Resources Policy 85 (2023) 103866

Contents lists available at ScienceDirect

Resources Policy
journal homepage: www.elsevier.com/locate/resourpol

Exploring the role of resource industry dependence and green finance in

green development efficiency in the context of post-Covid-19 period
Xiuli Wang
College of Tourism Management, Wuhan Business University, Hubei, 430056, China

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

Keywords: After the Covid-19 pandemic, the best chance for recovery is to prioritize clean, zero-carbon energy and achieve
Resource industry dependence green development efficiency gains. The research selects the super-efficient SBM model to measure the green
Green finance development efficiency (GDE) of 30 regions in China, and on this basis, explores resource industry dependence
Green development efficiency
and green finance impact on GDE, conclusions as follows: (1) GDE as a whole shows a steady upward trend, and
Natural resource policy
regional GDE shows a trend of first a small decline and then an increase; the value of eastern part is highest. (2)
In any case, resource industry dependence primary term is always negative, which proves that there is indeed a
“resource curse” phenomenon in GDE. The positive but insignificant coefficient of the quadratic term of resource
industry dependence does not prove that there is a conditional “resource curse” between resource industry
dependence and GDE. Green finance can promote GDE in China; it enhances GDE by allocating green credit
funds, supporting environmental protection projects, and improving technology and optimizing industrial
structure. (3) Environmental regulation and urbanization affect GDE shows non-linear relationship, and FDI
improves GDE confirms that the pollution heaven hypothesis is not valid. The rationalization of industrial
structure inhibits GDE. The research in this paper can provide a theoretical reference for regional policy choices
and has important practical significance for regions to achieve sustainable development goals.

1. Introduction being paid to green development. “The 13th Five-Year Plan emphasizes
the need to firmly establish and effectively implement the concept of
Since the 21st century, with the continuous economic development green development; the report of the 19th National Congress puts
and population expansion, the world’s ecosystem has suffered more “promoting green development and accelerating the reform of ecolog­
serious damage, and environmental problems such as resource depletion ical civilization system” as a special chapter for deployment, and also
and air pollution have emerged. In 2018, report of “Economic Conse­ proposes to " Form a green way of development and lifestyle, firmly
quences of Air Pollution” published by the OECD pointed out that global follow the path of civilized development of production development,
air pollution is more serious, and future air pollution will bring eco­ rich living and good ecology, build a beautiful China, create a good
nomic losses of $2600 billion; these environmental problems seriously production and living environment for the people, and contribute to
constrain the economic growth of countries, so promoting economic global ecological security. Green development is an inevitable choice for
green development has become an inevitable choice for countries to sustainable development of all countries, a negation of the traditional
achieve sustainable development (Doner and Schneider, 2016; Ross, economic development method of high energy consumption and high
2015). In 2011, the OECD pointed out in the report “Towards Green pollution, and a new task for China’s economic development in the new
Growth” that countries around the world were currently facing more era. Green development is not “GDP-only” economic growth and only
serious environmental problems, and the traditional economic growth considers ecological benefits, but a comprehensive economic benefit
model was no longer desirable. There is an urgent need to make a that takes into account resource and environmental constraints (Luuk­
transition to an economic green development model. As socialism with kanen et al., 2019; Lyeonov et al., 2019). Sustainable development
Chinese characteristics enters a new era, more and more attention is should takes into account both economy and environment. Therefore,

Abbreviations: GDE, green development efficiency; GDP, Gross Domestic Product; SBM, Slacks-Based Measure; SFA, Stochastic Frontier Approach; DEA, Data
Envelopment Analysis; GMM, Generalized Method of Moment; DMU, decision making unit; FDI, Foreign Direct Investment; LLC, Levin-Lin-Chu; ECD, Organization
for Economic Co-operation and Development.
E-mail address: wangxl_1998@163.com.

https://doi.org/10.1016/j.resourpol.2023.103866
Received 9 April 2023; Received in revised form 3 June 2023; Accepted 26 June 2023
Available online 5 July 2023
0301-4207/© 2023 Elsevier Ltd. All rights reserved.
X. Wang Resources Policy 85 (2023) 103866

the improvement of GDE is the core embodiment of the win-win balance gradually begun to give a richer meaning to the relatively new concept
between economic growth and ecological protection, as well as the of green growth, particularly in the Organization for Economic
inevitable choice to implement the strategic deployment of high-quality Co-operation and Development, which has continued to deepen and
economic (Kurniawan et al., 2022). refine its definition of green growth, redefining it in Towards Green
China’s economic construction has achieved big achievements, but it Growth as “promoting economic growth and development while
has formed a way of economic development that relies excessively on ensuring that natural resources meet the needs of human well-being and
the input of resource factors rather than the improvement of production environmental services”, and furthermore, from a systemic perspective,
efficiency, which has brought about problems such as environmental green growth is measured through the path of “production-re­
pollution and ecological damage, and how to coordinate them has sources-consumption-policy”, providing a set of relatively objective
become an inescapable reality. In this context, green development has criteria for the practical application of green growth theory. Unlike
become an inevitable choice for countries around the world, and its sustainable development, green growth is a way to promote economic
three basic connotations of economic growth, resource conservation and growth and development by seeking a long-term balance between
environmental friendliness have been widely accepted, and reducing environmental hazards and economic growth (Chen and Zheng, 2022).
resource industry dependence is also considered an effective way to At present, it should be more appropriate to use “GDE” to evaluate
change the sloppy economic growth, transform economic growth mode green development compared with many other concepts with the same
and enhance GDE (Khan et al., 2019). Therefore, the work attempts to meaning. Wang and Huang (2014), Zhao et al. (2016), and Lu et al.
investigate whether resource industry dependence is a “blessing” or a (2016) argued that GDE included the reduction of resource inputs,
“curse” on the improvement of GDE in China, and proposes policy to which can take into account the factors of resource saving, pollution
inhibit resource curse. control and economic growth. Che et al. (2018), Liu and Guan (2018),
Meanwhile, green development is the foundation and inherent He and An (2019) also argued that GDE could fully consider resource
requirement of high-quality development. Financial industry as a core and environmental factors, and take into account the maximization of
element of economy, can enhance GDE by virtue of its “clean” and economic benefits and the minimization of resource consumption and
“power” characteristics, and green finance can transform the economic environmental pollution. Thus, it can be seen that many scholars have a
growth model (Lee and Lee, 2022; Sadiq et al., 2022). Developing green relatively unified understanding of the connotation of GDE, and believe
finance, and bringing into play the realization of the financial system’s that GDE highlights the distinctive characteristics of the times of green
resource allocation function in supporting green development have development in the new era. Dong (2022) believed that the green re­
become important national strategies (Hsu et al., 2021). Green finance covery has largely contributed to the development of new green in­
through supports green technologies, promotes the realization of the dustries, which will greatly reduce CO2 emissions and lay the
value of ecological products, and guiding the flow of capital. As an foundation for future sustainable development planning while ener­
important measure to implement China’s green development strategy, it gizing the economy.
is important to explore how green finance affect GDE to achieve the The method of measuring GDE mainly includes two categories: first,
harmonization of China’s economic development, environmental pro­ bulid evaluation system and assign weights to relevant indicators to sum
tection and resource conservation. Therefore, the work first measures up. For example, Guan and Han (2019) selected 62 three-level indicators
the GDE of 30 regions through the super-efficient SBM model, and on from three dimensions: greenness of economic growth, resource and
this basis, explores resource industry dependence and green finance environment carrying potential, and government policy support and
impact on GDE, and on this basis, explores policy to improve GDE, which measured the green development index according to the average
can expand and enrich new research perspectives and contents. weighting method. Fang et al. (2020) constructed a synergistic evalua­
The main contributions of this paper as follows: first, it enriches the tion system from five dimensions and used a factor analysis model to
evaluation research on GDE by constructing a method which consid­ examine GDE. Since the selection of indicators and the setting of priority
ering resource and environmental constraints; second, it further en­ order are mostly based on the personal experience of researchers, the a
riches and deepens the understanding of the relationship between priori and subjective nature of this type of research is difficult to avoid,
resource industry dependence and green finance impact on GDE. which often leads to inaccurate, non-objective and unfair measurement
Thirdly, it provides decision support for improving the top-level design results.
of government policies and enhancing GDE. The other type is the use of parametric SFA or DEA to select a variety
The main structure of this paper is organized as follows: Part one is of economically relevant input-output indicators for efficiency evalua­
the introduction. This part mainly introduces the research background tion. The former measured GDE by estimating the specific form of the
of this paper, which presents an overview of the impact of resource in­ production frontier function (Ainger et al., 1977). For example, Herrala
dustry dependence and green finance on GDE, and raises corresponding and Goel (2012) revealed the trend of GDE in 170 countries worldwide
questions. The second part is the literature review, which reviews the during 1997–2007, and Cai et al. (2019) used SFA to compare the GDE of
research on the relationship between variables. The third part provides 280 cities across the country. However, SFA is also flawed in that it
an overview of the methods and the selection of variables in this paper. requires a specific functional form and is only applicable to
The fourth part is the empirical analysis section, which uses the super- single-output situations, making it can not effectively deal with
efficient SBM to measure GDE, and then verify the relationship be­ non-desired outputs. DEA has become the mainstream method for GDE
tween variables. The fifth part is the conclusion and policy recommen­ evaluation by virtue of its unparalleled advantages in dealing with
dations, which summarizes the main findings of this paper and proposes non-desired outputs. For example, Li et al. (2010) selected this mothod
countermeasures based on them. to assess the green development status of Chinese regions from 1990 to
2006; Zhao and Ren (2012) constructed an input-output index system
2. Literature review and adopted the DEA-Malmquist model to assess the green development
status of the Bohai Rim region; Ren et al. (2017) also used the same
2.1. Green development efficiency model to measure and analyze GDE and its spatial and temporal evo­
lution characteristics by incorporating non-desired outputs into the
One of the earliest theories of green development is the green growth index system; Ye (2021) adopted SBM model to analyze the GDE and its
theory, first proposed by the United Nations Economic and Social spatial characteristics of Chinese provincial capitals (capital cities) from
Commission in 2005 to explore the opportunities for introducing a new 2009 to 2019.
low-carbon sustainable development model for rapidly developing
Asian countries (UNESCAP, 2005). Subsequently, academics have

2
X. Wang Resources Policy 85 (2023) 103866

2.2. Impact of resource industry dependence on GDE development, therefore, the relationship between resource industry
dependence and GDE is studied. The work explores the existence of the
The resource curse hypothesis considers resource endowment as one “resource curse” in the area of GDE by systematic GMM. Meanwhile, the
of the reasons for long-term economic growth stagnation and argues that existing literature discusses the impact of green finance on socio-
resource abundance is instead detrimental to economic development. economics from different aspects such as micro and macro, however,
The empirical study by Sachs and Warner (1995) created the basic due to data and methodological limitations, the current research on how
analytical paradigm for analyzing resources and economic development, green finance affects GDE as a whole and the intrinsic relationship be­
and the study based on cross-country data revealed that a negative tween them needs to be further strengthened. Therefore, the work an­
relationship between resources and growth robustly exists. Subse­ alyzes the overall impact on GDE, and complements the existing
quently, the findings of both empirical studies at the cross-country level quantitative research.
and hypothesis testing at the regional level are quite controversial. By
comparing the carbon intensity of regions with different resource 3. Methodology
abundance, Friedrichs and Inderwildi (2013) found that higher resource
abundance regions with higher carbon intensity. Ahmadov and Borg 3.1. Super-efficient SBM model
(2019) argued that abundant natural resources facilitate renewable
energy production within a country. Zhang et al. (2018) found that 3.1.1. Model construction
resource endowment presents a significant green development inhibi­ Initially, scholars mostly used the data envelopment approach (DEA)
tory effect; similarly, Li and Xu (2018), Wang and Sun (2019) focused on for efficiency measurement. DEA is the use of linear programming in
resource abundance impact on green economic growth, and the results mathematics and data statistical methods to compare the relative effi­
similarly indicated that resource-based regions in China generally face a ciency of each decision unit with the reference decision unit, but the
significant green development “resource curse " dilemma; further, Zhang relative efficiency evaluation idea of DEA implies minimizing inputs and
et al. (2019) used dual indicators of resource reserves and production to expanding outputs. However, the real production process is obviously
assess regional resource endowment and found that manufacturing in­ different from the DEA measurement, that is, the real production process
dustries face a “resource curse” dilemma in the process of green trans­ often has obvious by-products, such as the “three waste” emissions in
formation; Du and Zhang (2019) concluded that 8%–15% level of industry, which we call “non-desired output”. Then from the perspective
resource abundance contributes to urban eco-efficiency improvement, of green development, the higher the economic efficiency, the less these
while the impact of resource abundance level beyond this interval is non-desired outputs must be, which is not reflected in the traditional
backwards. DEA model. GDE needs to consider resource and environmental factors
to maximize economic benefits and minimize resource consumption and
2.3. The impact of green finance on GDE environmental pollution. There are numerous scholars who take
pollutant emissions as non-desired outputs and use non-radial and non-
Finance is the center of resource allocation, and its role in guiding the angle DEA-SBM models that consider non-desired outputs to measure
direction of socio-economic resource allocation has received increasing GDE. However, this type of SBM models although considers the ineffi­
attention from scholars. In fact, there are still some differences in aca­ cient DMU slack variable part and strips the influence of the inefficient
demic circles on whether financial development can promote green slack part, there will still be multiple effective DMUs with 1, this is not
development. Some scholars believe that finance can optimize the effi­ conducive to the comparison of actual values in regional. Then, Tone
ciency of resource allocation and improve technology and promote in­ (2002) proposed the Super-SBM model, which solves the problem of
dustrial structure upgrading (Sachs et al., 2019; Taghizadeh-Hesary and cannot be ranked, and can effectively compare, and portray the char­
Yoshino, 2020), then promote green economic development. While acteristics and evolution process of green development of urban clusters
some scholars have found that financial development increases regional in a deeper way. And the efficiency measurement model can reasonably
CO2 emissions (Charfeddine and Kahia,2019), which inhibits green deal with undesirable outputs such as pollution. Therefore, in this paper,
factor productivity. Other scholars directly point out that the the Super-SBM model is chosen to measure GDE. The model with n in­
bank-dominated financial structure is an important cause of sloppy puts, s1 desired outputs and s2 non-desired outputs, and m decision
growth and environmental problems (Ahmad et al., 2021). The tradi­ making units, the DMUs model is:
tional financial organization system and financial structure can hardly ∑
n −
Si
reach green development requirements and green finance has begun to 1 + 1n xik
receive attention from all sectors of society and develop at a high speed. min ρ = (∑ i=1
∑S2 ) (1)
1 S1 Srg+ Stb−
Gao and Zhang (2021) demonstrated that green credit can achieve the 1− S1 +S2 r=1 yg
rk
+ t=1 yb
tk

upgrading of China’s industrial structure by inhibiting the development

of polluting industries and promoting the development of environmental ⎧ ∑
m
⎪ xik ≥ xij λj − Si−
industries through constructing a systematic GMM model. Pan (2022) ⎪
⎪
⎪
⎪ j=1,j∕
=k
⎪
found that green credit instruments can help reduce the pollution ⎪
⎪
⎪
⎪ ∑
emissions of highly polluting enterprises; Li and Huang (2022) found ⎪ m
⎪
⎪
⎪
⎪ ygrk ≤ yri λj + Srg+
that green credit can improve the environmental quality, and thus ⎪
⎪
⎪ j=1,j∕
=k
⎪
achieve high-quality regional development. ⎪
⎪
⎪
⎪
⎨ ∑
m
Through combing the existing literature, there are abundant studies ybtk ≤ ybtj λj − Stb−
s.t. (2)
related to the “resource curse” proposition, but most scholars focus their ⎪
⎪ j=1,j∕
= k
⎪
⎪
research perspectives on resource abundance, resource industry ⎪
⎪
⎪ ( )
⎪ ∑S1 Sg+ ∑S2 Sb−
dependence and economic growth, ignoring the comprehensive impact ⎪
⎪
⎪
⎪ 1 −
1 r
g + t
>0
⎪ t=1 yb
on resource conservation, environmental friendliness and economic S1 + S2 r=1 y
⎪
⎪ rk tk
⎪
⎪
development. However, most scholars focus on resource abundance, ⎪
⎪
⎪
⎪
⎪ S− , Sg+ , Sb− , λ ≥ 0, i = 1, 2, ⋯, n, r = 1, 2, ⋯, S1
resource industry dependence and economic growth, ignoring the ⎪
⎩
comprehensive impact on resource conservation, environmental t = 1, 2, ⋯, S2 , j = 1, 2, ⋯, m
friendliness and economic development. In particular, as the ecological
g
and environmental problems are becoming more and more prominent, it where xik , yrk and ybtk are the i-th input, r-th desired output and t-th non-
is necessary to extend the research on the “resource curse” in green

3
X. Wang Resources Policy 85 (2023) 103866

desired output of the k-th DMU, respectively, s is the slack variable, λ is Table 1
the weight vector and ρ is the efficiency value. Due to the different Input-output index system.
economic conditions and production situations in different cities, the Indicator Category Indicator Indicator Measurement
∑
constraint m j=1 λj = 1 is added to the model to measure the efficiency Composition
values under the variable payoffs of scale constraint. Input Elemental Capital Inputs Physical capital stock
Indicators Inputs (perpetual inventory
3.1.2. Input-output indicators selection method)
Labor input Number of employment at
the end of the year
(1) Input indicators Energy Inputs Total water consumption
Electricity consumption of
Capital input (K). Capital input we express in terms of physical the whole society
Output Desired Economic output Real GDP
capital stock, considering that the National Bureau of Statistics does not
Indicators output Social Total retail sales of social
publish capital stock data at present, so this paper draws on the practice Development consumer goods
of Zhang et al. (2004), which is calculated by perpetual inventory Green Ecology Greening coverage of built-
method. The calculation formula is: Kit = Kit− 1 (1 − δ) + Iit , Kit denotes up areas
the current physical capital stock, Kit− 1 refers to the previous capital Non-desired Environmental Solid waste
outputs Pollution SO2 Emission
stock, δ is the economic depreciation rate, which is assumed to be 9.6%, Chemical Oxygen Demand
Iit is total fixed capital, and the initial capital stock K0 is 10%. The
calculated physical capital stock is deflated using 2009 as the base
period. The unit is billion yuan. Data on social fixed asset investment as the model have dynamic explanatory power, while the variables in the
well as fixed asset price index are obtained from the statistical model cannot be considered as strictly exogenous. Using traditional
yearbooks. research methods such as differencing and OLS models or fixed effects
Labor input (L). Referring to Wang et al. (2020), labor input is models in conjunction, the model results are biased and the estimated
expressed by employees numbers at the end of the year for the whole equation results are biased because the covariance between the depen­
society in each region, in 10,000 people, the data is obtained from the dent variable and the random error term is not equal to 0. To address the
EPS Global Statistics Database. Data in this paper are obtained from above endogeneity drawbacks, the paper uses the One-Step Systematic
China Statistical Yearbook. Generalized Method of Moments estimation method to study. The dif­
Energy input (E). In terms of energy variables, considering that the ferential GMM is an estimation method conceived by Arellano and Bond
demand for energy such as water and electricity is the most huge and (1991) to derive estimation methods that satisfy the moment estimation
direct in the urbanization process, this paper selects the total social conditions through auxiliary variables. The basic principle of this
electricity consumption and total urban water consumption to measure method is to take first-order difference measures to offset individual
energy consumption. Data in this paper are obtained from China Sta­ differences in the model, and later to use lagged variables as auxiliary
tistical Yearbook. variables to overcome correlation problems. Arellano and Bover (1995)
proposed to construct new moment conditions to estimate the model by
(2) Output indicators associating the original level equation and the difference transformation
equation. The systematic GMM method for estimating the model with
Expected output. This paper selects indicators from three aspects, new moment conditions was proposed by Arellano and Bover (1995) to
namely, economic growth, social development and ecological green­ remedy this deficiency.
ness, to reflect the multiple objectives of green development. The GDP is Unlike the differential GMM model, sys-GMM method could effec­
converted into real GDP with 2009 as the base period; the total retail tively correct the bias of variables; overcome the correlation with the
sales can intuitively reflect the living consumption level of residents, so random error term and the insufficient information of instrumental
this paper selects it to measure social development level; the green variables. Using the weight matrix as the basis for classification, there
coverage rate of built-up area reflects the result of the city’s investment are one-step and two-step estimates for GMM estimation. In the case of
in and treatment of ecological environment, so this indicator is selected small sample size, the estimation of the coefficients of the variables
to measure the ecological green level. prefers to use one-step estimation, and one-step estimation can yield
Non-expected output. Non-desired output: The quantity of “three consistent estimates while there is a loss rate, so one-step estimation is
wastes” emitted from industrial production in each province and city is used in this paper. The basis for applying this method is that the residual
used to express the non-desired output. Among them, solid waste is terms are not linearly related to the explanatory variables, so the work
expressed by the general solid waste emission from industrial produc­ taking the perturbation term autocorrelation test and overidentification
tion; exhaust gas is expressed by the SO2 emission from industrial pro­ test for the variables. Accordingly, construct the model in this paper as
duction; because of lack of data on industrial wastewater emission, the follows:
chemical oxygen demand (COD) in industrial wastewater is used to
Yit = α + β1 Yit− 1 + β2 X′it + β3 Cit + θit + εit (3)
measure the wastewater output, following the practice of Jing and
Zhang (2014).
Where, Yit is the explanatory variable, indicating the value of province
The specific indexes are selected in Table 1 below. Data in this paper
and city i at time t; Yit− 1 is the explanatory variable lagged by one
are obtained from China Statistical Yearbook and China Environmental
period, indicating the value of province and city i at time t-1, with α as
Statistical Yearbook.
the constant term; X′it is core explanatory variable; Cit is control variable;
θit is the unobserved fixed effect for each province and city; β1, β2, β3 are
3.2. System GMM the impact coefficients to be estimated; εit is the random disturbance
term and εit ∼ N(0, σ2 ).
3.2.1. Model construction
Static panel estimation is prone to omitting dynamic factors and thus 3.2.2. Variable selection and data sources
leads to bias, which is related to the tendency to ignore lagged terms of
the explanatory variables. It is more effective in studying the dynamic (1) Explanatory variables
behavior of the economy. The introduction of lagged variables makes

4
X. Wang Resources Policy 85 (2023) 103866

Green development efficiency (GDE). GDE in this paper are structure deviation degree, and use it as a weight to comprehensively
measured by super-efficient SBM model, results as Table 4. portray IR, and measurement formula as follows:
∑ 3 ⃒ ⃒
⃒Yi /Li ⃒
(2) Core explanatory variables IR = (Yi / Y)⃒⃒ − 1⃒⃒ (5)
i=1
Y/L
Resource Industry Dependence (RID). The measure of resource in­
Environmental Regulation (ER). There is a large body of literature on
dustry dependence should be able to accurately measure the degree of
the measurement of environmental regulation, and many scholars use
dependence of the regional economy on natural resources. This study
emission fees to measure environmental regulation. It is important to
uses the share of mining employment in total employment to measure
note that sewage charges are cost-based data, and measuring the in­
the dependence of resource-based industries.
tensity of regional environmental regulation directly by sewage charges
Green Finance (GF). In recent years, the importance of green econ­
cannot eliminate price change situations, and also ignores the endoge­
omy has increased year by year, and China has been improving the
nous influence of regional GDP and industrial output. The payment of
relevant institutional construction and product development for green
emission fees involves pollutant emissions and emission charges, and
finance development. Chinese green financial includes green credit,
considering that China only raised the emission charge levy in 2015, the
green securities, green insurance, green investment and carbon finance,
impact on emission fees is not significant during the sample period.
which also have great contribution to green economy. Taking into ac­
Therefore, this paper measures environmental regulation by dividing
count Chinese actual situation, the evaluation system containing 9 index
the emission fee by the emissions of four major pollutants, i.e., the
layers (see Table 2) was created by drawing on the research results of
emission fee paid per unit of pollutant emissions, and the value of
Zeng et al. (2014) and improving them.
environmental regulation is directly proportional to the strength of
environmental regulation; In addition, considering that industrial en­
(3) Control variables
terprises, as the major contributors to pollution emissions, are also the
Urbanization level (UR). In the work, the population share
main targets of sewage charges, this paper deflates the sewage charges
method will be used to measure the level of urbanization.
by the ex-factory prices of industrial producers to eliminate the factor of
Foreign direct investment (FDI). Expressed as foreign direct
price increase.
investment (converted to billion yuan). FDI has a strong tech­
The data in this section are mainly obtained from various yearbooks
nology spillover effect and contributes to the improvement of
from 2009 to 2021, including China Industrial Statistical Yearbook and
local GDE, but some scholars hold the opposite view that FDI will
the Fourth National Economic Census Bulletin, Guotaian database, and
lead to the transfer of pollution-intensive industries and increase
China Carbon Emission Database. The descriptive statistics of variable as
environmental pollution in the receiving area, which means there
Table 3 below.
is a “pollution paradise hypothesis”. Therefore, the impact of FDI
on GDE cannot be determined yet. So the work uses the propor­
4. Results
tion of actual FDI to actual GDP to express the intensity of FDI.
The article uses the flow data of actual utilization of FDI from
4.1. Results of GDE measurement
2009 to 2021 and converts them into RMB based on the average
exchange rate of US dollars over the years.
The work measured the GDE values of cities from 2009 to 2021 using
Matlab2016 software and averaged the obtained data by time series
Industrial Structure Rationalization (IR). IR reflects the degree of
after dividing them into regions, and plotted the time change trend (see
coordination among industries, i.e. industrial structure deviation.
Table 4 and Fig. 1).
∑ 3 ⃒⃒Yi /Li
⃒ ∑
⃒ 3 ⃒⃒Yi /Y
⃒
⃒ The specific growth of GDE in Table 4 shows a turn around in 2013
Dev = ⃒
⃒ Y/L − 1 ⃒=
⃒
⃒
⃒ L /L − 1⃒
⃒ (4) and 2016. Among them, GDE generally showed an increasing trend from
i
2009 to 2013, probably because governments gradually began to realize
i=1 i=1

Where Y is output value, L is employment, Yi is output value in i-th the limitations of the current economic development model and the
industry, and Li is employment in i-th industry. In this paper, drawing on contradictions between economy, environment and resources, and
the study of Gan et al. (2011), we use the Theil index to construct the IR further recognized the importance of green and sustainable develop­
on the basis of the industrial structure deviation degree, which is given ment was gradually attached to environmental development; while a
to each of the three major industries as a proportion of the industrial small decline occurred in 2014 and 2015, probably because the country
proposed to put ecological civilization in a prominent position, and
green and sustainable development thus became the strategic direction
of regional development, while GDE was in the decline stage of the
Table 2
initial adjustment of the strategy in this period; after 2016, it has been in
Green financial development index.
a faster growth state, which is due to the fact that China put forward the
Primary Secondary Indicators
concept of green development for the first time, and the national
Indicators
development entered green development mode, and GDE steadily
Green Credit Green Credit Balance/Loan Balance of Financial Institutions increased. In addition, the regional differences in GDE cannot be
Interest expenses of six high-energy-consuming industrial
industries/total interest expenses of industrial industries
Green Securities Total market value of energy-saving and environmental
protection enterprises/Total market value of A-shares Table 3
Total market value of high-energy-consuming industries/Total Descriptive statistics results.
market value of A-shares
Variable Mean Standard deviation Minimum Maximum
Green insurance Environmental pollution liability insurance premium income/
property insurance premium income GDE 0.549 0.597 0.114 1.202
Environmental pollution liability insurance payout expenditure/ RID 0.0453 0.0428 0.0097 0.2135
environmental pollution liability insurance premium income GF 0.6674 0.4436 0.3897 0.5904
Green Fiscal expenditure on energy saving and environmental ER 0.1274 0.1103 0.0084 0.5463
investment protection industry/total fiscal expenditure FDI 0.0470 0.0491 0.0021 0.2137
Environmental pollution control investment/GDP IR 0.0734 0.0894 0.0123 0.5784
Carbon Finance Carbon Dioxide Emissions/GDP UR 0.5213 0.3458 0.4987 0.8846

5
X. Wang Resources Policy 85 (2023) 103866

∑ ∑ ∑
n xy− x y
calculated as: Rxy = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ∑ 2 ∑ 2̅, this paper uses
∑ 2 ∑ 2 √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
n x − ( x) n y − ( y)

Statel6.0 to do correlation test on each variable

In Table 5, the Pearson correlation coefficients among the variables
are small, and the problem of multicollinearity is excluded. In order to
further exclude the possible pseudo-regression problem in the model,
further smoothness tests are needed.

(2) Test for data smoothness

The research sample in this paper is a panel data of 30 provinces, and

there is a possibility of pseudo-regression problem if the panel data are
not smooth. Therefore, it is necessary to test whether the panel data are
smooth or not. The study used the LLC test to verify the smoothness of
the model indicators, and the results are shown in Table 6. As can be
Fig. 1. Trends in GDE in different regions. seen from the table, the test statistics of each sample in the regression
model passed the significance test, indicating that there is no unit root
ignored. Fig. 1 report the change trend, the GDE in the eastern region is for each sample variable, i.e., all variables are smooth series. Therefore,
highest, western regions is lowest, which is similar to the findings of the selected indicators can be included in the panel regression model.
previous scholars. The inter-regional differences in efficiency may be
related to many differential factors such as technology level, develop­ (3) Hausman test
ment stage and production factor allocation in different regions. The regression analysis includes three categories of fixed,

Table 5
4.2. Analysis of systematic GMM regression results
Correlation test results.
Variable RID GF ER FDI IR UR
4.2.1. Test of validity of estimation method
RID 1
(1) Correlation test GF 0.103 1
ER 0.023 0.121 1
FDI 0.135 0.227 − 0.005 1
In order to avoid the analysis results being affected by multi­ IR 0.245 0.034 0.013 0.219 1
collinearity, which leads to the phenomenon of pseudo-regression, the UR 0.001 0.112 0.034 0.109 0.177 1
most common is the Pearson correlation coefficient. The coefficient is

Table 4
Results of GDE measurement.
Region 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Mean

Eastern Beijing 1.025 1.032 1.052 1.064 1.075 1.065 1.054 1.058 1.257 1.327 1.478 1.502 1.631 1.202
Tianjin 1.019 1.029 1.076 1.113 1.023 1.023 1.016 1.009 1.193 1.226 1.376 1.408 1.526 1.140
Hebei 0.426 0.431 0.442 0.522 0.421 0.421 0.382 0.444 0.513 0.624 0.735 0.818 0.927 0.565
Liaoning 0.437 0.448 0.503 0.633 0.433 0.533 0.528 0.535 0.691 0.708 0.719 0.732 0.821 0.611
Shanghai 0.919 0.929 1.036 1.127 0.922 1.052 1.046 1.104 1.213 1.226 1.296 1.318 1.426 1.141
Jiangsu 0.657 0.668 0.695 0.751 0.662 0.762 0.675 0.601 0.889 0.943 1.036 1.128 1.288 0.850
Zhejiang 0.697 0.605 0.796 0.835 0.601 0.681 0.606 0.758 0.841 0.959 1.065 1.125 1.261 0.847
Fujian 0.626 0.647 0.622 0.779 0.644 0.644 0.612 0.681 0.712 0.861 0.985 1.011 1.072 0.757
Shandong 0.421 0.436 0.432 0.457 0.431 0.431 0.362 0.415 0.543 0.657 0.772 0.814 0.919 0.556
Guangdong 0.668 0.684 0.696 0.768 0.613 0.683 0.666 0.631 0.829 0.972 1.026 1.152 1.217 0.834
Hainan 0.428 0.435 0.429 0.458 0.425 0.475 0.418 0.456 0.583 0.616 0.741 0.868 0.932 0.584
Eastern mean 0.656 0.665 0.713 0.766 0.798 0.706 0.670 0.717 0.842 0.920 1.021 1.080 1.184 0.826
Central Shanxi 0.277 0.286 0.349 0.457 0.276 0.346 0.337 0.353 0.529 0.665 0.771 0.862 0.932 0.471
Jilin 0.265 0.297 0.355 0.461 0.281 0.361 0.359 0.326 0.433 0.506 0.658 0.727 0.891 0.478
Heilongjiang 0.259 0.273 0.326 0.435 0.264 0.364 0.248 0.316 0.466 0.573 0.613 0.743 0.768 0.438
Anhui 0.293 0.227 0.335 0.457 0.277 0.377 0.358 0.324 0.473 0.532 0.664 0.711 0.883 0.479
Jiangxi 0.245 0.255 0.343 0.495 0.258 0.358 0.262 0.311 0.474 0.531 0.738 0.855 0.895 0.456
Henan 0.282 0.281 0.338 0.471 0.291 0.391 0.323 0.417 0.559 0.664 0.706 0.828 0.982 0.513
Hubei 0.257 0.273 0.326 0.485 0.277 0.377 0.283 0.301 0.448 0.632 0.762 0.816 0.988 0.454
Hunan 0.253 0.282 0.336 0.511 0.387 0.387 0.288 0.302 0.415 0.622 0.734 0.853 0.974 0.456
Central mean 0.266 0.284 0.339 0.437 0.457 0.370 0.307 0.344 0.462 0.553 0.656 0.737 0.874 0.468
Western Neimenggu 0.249 0.259 0.251 0.264 0.257 0.277 0.231 0.243 0.357 0.462 0.567 0.664 0.727 0.377
Guangxi 0.248 0.253 0.267 0.278 0.263 0.263 0.229 0.253 0.342 0.459 0.572 0.681 0.783 0.379
Chongqing 0.247 0.267 0.271 0.297 0.252 0.252 0.228 0.237 0.459 0.593 0.622 0.748 0.852 0.416
Sichuan 0.373 0.388 0.392 0.413 0.385 0.385 0.296 0.353 0.495 0.503 0.611 0.716 0.858 0.481
Guizhou 0.179 0.199 0.242 0.259 0.198 0.198 0.147 0.156 0.238 0.342 0.444 0.558 0.661 0.295
Yunnan 0.327 0.377 0.381 0.392 0.351 0.351 0.253 0.322 0.372 0.426 0.531 0.655 0.762 0.431
Shanxi 0.187 0.193 0.268 0.253 0.188 0.238 0.189 0.162 0.215 0.313 0.422 0.537 0.642 0.293
Gansu 0.179 0.199 0.217 0.249 0.187 0.197 0.114 0.137 0.245 0.342 0.448 0.565 0.659 0.294
Qinghai 0.217 0.218 0.226 0.231 0.206 0.206 0.186 0.237 0.342 0.452 0.561 0.672 0.788 0.353
Ningxia 0.153 0.159 0.168 0.253 0.161 0.171 0.128 0.152 0.215 0.313 0.422 0.537 0.642 0.269
Xinjiang 0.183 0.193 0.211 0.236 0.193 0.193 0.155 0.182 0.215 0.352 0.412 0.527 0.632 0.290
Western mean 0.231 0.245 0.256 0.276 0.317 0.248 0.196 0.221 0.318 0.414 0.510 0.624 0.728 0.353
National mean 0.384 0.398 0.436 0.493 0.524 0.442 0.391 0.427 0.541 0.629 0.729 0.813 0.929 0.549

6
X. Wang Resources Policy 85 (2023) 103866

Table 6 Table 8
Results of the test for stationarity. Systematic GMM regression results.
Variable LLC Statistics P value Result Explanatory variables Elasticity coefficient P value t value

RID − 22.459 0.000 Smooth Yit− 1 0.203 0.0001 3.29

GF − 24.115 0.000 Smooth RID − 0.182 0.0000 3.02
ER − 9.778 0.000 Smooth RID2 0.004 0.2132 0.83
FDI − 10.490 0.000 Smooth GF 0.166 0.0013 4.68
IR − 11.104 0.000 Smooth ER − 0.0342 0.0000 2.99
UR − 9.234 0.000 Smooth (ER)2 0.1187 0.0012 5.13
UR − 0.0764 0.0000 3.78
(UR)2 0.2145 0.0312 2.16
random effects, and mixed regression. The premise of using FDI 0.1568 0.0001 7.43
mixed regression is that there is no individual effect, and the IR − 0.2237 0.0012 3.99
Hansen test 0.239
premise of random effects regression is that the random error AR(1) 0.006
term is independent of all explanatory variables. In order to avoid AR(2) 0.335
incorrect model settings, it is necessary to distinguish which
panel data model is suitable. Since the mixed regression model is
not suitable for the research content of this paper, Hausman test is Jing, 2006), thus the elasticity coefficient of resource industry depen­
then selected to test the fixed and random effects of the panel dence is significantly negative in this paper. Engaging in resource
data. According to the relevant theory, Hausman test is a valid exploitation and utilization industries can bring entrepreneurs and
method to test model form. The results are in Table 7, in this start-ups more rapid and lucrative resource returns, so they are attracted
study, the fixed effects model is estimated relatively well. The to engage in primary product production and processing activities, and
next step is to apply the fixed effects model for empirical analysis. the development of the innovation sector receives a squeeze.
Resource-based industries will squeeze out other new industries that are
4.2.2. Systematic GMM model regression results more conducive to GDE enhancement, thus generating the “resource
Table 8 reports the result; the squared terms of some of the variables curse ". This is another confirmation and development of the “resource
were included in the model to verify their nonlinear correlation, since curse” paradox, which extends the research from economic growth to
there may be a nonlinear relationship between the effects of some of the GDE field. The positive but insignificant coefficient of the quadratic term
variables. of resource industry dependence can’t prove the conditional “resource
Autocorrelation test, AR (1) is less than 0.1 passing the significance curse” phenomenon between resource industry dependence and GDE, so
test and AR (2) is greater than 0.1, means second-order serial autocor­ the work argues that resource industry dependence inhibit GDE in the
relation is not significant and it can be concluded that each model dif­ long term.
ference equation avoids residual autocorrelation. In addition, the P- Green finance can significant promote GDE, when green finance in­
values of Hansen’s test to investigate the validity of the proxy variables crease 1%, the GDE will increase by 0.166%. The way green finance
were 0.239 and greater than 0.1, respectively, and the instrumental promotes GDE as three points: first, it lies in the establishment of green
variables were selected to be valid because they passed the significance finance identification mechanism, changing the resource allocation
test. The estimated parameters of the lagged terms of the function values structure, regulating the amount of credit granted to different enter­
in the observed model were between the term factors obtained from the prises, specifically by reducing the cost of green environmental protec­
mixed regression and fixed effects model, and were stable and valid tion and energy saving projects by lowering the loan interest rate, or
using the systematic GMM regression analysis method. In summary, the raising the access threshold for loans based on the green credit identi­
choice of model settings and estimation methods in the article is fication mechanism for enterprises or projects with environmental
reasonable and robust. pollution emissions exceeding the standard. Second, financial in­
From the regression results, it is easy to find that the first-order lag stitutions support environmental protection projects through green
term of GDE is always positive, which indicates that GDE in each region credit to improve the ecological environment, which is a kind of pro­
is positively influenced by the GDE in the previous period, and the GDE duction and environmental resources. Secondly, financial institutions
is coherent. At the same time, resource industry dependence shows support environmental protection projects through green credit to
sufficient significance in all cases and is significantly negative, thus improve the ecological environment which can be transformed into
verifying that resource industry dependence inhibits GDE, and excessive monetary assets by using the market transaction mechanism and the
resource dependence has an obstructing effect on GDE. The negative production of green products and services can effectively promote eco­
impact of resource industry dependence on GDE may be because the nomic development and reduce social operation costs due to the vast
resource industry can generate greater economic benefits in the short market for green products and services. This is because the upgrade of
term, thus attracting excellent talents to the resource ministry sector, but industrial structure means the transformation from low efficiency to
the resource industry sector tends to be a sector with less technological high efficiency, which will improve GDE of the region. This study is
contribution, thus making other industry sectors lack more excellent consistent with the findings of most scholars, such as Wang et al. (2023),
talents than necessary, which is not conducive to innovation in other who found that green credit can improve GDE and promote the
industries, thus excessive resource industry dependence squeezes out harmonization of regional economic development and ecological envi­
innovation output and inhibits the improvement of green development ronment. Li et al. (2020) found that green credit can significantly pro­
efficiency, most scholars have reached the same conclusion (Zhang and mote the upgrading of industrial structure, but there are differences in
the degree of impact on the east and the central and western parts of the
country, due to the difficulty of obtaining data, there is no sub-regional
Table 7 discussion of China in this paper, which is a direction that needs to be
Hausman test results. further explored in the future.
Test method Test results Description The negative coefficient of the primary term and positive squared
F-test F = 10.136 Selection of fixed effects term of environmental regulation indicate that there is a U-shaped
P = 0.0001 relationship between environmental regulation and GDE, and there is an
Huasman test Chi2（2） = 41.22 Selection of fixed effects inflection point of the effect of environmental regulation on GDE, and
P = 0.0000 the implementation of environmental regulation before the inflection

7
X. Wang Resources Policy 85 (2023) 103866

point is unfavorable to the improvement of GDE, probably because conditions, so as to rationalize the structure and rationalize the alloca­
environmental regulation forces enterprises to upgrade technology and tion and effective use of resources among industries, thus promoting
reduce pollution, which greatly increases the cost of enterprises and thus economic growth. However, GDE is different from economic growth and
cannot concentrate capital on innovation, so inhibits the improvement focuses more on the coordinated development of economy and envi­
of GDE. When the inflection point is exceeded, environmental regula­ ronment. If the adjustment direction of production factors in the in­
tions can promote the improvement of GDE level, and environmental dustry turns to be environment-friendly, it will show the green
regulations force enterprises to upgrade their technology, which can development efficiency; on the contrary, it will show the inhibition of
effectively reduce pollution and emissions and enhance the efficiency of GDE, and now the rationalization of industrial inhibiti GDE.
green development. This conclusion is recognized by most scholars, such
as Huang and Wu (2020), Huang and Gao (2016). However, some 4.3. Robustness test
scholars have reached the opposite conclusion from this paper, for
example, Shen (2012) showed that the relationship between environ­ In order to further prove the robustness of the model, this paper uses
mental regulation and environmental efficiency in China showed an the replacement model method to conduct the robustness test, because
“inverted U-shaped” nonlinear relationship, and Ren et al. (2016) tested the green development efficiency value of the super-efficient SBM
the impact of environmental regulation on regional eco-efficiency also measure used in this paper is truncated data, so a large number of
showed an “inverted U-shaped relationship. The main difference with scholars also use the Tobit model for empirical analysis, this paper tests
this paper is that the objects of their analysis are environmental effi­ the stability of the model by transforming the form of the model, and the
ciency and eco-efficiency, which can be seen that environmental effi­ test results are shown in Table 9.
ciency and eco-efficiency can also reflect green development, but there The results show that the coefficients of green finance and resource
are still some differences with GDE. It is also a supplement and dependence variables on GDE are 0.193 and − 0.128, respectively, and
improvement of the research in this field. pass the 1% significance test, while the coefficients of other variables
The negative coefficient of the primary term and positive squared change slightly, but the significance level and the direction of influence
term of urbanization indicate that there is a U-shaped relationship be­ do not change, indicating that the results of the empirical analysis do not
tween urbanization and GDE, that is, urbanization has a significant change much under the systematic GMM and Tobit models. Therefore,
positive “U" shape trend of inhibiting and then promoting GDE. Ac­ after replacing the model, the results are consistent with the system
cording to the regression results, when the change of urbanization on GMM estimation results, proving the robustness of the model selection.
GDE is below the inflection point, urbanization development will inhibit
the improvement of regional development efficiency; while after the 5. Conclusions and policy recommendations
urbanization level exceeds the inflection point, urbanization develop­
ment can promote regional GDE. There are fewer nonlinear analyses of 5.1. Conclusions
the relationship, only some scholars take eco-efficiency as an example to
analyze the nonlinear relationship between them, such as Zheng et al. The work selects the super-SBM model to measure the GDE of 30
(2017) found that urbanization in eastern provinces has a nonlinear regions in China, and on this basis, explores the influence of resource
impact relationship; Zhao et al. (2019) found that there are stage dif­ industry dependence and green finance on GDE, conclusions as follows:
ferences in the impact between urbanization and regional green devel­
opment performance, and the overall shows a nonlinear impact (1) GDE generally showed an increasing trend from 2009 to 2013,
relationship. while a small decline occurred in 2014 and 2015, after 2016, it
FDI impact on GDE is significant and positive, indicating that the has been in a faster growth state, and the regional differences in
GDE increases when FDI flow in China. This indicates that the “Haven GDE cannot be ignored, the GDE in the eastern region is highest,
Hypothesis “does not hold true in China. FDI brings advanced man­ western regions is lowest.
agement experience and technology, which improves regional envi­ (2) In any case, the primary term of resource industry dependence is
ronmental quality, reduces resource consumption and pollution always negative, which proves that there is indeed a “resource
emissions, and improves GDE through spillover effects while promoting curse” phenomenon between resource industry dependence and
economic growth. The findings of this paper are consistent with some GDE. Meanwhile, the GDE in the current period is positively
scholars, such as Gong et al. (2018), Chaudhuri and Mukhopadhyay influenced by the previous period. The positive but insignificant
(2013) who argued that scale effects and technology spillover effects can coefficient of the quadratic term of resource industry dependence
improve the ecological environment of the host country and help means no conditional “resource curse” phenomenon between
improve GDE, but there are also scholars who come to the opposite resource industry dependence and GDE, so resource industry
conclusion from this paper, such as Tu (2008) found that FDI hinders dependence inhibits GDE in the long term. Green finance can
GDE, so that the entry of high energy-consuming and high-polluting improve GDE in China, and green finance enhances the level of
industries should be restricted, mainly because the sample selection
cycle of some scholars is before 2015, and China has proposed the Table 9
construction of ecological civilization and vigorously developed green Robustness test results.
economy in recent years, and higher demand for quality of foreign in­
Explanatory variables Elasticity coefficient P-value t-value
vestment, thus positive effect is reflected.
Industrial structure rationalization does not exist to achieve regional Yit− 1 0.113 0.0000 3.99
RID − 0.128 0.0001 6.23
GDE, but rather has a certain degree of inhibitory effect. Rationalization
RID2 0.017 0.1432 0.22
of industrial structure refers to the dynamic adjustment process of GF 0.193 0.0000 5.18
rationalizing the layout of factors between and within industries by ER − 0.0786 0.0002 4.58
adjusting the industrial structure that is not compatible with the existing （ER）2 0.1453 0.0000 3.48
economic development level under the constraints of existing resource UR − 0.0999 0.0012 4.33
（UR）2 0.1387 0.0477 2.22
conditions and technology level, and scientifically allocating production FDI 0.1318 0.0000 4.51
factors according to the existing demand structure and technology level. IR − 0.2543 0.0002 7.44
The rationalization of industrial structure requires that at a certain stage Hansen test 0.347
of economic development, the initial unsatisfactory industrial structure AR（1） 0.005
AR(2) 0.276
should be adjusted according to the consumption demand and resource

8
X. Wang Resources Policy 85 (2023) 103866

GDE by improving technology and optimizing industrial technology and promote financial development to maintain sus­
structure. tainable and healthy economic development, move from “high
(3) Environmental regulation and urbanization affect GDE shows a growth” to “high quality development".
U-shaped non-linear relationship. FDI promote GDE, which con­ (5) Enhance innovation capacity and accelerate economic trans­
firms the “pollution halo” hypothesis. Industrial structure formation. Technological innovation is very important for eco­
rationalization inhibits the GDE at 1% confidence level, and in­ nomic growth and environmental protection, while resource
dustrial structure rationalization suppresses the regional GDE. industry dependence will adversely affect the improvement of
GDE through technological innovation crowding out effect.
5.2. Recommendations Therefore, resource-based regions should improve innovation
capacity. On the one hand, resource-based regions should provide
In order to enhance the GDE and strengthen the promotion effect of incentives and support for enterprise technological innovation,
resource industry dependence and green finance on regional GDE, the and encourage enterprises to engage in technological innovation
work proposes the following countermeasure suggestions. activities by reducing taxes and fees, setting up special funds and
financial subsidies, etc. On the other hand, resource-based re­
(1) In the post-epidemic era, with the epidemic easing, green trans­ gions should introduce appropriate measures to introduce talents
formation and inclusive development have gradually become the to attract more innovative talents to invest in technological
focus of global recovery policies. The policy objectives of the innovation and enhance social innovation enthusiasm.
“green and inclusive” recovery include sustainable economic
growth, green and inclusive employment creation, green and low CRediT authorship contribution statement
carbon to help create a zero carbon future, and a fair and inclu­
sive social environment. During the recovery period, industrial Xiuli Wang: Conceptualization, Writing - review & editing, Writing -
investment policies are tilted toward promoting investment in the original draft; Methodology, Writing - review & editing, Formal analysis;
direction of green upgrading of industry stock and green devel­ Project administration, Supervision, Data curation.
opment of incremental scale, accelerating the construction of
infrastructure related to green industries, vigorously developing
Declaration of competing interest
renewable energy industries, and focusing on green quality
improvement of consumption policies, such as “Internet + ser­
The authors declare that they have no known competing financial
vice”, intelligent retail, intelligent products, etc. The new mode of
interests or personal relationships that could have appeared to influence
consumption has become a growth point for green consumption.
the work reported in this paper.
(2) Vigorously develop green finance, broaden the financing chan­
nels for green industrial, and promote industrial structure
References
upgrading. Support projects with environmental benefits through
green financial. Introduce relevant support policies and develop Ahmad, M., Jiang, P., Murshed, M., 2021. Modelling the dynamic linkages between eco-
environmental protection standards and risk rating standards for innovation, urbanization, economic growth and ecological footprints for G7
projects. Accelerate the development of laws and regulations on countries: does financial globalization matter? Sustain. Cities Soc. 70, 102881.
Ahmadov, A.K., Borg, C.V.D., 2019. Do natural resources impede renewable energy
energy emission reduction. production in the EU? A mixed-methods analysis. Energy Pol. 126, 361–369.
(3) Increase the investment of financial institutions in green-related Ainger, D., Lovell, C., Schmidt, P., 1977. Formulation and estimation of stochastic
industries. China’s green development is relatively low, mainly frontier production function models. J. Econom. 6, 21–37.
Arellano, M., Bond, S., 1991. Some tests of specification for panel data: Monte Carlo
because the financing of some environmental protection in­ evidence and an application to employment equations. Rev. Econ. Stud. 58 (2),
dustries has a certain degree of difficulty, and financial in­ 277–297.
stitutions are unwilling to invest in these green industries, and to Arellano, M., Bover, O., 1995. Another look at the instrumental variable estimation of
error-components models. J. Econom. 68 (1), 29–51.
remedy environmental pollution by carrying out environmental Cai, B.F., Guo, H.X., Ma, Z.P., 2019. Benchmarking carbon emissions efficiency in
pollution treatment after causing it, which is undoubtedly a waste Chinese cities: a comparative study based on high-resolution gridded data. Appl.
of resources and a loss of efficiency loss of efficiency. Therefore, Energy 242, 994–1009.
Charfeddine, L., Kahia, M., 2019. Impact of Renewable Energy Consumption and
the degree of investment in green-related industries by financial
Financial Development on CO2 Emissions and Economic Growth in the MENA
institutions must be maintained to a certain extent, and more Region: a Panel Vector Autoregressive (PVAR) Analysis. Renewable energy139,
credit support as well as policy inclination must be added to these pp. 198–213.
green development enterprises in environmental technology, Chaudhuri, S., Mukhopadhyay, U., 2013. Foreign direct investment, environmentally
sound technology and informal sector. Econ. Modell. 31, 206–213.
clean production technology, resource efficient utilization tech­ Che, L., Bai, Y.P., Zhou, L., Wang, F., Ji, X.P., Qiao, F.W., 2018. Spatial characteristics
nology, clean energy and raw material technology. If China wants and spillover analysis of green development efficiency in China. Geoscience 11,
to form a new type of society with low energy consumption, less 1788–1798.
Chen, Z., Zheng, J.H., 2022. Can green technology innovation promote high-quality
pollution and high economic development efficiency, the strict regional economic development? A discussion of the selection effects of
formulation and implementation of policies cannot be achieved environmental policies. Contemporary Econ. Sci. 44 (4), 43–58.
without the help of financial institutions behind the scenes. Doner, R.F., Schneider, B.R., 2016. The middle-income trap: more politics than
economics. World Polit. 68 (4), 608–644.
Therefore, in order to promote green development faster and Dong, Y.K., 2022. The impact of green recovery on the economy: opportunities and
better, it needs to increase investment of financial institutions in challenges. Sci. Technol. Energetic Mater. 12, 42–44.
green-related industries. Du, K.R., Zhang, N., 2019. Resource abundance and urban eco-efficiency in China: an
empirical analysis based on the conditional SBM model. J. Xi’an Jiaot. Univ.: Social
(4) Get rid of the single industrial structure dominated by resource- Science Edition 39 (1), 65–72.
based industries as soon as possible. Local governments in Fang, G.C., Wang, Q.L., Tian, L.X., 2020. Green development of yangtze river delta in
resource-based need to increase research subsidies, while estab­ China under population-resources-environment-development-satisfaction
perspective. Sci. Total Environ. 727, 138710.
lishing an effective talent incentive system and improving the
Friedrichs, J., Inderwildi, O.R., 2013. The carbon curse: AreFuel rich countries doomed
cooperation mechanism between industry, academia and to high CO2 intensities? Energy Pol. 62, 1356–1365.
research institutes to put green technologies into the actual pro­ Gan, C.H., Zheng, R.G., Yu, D.F., 2011. The impact of industrial structure change on
duction and processing process as early as possible. Resource- economic growth and volatility in China. Econ. Res. 5 (4), 4–16+31.
Gao, J.J., Zhang, W.W., 2021. Research on the impact of green finance on the ecological
based cities should also continue to put development as their structure of China’s industry–an empirical test based on systematic GMM model.
first priority, accelerate the construction of information Econ. Rev. 2, 105–115.

9
X. Wang Resources Policy 85 (2023) 103866

Gong, X.S., Wang, M., Zhang, H.Z., 2018. FDI, market segmentation and regional eco- Ross, M.L., 2015. What have we learned about the resource curse? Annu. Rev. Polit. Sci.
efficiency: direct effects and spillover effects. China Popul. - Res. Environ. 8, 95–104. 18, 239–259.
Guan, C.H., Han, J., 2019. 2017-2018 China Green Development Index Report. Economic Sachs, J., Warner, A., 1995. Natural Resource Abundance and Economic Growth. NBER
Daily Press, Beijing. Working Papers No, p. 5398.
He, A.P., An, M.T., 2019. Local government competition, environmental regulation and Sachs, J.D., Woo, W.T., Yoshino, N., 2019. Importance of green finance for achieving
green development efficiency. China Popul. - Res. Environ. 3, 21–30. sustainable development goals and energy security. In: Handbook of Green Finance.
Herrala, R., Goel, R.K., 2012. Global CO2 efficiency: country-wise estimates using a Springer, Singapore.
stochastic cost frontier. Energy Pol. 45, 762–770. Sadiq, M., Nonthapot, S., Mohamad, S., 2022. Does green finance matter for sustainable
Hsu, C.C., Quang-Thanh, N., Chien, F.S., 2021. Evaluating green innovation and entrepreneurship and environmental corporate social responsibility during COVID-
performance of financial development: mediating concerns of environmental 19? China Finance Rev. Int. 12 (2), 317–333.
regulation. Environ. Sci. Pollut. Control Ser. 28 (40), 57386–57397. Shen, N., 2012. Environmental efficiency, industry heterogeneity and optimal regulation
Huang, Q.H., Gao, M., 2016. Research on the effect of energy saving and emission intensity: a nonlinear test on panel data of Chinese industrial sectors. China Indus.
reduction of environmental regulation tools in China. Sci. Res. Manag. 6, 19–27. Econ. 3, 56–68.
Huang, L., Wu, C.Q., 2020. Research on the impact of environmental regulation on the Taghizadeh-Hesary, F., Yoshino, N., 2020. Sustainable solutions for green financing and
efficiency of green development of urban industry in the Yangtze River Economic investment in renewable energy projects. Energies 13 (4), 788.
Belt. Yangtze River Basin Resources and Environment 30 (5), 1075–1084. Tone, K., 2002. A slacks-based measure of super-efficiency in data envelopment analysis.
Jing, W.M., Zhang, L., 2014. Environmental regulation, openness and green Eur. J. Oper. Res. 143 (1), 32–41.
technological progress of Chinese industry. Econ. Res. 49 (9), 34–47. Tu, Z.G., 2008. The harmony of environment, resources and industrial growth. Econ. Res.
Khan, S.A.R., Sharif, A., Golpîra, H., 2019. A green ideology in Asian emerging 2, 93–105.
economies: from environmental policy and sustainable development. Sustain. Dev. UNESCAP, 2005. Implementing the monterrey consensus in the Asian and Pacific region:
27 (6), 1063–1075. achieving coherence and consistency. Asia Pac. Dev. J. 12, 141–143.
Kurniawan, T.A., Maiurova, A., Kustikova, M., 2022. Accelerating sustainability Wang, B., Huang, R.J., 2014. Regional green development efficiency and green total
transition in St. Petersburg (Russia) through digitalization-based circular economy in factor productivity in China:2000-2010 - an empirical study based on parametric
waste recycling industry: a strategy to promote carbon neutrality in era of Industry common boundaries. Indust. Econ. Rev. 1, 16–35.
4.0. J. Clean. Prod. 363, 132452. Wang, C., Sun, B.X., 2019. The impact of energy endowment, trade openness on the
Lee, C.C., Lee, C.C., 2022. How does green finance affect green total factor productivity? green efficiency of resource utilization. J. Dalian Univ. Technolgy 40 (2), 10–16.
Evidence from China. Energy Econ. 107, 105863. Wang, Z.J., Wei, J., Ren, B.P., 2020. The impact of two-way FDI on green economy
Li, J.T., Huang, H.Y., 2022. The ecological and environmental effects of green finance: an efficiency in the context of heterogeneous environmental regulations. Res. Fin. Trade
empirical review of Guangdong, Hong Kong and Macao in the Greater Bay Area 31 (3), 1–16.
under the dual carbon target. J. Guangdong Univ. Fin. Econ. 1, 87–95. Wang, Y.T., Zhang, Y.Y., Zhang, X., 2023. Research on the impact of green credit on
Li, J.L., Xu, B., 2018. "Curse" or "Gospel": how does resource abundance affect China’s green development efficiency - a spatial econometric analysis based on super-
green economic growth? Econ. Res. 53 (9), 151–167. efficient SBM model. J. Dalian Univ. Technolgy: Social Science Edition 44 (2),
Li, J., Tan, Q.M., Bai, J.H., 2010. Spatial econometric analysis of regional innovation 37–44.
production in China :An empirical study based on static and dynamic spatial panel Ye, W.X., 2021. Analysis of green development efficiency of provincial capital cities in
models. Manag. World 26 (7), 43–55. China. Res. Dev. Market 37 (11), 1281–1287.
Li, Y., Hu, H.Y., Li, H., 2020. An empirical analysis of the impact of green credit on Zeng, X.W., Liu, Y.Q., Man, M.J., 2014. Measurement analysis of the degree of green
industrial structure upgrading in China–based on provincial panel data in China. finance development in China. Journal of China Yan’an Cadre Institute 6, 112–121.
Econ. Issues 1, 37–43. Zhang, F.M., Jing, P.Q., 2006. A review of resource-based economy and its
Liu, X.P., Guan, K., 2018. Measurement and evaluation of green development efficiency transformation. China Soc. Sci. 6, 78–87.
in hubei yangtze river economic belt. Statistics and Decision Making 34 (18), Zhang, F., Xue, H.F., Shi, Z.W., 2018. Will resource endowment and environmental
103–106. regulation promote green development in manufacturing? Scientific Decision
Lu, L.W., Song, D.Y., Li, S.F., 2016. Research on green efficiency of urban development in Making 5, 60–78.
yangtze river economic zone. China Popul. - Res. Environ. 26 (6), 35–42. Zhang, F., Song, X.N., Dong, H.Z., 2019. The driving mechanism of resource endowment
Luukkanen, J., Kaivo-Oja, J., Vähäkari, N., 2019. Green economic development in Lao on green transformation and upgrading of manufacturing industry–explanation
PDR: a sustainability window analysis of green growth productivity and the based on spatial Durbin model. East China Econ. Manag. 33 (7), 111–119.
efficiency gap. J. Clean. Prod. 211, 818–829. Zhao, L.L., Ren, C.Y., 2012. Re-evaluation of economic efficiency of Bohai Economic
Lyeonov, S., Pimonenko, T., Bilan, Y., 2019. Assessment of green investments’ impact on Circle from the perspective of green economy: based on entropy weight method and
sustainable development: linking gross domestic product per capita, greenhouse gas DEA-Malmquist method. Res. Dev. Market 28 (1), 45–47.
emissions and renewable energy. Energies 12 (20), 3891. Zhao, L.D., Zhang, L., Xu, L., Hu, M.Z., 2016. The mechanism of human capital, industrial
Pan, Y., 2022. Research on the effect of green credit instruments on pollution reduction restructuring and green development efficiency. China Population-Res. Environ. 11,
and emission reduction of highly polluting enterprises–an analysis of the constraint 106–114.
mechanism of limiting loans to highly polluting enterprises. Price Theory Pract. 3, Zhao, L.D., Yuan, T., Zhao, Z.B., 2019. Study on the threshold effect of urbanization on
131–134. green development performance : an example of cities in two economic zones in the
Ren, S.G., Jiang, T.T., Li, X.L., 2016. A study on the differentiated mechanism of the Great Northwest and the Middle reaches of the Yellow River. Arid Zone Res. Environ.
impact of environmental regulation types on regional eco-efficiency in China. Econ. 33 (9), 10–16.
Manag. 1, 157–165. Zheng, H., Jia, S., Zhao, X., 2017. Analysis of regional eco-efficiency in China in the
Ren, Y.F., Fang, C.L., Lin, X.Q., 2017. Evaluation of eco-efficiency of four major urban context of new urbanization. Resour. Sci. 39 (7), 1314–1325.
agglomerations in the eastern coastal region of China. J. Geogr. 72 (11), 2047–2063. Zhang, J., Wu, G.Y, Zhang, J.P., 2004. Interprovincial physical capital stock estimation in
China:1952-2000. Econ. Res.,10,35-44.

10