic emerging industries and the reduction of the proportion of heavy industries. Regional central cities should rely on the advantages of resources, talents, and capital to continue to promote the development of modern service industries.

Realize the transformation and improvement of urban agglomeration industry by optimizing the industrial structure. First, it is possible to develop regional characteristic industries and advantageous industries, and build a modern industrial system with complementary advantages and close coordination jointly. Second, establish a new industrial system with low carbon cycle and regional integration, and cultivate a clean energy industry. For example, Nantong, the eastern city, can rely on its coastal advantages to the sea to establish a marine industrial base. Then promote the industrial structure to the direction of rationalization and high level to promote the improvement of energy–environmental efficiency in transformation of industrial structure.

Strengthen the government environmental protection to improve efficiency at the macro level. The government should increase investment in environmental protection and restoration, and explore inter-regional ecological compensation mechanism actively, through the way of establishing ecological compensation fund, improving taxation policy, and exploring market-oriented model to expand the coverage of ecological compensation and improve the benefits of it. Government departments should explore the “Green GDP” accounting system actively, strengthen the corporate environmental credit management, and improve the entry barriers for energy and environmental projects, thereby improving the ecological environment at a macro level and promoting the energy–environmental efficiency of the Yangtze River urban agglomeration.

Establish a regional coordination mechanism to improve the coordinated development of various regions. First of all, we should increase the policy tilt of human resources, finance, and technology in cities with lower energy–environmental efficiency levels. At the same time, the high efficient areas can promote the development of low efficient areas through external collaboration and factor transfer. Second, it is necessary to strengthen the cooperation among municipal governments and break down administrative barriers and factor barriers. It is also necessary to build a coordinated management mechanism for energy consumption and pollution emissions to achieve coordinated and integrated development of energy–environmental efficiency in urban agglomerations.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was financially supported by the National Natural Science Foundation of China (NSFC), Peoples’ Republic of China (Grant number 71263040, 71850410541, 91546117), Humanities and social sciences research project of the Ministry of Education (20YJAZH096), the Startup Foundation for Introducing Talent of Nanjing University of Information Science and Technology (NUIST), Peoples’ Republic of China (Grant number 2017r101), Key Project of National Social and Scientific Fund Program (18ZDA052), Project of National Social and Scientific Fund Program (17BGL142).

ORCID iD

Zhaoqiang Zhong

Zhaoqiang Zhong is a Master student at the School of Management Science and Engineering, Nanjing University of Information Science and Technology, Nanjing, China.

Benhong Peng is a Professor at School of Management Science and Engineering, Nanjing University of Information Science and Technology, Nanjing, China and Binjiang College, Nanjing University of Information Science and Technology, Wuxi, China.

Ehsan Elahi is a Professor at School of Business, Nanjing University of Information Science and Technology, Nanjing, China.

References

Sun

Edziah

Sun

, et al. Institutional quality, green innovation and energy efficiency. Energy Policy 2019; 135: 111002.

Chang

Wen

Zheng

, et al. Is higher government efficiency conducive to improving energy use efficiency? Evidence from OECD countries. Econ Model 2018; 72: 65–77.

Elahi

Weijun

Jha

, et al. Estimation of realistic renewable and non-renewable energy use targets for livestock production systems utilising an artificial neural network method: a step towards livestock sustainability. Energy 2019; 183: 191–204.

Elahi

Weijun

Zhang

, et al. Agricultural intensification and damages to human health in relation to agrochemicals: application of artificial intelligence. Land Use Policy 2019; 83: 461–474.

Dorian

Franssen

Simbeck

DR.

Global challenges in energy. Energy Policy 2006; 34: 1984–1991.

Nejat

Jomehzadeh

Taheri

, et al. A global review of energy consumption, CO₂ emissions and policy in the residential sector (with an overview of the top ten CO₂ emitting countries. Renew Sustain Energy Rev 2015; 43: 843–862.

BP. BP statistical review of world energy, June 2018, https://www.bp.com/zh_cn/china/home/news/reports/statistical-review-2018.html (2018, accessed 20 October 2018).

Moore

Manring

SL.

Strategy development in small and medium sized enterprises for sustainability and increased value creation. J Clean Prod 2009; 17: 276–282.

Zhou

, et al. Energy consumption model and energy efficiency of machine tools: a comprehensive literature review. J Clean Prod 2016; 112: 3721–3734.

10.

Lin

Liu

HX.

Is foreign trade conducive to improving energy and environmental efficiency: taking China’s industrial industry as an example.

Econ Res 2015; 50: 127–141.

11.

Bhadbhade

Yilmaz

Zuberi

, et al. The evolution of energy efficiency in Switzerland in the period 2000–2016. Energy 2019; 191: 116526.

12.

Cao

Xiao

, et al. Finding of urban rainstorm and waterlogging disasters based on microblogging data and the location-routing problem model of urban emergency logistics. Ann Oper Res 2018; 1–32.

13.

Liu

Comprehensive carrying capacity of the urban agglomeration in the Yangtze River Delta, China. Habitat Int 2012; 36: 462–470.

14.

Wang

Total-factor energy efficiency of regions in China. Energy Policy 2006; 34: 3206–3217.

15.

Chang

MC.

Room for improvement in low carbon economies of G7 and BRICS countries based on the analysis of energy efficiency and environmental Kuznets curves. J Clean Prod 2015; 99: 140–151.

16.

Borozan

Technical and total factor energy efficiency of European regions: a two-stage approach. Energy 2018; 152: 521–532.

17.

Ouyang

Mao

Sun

, et al. Industrial energy efficiency and driving forces behind efficiency improvement: evidence from the Pearl River Delta urban agglomeration in China. J Clean Prod 2019; 220: 899–909.

18.

Zhou

DQ.

Total-factor energy efficiency with congestion. Ann Oper Res 2017; 255: 241–256.

19.

Mandal

SK.

Do undesirable output and environmental regulation matter in energy efficiency analysis? Evidence from Indian cement industry. Energy Policy 2010; 38: 6076–6083.

20.

Wang

Kong

FY.

Energy efficiency analysis of Beijing using the DEA-Tobit two stage method. Adv Mater Res 2013; 608: 1210–1214.

21.

JL.

Ecological total-factor energy efficiency of regions in China. Energy Policy 2012; 46: 216–224.

22.

Fan

Zhou

, et al. Industrial energy efficiency with CO₂ emissions in China: a nonparametric analysis. Energy Policy 2012; 49: 164–172.

23.

Lin

. Theoretical exploration of the scientific development concept from the perspective of energy economics—Comment on “Research on Energy Saving, Emission Reduction, Structural Adjustment and Industrial Development Mode Transformation”. Econ Res 2012; 47(3): 154–159.

24.

Song

Zhou

, et al. Does environmental regulation affect energy efficiency in China’s thermal power generation? Empirical evidence from a slacks-based DEA model. Energy Policy 2014; 66: 537–546.

25.

Lin

Tan

Ecological total-factor energy efficiency of China’s energy intensive industries. Ecol Indic 2016; 70: 480–497.

26.

Pan XF, Li LY and

Yang

Study on the dynamic evolution of the spatial and temporal pattern of regional differences in energy efficiency in China. Manag Rev 2012; 24(11): 13–19.

27.

Yang

Liu

Study on china’s energy efficiency and its spatio-temporal variation from 1990 to 2010 based on DEA-ESDA. J Nat Resour 2014; 29: 1815–1825.

28.

Guan

ST.

Spatial pattern and spatial effect of China’s energy eco-efficiency. Acta Geogr Sin 2015; 70: 980–992.

29.

Cui

Kuang

, et al. The changing trend and influencing factors of energy efficiency: the case of nine countries. Energy 2014; 64: 1026–1034.

30.

Yao

Zhou

Zhang

, et al. Regional energy efficiency, carbon emission performance and technology gaps in China: a meta-frontier non-radial directional distance function analysis. Energy Policy 2015; 84: 142–154.

31.

Chen

Zhao

Wang

The evaluation of energy–environmental efficiency of China’s industrial sector: based on super-SBM model. Clean Tech Environ Policy 2019; 1–18.

32.

Xiao

Wang

Shi

, et al. Sectoral energy-environmental efficiency and its influencing factors in China: based on SU-SBM model and panel regression model. J Clean Prod 2018; 182: 545–552.

33.

Huang

The spatial correlation network structure of China’s energy-environmental efficiency and its influencing factors. Resour Sci 2018; 40: 759–772.

34.

Tone

A slacks-based measure of efficiency in data envelopment analysis. Eur J Oper Res 2001; 130: 498–509.

35.

Tu ZG and Liu LK. Efficiency evaluation of industrial sectors in China accounting for the energy and environment factors: based on provincial data by a SBM approach. Econ Rev 2011; 2: 55–65.

36.

Jiang

Huang

Yang

, et al. A novel probabilistic wind speed prediction approach using real time refined variational model decomposition and conditional kernel density estimation. Energy Convers Manag 2019; 185: 758–773.

37.

Elahi

Abid

Zhang

, et al. Domestic water buffaloes: access to surface water, disease prevalence and associated economic losses. Prev Vet Med 2018; 54: 102–112.

38.

Elahi

Abid

Zhang

, et al. Agricultural advisory and financial services; farm level access, outreach and impact in a mixed cropping district of Punjab, Pakistan. Land Use Policy 2018; 71: 249–260.

39.

Shan

HJ.

Re-estimation of China’s capital stock K: 1952–2006.

Quant Econ Technol Econ Res 2008; 25: 17–31.

40.

Young

Gold into base metals: productivity growth in the People’s Republic of China during the reform period. J Polit Econ 2003; 111: 1220–1261.

41.

Sun

Dynamic evolution of China’s marine economic development based on kernel density estimation. Econ Geogr 2015; 35: 96–103.

42.

Pantaleo

Giarola

Bauen

, et al. Integration of biomass into urban energy systems for heat and power. Part II: sensitivity assessment of main techno-economic factors. Energy Convers Manag 2014; 83: 362–376.

43.

Liu

Shen

Analysis on the spatial and temporal evolution pattern of regional energy efficiency in China and its influencing factors. J Nat Resour 2010; 25: 2142–2153.

44.

Cheng

Ren

Chen

, et al. The dynamic evolution of spatial pattern of environmental regulation efficiency in china and its driving mechanism. Geogr Res 2016; 35: 123–136.

45.

Liu

, et al. What are the impacts of tropical cyclones on employment? An analysis based on meta-regression. Weather Clim Soc 2019; 11(2): 259–275.

46.

Zuo

Wang

The influence of environmental decentralization on total factor energy efficiency – based on the analysis of Spatial Dubin Model. Regional Res Dev 2019; 38(1): 123–127.

47.

Dong

Research on total energy efficiency of the Yangtze River economic belt under environmental constraints. China Soft Sci 2016; 03: 73–83.

48.

Vaninsky

Energy-environmental efficiency and optimal restructuring of the global economy. Energy 2018; 153: 338–348.

49.

Luo

Analysis of the differences between economic growth and energy consumption intensity in China. Econ Res 2007; 74–81.

50.

Bai

Niu

Hao

Research of regional energy efficiency based on undesirable outputs and its influential factors: a case of western China. Energy Procedia 2012; 16: 802–809.

51.

Meng

MY.

Evaluation of China’s energy efficiency and its influencing factors based on optimal combination weighting method.

Oper Res Manag Sci 2013; 22(6): 153–160.

52.

Zhou

Can technological advancement improve energy efficiency? – Based on the empirical test of China’s industrial sector.

Manag World 2006; 10：82–89.

53.

Shi

JF.

Energy efficiency analysis on Chinese industrial sectors: an improved super-SBM model with undesirable outputs. J Clean Prod 2014; 65: 97–107.

54.

Zeng

XG.

China’s regional environmental efficiency and its influencing factors. Econ Theory Bus Manag 2011; 103–110.