Journal of Environmental Management 261 (2020) 110227

Contents lists available at ScienceDirect

Journal of Environmental Management

journal homepage: http://www.elsevier.com/locate/jenvman

Research article

Modeling the role of environmental regulations in regional green economy

efficiency of China: Empirical evidence from super efficiency
DEA-Tobit model
Su Shuai , Zhang Fan *
Business School, Shandong University of Political Science and Law, Jinan, 250014, China

A R T I C L E I N F O A B S T R A C T

Keywords: This paper used the panel data of various regions in China from 2007 to 2018 and constructs a green economic
Super efficiency DEA efficiency measurement index system. Following this, the super-efficient DEA model has been employed to
Environmental regulation measure the efficiency of China’s green economy. Then, the Tobit model is used to verify the environmental
Green economy efficiency
regulation influence on efficiency of China’s regional green economy extent and direction. The results show that:
(1) in 2007–2018, the green economy efficiency level of China’s eastern, central and western regions is on the
rise, accompanied by more obvious spatial differences. The green economy efficiency basically shows the spatial
differentiation characteristics of the highest in the eastern region and the lowest in the western region. (2) From
the national perspective, environmental regulation influence on efficiency of green economy presents a “U”
shaped curve that promotes and then suppresses. At the eastern and national levels, environmental regulation has
the same characteristics for green economy efficiency, and both exhibit U-shaped curve characteristics. Whereas,
in the central and western regions shows negative correlation. (3) From the perspective of control variables,
there are significant differences in variables at the national and regional levels. Finally, the study concludes with
some policy suggestion for future green development and the formulation of environmental regulations in China.

1. Introduction main evaluation indicator of sustainable development, the efficiency of

the green economy has become the focus of scholars around the globe.
From November 30 to December 11, 2015, the Twenty-first Meeting Since reform and opening up of China, its economy has developed
of the Parties to the United Nations Framework Convention on Climate rapidly in both speed and quality. GDP has grown from 367.87 billion
Change was held in Paris, France. The primary goal of this conference yuan in 1978 to 90.03 trillion yuan in 2018, an increase of more than
was to reach a new “global legally binding and applicable” global 3000 times than in 1978 (Yang 2019). However, it can also be found that
agreement to reduce emissions. At the core of this meeting was the China’s early development has a negative impact on the environment
suppression or control of carbon emissions and it aimed at fulfilling the while pursuing speed of economic development. Many industries
goal set by the 2009 Copenhagen Climate Conference: reaching an absorbing high input, consume higher energy and of course results in the
agreement to curb global warming and ensuring that the temperature of term of high emissions, which have caused seriously negative impacts on
the earth does not exceed 2 � C before the era of industrial revolution. the ecological environment. Similarly, it can be stated that the country
The Paris Agreement (COP21) further defines the concept of global obtained economic benefits by sacrificing the environment. However,
climate governance as low-carbon green development. The adoption of this mechanism is not sustainable for China as well as the global com­
the Paris Agreement (COP21) demonstrates the clear commitments of all munity. In recent years, over-exploitation of resources and environ­
countries to the development of a low-carbon green economy, and sends mental pollution has become increasingly prominent. Scholars and
a clear and strong signal to the world. The road to low-carbon green policy makers have begun to pay attention to environmental issues. Such
development is the only option for human development in the future and as the study of Yasmeen et al. (2019) emphasized on the modeling the role
it became the core idea of global climate governance. Therefore, the of the government, firms and civil society of the environmental sustain­
development of the green economy has become a strategic goal for ability. Zameer et al. (2020) highlighted the role of cleaner production is
countries around the world to ensure sustainable development. As the China. Wang et al. (2019) confirmed the existence of resource curse

* Corresponding author.
E-mail address: [email protected] (Z. Fan).

https://doi.org/10.1016/j.jenvman.2020.110227
Received 27 November 2019; Received in revised form 12 January 2020; Accepted 29 January 2020
Available online 3 February 2020
0301-4797/© 2020 Elsevier Ltd. All rights reserved.
S. Shuai and Z. Fan Journal of Environmental Management 261 (2020) 110227

hypothesis in the context of China. Overall, they have realized that it is no doctrine”, which is represented by the promotion of enterprise technol­
longer suitable for China’s national conditions by sacrificing the envi­ ogy innovation.
ronment for economic development. Therefore, sustainable development
has become a contemporary issue (Li, 2019; Tian and Li, 2019). The main 2.1. Significant inhibition: “following the cost theory”
goal of economic growth and to achieve sustainable development, one
must unswervingly develop a green economy (Pan et al., 2018; Chen Following “the cost theory” appeared in the context of the slowdown
et al., 2019). In order to achieve the goal of sustainable development, of the US economic growth in the late 1970s. A group of scholars began
China has raised ecological civilization as national strategy. At the same to explore about the factors affecting economic growth. Among them, a
time, it has issued a number of documents to deal with environmental large number of researches were carried out on the impact of environ­
issues, such as the “Overall Plan for the Reform of Ecological Civilization mental institutions on the industry’s total factor productivity and eco­
System” issued by the government of China. The key point of the said nomic growth. Accordingly, the formulation and emphasis over
document is to prevent environmental pollution. China puts forward five environmental policies has significantly increased the production costs
concepts such as innovation, coordination, green, openness and sharing of enterprises and caused the decline of corporate benefits. Gollop and
under the sustainable development strategy. Green is the necessary Roberts (1983) measured and analyzed the impact of sulfur dioxide
condition for sustainable development and it is the basis for achieving emission restrictions on the productivity growth rate of the power in­
high-quality economic development (Deng et al., 2019; Aresta, 2019). dustry during the 1973–1979 business cycles, and found that emissions
Therefore, in the process of resolving the contradiction between the regulations caused significant increases in power generation costs.
economy and the environment, exploring the relationship between Subsequently, Gray (1987) used the data of 450 manufacturing in­
environmental regulation and green economy under macroeconomic dustries in the United States to measure total factor productivity. He also
policies from microeconomic subjects has become a crucial and impor­ argued that government supervision and environmental regulation
tant research topic, and it is also an economy under the characteristics of increased the environmental costs of enterprise production to a certain
socialism in the new era. The objective requirements of the new normal extent, resulting in lower productivity in the manufacturing industry.
provide a basis for decision-making on the strategy for building an Since then, in the light of the cost theory, the role of environmental
innovation-driven economy (Lee and Rodrigue, 2019; Glazyrina and regulations in production cost has been continuously revealed in
Zabelina，2018). empirical research by selecting different research methods, research
The key motivations and contributions of the paper are as follows: (1) angles, and research objects (Jorgeson and Wilcoxen, 1990; Joshi, 2001;
under the new normal state of economic development, it is the key to Xu et al., 2019). Zhao (2019) argued that environmental protection
change the mode of economic development and improve the quality of policies require companies to pay corresponding costs for the environ­
economic development. The traditional method of measuring economic mental pollution and waste of resources and energy they cause, which
volume and development prospects using GDP and economic growth rate leads to the increase of corporate environmental costs. Similarly, envi­
is no longer applicable under the global trend of advocating sustainable ronmental cost management has become the focus of corporate
development and green development. Consider the impact of resource attention.
constraints and environmental pollution on economic efficiency, and
build a correct assessment of green economic efficiency indicators for 2.2. Indirectly promoting impact: “innovative compensation theory”
economic growth performance are impending. Therefore, to deal with it,
it is of great significance to study the impact of green economic efficiency Porter (1991) proposed that although environmental regulation
and environmental regulation on green economic efficiency in China’s would increase the cost of the enterprise, but it would also stimulate
future green development and the formulation of environmental regu­ technological innovation of the enterprise, thereby enhancing the
latory policies. (2) This article attempts to answer the following ques­ competitiveness of the enterprise. Therefore, Porter (1991) thought that
tions: What is the current level of China’s green economy efficiency? How well-designed and well-enforced environmental supervision would be
and to what extent the impact of environmental regulation tools on the beneficial to the environment and the company. Ambec (2013) examined
efficiency of the green economy is different in various regions of China? the key theoretical basis and put-forward empirical evidence about the
Are there significant spatial differences between environmental regula­ Porter hypothesis, and believed that the impact mechanism of environ­
tions and regional green economic efficiency? (3) In view of aforemen­ mental regulation by stimulating enterprises to carry out technological
tioned questions and based on the existing research and its shortcomings, innovation and enhance their comprehensive competitiveness does exist.
this study attempts to expand from the following aspects: The use of the Albrizio (2017) used the new Schumpeter productivity model to
super-efficiency SBM model to measure the green economic efficiency construct a dynamic environmental policy strict index (EPS) that change
considering undesired output, and to compare and analyze the space with changes in industry pollution levels and technological progress, and
distribution and differences of green economic efficiency in various re­ found that tighter environmental policies are related to short-term
gions. The Tobit econometric model was used to examine the impact of growth in industry productivity in the context of technologically
environmental regulations on China’s green economy efficiency. advanced countries. Zhao (2007) empirically analyzed the impact of
environmental regulation on industrial technological innovation using
2. Literature review panel data from 18 industries in China from 1996 to 2004, and found that
environmental regulation had a significant positive effect on R&D
The literature review is based on the public goods theory proposed by expenditure and the number of patent applications lagging behind 3
Coase RH, (1988). Environmental resources without property rights have periods. It shows that environmental regulation has a certain incentive
resulted in market failures and negative externalities. Environmental effect on technological innovation in the medium and long term, and
regulation is an effective means for the government to regulate envi­ confirms the applicability of the “Porter Hypothesis” in China. Kong
ronmental pollution and has an indispensable role in improving the ef­ Fanbin, Yuan and Liu (2013) conducted a correlation analysis of envi­
ficiency of the green economy (Geng and Li, 2019; Du, 2019). Existing ronmental regulation, environmental research and development, and
research on environmental regulation mainly focused on the analysis of green technological progress by establishing a duopoly simultaneous
its mechanism of economic growth and empirical verification using game model. Their research shows that environmental research and
relevant regional or industrial data (Guan, 2019). Mainly, the views can development is an important transmission path for environmental
be summarized into two categories: one is to increase the “production regulation to promote green technological progress. The strength of
cost doctrine”, it is the representative view of the production cost of environmental subsidies will significantly affect the technological prog­
enterprises. However, the second is the “innovation compensation ress of environmental regulation policy portfolios. Further, it is found

2
S. Shuai and Z. Fan Journal of Environmental Management 261 (2020) 110227

that there is a highly universal application strategies worldwide. The between the explanatory variables and the dependent variables, thus
combination of environmental tax and environmental subsidy regulatory avoiding the inadequacy of the wrong conclusions by using the wrong
policies can effectively improve social welfare, which is specifically re­ function form. In economics, DEA methods are often used to solve
flected in the increase in production, reduction in production costs, problems of cost, benefit, and profit, and are also used to seek out the
reduction in pollution generation, and pollution treatment volume distribution of effectiveness, estimation of technology and productivity
through the coordinated progress of pollution prevention technology and progress. The DEA analysis method is evaluated from the perspective of
pollution treatment technology increase. the most favorable decision-making unit (DMU), focusing on the opti­
Xiao and Wu (2019) research shows that environmental regulation mization of the indicators of each decision-making unit, and indicating
policies can alleviate the adverse impact of FDI on regional green the adjustment direction of the relevant metrics. Therefore, this paper
innovation efficiency. Through further research on intermediate mech­ used the DEA principle to construct an analytical model for the evalu­
anisms, it is found that environmental regulation policies mainly ach­ ation of green economic efficiency. The DEA model refers to the “unit”
ieve FDI and regional development by improving the structure of factors or “department” to be evaluated as DMU, and each UDMj ðj ¼ 1; 2; 3; ⋯; nÞ
which indicate the positive regulation effect of green innovation effi­ has r inputs.
ciency. Whereas, in the regional robustness test on the regulation effect The following steps have been followed for the comprehensive
and internal mechanism, it was found that the positive regulation effect analysis.
of environmental regulation is not obvious in the western region of Step1: Let suppose Xj ¼ ðx1j ; x2j ; ⋯; xrj Þ and s term output Yj ¼
China. Yu et al. (2019) found that there is heterogeneity in the forced ðy1j ; y2j ; ⋯; ysj ÞT , where xmj represents the m-th type of input of the j-th
effect of environmental tax on green transformation of enterprises, that UDMj the quantity, ylj denotes the input quantity of the j-th UDMj ,
is, the forced effect of environmental tax on green transformation of xmj > 0，ylj > 0,m ¼ 1,2,3 …,r,l ¼ 1,2, …,s.
enterprises is mainly reflected in private enterprises. Whereas, enter­ Step 2: Model construction
prises with high marketization levels, low financing constraints and with
high information transparency, their performance is not significant in minθi :
the context of the state-owned enterprises. In contrast, in the context of
enterprises with low marketization levels, high financing constraints,
8X n

and with low information transparency their performance is significant. >> λ X þ s ¼ θ i Xi i ¼ 1; 2; :::n:
In summary, the “following the cost theory” and the “innovation
<j¼1
j j

Xn
compensation theory” form a paradox and are stalemate. The analysis λj Xj sþ ¼ Yi i ¼ 1; 2; :::n: (1)
found that: in empirical studies, studies that reached the conclusion of >> j¼1

promoting environmental regulation mostly used economic investment- : λ � 0; j ¼ 1; 2; :::n:

j

type index data, and the literature that reached the “inhibition” sþ � 0; s � 0
conclusion mostly used government-regulated data to characterize
environmental regulation. At the same time, according to the theory of Where (equation (1)), θi is the effective value of UDMj , and the effective
comparative advantage, environmental regulation directly affects the value is closer to 1, the more effective the input of this DMU is. The
flow of investment in the market by acting on the economic structure validity judgment method is: if θi ¼ 1, UDMj is called valid or weak
and economic policy environment, triggering competition in the envi­ effective for DEA. When sþ ¼ s ¼ 0 , ​ UDMj is called DEA is valid; if
ronment (Wang, 2013), and its space spillover effect cannot be ignored. θi <1, then UDMj is invalid for weak DEA. The input data slack variable is
Existing research focused on the role of single environmental regulation s , and s represents an input surplus, that is, an unused resource. If
(Gong, 2018; Xu and Lin, 2019), and scholars discussed the impact of s 6¼0 indicates that the output is unchanged, the input can also reduce
multiple environmental regulations on traditional economic growth s ; the output slack variable for sþ , sþ means that there is insufficient
(Chen, 2019). And also, the geospatial effects of environmental regula­ output, and s 6¼0 means that if the input is constant, the output can also
tion (Qin and Ge, 2019), without combining multiple environmental increase ​ sþ . Therefore, if a DMU is not valid, DEA can be effectively
regulations with green economic efficiency, there is a lot of room for adjusted by not writing input and output indicators. Assuming a fixed
expansion in this field. Similarly, fresh evidence is the need of time to output level, the input variable is adjusted to ​ Xi ¼ θi Xi s . If assuming
understand the role of environmental regulations in regional green
a fixed input level, the output variable is adjusted to Yi ¼ Yi þsþ .
economy efficiency.
Step3: Super-efficient DEA model construction.
With traditional DEA model when analyzing the green economic
3. Methods
efficiency, multiple DEAs may be effective. At this time, their compre­
hensive technical efficiency index is θ ¼ 1, which makes it impossible to
3.1. Model construction and variable selection
further evaluate the effective of DEA. Therefore, the use of a super-
efficiency model allows for a more in-depth production efficiency
3.1.1. Model construction
ranking of all DEA effective decision making units. The super-efficient
On the basis of discussion in the previous part and using the quan­
DEA (SE-DEA) model is as follows.
titative evaluation model of green economic efficiency, in this part, an
8
evaluation model suitable for the evaluation of green economic effi­ >
>
h �X m Xs �i
> min θ ε sþ
ciency is being established. Fuzzy evaluation method and analytic hi­ >
>
>
si þ r
>
>
erarchy process are commonly used methods for enterprise financing
i¼1 r¼1
>
>
>
> X n
efficiency evaluation. Because it is necessary to determine the priority of >
>
> λj Xij þ s � θX0
>
>
each indicator, it is difficult to avoid subjectivity, which leads to the >
>
> j¼1
>
unfairness of evaluation results. The DEA method proposed in this paper <
(2)
solves this problem well. Similarly, it does not need to assume a specific
j6¼k
>
>
>X
functional form. Thus, the weight in the model is generated by mathe­ > n
>
>
> λj Yj sþ ¼ Y0
matical programming, which eliminates the subjective factors. >
>
>
>
> j¼1
The DEA method is a non-parametric, multi-factor productivity >
>
>
>
>
analysis tool that estimates the relative effectiveness of multiple inputs >
>
>
j6¼k
>
and multiple outputs. The DEA method does not need to pre-estimate the : λ � 0; j ¼ 1; 2; :::; n sþ � 0; s � 0
j
parameters, nor does it need to establish a functional relationship

3
S. Shuai and Z. Fan Journal of Environmental Management 261 (2020) 110227

Among them, λ is the weight variable of DMU, θ is the parameter to the DEA efficiency boundary. Efficiency data may be truncated. For such
be determined, slack variable sþ r sr , X is the input quantity, and Y is the a situation where multiple samples become a certain limit value within a
output quantity. The solution to the model is denoted by θ* . If θ* <1, it specific range, the general regression method cannot explain the dif­
indicates that there is a virtual decision unit whose output is not lower ference in properties between the limit value and the non-limiting
than the output of the first j0 decision unit, and the input ratio is the observation value to ensure the consistency of the estimates. To esti­
input of the j0 decision units. Below, this shows that j0 is non-DEA valid. mate the panel data Tobit model have two types of methods, fixed effect
If θ* ¼ 1 and the slack variables are all 0, then the j0 decision unit is DEA and random effect and both can be used. If a fixed effect model is set, the
valid; and θ* <1, but the slack variable is not 0, the j0 decision unit is estimated value obtained using the maximum likelihood estimation
valid for weak DEA. method is biased and inconsistent. False and marginal effects are
underestimated, so we use random effects models for maximum likeli­
3.1.2. Variable selection hood estimation. Therefore, the research in this article will use the Tobit
The green economic efficiency index pursues the balance between model to analyze the influencing factors of China’s green economy ef­
social and economic benefits and eco-environmental benefits. It needs to ficiency, and then extract the main factors affecting China’s green
consider the efficiency of using input factors in production operations, economy efficiency, and provide sufficient empirical data for subse­
that is, the ability to obtain expected output, and at the same time, the quent policy recommendations. The development of the model is as
additional production acquired in production and operation activities. follows:
Similarly, it means that the resource environment at the expense of the �
αit þ βT xit þ εit yit � 0
desired output is taken into account. yit ¼ (3)
0 yit < 0
This paper builds an indicator system for inputs and outputs required
for green economy efficiency (Table 1). Since the data on the capital Among them, the interpreted variable yit is the green economy effi­
stock of each province cannot be obtained directly from the statistical ciency of the t-th year of the i-th region. Xit is the explanatory variable,
yearbook, the paper refers to the method of Pittman (1983) to measure βT is the unknown parameter, εit ～ N (0，σ2) is the intercepted
the capital stock through the application of the perpetual inventory regression model of the panel data, the explanatory variable xit takes the
method. For the energy consumption input, this paper will total the actual observation value, and the interpreted variable yit is in a
primary energy consumption. Input as an energy input, using linear restricted manner. When the value yit �0, it takes the actual observa­
interpolation to fit missing data in some years; for labor, use partici­ tion; when yit < 0, the observation is truncated to 0. αit is the fixed effect
pating of employment in each province to estimate the labor input in the of the t-year of region i, which is an unknown constant.
model.
Outputs have both expected and undesired outputs, where expected 3.2.1. Variable selection
output refers to regional GDP, used to measure the actual situation of the
regional economy. Domestic ecological pollution is mainly based on (1) The explanatory variable (dependent variable). The explanatory
industrial pollution. As a national key monitoring target, its relevant variable in this paper is Green Economic Efficiency (GF).
data is relatively complete. Industrial wastewater, industrial waste gas (2) Independent variables. This paper mainly reveals that the effi­
and solid waste were selected as undesired output indicators. Relevant ciency of green economy is affected by environmental regulation
data were collected from China Environmental Statistics Yearbook. (ER). Therefore, the author chooses environmental regulation
The paper studies the annual data of 30 provinces (excluding Tibet, (ER) index. The cost of using industrial wastewater, industrial
Hong Kong, Macao and Taiwan) from 2007 to 2018. The research object waste gas and solid waste in various provinces and municipalities
is for industrial enterprises listed companies, excluding ST Company, represents the environmental regulation in the proportion of the
relevant data are collected from China Statistical Yearbook, China Sci­ main business costs and sales value of large and medium-sized
ence and Technology Statistical Yearbook and provincial statistical industrial enterprises in the region. Data on pollution control
yearbooks. For minimize the impact of price factors on the conclusions costs in various regions were collected from China’s environ­
of the measurement analysis, the economic data is converted into mental statistics yearbook. The main business costs and sales
comparable price indices using historical price indices. value data of large and medium-sized industrial enterprises were
collected from the website of the State Administration of
Statistics.
3.2. Tobit model (3) Control variables: The following indicators are selected as the
control variables: ① GDP per capita. The per capita GDP of each
The regional green economy efficiency measured by the super- province and its square term verify the environmental Kuznets
efficiency DEA model is affected not only by input-output indicators, curve (KEC). ② Industrial structure (PC). Because the secondary
but also by many external factors. Therefore, the degree of influence of industry is more polluting and usually has a suppressive effect on
these external factors on green economic efficiency needs to be the green economy, this article uses the proportion of the sec­
measured. This article uses the truncated regression method Tobit model ondary industry to the entire industry to represent the industrial
to analyze the influencing factors of regional green efficiency. The Tobit structure. ③ Urbanization (URB). To some extent, urbanization
model is used because when using the DEA method to estimate the green reflects the process and degree of population concentration in
economic efficiency of each region, there may be one or more regions at cities. In this process, the economy and environment will inevi­
tably be affected. This article uses the proportion of the perma­
Table 1 nent population in each province and city to the total population
Green economic efficiency measurement index system. in the region to indicate urbanization. ④ Investment in fixed
Indicator index classification Index composition
assets (K). Use the ratio of investment in fixed assets in different
provinces to the GDP of each province. Increasing society’s fixed
Input indicators Labor input Employment in various regions
assets and expanding production scale are important means of
Capital investment Capital stock
Energy input Total energy consumption economic development. ⑤ Energy consumption structure (CS).
Output indicator Expected output GDP Considering that the relationship between pollutant emissions
Output indicator Industrial exhaust emissions and energy consumption is very close, this article reflects the
Industrial wastewater discharge
Industrial solid waste discharge

4
S. Shuai and Z. Fan Journal of Environmental Management 261 (2020) 110227

consumption energy structure in terms of the proportion of coal the main affecting factors.
consumption in total energy consumption.

3.2.2. Tobit model construction 4.2. Tobit regression results and analysis
In this paper, sample of 30 provinces are divided into the eastern
region, the central region and the western region according to the pro­ 4.2.1. Tobit regression results
vincial and municipal regional standards. The existing research shows In this paper, the estimated results obtained by using the software
that the relationship between the two is not certain, but presents Stata14.0 are shown in Table 3.
nonlinear characteristics. Therefore, the environmental regulation
quadratic item is introduced into the analysis model. Because the en­ 4.2.2. Analysis of regression results
terprise green economy efficiency is in the process of gradual accumu­ (1) Environmental regulation perspective. From the national sample,
lation, so the lag is added in model. by measuring the lag term of the explanatory variable lnGF, it is found
that the influence coefficient is significantly positive, which is consistent
lnGFit ¼ β0 þ β1 lnGFit 1 þ β2 lnERit þ β3 ðlnERÞ2it þ β4 lnGDPit þ β5 ðlnGDPÞ2it þ with the Porter hypothesis analyzed above. Porter (1991) also believes
β6 lnPCit þ β7 lnURBit þ β8 lnKit þ β9 lnCSit þ μit that the environment and economy can achieve a “win-win”. This situ­
(4) ation means that green economic efficiency is in the process of contin­
uous accumulation, and at the same time, it can verify that the setting of
4. Results and findings the dynamic model is reasonable. Further, by measuring the quadratic
term of the core explanatory variable lnER, it is found that the influence
4.1. China’s regional green economy efficiency calculation results coefficient is significantly positive, which means that the impact of
environmental regulation on green economy efficiency is revealed as
Using the DEAP 2.1 software, the determined input and output in­ “U” shaped. When environmental regulatory efforts remain within a
dicators are substituted into the model and the results are shown in certain “critical point”, its impact on green economic efficiency is
Table 2. inhibiting, but if this intensity exceeds this after the “critical point”, its
In terms of regions, the overall green economy efficiency is higher in influence is promoted. In a sense, it can be stated that the relationship
Beijing, Shanghai, Tianjin, and Guangdong, both higher than 0.9, and between the two is not an absolute positive or negative correlation, but
the green economy efficiency is below 0.5 in Hebei, Liaoning, Guizhou rather the relationship is complex.
and Gansu. However, in the context of Qinghai, Ningxia and Xinjiang, as From the sub-samples of the three regions, the conclusion is obtained
these areas are less green, and five of them are located in the west, by estimating the core variable lnER: the coefficient values of the pri­
indicating that most of the green economy in these areas is less efficient. mary and secondary terms of lnER in the eastern region are significantly
The values of many areas under investigation is found between 0.5 and positive, and the coefficients of the primary terms of lnER in the central
0.9. It can be seen that the green economy efficiency shows certain and western region are significantly negative. This means that the
heterogeneity in different regions, so need to further analyze which is impact of environmental regulation on green economic efficiency in the
central and western regions is inhibiting. But, the impact of the two in

Table 2
Different regional green economy efficiency in China.
Region 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Mean

East Beijing 2.331 2.335 2.887 3.112 3.234 3.351 3.457 3.503 3.543 3.588 3.598 3.624 3.214
Tianjin 0.784 0.791 0.799 0.808 0.816 0.826 0.838 0.869 0.921 1.101 1.144 1.168 0.905
Hebei 0.289 0.303 0.321 0.349 0.367 0.383 0.422 0.438 0.449 0.479 0.498 0.522 0.402
Liaoning 0.301 0.313 0.324 0.336 0.341 0.347 0.356 0.359 0.367 0.372 0.378 0.385 0.348
Shanghai 2.773 2.798 2.816 2.845 2.856 2.885 2.912 2.934 2.947 2.968 2.977 2.991 2.892
Jiangsu 0.801 0.813 0.816 0.821 0.833 0.825 0.817 0.834 0.841 0.856 0.868 0.897 0.835
Zhejiang 0.811 0.824 0.832 0.836 0.839 0.842 0.854 0.863 0.867 0.873 0.878 0.811 0.844
Shandong 0.667 0.678 0.689 0.695 0.709 0.711 0.725 0.734 0.746 0.758 0.765 0.772 0.721
Shandong 0.478 0.485 0.496 0.506 0.518 0.522 0.526 0.537 0.546 0.552 0.558 0.563 0.524
Guangdong 0.894 0.906 0.915 0.927 0.937 0.948 0.952 0.957 0.963 0.969 0.975 0.982 0.944
Hainan 0.789 0.794 0.802 0.812 0.819 0.825 0.836 0.847 0.852 0.861 0.873 0.879 0.832
Eastern region mean 0.993 1.004 1.063 1.095 1.115 1.133 1.154 1.170 1.186 1.216 1.229 1.236 1.133
Central Shanxi 0.401 0.411 0.416 0.424 0.431 0.442 0.446 0.454 0.462 0.471 0.484 0.491 0.444
Jilin 0.511 0.513 0.516 0.524 0.531 0.536 0.547 0.554 0.564 0.569 0.574 0.579 0.543
Heilong jiang 0.589 0.593 0.609 0.616 0.627 0.636 0.642 0.648 0.655 0.667 0.672 0.679 0.636
Anhui 0.689 0.698 0.708 0.714 0.726 0.745 0.747 0.759 0.765 0.772 0.777 0.781 0.740
Jiangxi 0.783 0.789 0.794 0.801 0.812 0.819 0.825 0.829 0.837 0.856 0.864 0.869 0.823
Henan 0.501 0.513 0.518 0.525 0.529 0.537 0.544 0.549 0.555 0.568 0.574 0.583 0.541
Hubei 0.488 0.496 0.501 0.516 0.528 0.534 0.543 0.548 0.554 0.559 0.567 0.575 0.534
Hunan 0.653 0.658 0.664 0.674 0.682 0.693 0.709 0.717 0.729 0.736 0.747 0.758 0.702
Central region mean 0.577 0.584 0.591 0.599 0.608 0.618 0.625 0.632 0.640 0.650 0.657 0.664 0.620
West Neimenggu 0.583 0.591 0.599 0.609 0.615 0.623 0.636 0.648 0.658 0.669 0.672 0.686 0.632
Guangxi 0.611 0.617 0.621 0.634 0.639 0.647 0.658 0.669 0.674 0.687 0.696 0.711 0.655
Chongqing 0.501 0.515 0.521 0.533 0.539 0.548 0.552 0.559 0.567 0.575 0.588 0.604 0.550
Sichuan 0.512 0.518 0.524 0.533 0.549 0.556 0.564 0.571 0.581 0.587 0.598 0.616 0.559
Guizhou 0.312 0.314 0.316 0.319 0.323 0.328 0.337 0.341 0.353 0.358 0.359 0.366 0.335
Yunnan 0.501 0.509 0.513 0.517 0.523 0.526 0.527 0.531 0.533 0.537 0.538 0.542 0.525
Shanxi 0.516 0.519 0.526 0.529 0.532 0.537 0.544 0.549 0.559 0.61 0.556 0.573 0.546
Gansu 0.211 0.218 0.225 0.236 0.243 0.256 0.261 0.277 0.289 0.301 0.322 0.344 0.265
Qinghai 0.361 0.365 0.372 0.381 0.379 0.384 0.389 0.388 0.396 0.398 0.392 0.399 0.384
Ningxia 0.255 0.263 0.268 0.273 0.281 0.287 0.289 0.298 0.299 0.301 0.303 0.309 0.286
Xinjiang 0.347 0.351 0.359 0.364 0.371 0.379 0.385 0.391 0.398 0.407 0.424 0.427 0.383
Western region mean 0.428 0.435 0.440 0.448 0.454 0.461 0.467 0.475 0.482 0.494 0.495 0.507 0.465

5
S. Shuai and Z. Fan Journal of Environmental Management 261 (2020) 110227

Table 3 structure is achieved, it will also significantly improve the efficiency of

Regression results. regional green economy. This article is consistent with its conclusions. .
Variable National East Central West From the perspective of urbanization, the impact of the urbanization
rate on the national green economy efficiency is negative, but failed to
α 0.201** (0.000 NA 1.128* (0.000 1.172* (0.000
1) 1) 0) pass the significance test in the central and western regions, that is, the
lnGFit-1 0.741*** 0.813*** 0.819*** 0.836*** impact of urbanization on the green economic efficiency of the central
(0.000 1) (0.000 1) (0.000 5) (0.0011) and western regions is not obvious enough, the main reason is that the
lnER 0.221*** 0.319*** 0.216*** 0.077*** level of urbanization in the west is relatively low and has not yet a
(0.000 51) (0.0006) (0.0011) (0.0013)
(lnER)2 0.011** 0.032* 0.036** 0.024***
significant impact on the efficiency of the regional green economy, but
(0.0031) (0.0711) (0.0512) (0.0012) further urbanization will cause a decline in the efficiency of the green
GDP 0.128*** 0.121* (0.052 0.148* 0.259*** economy. Urbanization in the eastern region has a significant negative
(0.000 1) 1) (0.0821) (0.0011) impact on the efficiency of the green economy. The level of urbanization
(lnGDP)2 0.014*** 0.012*** 0.026*** 0.032***
in the east is rapidly advancing. The increase in urbanization is
(0.000 1) (0.000 2) (0.0001) (0.000 5)
0.011** 0.023* 0.128** 0.173** accompanied by excessive development of resources and energy con­
PC
(0.0531) (0.0911) (0.0322) (0.023 2) sumption, resulting in a decline in the efficiency of the green economy.
URB 0.115*** 0.012* 0.002 0.001 This point is quite different from other scholars. For example, Wang and
(0.000 1) (0.0644) (0.4044) (0.2323) Ren (2016) research believes that the green economy efficiency in the
K 0.034*** 0.118 0.026** 0.366***
(0.0012) (0.2331) (0.0421) (0.0011)
eastern region is higher, and the green economy efficiency of some
CS 0.339*** 0.014*** 0.214*** 0.361*** provinces and cities in the region is already in the rising stage of the
(0.0012) (0.0011) (0.0004) (0.001 2) U-shaped curve. The green economic efficiency of most provinces and
cities is still in the decline stage of the U-shaped curve, that is, the green
economic efficiency decreases with the increase of urbanization level.
the eastern region is facilitating, because the result of the estimation
From the perspective of energy consumption structure, the energy
(lnER) is a significant positive value, which can verify the national level
consumption structure has a negative correlation with the efficiency of
empirical analysis results, which are basically the same, and there is a
the green economy in the country and the three regions, indicating that
significant “U” -type relationship between them. This is in line with
the higher the proportion of coal consumption in energy consumption,
domestic scholars Li and Tao (2012), Song et al. (2017). The results were
the greater the impact on the environment. Fixed assets have a negative
consistent, and they all came to U-shaped characteristics of environ­
correlation with the efficiency of the green economy across the country
mental regulatory impacts. It is concluded through the regression
and the central and western regions, which indicate that excessive in­
analysis by region that the environmental regulations adopted in the
creases in fixed assets increase energy consumption and have a negative
process of optimizing the adjustment of the supply-side structure will
impact on the efficiency of the green economy. Although this effect is
help to improve the efficiency of the green economy in the eastern re­
negatively correlated in the east, but it is not significant. This is
gion, but have a suppressive effect on the efficiency of the green econ­
consistent with the findings of most scholars (Su, 2018; Yu and Zhang,
omy in the central and western regions (Yang Shouyun, 2019). The
2016; Tang and Zhang, 2015).
reason is that, on the one hand, different regions have large differences
in regional structure, level of economic development, and level of
5. Conclusions and implications
openness; on the other hand, it may be related to a higher degree of
economic development, a better market mechanism, and the degree of
5.1. Conclusions
understanding and grasp of policies. The effectiveness of environmental
regulations in developed regions in improving the efficiency of the green
Based on the inter-provincial panel data of 30 provinces in China
economy is even more pronounced.
from 2007 to 2018, the paper analyzes the impact of environmental
(2) Control variables analysis. The positive impact of GDP on the
regulation on the efficiency of green economy and verifies the extent and
whole country indicates that overall economic growth has promoted the
direction of impact. Firstly, the paper used Super-SBM model to incor­
increase of national income, and more funds can be used for environ­
porate undesired output into the non-parametric efficiency model. The
mental governance, which is conducive to environmental improvement.
measurement considers the green economy efficiency under the
In terms of regions, the impact on the east is positive and the impact on
resource environment constraints. Then, the panel Tobit model is used
the Midwest is negative, indicating that the economic growth of the
to further explore the relationship of them. From the characteristics of
Midwest depends on the excessive use of energy, which exceeds the
time evolution, green economy efficiency of China’s eastern, central and
environmental capacity and reduces the efficiency of the green econ­
western regions showed a volatility upward trend from 2007 to 2018,
omy. At the same time, its square term is positively verifying the
accompanied by more obvious spatial differences. From the perspective
inverted “U” relationship of GDP. Zhang et al. (2018) also found that
of spatial characteristics, the green economy efficiency basically shows
there is a positive relationship between economic development and
that the eastern region is largest, the western region is lowest. Under the
green economic efficiency, but no U-shaped relationship was found. It is
general trend of green development and the common action of the whole
different from this article.
society, although China’s green economy efficiency and environmental
From the perspective of industrial structure, the impact on the effi­
regulation level show obvious spatial differences, the green economy
ciency of the green economy is negative, indicating that the higher the
efficiency of each region has shown a steady upward trend with the
proportion of the secondary industry, the more it is not conducive to the
passage of time.
improvement of the efficiency of the green economy, mainly because the
On the basis of the Tobit model verification, it is found that the
secondary industry is industrial, and the energy consumption of the
process of environmental regulation at the national level and the eastern
industry is relatively large. The pollution is also large, which is not
region has an inverted “U” shape, before the turning point, environ­
conducive to environmental improvement. However, the impact on the
mental regulation can improve green economy efficiency. After the
east is significantly smaller than that in the middle and west, which
turning point, it will hinder the improvement of the efficiency of the
indicates that the east has a stronger ability to handle pollution. Yu and
green economy. In the central and western regions environmental
Zhang (2016) concluded that the upgrading of industrial structure has
regulation has a significant negative correlation with the efficiency of
significant direct and indirect effects on the growth of green economic
the green economy. Meanwhile, from the perspective of control vari­
efficiency. This article believes that if the upgrading of industrial
ables, the influences of GDP at the national and eastern regions are

6
S. Shuai and Z. Fan Journal of Environmental Management 261 (2020) 110227

positive, and the impact on the central and western regions is negative. 5.3. Limitations
And the squared term is the inverted “U” relationship that validates the
impact of GDP. Considering the industrial structure, the influences on Due to the limited research capacity, the following deficiencies exist
the efficiency of the green economy is negative, but the impact on the in this paper: (1) when measuring the efficiency of the green economy,
east region is significantly smaller than central and western regions, the calculation results are different due to the different inputs in the
indicating that the eastern region has a stronger ability to deal with input-output function, different settings of undesired outputs, and ex­
pollution; The urbanization rate is a negative index when in the national pected outputs. It may make the calculation results inaccurate. (2) When
and eastern regions, but it not passed the significant test in the central verifying the impact of environmental regulations on the efficiency of
and western regions, that is, the impact of urbanization on the green green economy, although this article measures different environmental
economy efficiency of the central and western regions is not obvious. In regulations, it can be further improved from other perspectives. (3) In
terms of consumption structure, the energy consumption structure is the “Porter hypothesis” certification process, the selection of the mea­
negatively correlated with the efficiency of green economy in the surement of technological innovation indicators is too single, and the
country and the three major regions, indicating that the higher the problem is more obvious from the national perspective. It may be more
proportion of coal consumption in energy consumption, the greater reasonable to conduct research from a specific industry. (4) The paper
impact on the environment; the fixed assets are negatively related to did not reach the optimal R2 in the benchmark regression, mainly
green economic efficiency for the national, central and western regions, because of the green economy. The measurement of efficiency is based
and negatively correlated in the east, but not significant. on the variables synthesized under the SBM-DDF model, which makes
the study of the relationship between environmental regulation and
5.2. Policy suggestions green economic efficiency less well-fitted. According to Woodridge’s
book, R2 is only one of the indicators for judging the quality of the
Strengthen the seriousness of policy implementation and reflect the dif­ model, not all of them. Especially when using microdata and a large
ferences. From a regional perspective, it is necessary to strengthen the sample size, the absolute coefficient can be small, but this does not
enforcement of environmental regulations, eliminate the blind pursuit of represent the model bad. However, it can be further tested.
GDP, and pursue its lifetime responsibility for causing damage to the
ecological environment. At the same time, when implementing envi­
ronmental regulations, it is necessary to pay attention to the regional 5.4. Future outlook
differences in the impact of environmental regulations on the efficiency
of green economy, to reflect regional differences, and to introduce more (1) Draw on more scientific methods to calculate the green economy
precise incentives and restraints for different regions on the premise of and assess the degree of environmental and economic coordina­
following unified environmental standards. Make environmental regu­ tion more accurately. As China’s research on the coordinated
lation actionable, not bypass it. In view of the impact of environmental development of the green economy is still immature, there is less
regulations on green economy efficiency, it has a U-shaped character­ literature on quantitative analysis of green economy efficiency. In
istic, so different measures must be taken at different stages. For the the future, it can be studied that developed countries have
backward areas of the Midwest, we must strictly control from the established a good green economy system, thereby establishing a
beginning to avoid the old road of “polluting first, and then treating it”. more scientific green economy evaluation system and measuring
We must focus on raising the entry threshold. When environmental green economic efficiency more accurately.
regulations cannot be met, we must introduce green technology instead (2) Improve the selection of variable indicators for measuring green
of replace with “black” technology in order to narrow the gap with economic efficiency. For the input variables for measuring the
developed regions. green economic efficiency, the selection of the undesired output
Establish a regional environmental protection compensation mechanism, value can be more diversified. Compared with the measurement
and reasonably allocate regional environmental protection costs. Regional of the efficiency gap, the more suitable input and output are
compensation for environmental protection is an important guarantee selected.
for effectively regulating regional environmental behavior. According to (3) For the study of environmental regulation and green economy
the “polluter pays” and “beneficiary compensation” principles, through efficiency mechanism, the scope of research can be reduced.
the horizontal or vertical fiscal transfer payments, the regional ecolog­ Select a certain type of enterprise to conduct a long-term research
ical environmental protection costs should be reasonably reallocated on it, study the key mechanism of the company’s production
between relevant administrative areas. This may include the definition products, and provide targeted suggestions for the enterprise.
of compensation subjects and objects and their negotiation mechanism,
the calculation and compensation mechanism of compensation funds, CRediT authorship contribution statement
the basic elements of compensation fund collection, use and manage­
ment. The design of the environmental protection area compensation Su Shuai: Conceptualization, Methodology, Software, Data curation,
mechanism can effectively encourage the full provision of regional Writing - original draft. Zhang Fan: Visualization, Investigation, Su­
environmental public goods, solve the reasonable returns of environ­ pervision, Software, Validation, Writing - review & editing.
mental investors, and help to encourage people to invest in environ­
mental protection and increase the value of environmental capital. References
Give full plays to environmental regulations. The positive impact of
environmental regulations on economic efficiency must also recognize Albrizio, S., 2017. Environmental policies and productivity growth: evidence across
industries and firms. J. Environ.Econ.Mana. 81, 209–226.
the staged nature of economic development, and take advantage of the
Ambec, S., 2013. The porter hypothesis at 20: can environmental regulation enhance
positive externalities brought about by regional economic advantages, innovation and competitiveness? Rev. Environ. Econ. Pol. 1, 2–22.
make full use of the advantages of talents, technologies and capital in the Aresta, M., 2019. Carbon dioxide utilization: the way to the circular economy. Gree.
region, be people-oriented, innovation-oriented, and actively promote Gase.: Sci. Tech 4, 610–612.
Chen, H., Yao, M., Chong, D., 2019. Research on institutional innovation of China’s green
the industry. Similarly, the rationalization and upgrading of the struc­ insurance investment. J. Indu. Inte. Mana. 1, 67–70.
ture promotes the improvement of the efficiency of the green economy. Coase RH, 1988. The Coase theorem and the empty core: a comment. J. Law Econ. 24 (1),
183–187.
Deng, Z., Lv, L., Huang, W., 2019. A high efficiency and low carbon oriented machining
process route optimization model and its application. Int. .J. Precis. Eng. Manuf.
Tech. 1, 23–41.

7
S. Shuai and Z. Fan Journal of Environmental Management 261 (2020) 110227

Du, W.J., 2019. Can environmental regulation policies promote the governance of excess Wang, C.X., Ren, Y.J., 2016. Spatial spillover effects of green innovation efficiency in
capacity? - micro-evidence from Chinese industrial enterprises. J. China Univ. high-tech industries. Indu. Econ. Rev. 6, 76–84.
Geosci.: Soc. Sci. Ed. 4, 57–71. Wang, W.P., 2013. Environmental regulation, space spillover and regional industrial
Geng, Y.Q., Li, Y.Y., 2019. Environmental regulation, government subsidies and competitiveness. Chinese Popul. Res. Environ. Terr 8, 123–130.
corporate R & D investment-Empirical Analysis Based on Chinese manufacturing Wang, R., Zameer, H., Feng, Y., Jiao, Z., Xu, L., Gedikli, A., 2019. Revisiting Chinese
enterprises. Jian. Forum 7, 11–20. resource curse hypothesis based on spatial spillover effect: a fresh evidence. Resour.
Gong, Y.Y., 2018. Impact of environmental regulations on China’s green economy Pol. 64, 101521.
efficiency: based on panels in 30 provinces analysis of data. Urba. Prob. 8, 68–78. Xiao, Y.F., Wu, Y., 2019. Green innovation efficiency FDI, environmental regulation
Gollop, F.M., Roberts, M.J., 1983. Environmental regulations and productivity growth: policies and regional. Chon. Qing. Univ. Tech. (Soc. Sci.) 6, 50–62.
the case of fossil-fueled electric power generation. J.Poli. Econ. 4, 654–674. Xu, L.P., Wang, Y.Z., Liu, N., 2019. Local government environmental regulation and
Gray, W.B., 1987. The cost of regulation: OSHA, EPA and the productivity slowdown. corporate relocation behavior - based on the database of samples of Chinese
Am. Econ. Assoc. 5, 998–1006. industrial enterprises in the cost. Finance Pract. 4, 81–87.
Guan, C., 2019. Research on the impact of environmental regulations on the Xu, Y.X., Lin, R., 2019. Green development path of forestry carbon sequestration industry
development of China’s industrial economy. Xinj. Fina. Econ. 3, 38–49. under environmental regulation. Mark Res. 8, 20–21.
Jorgeson, D.W., Wilcoxen, P.J., 1990. Environmental regulation and U.S. economic Yang, S.Y., 2019a. Promoting industrial structure upgrade with environmental
growth. J. Econ. 2, 314–340. regulations. People Forum 16, 90–91.
Joshi, S., 2001. Estimating the hidden costs of environmental regulation. Account. Rev. Yang, Y.Y., 2019b. The Source of the Heart of the Heart-The Traditional Cultural Gene of
2, 171–198. the Communist Party of China, Tsin. Univ. Pres.
Lee, J.G., Rodrigue, H., 2019. Efficiency of origami-based vacuum pneumatic artificial Yasmeen, H., Wang, Y., Zameer, H., Ismail, H., 2019. Modeling the role of government,
muscle for off-grid operation. J. Preci. Eng. Manu. Tech. 4, 789–797. firm, and civil society for environmental sustainability. Int. J. Agric. Environ. Inf.
Li, H.R., 2019. Analysis of green economy efficiency of resource-based cities based on Syst. 10 (2), 82–97.
non-estimated parameters. Ind. Tech. Econ. 2, 52–58. Yu, L., ZhangBi Qian, W., 2019. Study on the inverse effect of environmental tax on
Li, L., Tao, F., 2012. The choice of the optimum environmental regulation strength of enterprise green transformation. China population. Resour. Environ. 7, 112–120.
China’s manufacturing industry-based on the perspective of green total factor Yuan, Y.J., Liu, L., 2013. Environmental regulation and economic growth: a study based
productivity. China Indu. Econ 5, 70–82. on the classification of economic regulations. Econ. Rev. 1, 27–33.
Pan, S.Y., Gao, M., Kim, H., 2018. Advances and challenges in sustainable tourism Yu, W., Zhang, P., 2016. Urbanization process, space spillover and green economy
toward a green economy. Sci. Tota.Environ. 635, 452–469. efficiency growth-a study of spatial econometrics based on provincial units from
Pittman, R.W., 1983. Multilateral productivity comparisons with unwanted outputs. 2002 to 2012. Econ. Issues 402 (1), 81–86.
Econ. J. 372, 883–891. Zameer, H., Wang, Y., Yasmeen, H., 2020. Reinforcing green competitive advantage
Porter, M.E., 1991. America’s green strategy. Sci. Am. 264, 168. through green production, creativity and green brand image: implications for
Qin, B.T., Ge, L.M., 2019. Chinese high pollution industry transfer and the overall cleaner production in China. J. Clean. Prod. 247 (February), 119119.
environment pollution - based on the regional threshold model of the relative Zhang, Y.H., Chen, J.L., Cheng, W., 2018. Research on the influence mechanism of
environmental regulation. China Environ. Sci. 8, 3572–3584. environmental regulation on China’s regional green economy efficiency-an empirical
Song, D.Y., Deng, J., Gong, Y.Y., 2017. Analysis of the impact of environmental analysis based on super efficiency model and spatial panel measurement model. Res.
regulation on green economy efficiency in China. Learn Prac. 3, 23–33. Environ. 11, 15–26.
Su, J.Y., 2018. Research on green economy efficiency of the yangtze river economic belt Zhao, H., 2007. Impact of environmental regulation on China’s industrial technology
based on DEA and malmquist index. J. Xiangnan Univ. 175 (2), 47–52. innovation. Econ. Manag. 21, 57–61.
Tian, L.F., Li, Y.P., 2019. Obstacles and path choices for energy efficiency improvement Zhao, H., 2019. Application of material flow cost accounting in enterprise environmental
in China-based on the concept of green growth. Tech. Econ. Manage. Res. 8, 21–25. cost management-taking zhengzhou fuxing paper Co., ltd. as an example. Accos.
Tang, X.L., Zhang, H.M., 2015. Research on construction land use efficiency in xi’an city Commun. 20, 85–89.
from the perspective of green economy. Mod. City Res. (5), 75–79.