1 s2.0 S0140988322000469 Main

Energy Economics 107 (2022) 105863
Contents lists available at ScienceDirect
Energy Economics
journal homepage: www.elsevier.com/locate/eneeco
How does green finance affect green total factor productivity? Evidence
from China☆
Chi-Chuan Lee a, Chien-Chiang Lee b, c, *
a
School of Public Administration, Southwestern University of Finance and Economics, Chengdu, China.
b
Research Center of the Central China for Economic and Social Development, Nanchang University, Nanchang, China
c
School of Economics and Management, Nanchang University, Nanchang, China
A R T I C L E I N F O A B S T R A C T
JEL classification: While the literature has studied various factors affecting green productivity growth, there is a relative dearth of
G28 empirical studies quantitatively analyzing the linkage between green finance development and green produc
G32 tivity. Based on a comprehensive index of green finance development, this research thus employs panel data of
O13
30 China's provinces for the period 2006–2018 to explore the influence of green finance on green total factor
O44
Q56
productivity, revealing estimation results that green finance development significantly improves the level of
green productivity. This beneficial effect tends to be stronger in provinces with higher levels of economic and
Keywords:
Green finance
social conditions, less public participation in environmental protection, and high pollution levels. We also find
Green total factor productivity that implementing a green finance policy can further enhance the impact of green finance development. The
Green economy empirical results herein offer policy implications to China's green finance planning and environmental policy.
Environmental policy
China
1. Introduction models postulate that a sustainable economy only depends on technol

ogy advancement, which is commonly measured by total factor pro
The ecological crisis and resource dilemma characterized by global ductivity (TFP) growth. However, the associated measurements ignore
warming, land desertification, resource depletion, and environmental the consideration of several social benefits regarding resource and
degradation since the 1960s have become challenges that need to be environmental factors (Chen and Golley, 2014; Tian and Lin, 2017). The
urgently resolved for the whole world (Chiu and Lee, 2020; Hao et al., absence of them is particularly evident when energy and the environ
2018; Wang et al., 2022; Wen et al., 2022). To cope with the changing ment become the main restraints on economic growth, because a portion
global climate and increasingly severe environmental risks and to ach of the inputs that were initially employed in production must be trans
ieve environmentally inclusive growth of sustainable development goals ferred to the activities of environmental governance, leading to
(SDGs), many countries have begun to develop their own green econ misleading policy guidance (Liu et al., 2020b). This limitation can be
omy, which denotes coordination in the development of the economy, readily solved by the notion of green productivity, which was first
resources, and ecological environment. In this regard, green produc proposed by the Asian Productivity Organization in the 1990s.
tivity is an important measure to assess the degree of how a country Accordingly, green productivity refers to a dynamic strategy that pro
mitigates its environmental concerns, improves environmental perfor motes productivity and improves environmental performance simulta
mance, and still ensures sustainable economic development. Therefore, neously. Since then, researchers have performed extensive studies on the
how to promote green productivity growth is largely considered as an measurement of green productivity growth to assess the sustainability of
important topic in the new phase of economic development, thus an economy (Liu et al., 2020b; Xia and Xu, 2020; Yan et al., 2020).
attracting broad attention from policy-makers and scholars. As the largest developing and transition economy in the world, China
The traditional Solow growth model and the endogenous growth has made substantial progress and considerable achievements ever since
Chien-Chiang Lee is grateful to the Social Science Foundation of Jiangxi Province of China for financial support through Grant No: 21JL02. Chi-Chuan Lee is
☆
grateful for the support from the Fundamental Research Funds for the Central Universities of China.
* Corresponding author at: School of Economics and Management, Nanchang University, Nanchang, Jiangxi, China.
E-mail address: cclee6101@gmail.com (C.-C. Lee).
https://doi.org/10.1016/j.eneco.2022.105863
Received 28 October 2021; Received in revised form 20 December 2021; Accepted 23 January 2022
Available online 29 January 2022
0140-9883/© 2022 Elsevier B.V. All rights reserved.
C.-C. Lee and C.-C. Lee Energy Economics 107 (2022) 105863
its reforms and opening up began in the late 1970s (Lee et al., 2021b). 2021). While another strand investigates the role of government policies
Along with accelerated industrialization and fast economic expansion, and regulatory frameworks in shaping green productivity (e.g., Li and
China's extensive use of energy and natural resources and accompanying Lin, 2016; Wu et al., 2020a). These valuable efforts are important to our
concerns over environmental pollution have also increased dramati understanding the issue of what causes changes in green productivity
cally. According to statistics of the U.S. Energy Information Adminis and guides governments and policymakers to improve green economy
tration (EIA), as illustrated in Fig. 1, China's energy consumption in development.
2018 was 3719 million metric tons of oil equivalent (Mtoe), accounting This paper assesses GTFP by applying a super-SBM model with un
for 24.6% of global consumption. China is currently the world's largest desirable outputs proposed by Tone (2002). This approach is widely
emitter of carbon emissions at 29.1% of total global carbon dioxide applied in the literature for the case of treating resource and environ
emissions. These statistics clearly show that China is facing huge mental constraints during the process of sustainable development (e.g.,
development and environmental challenges that have concerned the Guo and Yuan, 2020; Liu et al., 2021; Wu and Zhang, 2020; Yu et al.,
international community. The ways to improve green productivity 2019). The advantages of this method have the following three points:
growth through reducing energy consumption, protecting the environ (1) imposing the weak disposability assumption indicates that
ment, and sustaining economic development at the same time have decreasing bad outputs is costly, which is consistent with the actual
become a serious challenge for China. process of production. (2) Unlike the traditional Data Envelopment
The current research aims to gain further insights on the develop Analysis (DEA), which neglects the input or output slack leading to
ment of green finance policy and the resulting green productivity change biased estimation, the slack variables are inserted into the objective
of 30 China's provinces for the period 2006–2018. During the Paris function to eliminate radial and oriented deviation (Song et al., 2012).
Climate Change Conference in 2015, China pledged to achieve the goal (3) Introducing Super-efficiency DEA allows us to rank multiple efficient
of a carbon emissions peak before 2030. On September 22, 2020, China DMUs effectively (Guo and Yuan, 2020).
announced that it intends to achieve the target of carbon neutrality The recent progress in China's construction of green financial mar
before 2060, which will help further accelerate the mitigation of CO2 kets and the formation of a green financial system have raised some new
emissions. Exploring a viable green development mode will not only issues on the role of green finance development in affecting green pro
solve the severe economic and environmental problems to meet the ductivity. To promote green economy transformation, green finance has
above commitments, but also offer solutions enacted in China for other become an important tool to realize green growth. For example, the 13th
countries to realize their own green economy transformation and to Five-year Plan (2016–2020) in China calls for constructing a green
promote green development. This motivates us to assess green total financial system, introducing green credit and green securities, as well
factor productivity (GTFP) and investigate the driving factors affecting as establishing a green development fund. Based on the statistics of
GTFP in the context of the twin objectives of economic development and Climate Bond Initiative (CBI), the total amount of green bonds issued has
carbon neutrality. reached US$55.8 billion in 2019, accounting for 22% of global issuance.
Abundant efforts in the literature have been made to explore the With the increasing development of green finance, there has been
issue of the green economy over the past few decades. One strand con greater interest in exploring the concept and the assessment of green
structs an indicator to assess the development of a green economy. finance and its influence on the economy and environment (He et al.,
Earlier researches use the traditional TFP index evaluate the sources of 2019a, 2019b; Liu et al., 2019; Hu et al., 2021; Lee et al., 2021a). As one
economic growth without consideration of undesirable output (Feng of the largest sources of global green bond issuances, the influence of
et al., 2018; Walker et al., 2020), while more recent studies propose green finance development cannot be ignored in the exploration of green
GTFP index which takes energy constraint and environmental perfor productivity. From an economic perspective, the capital support func
mance into account (Xia and Xu, 2020; Yan et al., 2020; Gao et al., tion of green finance can effectively accelerate the rationalization of
2021). Among these discussions, GTFP is largely considered as a better economic structure adjustment and improve supply-side investment and
indicator to reflect sustainable economic growth (Feng et al., 2018; Liu its quality, thus stabilizing economic growth (He et al., 2019a, 2019b;
et al., 2020b; Wang and Lee, 2022). For example, Xia and Xu (2020) use Lee et al., 2022; Xie et al., 2020; Zhou et al., 2020). From the environ
GTFP to explore the quality of growth in China and show different trends mental viewpoint, the ‘going green’ feature can guide enterprises to
between GTFP and standard TFP indexes ignoring environmental out engage in green innovation and environmental protection activities and
comes. Some provinces in northeast region performed better, while to promote corporate social responsibility and environmental perfor
others in western region performed worse. Another branch of literature mance, thereby achieving sustainable economic and environmental
investigates the influencing factors of green productivity growth in development (Xu and Li, 2020; Zhou et al., 2020; Hu et al., 2021).
China. Such research is mainly concentrated on exploring the impacts The contributions of the current research can be summarized as
from economic, resource, and ecological elements on a country's green three-fold. First, we fill the gap in the literature on increasing awareness
economy development. One prominent strand examines the effect of as to the importance of the green economy by exploring the connection
economic scales, structure, and technological progress and innovation between green finance and green productivity. Compared with the
(e.g., Yan et al., 2020; Yuan et al., 2020; Wu et al., 2020b; Gao et al., standard approach that considers green productivity is a function of
scale, technology, and structure, we expand the theoretical framework
12,000.00 30.00 by adding green finance development as an important influencing factor
on GTFP. By doing so, more complete information about the causes of
10,000.00 25.00
green productivity can be drawn not only from the real sector, but also
8,000.00 20.00 from the financial sector. Second, compared to previous single-
dimensional measures of green finance such as low-carbon financial
6,000.00 15.00
flows and the amount of green credit or green investment, we extend
4,000.00 10.00 existing research by introducing a thorough index system of green
finance from the perspectives of green-oriented credit, securities, in
2,000.00 5.00
surance, and investment to explore the role of green finance develop
0.00 0.00 ment on the green economy. The multidimensional index could provide
a more complete evaluation on the development of green finance. Third,
unlike previous works on traditional productivity that mainly considers
Energy intensity Energy consumption CO2 emissions
good output, we apply the super-SBM model with undesirable output
Fig. 1. Energy use and emissions in China over 1980–2018. and solve the potential omission problem of energy and environmental
2
constraints. consumer and investment behavior guidance, the implementation of

The remainder of the paper is arranged as follows. Section 2 presents green credit policy not only promotes the adjustment of industry
a summary of relevant prior research. Section 3 introduces the empirical structure, but also improves the ecological environment.
model and methodology. Section 4 describes the data used in this study. Compared with previous qualitative analysis, there is a relative
Section 5 presents and discusses the estimation results. Finally, Section 6 dearth of research conducting quantitative analysis in measuring the
concludes and draws some policy implications. level of green finance development. Existing studies mainly use a single-
index evaluation such as progress on low-carbon financial flows (Bodnar
2. Literature review et al., 2017; Hafner et al., 2020) and the amount of green credit and
green investment (He et al., 2019b; Xie et al., 2020) to proxy for the
2.1. The measure of green productivity growth development of green finance. Although those single-index indicators
can reflect the development of green finance to a certain extent, it is
As mentioned in the prior section, it is essential to consider resource difficult for them to reveal the overall picture of green finance. To solve
and environmental constraints when assessing green productivity the limitations of current indicators, some scholars further construct
growth. Most earlier works rely on either treating undesirable outputs as evaluation systems to measure the comprehensive level of green finance
inputs (Ramanathan, 2005; Reinhard et al., 2000) or using the data development. Referring to the composite system of green financial
transformation approach for the undesirable outputs (Scheel, 2001; development established by Zeng et al. (2014) and He et al. (2019a), Liu
Seiford and Zhu, 2002), which contradict the actual production process et al. (2019) propose the index system via five levels (i.e., green-oriented
and may provide a misleading evaluation of green productivity growth. credit, securities, insurance, investment, and carbon finance) to measure
The third approach is to apply the directional distance function (DDF) China's green finance development.
with weak disposability assumption of undesirable outputs. For
example, based on the DDF approach, Gao et al. (2021) explore China's 2.3. Green finance and economic development
provincial TFP under the consideration of carbon emissions. Their
empirical results indicate that green productivity tends to be stable with The theoretical literature suggests that financial development can
smaller variance after accounting for carbon emissions. They also find contribute to economic growth through mobilizing savings, promoting
regional differences in that eastern provinces have higher levels of capital accumulation, improving capital allocation, and facilitating risk
productivity growth. However, the DDF approach implicitly assumes management and trade (see Levine, 2005, for a comprehensive survey).
identical changes in the proportions of desirable and undesirable out As a new type of financial innovation, the functions and characteristics
puts, which appear to be strong in reality (Chen and Golley, 2014). of green finance are not inherently distinct from traditional finance, but
To further address the radial and oriented deviation in evaluating their environmental purpose appears to be the key distinction between
green productivity growth, many recent studies shift their attention on them. As noted by He et al. (2019a, 2019b), on the one hand, green
the use of the slack-, non-radial, and non-oriented SBM models proposed financial development can help strengthen the financial structure of the
by Tone (2001). Following this vein, Tone (2002) develops a new super- macro-environment and guide micro-entities to be involved in envi
SBM model through combining the advantages of Super-efficiency DEA ronmental and economic benefits for the development of green econ
and the SBM model to further rank the efficient decision-making units. omy. On the other hand, through changing traditional “financing-
Since then, the super-SBM model has evolved into the most powerful and investment” business, green finance plays a crucial role in affecting
widely used tool for evaluating green productivity growth. For example, enterprises' environmentally friendly investments and becomes the
Wu and Zhang (2020) examine the influence of Internet technology on major channel to develop a green financial system, thus achieving sus
China's forestry GTFP. Using the super-SBM model, the results reveal tainable development. Their empirical evidence suggests in the long run
that technical progress constitutes the primary source of forestry GTFP. that renewable energy investment effectively promotes the growth of
green economy. Following this vein, Zhou et al. (2020) explore how
2.2. The concept and the measurement of green finance green finance impacts economic development and environmental qual
ity. The results reveal that green finance can fulfill a win-win situation
As a new financial mode, there is no unified definition of green between economic growth and environmental optimization. Zhang
finance, but the existing literature offers three main concepts. The first (2021) also indicates that green finance contributes to green produc
refers to environmental finance, which deals with practical issues of tivity through reducing pollution emissions and increasing cleaner
environmental protection, pollution control, resource conservation, and production. By dividing R&D into either environmental-induced R&D or
other green projects through providing financial services (Salazar, 1998; production R&D, evidence reveals that green productivity is positively
Cowan, 1999; Gray, 2002). The second view considers it as a kind of impacted by a green credit policy.
financial innovation that prevents environmental risks through a variety
of financial instruments and products (Labatt and White, 2002). The 2.4. Green finance and environmental performance
third and more recent perspective suggests that it is a financial form to
promote environment-friendly investments and foster an ecologically The influence of green finance on environmental performance runs
minded society through green-oriented credit, securities, insurance, and along various channels: capital support effect, resource allocation effect,
investment, as well as carbon finance (He et al., 2019a; Liu et al., 2019; and technological innovation effect (He et al., 2019b; Liu et al., 2019).
Hu et al., 2021). The capital support effect means that green finance can provide capital
Based on the above concepts, increasing efforts in the literature have to low-energy consuming, low-polluting, and low-carbon enterprises
been made to explore the issue of green finance over the past decades. and restrain high-pollution and high-emission production behavior,
These studies mainly focus on qualitative analysis and a conceptual leading to better environmental performance. For example, from the
understanding of green finance. For example, Salazar (1998) argue that micro-mechanisms, Xu and Li (2020) assess the impact of a green credit
green credit policies are helpful for promoting allocation efficiency of policy and its development on enterprises' financing cost in China. Their
financial resources and guiding capital flow from high-energy con empirical results suggest that green credit increases the cost of financing
sumption and high-pollution industries to environmental protection for high-polluting and high-emission enterprises, while it reduces the
industries, thereby improving environmental quality. Cowan (1999) cost of green firms. From the macro-mechanisms, Liu et al. (2019)
also indicates that green financial capital and environmental gover consider the credit support function of green financial development and
nance investment play a pivotal role in mitigating the derived problems examine the effect of green finance on the green economy. They find that
of environmental degradation. Through credit resource allocation and the green finance effect on ecological efficiency exhibits regional
3
differences. variables are equal to 0. When ρ* < 1, the DMU is inefficient, but can be
The resource allocation effect states that green finance can promote efficient through decreasing the excess inputs and undesirable outputs
the efficiency of capital utilization and guide the flow of financial re and increasing the shortage of desirable outputs. However, there may be
sources from low-efficiency and high-pollution industries to high- multiple efficient DMUs with an efficiency value of 1 simultaneously.
efficiency and low-polluting industries, thereby boosting industrial Given the efficiency value is bounded from 0 to 1, further distinguishing
upgrading, optimizing energy structure, and improving environmental the efficiency among these efficient DMUs is impossible. To solve this
quality. From the micro-mechanisms, Fan et al. (2021) conduct an difficulty, Tone (2002) proposes a super-SBM model with undesirable
investigation of how a green credit policy affects enterprises' loan con outputs as:
duction. The results reveal that non-compliant firms are more likely to 1
∑M x
have higher interest rates and encounter more difficulty in accessing ρ** = min (
m j=1 xj0
) (4)
loans. From the macro-mechanisms, a developed financial system can 1
∑ N1 yd ∑ N2 yu
1 + N1 +N +
contribute to environmental quality by allocating idle funds to relieve 2 p=1 yd
p0
q=1 yu
q0
the financial constraints of environmentally friendly enterprises and

promoting green upgrading (Acheampong et al., 2020; Zhou et al., 2020; ⎧
⎪
Lee et al., 2021; Wang et al., 2021; Yu et al., 2021). ⎪
⎪
⎪ ∑M
⎪
⎪ x≥ xjk λk
The technological innovation effect means that green finance can ⎪
⎪
⎪ k=1,∕
=0
⎪ ∑N1
provide external credit support to enterprises involved in green tech ⎪
⎪
⎨ yd ≤ ydpk λk
nology innovation activities, improve the efficiency of energy use, thus s.t. k=1,∕
=0
(5)
⎪ ∑N2
ultimately promoting the development of the green industry and ⎪
⎪
⎪ yu ≥ yu λk
⎪ =0 qk
reducing ecological damage and environmental pollution. Li et al. ⎪ k=1,∕
⎪
⎪
⎪
⎪ x ≥ xjk , yd ≤ ydpk , yu ≥ yuqk , λk > 0
(2018) construct a green loan theory among enterprises, banks, and ⎪
⎪
⎩
government to assess the performance of green loan and subsidy on
green production innovation. They conclude that government subsidy Here, ρ** represents the super efficiency value of the evaluated DMU
can decrease enterprises' financing cost, thus improving the intention of
whose value can be larger than 1; and x, yd , and yu are the mean vector
executing technological innovation. Hu et al. (2021) explore the role of a
of inputs, desirable output, and undesirable output, respectively.
green credit policy at affecting green innovation for high-polluting en
Table A1 of Appendix presents the input-output indicators used in the
terprises. They find that the enforcement of the ‘Green Credit Guide
super-SBM model.
lines’ in 2012 did exert a significantly positive effect on the green patent
output of enterprises.
3.2. Econometric models
3. Empirical model and methodology
According to the standard approach of Grossman and Krueger (1995)
3.1. Super-SBM model with undesirable output and Copeland and Taylor (2004) in hypothesizing the nexus between
economic activities and the environment, sustainability and green
Suppose there are k decision-making units (DMUs) at time t, and development are viewed as a function of scale, technology, and struc
each DMU uses M-vector of inputs Xit = (x1it, ⋯, xMit) ∈ ℝM+ to produce ture. In this study we expand the theoretical framework by adding other
N1-vector of desirable outputs yitd = (y1itd, ⋯, yN1itd) ∈ ℝN1+ and N2- important influencing factors on GTFP. The basic function is:
vector of undesirable outputs yitu = (y1itu, ⋯, yN2itu) ∈ ℝN2+. All input-
GTFP = F(Scale, Technique, Structure, other influencing factors) (6)
output factors are under the non-zero assumption. The production
possibility set is then defined as: Here, Scale denotes the scale effect, reflecting the scale of economic
{( )⃒ } activity on green productivity; Technique stands for the technology ef
P = x, yg , yb ⃒x ≥ Xλ, yg ≤ Y g λ,yb ≥ Y b λ,λ ≥ 0 (1) fect, reflecting technological progress on green productivity; Structure
represents structural effect, reflecting changes in economic structure.
where λ is a non-negative intensity vector of weights assigned to all
Referring to Cheng et al. (2018), Wu et al. (2020a, 2020b), and Yuan
input-output factors. Under resource and environmental constraints, the
et al. (2020), GDP per capita (PGDP) and financial support (FS) are used
mathematical form of the SBM model with undesirable output (Tone,
to measure the scale effect, while R&D investment (RD) and foreign
2001) for GTFP can be expressed as:
direct investment (FDI) are used to measure the technique effect, and
∑ s−j0
1 − m1 M industrial structure (INS) and energy consumption structures (ES) are
ρ* = min (2) used to measure the structural effect. Among other influencing factors,
j=1 xj0
( )
1
1 + N1 +N2
∑ N1 dp
s ∑N2 suq green finance (GF) is considered as the key variable of interest in this
p=1 yd + q=1 yu
p0 q0
study. Based on previous studies, we also incorporate urbanization
(URB), education level (EDU), and trade openness (TRADE) as other
⎧
⎪ control variables. Based on the above analysis, the specific model can be
⎪ ∑M
⎪
⎪
⎪
⎪ xj0 = xj λj − s−i described as:
⎪
⎪ j=1
⎪
⎪ ∑N1
⎪ GTFPit = β0 + β1 GF it + β2 PGDPit + β3 FSit + β4 RDit + β5 FDI it
⎪
⎨ ydp0 = yd λp − sdp (7)
s.t. j=1 p
(3) +β6 INSit + β7 ESit + β8 URBit + +β8 EDU it + β8 TRADEit + αi + εit
⎪
⎪ ∑ N2 u
⎪
⎪
⎪ yuq0 = y λq + suq
j=1 q Here, i identifies the cross-sectional unit that involves 30 provinces
⎪
⎪
⎪
⎪
⎪ in China; t denotes the time period from 2006 to 2018; αi is the indi
⎪ s−i ≥ 0, sdp ≥ 0, suq ≥ 0, λ ≥ 0
⎪
⎩ vidual fixed effect; and εit is the error term.
In Eq. (2) the optimal solution of the objective function is ρ*, which 4. Data description
indicates the target efficiency value in the range [0,1]; and s = (si− , spd,
squ) is a slack vector corresponding to the abundance of inputs, scarcity 4.1. Sample selection and data sources
of desirable outputs and excess of undesirable outputs, respectively. For
a given evaluated DMU, it is efficient if and only if ρ* = 1 and all slack Considering the availability and validity of data, we construct a
4
balanced panel dataset from 30 provinces in China over the period ratio of financial institutions' loan balances to GDP to proxy economic
2006–2018, with a total of 360 samples (Tibet is excluded from our development level (PGDP) and the degree of financial support (FS),
sample due to a lot of missing values). Our sample stops in 2018, because respectively. Wang et al. (2019) and Zheng et al. (2021) indicate that the
the statistical data in most recent years are incomplete. All data are level of economic development reflects its capability to pursuit green
collected from the China Statistical Yearbook and the provincial statis innovation and productivity growth. The ratio of R&D expenditure to
tical yearbooks, China Energy Statistical Yearbook, China Statistical GDP and the ratio of actual use of FDI to GDP measure technological
Yearbook of Environment, and China Statistical Yearbook of Science and progress caused by local firms (RD) and international technology spill
Technology. Table B1 of Appendix summarizes the variables' definitions, overs (FDI), respectively. These factors are largely considered as the
and the descriptive statistics are reported in Table 1. All nominal vari primary driver of technological progress, which contributes to
ables are deflated by the GDP deflator index with base year 2000. improving productivity (Gao et al., 2021; Shen et al., 2019).
Considering that the industrial sector and coal consumption are both
the major contributors to pollutant emissions (Wen and Lee, 2020; Yuan
4.2. Core explanatory variable
et al., 2020), we employ the ratio of the secondary industry to GDP and
the proportion of coal consumption in total energy consumption to
Following Zeng et al. (2014), He et al. (2019a), and Liu et al. (2019),
quantify industrialization level (IND) and energy consumption struc
we use four indicators of green-oriented credit, securities, insurance,
tures (ES), respectively. Finally, following Li and Lin (2017), Gao et al.
and investment to construct a comprehensive index of green finance.
(2021), Yan et al. (2020), Yuan et al. (2020), and Zhao et al. (2020), we
The related indicators and corresponding weights in the evaluation
use the proportion of urban population to total population, average
system are described in Table B2 of Appendix.
years of education, and the ratio of total trade amount to GDP to reflect
The index calculation proceeds in three steps. First, each signal in
social-economic changes such as urbanization (URB), education level
dicator is standardized through mathematical transformation to elimi
(EDU), and trade openness (TRADE), respectively.
nate any measurement problem, such as incommensurability and
contradiction among the indicators. To this end, the indicators are
divided into positive and negative indicators, and we use the formula of 5. Results and discussion
the inter-group range value for processing. Therefore, the standardized
value of Xijs+ and Xijs− can be specified as: 5.1. Spatiotemporal variations of green total factor productivity and their
( ) components
Xij − min Xj
Positive indicator : Xijs+ = ( ) ( ) (8)
max Xj − min Xj Based on a Super-SBM model with undesirable output, we measure
and compare GTFP and its components – namely, green technical change
( )
max Xj − Xij (GTC) and green efficiency change (GEC) – for 30 provinces in China,
Negative indicator : Xijs− = ( ) ( ) (9)
max Xj − min Xj spanning 2006 to 2018. Table A2 of Appendix summarizes the average
values of GTFP, GTC, and GEC, and the temporal variations of GF and
Here, Xij represents the original value of the jth indicator in the ith GTFP are drawn in Fig. 2. It can be seen that both GF and GTFP exhibit
province; and max(Xj) and min(Xj) are the maximum and minimum an overall upward trend with periodic variations, suggesting that pro
values of the jth indicator for all provinces in China, respectively. Sec duction activity tends toward less pollution given the same inputs. This
ond, after standardizing each indicator, the weight matrix (Wj) is characteristic is consistent with the findings of Wu et al. (2020a), Xia
established to assign a weight for each indicator according to its relative and Xu (2020), and Gao et al. (2021) and can be attributed to the
importance in the process of evaluating objects. Third and finally, using implementation of the 11th Five-year Plan (2006–2010) and 12th Five-
the standardized value and the weight of each indicator, the compre year Plan (2011–2015) with specific goals of energy conservation and
hensive index of green finance in each province is then obtained by: emission reduction. The various corresponding reform measures effec
∑J tively control pollutant discharge and enhance the effectiveness of
GF i = Wj × Xijs (10)
j=1 environmental governance, thereby increasing its green productivity.
Here, GFi is the comprehensive index of green finance in the ith Table A3 of Appendix presents regional heterogeneity among prov
province; and J is the sum of indicators selected for measuring the inces for these productivity measures, and the ranking of GTFP for each
development of green finance. province is depicted in Fig. 3. The average GTFP values range between
0.9520 and 1.1645, with an overall mean value of 1.0693. As shown at
the bottom of the table, green productivity is the highest in the eastern
4.3. Control variables region, followed by the central and western regions. These results
confirm the findings of prior studies and suggest high coordination be
Referring to the literature, we also introduce other factors as control tween economic and environmental performance in the eastern region
variables. As suggested by Gao et al. (2021), Wu et al. (2020a), and Yan
et al. (2020), we use the logarithmic value of GDP per capita and the
1.20 0.45
0.40
Table 1 1.15
0.35
Descriptive statistics.
1.10 0.30
Variable Obs. Mean Std. Dev. Min. Max.
0.25
GTFP 390 1.520 0.698 0.446 5.639 1.05
GF 390 0.302 0.0746 0.159 0.621 0.20
PGDP 390 10.13 0.576 8.528 11.53 1.00 0.15
ES 390 94.98 40.65 3.800 242.3
0.10
URB 390 54.11 13.65 27 90 0.95
IND 390 38.74 8.484 11.84 53.04 0.05
TRADE 390 30.34 36.54 1.690 172.2 0.90 0.00
FDI 390 2.258 1.728 0.0100 8.190 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
FS 390 121.8 43.41 53.72 258.5
GF GTFP
EDU 390 8.829 0.985 6.590 12.68
R&D 390 1.470 1.061 0.200 6.010
Fig. 2. Temporal variation of China's GTFP.
5
Hebei
Henan
Shandong
Hunan
Jiangsu
Inner Mongolia
Fujian
Liaoning
Sichuan
Shanghai
Tianjin
Hubei
Beijing
Guangdong
Heilongjiang
Zhejiang
Chongqing
Jilin
Anhui
Guangxi
Yunnan
Shaanxi
Jiangxi
Shanxi
Guizhou
Gansu
Xinjiang
Hainan
Ningxia
Qinghai
0.95 1 1.05 1.1 1.15 1.2
Fig. 3. Rankings of provinces by average GTFP over 2006–2018.
(Wu et al., 2020a; Gao et al., 2021). Furthermore, provinces located in Table 2 present the estimated results of the pooled ordinary least squares
the eastern region such as Hebei, Henan, and Shandong are the top three (POLS), the random effect (RE), the fixed effect (FE), the two-way fixed
is greatest green productivity, while those located in the western region effect (Two-way FE), and the generalized method of moments (GMM)
such as Qinghai, Ningxia, and Hainan are the bottom three with the least models. As to the core explanatory variable of interest, the results in the
green productivity. table reveal that the coefficients of GF are significantly positive at the
To gain more insight into the regional differences and dynamic 5% level, suggesting that the development of green finance can improve
changes of GTFP derived from the green credit policy, Fig. 4 presents a the level of green productivity. This result is consistent with and en
direct visualization of the results in the form of topographic maps, where riches the findings of He et al. (2019b) and Zhang (2021), who indicate
the darker red colors represent the highest GFTP level in 30 provinces of that green investment and credit can create a win-win situation between
China. It can be seen that GTFP has been constantly changing for 2006 to environment and economy, thereby promoting green economic devel
2018 with an overall upward trend. We also find remarkable evolution opment and green productivity. Therefore, in order to achieve the target
trends in 2012 and 2015, which can be attributed to the implementation of carbon neutrality in the second half of the century, China's central
of the ‘Green Credit Guidelines’ in 2012 and the ‘Energy Efficiency government should focus its attention on solving the problems of
Credit Guidelines’ in 2015. In addition, higher green productivity was financial constraints faced by enterprises during the process of con
concentrated in the eastern region in 2006 and gradually transferred to ducting environmental-friendly activities, which requires a well-
central and western provinces. This result runs in line with the findings functioning green finance system. By doing so, enterprises would more
of Yuan et al. (2020) and Wu et al. (2020a). likely to be involved in environmental and economic activities that
benefit the development of the green economy.
5.2. The influence of green finance on green total factor productivity As to the influences of control variables, evidence reveals that most
variables reach statistical significance. The coefficients of PGDP and
To assess the influence of GF on GTFP, we first estimate the baseline R&D are significantly positive, indicating that the higher the levels of
model as demonstrated in Eq.(7) by utilizing the panel data regressions economic development and technological progress are, the more green
with five different specifications. Following Yan et al. (2020) we use productivity can be improved. These results confirm the findings of Wu
accumulated GTFP which is obtain by the level of accumulated GTFP in et al. (2020a) and Yuan et al. (2020). On the one hand, R&D reflects the
the previous year multiplied by current GTFP (accumulated GTP for the local capacity of science and technology, which exert a beneficial impact
first year is set to be 1) as our dependent variable. Columns (1)–(5) in on economic development (Gao et al., 2021). On the other hand,
6
Fig. 4. Spatial distribution of China's GTFP over 2006–2018.
engaging in R&D activities can create conditions to improve environ growth group (HGDP), low foreign investment group (LFDI) and high
mental quality (Huang et al., 2020). The coefficients of ES and IND are foreign investment group (HFDI), and low financial support group (LFS)
significantly negative, indicating that green productivity decreases with and high financial support group (HFS). Table 3 summarizes the corre
energy consumption structure and industrialization level. These results sponding estimates for these sub-sample groups. Evidence reveals that
match our expectations and indicate the unreasonable structure of the beneficial impact of green finance stated in the prior section is much
China's high industrialization and strong dependence on coal con stronger in provinces with higher levels of economic growth, foreign
sumption (Li and Lin, 2016; Shen et al., 2019; Liu et al., 2020a; Yan investment, and financial support. These results can be attributed to its
et al., 2020). The coefficients of FS and EDU are significantly negative, absorptive capacity, as reflected by economic and financial conditions
implying that the high degrees of traditional financial support and ed (Aluko and Obalade, 2020; Dai et al., 2021; Du et al., 2019; Hu et al.,
ucation are associated with low green productivity. This can be attrib 2021), and are consistent with a well-known hypothesis in cross-country
uted to the wealth effect in emerging economies. The increased wealth studies, which posits that a country with strong absorptive capacity can
caused by higher levels of financial support and education tends to enhance the positive impact on growth (Borensztein et al., 1998; Lee
stimulate economic growth, which ultimately increases the demand for et al., 2017; Shen et al., 2010). Therefore, local governments in China
energy use, which is thus harmful for green productivity (Sardosky, should be fully aware of the importance of their economic activities in
2010, 2011; Acheampong, 2019). order to maximize the benefits of green finance.
As to the influence of social conditions, the total sample is divided
into two categories: low urbanization group (LURB) and high urbani
5.3. Heterogeneity analysis zation growth group (HURB) as well as low education group (LEDU) and
high education group (HEDU). The estimation results in Table 4 reveal
As the effect of green finance on green productivity may vary in that green finance in provinces with higher levels of urbanization and
different areas, we further split the whole sample into different levels of education is more likely to have a larger positive effect on green pro
economic activity, social condition, environmental protection, and ductivity. One possible explanation is regarding the innovation capacity,
pollution status. When economic activity is of concern, the total sample which is reflected by infrastructure and human capital in these provinces
is divided into three categories: low growth group (LGDP) and high
7
Table 2
Estimation results of the basic model.
Variable (1) (2) (3) (4) (5)
POLS RE FE Two-way GMM
GF 0.9200 1.3438*** 1.6632*** 1.1682** 1.8141***

(1.3237) (3.5432) (4.3742) (2.4481) (4.5021)
PGDP 1.5369*** 1.4552*** 0.6480*** 0.3077 0.7631***
(13.8990) (10.6800) (3.0724) (0.7033) (3.1086)
ES 0.0016** 0.0011 − 0.0055*** − 0.0031* − 0.0043**
(2.1914) (1.0439) (− 3.2563) (− 1.8979) (− 2.4500)
URB − 0.0226*** − 0.0127 0.0113 − 0.0148 0.0081
(− 4.3399) (− 1.3809) (0.8897) (− 1.1852) (0.5791)
IND − 0.0259*** − 0.0403*** − 0.0551*** − 0.0284*** − 0.0583***
(− 5.8696) (− 8.0666) (− 9.4760) (− 4.1349) (− 9.3698)
TRADE − 0.0042*** − 0.0057*** − 0.0010 0.0022 − 0.0020
(− 4.3583) (− 3.8519) (− 0.5003) (1.0003) (− 0.8926)
FDI − 0.0223* − 0.0061 0.0343 0.0245 0.0313
(− 1.8273) (− 0.2889) (1.5043) (1.1475) (1.3192)
FS − 0.0078*** − 0.0086*** − 0.0067*** − 0.0086*** − 0.0075***
(− 7.6184) (− 7.6344) (− 5.0888) (− 6.2971) (− 5.4303)
EDU − 0.2243*** − 0.3299*** − 0.1903** − 0.1026 − 0.2102**
(− 4.0208) (− 4.6914) (− 2.1975) (− 0.9625) (− 2.3650)
R&D 0.3712*** 0.4165*** 0.6818*** 0.7081*** 0.6923***
(8.7277) (7.2694) (6.4685) (6.6121) (6.2249)
cons − 9.6839*** − 7.9518*** − 2.0513 0.6746 –
(− 11.6123) (− 7.7829) (− 1.5422) (0.1709) –
R2 0.6406 0.6447 0.6929 0.7454 0.6910
F / Wald 54.7946*** 652.112*** 78.9516*** 44.9915*** 71.5562
Obs. 390 390 390 390 360
Notes: t statistics are in parentheses. * p < 0.1, ** p < 0.05, and *** p < 0.01.
Table 3
Heterogeneity analysis: economic activity.
Variable (1) (2) (3) (4) (5) (6)
LGDP HGDP LFDI HFDI LFS HFS
GF 0.7589 2.7525*** 0.1187 3.0207*** 0.9020** 2.0104***

(1.6491) (4.3293) (0.3915) (5.3051) (1.9864) (3.4659)
PGDP 0.8503*** 0.8802* 0.9370*** 0.8615** 1.2129*** − 0.2230
(3.7396) (1.9417) (6.0579) (2.0975) (5.1721) (− 0.6370)
ES − 0.0068* − 0.0009 − 0.0041*** 0.0007 − 0.0047** − 0.0023
(− 1.9054) (− 0.3559) (− 3.6010) (0.1905) (− 2.0848) (− 0.9328)
URB 0.0101 − 0.0179 − 0.0127 0.0215 − 0.0124 0.0568**
(0.6869) (− 0.6293) (− 1.1691) (1.0633) (− 0.8796) (2.5229)
IND − 0.0429*** − 0.0571*** − 0.0308*** − 0.0824*** − 0.0670*** − 0.0539***
(− 5.8003) (− 5.4444) (− 6.1915) (− 7.5590) (− 7.8011) (− 6.9910)
TRADE − 0.0016 − 0.0024 0.0090* − 0.0007 − 0.0017 0.0039
(− 0.6052) (− 0.7550) (1.7678) (− 0.2715) (− 0.8816) (0.4195)
FDI − 0.0151 0.1692*** 0.1541*** 0.0803*** 0.0229 0.1351**
(− 0.5583) (3.4255) (5.1133) (2.6341) (0.9085) (2.5539)
FS − 0.0052*** − 0.0076*** − 0.0043*** − 0.0032 − 0.0030 − 0.0066***
(− 3.1925) (− 3.3535) (− 4.0518) (− 1.4716) (− 1.6183) (− 3.5854)
EDU − 0.2497** − 0.2004 − 0.1347* − 0.1258 − 0.1388 − 0.2339*
(− 2.0503) (− 1.4969) (− 1.8951) (− 0.9998) (− 1.3053) (− 1.8621)
R&D 0.5060*** 0.9876*** 0.3046** 0.1041 0.0731 1.1854***
(3.8578) (5.0645) (2.5348) (0.7172) (0.6134) (5.6941)
cons − 3.4471** − 4.0608 − 4.6967*** − 5.1926* − 5.9384*** 3.8149*
(− 2.2758) (− 1.4506) (− 4.8860) (− 1.8000) (− 4.1461) (1.6721)
R2 0.7925 0.6200 0.6912 0.7836 0.8011 0.6812
F-value 60.3481*** 25.7783*** 36.2688*** 59.3877*** 67.2443*** 35.9052***
Obs. 194 196 195 195 195 195
(Balaguer and Cantavella, 2018; Chen et al., 2019; Yuan et al., 2020). divided into three categories: low industrial water emission group
With respect to the role of environmental protection, the total sample (Lwater) and high industrial water emission group (Hwater), low in
is divided into two categories: low green finance group (LGF) and high dustrial gas emission group (Lgas) and high industrial gas emission
green finance group (HGF) as well as low environmental regulation group (Hgas), and low industrial solid waste group (Lsolid) and high
group (LER) and high environmental regulation group (HER). The industrial solid waste group (Hsolid). The estimation results in Table 6
estimation results in Table 5 reveal that the effect of green finance is demonstrate that compared with low-pollution provinces, green finance
greater in provinces with lower levels of green finance and environ development in high-pollution provinces tends to have stronger ability
mental regulation than those of higher ones. When focusing our atten to improve green productivity. Therefore, to deal with environmental
tion on the influence of industrial pollutant emission, the total sample is inequality and injustice, policymakers in provinces with lower levels of
8
Table 4 has proclaimed a series of energy conservation, emission reduction

Heterogeneity analysis: social condition. green finance policies. From the observations of the spatiotemporal
Variable (1) (2) (3) (4) variation of GTFP, we find considerable evolution trends after the
implementation of the ‘Green Credit Guidelines’ in 2012 and the ‘Energy
LURB HURB LEDU HEDU
Efficiency Credit Guidelines’ in 2015. To gain better insight into the
GF 0.5888 1.4234*** 0.2549 2.1053*** effectiveness of the government's green finance policy, we further
(1.2595) (2.8710) (0.5746) (3.9226)
PGDP 1.0643*** 0.1144 0.4006 1.0341**
consider a policy dummy and the interaction term with green finance in
(4.2728) (0.3998) (1.5659) (2.2529) our baseline model to investigate how these policies affect the relation
ES − 0.0038** − 0.0140*** − 0.0035 − 0.0083*** between green finance and green productivity.
(− 2.2464) (− 4.1749) (− 1.6167) (− 2.8186) Table 7 summarizes the corresponding estimates when a policy
URB − 0.0243 0.0463*** 0.0319* 0.0033
dummy gp12 (or gp15) and the interaction term GF × gp12 (or GF ×
(− 1.4513) (2.9876) (1.6971) (0.1386)
IND − 0.0462*** − 0.0633*** − 0.0443*** − 0.0558*** gp15) are respectively added into the model. The estimation results
(− 5.4830) (− 8.8713) (− 6.2559) (− 5.3984) mostly support our prior findings. The development of green finance still
TRADE − 0.0018 − 0.0005 − 0.0011 0.0023 has significantly positive effects on green productivity. The positive sign
(− 0.3001) (− 0.2553) (− 0.3071) (0.7386) of the interaction term also notes that the implementation of a green
FDI 0.1331* 0.0833*** 0.1097*** 0.0037
(1.7456) (2.9340) (3.1492) (0.1203)
finance policy would further enhance the beneficial impact of green
FS − 0.0058*** − 0.0053*** − 0.0070*** − 0.0054** finance on green productivity. On the one hand, policies aimed at pro
(− 3.6803) (− 2.7944) (− 2.8817) (− 2.5757) moting green financial support and increasing energy efficiency through
EDU − 0.2234** − 0.0822 − 0.2261*** − 0.0882 credit resource allocation and environmental-friendly investment
(− 2.1093) (− 0.8169) (− 2.6201) (− 0.6538)
behavior guidance should accelerate industry structure adjustment and
R&D 0.7678*** 0.4796*** 0.5469*** 0.3501**
(4.3251) (3.9770) (3.6377) (2.2542) reduce environmental damage (Xu and Li, 2020). On the other hand,
cons − 4.2672*** 1.0461 − 0.2245 − 6.2763* these policies can further enhance the development of the green econ
(− 2.7923) (0.5376) (− 0.1472) (− 1.8719) omy to better absorb the beneficial effect of green finance (Liu et al.,
R2 0.6105 0.7996 0.6535 0.7134 2019)..
F-value 25.0814*** 65.4207*** 28.2852*** 41.0711***
Obs. 194 196 186 204
5.5. Robustness tests
In our previous empirical investigation, green finance development
is assessed by a composite index system using the levels of green-
Table 5
Heterogeneity analysis: environmental protection.
oriented credit, securities, insurance, and investment. To further
ensure whether the findings are sensitive to different measures of green
Variable (1) (2) (3) (4)
finance, we also replicate the analysis by using four subcomponents of a
L-GF H-GF L-ER H-ER composite green finance development index. Table 8 presents the
GF 1.8991*** 1.3487** 2.0185*** 1.8803*** robustness results for the corresponding estimation. The empirical re
(3.5106) (2.1937) (3.7431) (3.3867) sults are basically consistent with our previous conclusions, proving that
PGDP 0.3514* 0.5524 0.4568 0.3161 our results are robust from this angle, but different components have
(1.6933) (1.3463) (1.5465) (0.8888)
varying degrees of impact. When the influences of different categories of
ES − 0.0088*** − 0.0027 − 0.0103** − 0.0026
(− 4.6343) (− 1.0165) (− 2.4799) (− 1.1301) green finance are concerned, evidence further indicates that green credit
URB 0.0412*** 0.0004 0.0361* 0.0186 and green securities have a significantly positive impact on green pro
(3.0924) (0.0197) (1.9696) (0.8081) ductivity, while green insurance has a significantly negative effect on
IND − 0.0358*** − 0.0667*** − 0.0560*** − 0.0552***
green productivity.
(− 6.6758) (− 5.4604) (− 6.5129) (− 5.8539)
TRADE − 0.0048* 0.0006 − 0.0054* 0.0075*
Considering that the impact of green finance may vary with different
(− 1.8721) (0.2053) (− 1.9468) (1.7261) dimensions of green productivity measures, we also conduct an alter
FDI − 0.0069 0.0956** 0.0218 0.0562 native analysis by using several environmental indicators: environ
(− 0.3209) (2.4282) (0.7878) (1.0249) mental pollution, governance, and quality. The related indicators and
FS 0.0010 − 0.0085*** − 0.0071*** − 0.0055**
corresponding weights in the evaluation system are described in
(0.5253) (− 4.2529) (− 3.9097) (− 2.2551)
EDU − 0.1978** − 0.0780 − 0.2354** − 0.2038 Table B3 of Appendix. Table 9 reports the corresponding estimation
(− 2.1033) (− 0.5861) (− 2.0776) (− 1.4724) results. The evidence still suggests that green productivity is signifi
R&D 0.2189* 1.0504*** 0.4411*** 1.0646*** cantly affected by green finance development, particularly with regard
(1.8346) (5.9587) (3.1548) (5.8142) to environmental governance and environmental quality. Strengthening
cons − 0.9186 − 1.8455 − 0.4130 0.0218
(− 0.7169) (− 0.7033) (− 0.1981) (0.0096)
environmental governance or improving environmental quality can
R2 0.8306 0.6388 0.7544 0.6359 promote the development of a green economy, resulting in greater green
F-value 79.4404*** 28.4712*** 44.5427*** 29.6856*** productivity. Overall, the robustness testing results mostly support our
Obs. 195 195 180 210 main findings.
6. Conclusions
environmental protection should not only pay attention to strengthening
environmental protection monitoring and improving the standards of The concept of a green economy that coordinates multiple de
environmental regulations, but also accelerate the establishment of a velopments of the economy, resources, and ecological environment has
sophisticated green finance system. undoubtedly become one of the major focuses in the economic and
environmental literature over recent decades. While considerable efforts
have been made to explore ways to improve green productivity growth
5.4. Further analysis from the perspectives of economic condition, resource utilization, and
the quality of ecological environments, knowledge is rather limited
To cope with the problems of environmental degradation and at the regarding the role played by recently developed green finance in
same time achieve high-quality development in China, the government affecting green total factor productivity. This research therefore
9
Table 6
Heterogeneity analysis: industrial pollutant emission.
Variable (1) (2) (3) (4) (5) (6)
Lwater Hwater Lgas Hgas Lsolid Hsolid
GF 0.2486 4.0426*** 0.6828* 3.0383*** 0.0625 3.5022***

(0.6468) (6.8370) (1.8257) (4.9085) (0.2302) (6.7398)
PGDP 1.0050*** 1.0334** 0.7680*** 0.8988* 1.0742*** 0.8462**
(5.4968) (2.2315) (4.3057) (1.7489) (7.6952) (2.1104)
ES − 0.0087*** − 0.0063** − 0.0138*** 0.0002 − 0.0074*** − 0.0081***
(− 4.4284) (− 2.3726) (− 6.8077) (0.0671) (− 5.8711) (− 2.9921)
URB − 0.0345*** − 0.0185 0.0054 − 0.0146 − 0.0450*** 0.0066
(− 2.8230) (− 0.6610) (0.4417) (− 0.4672) (− 4.7284) (0.3026)
IND − 0.0373*** − 0.0490*** − 0.0384*** − 0.0398*** − 0.0305*** − 0.0456***
(− 6.1592) (− 4.4491) (− 5.8050) (− 3.7494) (− 6.3788) (− 5.0122)
TRADE − 0.0028 0.0043 − 0.0045*** 0.0110* − 0.0047*** 0.0089**
(− 1.4737) (0.9849) (− 2.6369) (1.8034) (− 3.6397) (2.3319)
FDI 0.0601** 0.0363 0.0175 − 0.0098 0.0078 0.0416
(2.3057) (1.0191) (0.6870) (− 0.2449) (0.4277) (1.3244)
FS − 0.0050*** 0.0029 − 0.0059*** − 0.0004 − 0.0042*** 0.0034
(− 4.2508) (0.9859) (− 4.8181) (− 0.1381) (− 5.1049) (1.2543)
EDU − 0.0574 − 0.1220 − 0.1534* − 0.2482* − 0.0613 − 0.1510
(− 0.6688) (− 0.8563) (− 1.8774) (− 1.6866) (− 1.1662) (− 1.0082)
R&D 0.4222*** 0.5056** 0.3560*** 1.1427*** 0.2584*** 0.6617***
(4.1689) (2.4546) (3.6340) (5.3851) (3.3777) (3.5251)
cons − 4.4286*** − 6.4545** − 2.6748** − 5.4506* − 4.5067*** − 5.9983**
(− 3.6402) (− 2.1684) (− 2.2228) (− 1.6952) (− 5.2787) (− 2.2998)
R2 0.7204 0.7995 0.7997 0.7340 0.7700 0.8285
F-value 41.7445*** 65.3982*** 65.4837*** 45.5400*** 55.2309*** 80.1673***
Obs. 195 195 195 195 195 195
Table 7
Further analysis: the role of green debt policy.
Variable (1) (2) (3) (4) (5) (6)
GTFP GTFP GTFP GTFP GTFP GTFP
GF 1.6104*** 0.7696 1.7308*** 1.9681*** 1.0077** 1.1842***

(4.2202) (1.3535) (4.3041) (5.3086) (2.1118) (3.2046)
PGDP 0.7339*** 0.8106*** 0.6698*** 0.5459*** 0.6857*** 0.6574***
(3.3448) (3.6534) (3.1120) (2.6751) (3.3225) (3.2797)
ES − 0.0053*** − 0.0051*** − 0.0054*** − 0.0035** − 0.0028* − 0.0028*
(− 3.1621) (− 3.0534) (− 3.2309) (− 2.0932) (− 1.6798) (− 1.6981)
URB 0.0099 0.0073 0.0111 − 0.0027 − 0.0084 − 0.0081
(0.7809) (0.5751) (0.8722) (− 0.2187) (− 0.6773) (− 0.6525)
IND − 0.0563*** − 0.0561*** − 0.0556*** − 0.0397*** − 0.0389*** − 0.0384***
(− 9.5911) (− 9.6004) (− 9.4475) (− 6.2972) (− 6.2417) (− 6.2324)
TRADE − 0.0009 − 0.0009 − 0.0010 0.0016 0.0019 0.0019
(− 0.4473) (− 0.4313) (− 0.4830) (0.7812) (0.9249) (0.9645)
FDI 0.0354 0.0321 0.0350 0.0260 0.0188 0.0196
(1.5548) (1.4114) (1.5318) (1.1821) (0.8587) (0.9003)
FS − 0.0067*** − 0.0075*** − 0.0067*** − 0.0077*** − 0.0090*** − 0.0089***
(− 5.1061) (− 5.4702) (− 5.0077) (− 5.9926) (− 6.7464) (− 6.7974)
EDU − 0.1500 − 0.1732* − 0.1740* − 0.1287 − 0.1535* − 0.1465*
(− 1.6453) (− 1.8923) (− 1.8888) (− 1.5282) (− 1.8372) (− 1.7732)
R&D 0.6992*** 0.7504*** 0.6825*** 0.7702*** 0.8311*** 0.8249***
(6.5966) (6.9068) (6.4684) (7.4864) (8.0347) (8.0246)
gp12 − 0.1020 − 0.4464**
(− 1.3968) (− 2.3753)
gfgp12 1.1612** − 0.1186
(1.9875) (− 0.5213)
gp15 0.3709*** − 0.0960
(5.3098) (− 0.5854)
gfgp15 1.6010*** 1.3300***
(3.1384) (6.2038)
cons − 3.1322** − 3.3291** − 2.4041 − 1.8663 − 2.4993* − 2.3837*
(− 2.0375) (− 2.1702) (− 1.6096) (− 1.4561) (− 1.9499) (− 1.8840)
R2 0.6946 0.6980 0.6931 0.7158 0.7236 0.7234
F-value 72.1465*** 67.0225*** 71.6495*** 79.9138*** 75.9325*** 82.9606***
Obs. 390 390 390 390 390 390
addresses this void in the literature by constricting a comprehensive constraints, and examining the influence of the development of green
index of green finance development, assessing green total factor pro finance on green productivity.
ductivity with the consideration of energy and environmental Using a balanced panel dataset of 30 provinces in China for the
10
Table 8 productive, and less pollutive. Second, the regression results indicate
Robust checks: alternative proxies for green finance. that the development of green finance significantly increases the level of
Variable (1) (2) (3) (4) green productivity. Third, green finance in provinces with higher levels
of economic and social conditions, less public participation in environ
GF GF GF GF
mental protection, and high pollution levels are more likely to have a
grecre 1.0564*** larger positive effect on green productivity. Finally, the implementation
(3.0600)
gresec 0.7471***
of a green finance policy can further enhance the beneficial impact of
(5.2169) green finance development on green productivity.
greins − 0.5482** Our study offers several implications to China's sustainable devel
(− 1.9691) opment planning and environmental policy. From the aspect of macro-
greinv − 0.1903
mechanisms, the government should continue to strengthen its green
(− 0.6937)
PGDP 0.7529*** 0.6919*** 0.7407*** 0.7583*** finance policy and accelerate the establishment of a sophisticated green
(3.5209) (3.3230) (3.4347) (3.3184) finance system. In addition, the policymakers should reinforce the
ES − 0.0046*** − 0.0060*** − 0.0056*** − 0.0053*** development of the country's economic, financial, and social conditions
(− 2.6420) (− 3.6154) (− 3.2479) (− 3.0355) in order to improve the absorptive capacity and innovation capacity via
URB 0.0070 0.0133 0.0104 0.0080
(0.5414) (1.0620) (0.8031) (0.6086)
green finance. As to the viewpoint of meso-mechanisms, the government
IND − 0.0581*** − 0.0523*** − 0.0590*** − 0.0583*** should substantially promote the development of green industry
(− 9.9462) (− 8.9598) (− 10.0177) (− 9.7940) through financial and environmental policies in order to achieve the
TRADE − 0.0005 − 0.0024 − 0.0017 − 0.0018 purpose of promoting productivity growth and improving environ
(− 0.2558) (− 1.2055) (− 0.8250) (− 0.8741)
mental performance simultaneously. From the perspective of micro-
FDI 0.0364 0.0370 0.0342 0.0393*
(1.5766) (1.6450) (1.4679) (1.6577) mechanisms, authorities should increase the intensity of environ
FS − 0.0069*** − 0.0060*** − 0.0061*** − 0.0063*** mental regulation, which would force enterprises to improve their
(− 5.1035) (− 4.6242) (− 4.5256) (− 4.6582) production process, engage in green technological innovation, and
EDU − 0.1933** − 0.2047** − 0.2040** − 0.1741* promote energy-saving and emission-reduction activities.
(− 2.1982) (− 2.3861) (− 2.2741) (− 1.9604)
R&D 0.6880*** 0.6505*** 0.6542*** 0.6450***
Finally, although this paper fills the gap in the extant research on
(6.4123) (6.2537) (6.0905) (5.9637) how green finance development affects green total factor productivity,
cons − 2.6793* − 2.2451* − 2.0915 − 2.4258 some questions remain unanswered and should be specified, offering
(− 1.9628) (− 1.7054) (− 1.5391) (− 1.6349) other directions for future research. First, as preliminary research we
R2 0.6845 0.6994 0.6796 0.6765
only quantitatively investigate the linkage between green finance
F-value 75.9355*** 81.4464*** 74.2418*** 73.1934***
Obs. 390 390 390 390 development and green productivity in a linear framework. However,
non-linearity in the relationship between economic and the environ
ment has attracted quite a bit of research more recently (Chang et al.,
2018; Chen et al., 2021). To get a more complete picture on green
Table 9
productively growth, it would also be interesting to explore how green
Robust checks: alternative proxies for green productivity. finance affects green total factor productivity under different market
conditions or states. Second, sustainable development is an all-
Variable (1) (2) (3)
encompassing term in a wide range of economic, environmental, and
envpol envgov envqua social circles. Owing to the difficulty of assessing social benefit, this
GF − 0.0019 0.0890* 0.0657** research mainly focuses on economic and environmental performances.
(− 0.0708) (1.9140) (2.3106) Future research could further consider this dimension to enrich the
PGDP 0.0906*** 0.3246*** 0.1175***
literature on green productivity growth.
(6.2011) (12.5878) (7.4453)
ES 0.0000 0.0014*** 0.0006***
(0.1533) (6.7322) (4.5665) Data availability statement
URB 0.0005 − 0.0007 0.0013
(0.6250) (− 0.4634) (1.3616) Data are available from the authors upon request.
IND 0.0017*** 0.0008 − 0.0001
(4.2829) (1.1649) (− 0.1759)
TRADE 0.0001 − 0.0007*** 0.0005*** Funding information
(1.0468) (− 2.8595) (3.3075)
FDI − 0.0011 0.0053* − 0.0018 Chien-Chiang Lee is grateful to the Social Science Foundation of
(− 0.7124) (1.8917) (− 1.0745)
Jiangxi Province of China for financial support through Grant No:
FS − 0.0000 0.0001 0.0001
(− 0.1160) (0.5318) (0.5688)
21JL02.
EDU − 0.0222*** − 0.0108 0.0161**
(− 3.7015) (− 1.0220) (2.4845) CRediT author statement
R&D − 0.0023 − 0.0237* − 0.0139*
(− 0.3189) (− 1.8384) (− 1.7589)
Chi-Chuan Lee,
cons 0.1213 − 2.5743*** − 1.1904***
(1.3167) (− 15.8289) (− 11.9627) Investigation, Software, Data curation, Writing- Original draft
R2 0.4707 0.8693 0.8221 preparation.
F-value 31.1217*** 232.8218*** 161.7843*** Chien-Chiang Lee.
Obs. 390 390 390
Corresponding author, Visualization, Supervision, Writing- Review
Notes: t statistics are in parentheses. * p < 0.1, ** p < 0.05, and *** p < 0.01. ing and Editing.
Both authors provided critical feedback and helped shape the re-
period 2006–2018 and the super-SBM model with undesirable output, search, analysis and manuscript.
our empirical results first reveal that both green finance development
and green total factor productivity exhibit an overall upward trend, Declaration of Competing Interest
suggesting that economic activities tend to be more efficient, highly
The authors declare that they have no conflict of interest. This article
11
does not contain any experiments with human participants or animals Acknowledgements
performed by any of the authors. Both authors contributed equally to
this study and share first authorship. We would like to thank the editor and the anonymous referees for
their highly constructive suggestions.
Appendix A
See Table A1 to Table A3
Table A1
List of input-output indicators and definitions.
Type Variable Definition
Input Labor force Number of employees

(10,000 people)
Capital stock Total fixed assets investment
(100 million yuan)
Energy input Total energy consumption
(104 tons)
Desirable output Economic output Real gross domestic product
(100 million yuan)
Undesirable output Environmental pollution Industrial wastewater emissions
(104 tons)
Industrial waste gas emissions
(100 million cu.m)
Industrial solid waste emissions
(104 tons)
Table A2
Year GTFP GTC GEC
2006 1.0310 0.9974 1.0402

2007 1.0358 1.0017 1.0352
2008 1.0344 1.0082 1.0553
2009 1.0600 0.9933 1.0723
2010 1.0457 0.9811 1.0688
2011 1.0000 1.0198 1.0118
2012 1.0712 0.9730 1.3372
2013 1.0595 1.0142 1.0642
2014 1.0784 0.9770 1.1075
2015 1.1127 1.0030 1.1112
2016 1.1303 1.0135 1.1164
2017 1.0848 1.1141 1.0079
2018 1.1574 1.0326 1.1281
Table A3
GTFP GTC GEC
Beijing 1.0842 1.0012 1.0856

Tianjin 1.0861 1.0284 1.0621
Hebei 1.1645 1.0928 1.1097
Shanxi 1.0359 0.9716 1.0686
Inner Mongolia 1.1077 1.0012 1.1067
Liaoning 1.0982 1.0292 1.0735
Jilin 1.0701 1.0197 1.0506
Heilongjiang 1.0769 0.9917 1.1329
Shanghai 1.0915 0.9972 1.0943
Jiangsu 1.1083 1.0036 1.1111
Zhejiang 1.0740 1.0021 1.0724
Anhui 1.0574 0.9982 1.0594
Fujian 1.0982 1.0387 1.0622
Jiangxi 1.0409 0.9620 1.1014
Shandong 1.1292 1.0216 1.1510
Henan 1.1398 1.0731 1.0912
Hubei 1.0854 1.0228 1.0615
Hunan 1.1237 1.0592 1.0610
(continued on next page)
12
Table A3 (continued )
GTFP GTC GEC
Guangdong 1.0833 0.9991 1.0840

Guangxi 1.0497 0.9968 1.0547
Hainan 1.0008 0.7852 2.3315
Chongqing 1.0704 1.0267 1.0441
Sichuan 1.0981 1.0329 1.0636
Guizhou 1.0337 1.0013 1.0328
Yunnan 1.0433 0.9940 1.0500
Shaanxi 1.0421 0.9955 1.0479
Gansu 1.0285 0.9960 1.0326
Qinghai 0.9520 0.9925 0.9589
Ningxia 0.9868 1.0867 0.9627
Xinjiang 1.0189 0.9772 1.0427
Eastern region 1.0926 1.0093 1.1540
Central region 1.0820 1.0111 1.0815
Western region 1.0323 1.0100 1.0290
Appendix B
See Table B1 to Table B3
Table B1
List of variables and definitions.
Variable Definition
Green TFP
Green productivity index calculated by the super-efficiency SBM model
(GTFP)
Green finance
Comprehensive index of green finance
(GF)
Economic development
Logarithmic value of GDP per capita
(PGDP)
Energy structure
Proportion of coal consumption in total energy consumption
(ES)
Urbanization
Proportion of urban population to total population
(URB)
Industrialization
Ratio of secondary industry to GDP
(IND)
Trade openness
Ratio of total trade amount to GDP
(TRADE)
Foreign capital
Ratio of actual use of FDI to GDP
(FDI)
Financial support
Ratio of financial institutions' loan balances to GDP
(FS)
Education
Average years of education
(EDU)
R&D expenditure
Ratio of R&D expenditure to GDP
(RD)
Table B2
Comprehensive evaluation system of green finance.
1st-class index 2nd-class index
Green credit Total amount of green credit

(45%) Ratio of interest expenditure in energy-intensive industries
Green securities Ratio of the market value of environmental protection enterprises
(25%) Market share of high-energy-consuming industries
Green insurance Ratio of agricultural insurance scale
(15%) Agricultural insurance loss ratio
Green investment Ratio of public expenditure on energy conservation and environmental protection
(15%) Ratio of investment in environmental pollution control
13
Table B3
Comprehensive evaluation index system of environmental assessment.
1st-class index 2nd-class index
Environmental pollution index Industrial wastewater discharge per unit of industrial output value
Industrial smoke emissions per unit of industrial output value
Industrial SO2 emissions per unit of industrial output value
Production of industrial solid waste per unit of industrial output value
Comprehensive utilization rate of industrial solid waste
Sulfur dioxide removal rate
Environmental governance index Industrial dust removal rate
Standard rate of industrial wastewater discharge
Investment in environmental pollution control
Forest coverage
Environmental quality index Proportion of nature reserve area to area under jurisdiction
Public green space per capita
References Guo, R., Yuan, Y., 2020. Different types of environmental regulations and heterogeneous
influence on energy efficiency in the industrial sector: evidence from Chinese
provincial data. Energy Policy 145, 111747. https://doi.org/10.1016/j.
Acheampong, A.O., 2019. Modelling for insight: does financial development improve
enpol.2020.111747.
environmental quality. Energy Econ. 83, 156–179. https://doi.org/10.1016/j.
Hafner, S., Jones, A., Anger-Kraavi, A., et al., 2020. Closing the green finance gap – a
eneco.2019.06.025.
systems perspective. Environ. Innov. Soc. Transit. 34, 26–60. https://doi.org/
Acheampong, A.O., Amponsah, M., Boateng, E., 2020. Does financial development
10.1016/j.eist.2019.11.007.
mitigate carbon emissions? Evidence from heterogeneous financial economies.
Hao, Y., Wang, L., Zhu, L., Ye, M., 2018. The dynamic relationship between energy
Energy Econ. 88, 104786. https://doi.org/10.1016/j.eneco.2020.104768.
consumption, investment and economic growth in China's rural area: new evidence
Aluko, O.A., Obalade, A.A., 2020. Financial development and environmental quality in
based on provincial panel data. Energy 154, 374–382. https://doi.org/10.1016/j.
sub-Saharan Africa: is there a technology effect? Sci. Total Environ. 747, 241515.
energy.2018.04.142.
https://doi.org/10.1016/j.scitotenv.2020.141515.
He, L., Liu, R., Zhong, Z., Wang, D., Xia, Y., 2019a. Can green financial development
Balaguer, J., Cantavella, M., 2018. The role of education in the environmental Kuznets
promote renewable energy investment efficiency? A consideration of bank credit.
curve. Evidence from Australian data. Energy Econ. 70, 289–296. https://doi.org/
Renew. Energy 143, 974–984. https://doi.org/10.1016/j.renene.2019.05.059.
10.1016/j.eneco.2018.01.021.
He, L., Zhang, L., Zhong, Z., Wang, D., Wang, F., 2019b. Green credit, renewable energy
Bodnar, P., Ott, C., Thwaites, J., De Marez, L., et al., 2017. Net Climate Finance:
investment and green economy development: empirical analysis based on 150 listed
Reconciling the Clean and Dirty Sides of the Finance Ledger. Available at:
companies of China. J. Clean. Prod. 208, 363–372. https://doi.org/10.1016/j.
https://rmi.org/wp-content/uploads/2017/07/RMI_Net_Climate_Finance_Discuss
jclepro.2018.10.119.
ion_Paper_2017-1.Pdf.
Hu, G., Wang, X., Wang, Y., 2021. Can the green credit policy stimulate green innovation
Borensztein, E., De Gregorio, J., Lee, J., 1998. How does foreign direct investment affect
in heavily polluting enterprises? Evidence from a quasi-natural experiment in China.
growth? J. Int. Econ. 45 (1), 115–135. https://doi.org/10.1016/S0022-1996(97)
Energy Econ. 98, 105134. https://doi.org/10.1016/j.eneco.2021.105134.
00033-0.
Huang, J., Luan, B., Cai, X., Zou, H., 2020. The role of domestic R&D activities played in
Chang, C.P., Wen, J., Dong, M., Hao, Y., 2018. Does government ideology affect
carbon intensity: evidence from China. Sci. Total Environ. 708, 135033. https://doi.
environmental pollutions? New evidence from instrumental variable quantile
org/10.1016/j.scitotenv.2019.135033.
regression estimations. Energy Policy 113, 386–400. https://doi.org/10.1016/j.
Labatt, S., White, R.R., 2002. Environmental Finance: A Guide to Environmental Risk
enpol.2017.11.021.
Assessment and Financial Products. John Wiley & Sons Ltd., Chichester, England.
Chen, S., Golley, J., 2014. ‘Green’ productivity growth in China’s industrial economy.
Lee, C.C., Lee, C.C., Chiou, Y.Y., 2017. Insurance activities, globalization, and economic
Energy Econ. 44, 89–98. https://doi.org/10.1016/j.eneco.2014.04.002.
growth: new methods, new evidence. J. Int. Financ. Mark. Inst. Money 51, 155–170.
Chen, L.L., Zhang, X.D., He, F., Yuan, R.S., 2019. Regional green development level and
https://doi.org/10.1016/j.intfin.2017.05.006.
its spatial relationship under the constraints of haze in China. J. Clean. Prod. 210,
Lee, C.C., Wang, C.W., Ho, S.J., 2021. The impact of natural disaster on energy
376–387. https://doi.org/10.1016/j.jclepro.2018.11.037.
consumption: international evidence. Energy Econ. 97, 105021. https://doi.org/
Chen, X., Fu, Q., Chang, C.P., 2021. What are the shocks of climate change on clean
10.1016/j.eneco.2020.105021.
energy investment: a diversified exploration. Energy Econ. 95, 105136. https://doi.
Lee, C.C., Lee, C.C., Li, Y.Y., 2021a. Oil price shocks, geopolitical risks, and green bond
org/10.1016/j.eneco.2021.105136.
market dynamics. North Am. J. Econ. Financ. 55, 101309. https://doi.org/10.1016/
Cheng, Z., Li, L., Liu, J., 2018. Industrial structure, technical progress and carbon
j.najef.2020.101309.
intensity in China’s provinces. Renew. Sust. Energy Rev. 81, 2935–2946. https://doi.
Lee, C.C., Lee, C.C., Xiao, S., 2021b. Policy-related risk and corporate financing behavior:
org/10.1016/j.rser.2017.06.103.
evidence from China’s listed companies. Econ. Model. 94, 539–547. https://doi.org/
Chiu, Y.B., Lee, C.C., 2020. Effects of financial development on energy consumption: the
10.1016/j.econmod.2020.01.022.
role of country risks. Energy Econ. 90, 104833. https://doi.org/10.1016/j.
Lee, C.C., Wang, C.W., Ho, S.J., 2022. Financial aid and financial inclusion: does risk
eneco.2020.104833.
uncertainty matter? Pac-Basin Financ. J. 71, 101700. https://doi.org/10.1016/j.
Copeland, B.R., Taylor, M.S., 2004. Trade, growth, and the environment. J. Econ. Lit. 42
pacfin.2021.101700.
(1), 7–71. https://doi.org/10.1257/002205104773558047.
Levine, R., 2005. Finance and growth: theory and evidence. In: Aghion, P., Durlauf, S.N.
Cowan, E., 1999. Topical issues in environmental finance. In: Research Paper was
(Eds.), Handbook of Economic Growth, vol. 1A. North Holland, Amsterdam,
Commissioned by the Asia Branch of the Canadian International Development
pp. 865–934.
Agency (CIDA), 1, pp. 1–20.
Li, K., Lin, B., 2016. Impact of energy conservation policies on the green productivity in
Dai, L., Mu, X., Lee, C.C., Liu, W., 2021. The impact of outward foreign direct investment
China’s manufacturing sector: evidence from a three-stage DEA model. Appl. Energy
on green innovation: the threshold effect of environmental regulation. Environ. Sci.
168, 351–363. https://doi.org/10.1016/j.apenergy.2016.01.104.
Pollut. Res. 28, 34868–34884. https://doi.org/10.1007/s11356-021-12930-w.
Li, K., Lin, B., 2017. Economic growth model, structural transformation, and green
Du, K., Li, P., Yan, Z., 2019. Do green technology innovations contribute to carbon
productivity in China. Appl. Energy 187, 489–500. https://doi.org/10.1016/j.
dioxide emission reduction? Empirical evidence from patent data. Technol. Forecast.
apenergy.2016.11.075.
Soc. Change 146, 297–303. https://doi.org/10.1016/j.techfore.2019.06.010.
Li, Z., Liao, G., Wang, Z., Huang, Z., 2018. Green loan and subsidy for promoting clean
Fan, H., Peng, Y., Wang, H., Xu, Z., 2021. Greening through finance? J. Dev. Econ. 152,
production innovation. J. Clean. Prod. 187, 421–431. https://doi.org/10.1016/j.
102683. https://doi.org/10.1016/j.jdeveco.2021.102683.
jclepro.2018.03.066.
Feng, C., Huang, J.B., Wang, M., 2018. Analysis of green total-factor productivity in
Liu, R., Wang, D., Zhang, L., Zhang, L., 2019. Can green financial development promote
China’s regional metal industry: a meta-frontier approach. Resour. Policy 58,
regional ecological efficiency? A case study of China. Nat. Hazards 95, 325–341.
219–229. https://doi.org/10.1016/j.resourpol.2018.05.008.
https://doi.org/10.1007/s11069-018-3502-x.
Gao, Y., Zhang, M., Zheng, J., 2021. Accounting and determinants analysis of China's
Liu, G., Liu, Y., Lee, C.C., 2020a. Growth sources of green economy and energy
provincial total factor productivity considering carbon emissions. China Econ. Rev.
consumption in China: new evidence accounting for heterogeneous regimes. Energy
65, 101576. https://doi.org/10.1016/j.chieco.2020.101576.
J. 41 (6), 33–63. https://doi.org/10.5547/01956574.41.6.gliu.
Gray, R., 2002. Of messiness, systems and sustainability: towards a more social and
Liu, G., Wang, B., Cheng, Z., Zhang, N., 2020b. The drivers of China’s regional green
environmental finance and accounting. Br. Account. Rev. 34 (4), 357–386. https://
productivity, 1999–2013. Resour. Conserv. Recycl. 153, 104561. https://doi.org/
doi.org/10.1006/bare.2002.0217.
10.1016/j.resconrec.2019.104561.
Grossman, G.M., Krueger, A.B., 1995. Economic growth and the environment. Q. J. Econ.
110 (2), 353–377. https://doi.org/10.2307/2118443.
14
Liu, H., Yang, R., Wu, D., Zhou, Z., 2021. Green productivity growth and competition Wen, H., Lee, C.C., 2020. Impact of environmental labeling certification on firm
analysis of road transportation at the provincial level employing Global Malmquist- performance: empirical evidence from China. J. Clean. Prod. 255, 120201. https://
Luenberger Index approach. J. Clean. Prod. 279 https://doi.org/10.1016/j. doi.org/10.1016/j.jclepro.2020.120201.
jclepro.2020.123677, 123677. Wen, H., Lee, C.C., Zhou, F., 2022. How does fiscal policy uncertainty affect corporate
Ramanathan, R., 2005. An analysis of energy consumption and carbon dioxide emissions innovation investment? Evidence from China’s new energy industry. Energy Econ.
in countries of the Middle East and North Africa. Energy 30, 2831–2842. https://doi. 105, 105767. https://doi.org/10.1016/j.eneco.2021.105767.
org/10.1016/j.energy.2005.01.010. Wu, L., Zhang, Z., 2020. Impact and threshold effect of internet technology upgrade on
Reinhard, S., Lovell, C.A.K., Thijssen, G.J., 2000. Environmental efficiency with multiple forestry green total factor productivity: evidence from China. J. Clean. Prod. 271,
environmentally detrimental variables; estimated with SFA and DEA. Eur. J. Oper. 122657. https://doi.org/10.1016/j.jclepro.2020.122657.
Res. 121 (2), 287–303. https://doi.org/10.1016/S0377-2217(99) 00218-0. Wu, H., Hao, Y., Ren, S., 2020a. How do environmental regulation and environmental
Salazar, J., 1998. Environmental finance: linking two world. In: A Workshop on Financial decentralization affect green total factor energy efficiency: evidence from China.
Innovations for Biodiversity, vol 1, pp. 2–18. Bratislava. Energy Econ. 91, 104880. https://doi.org/10.1016/j.eneco.2020.104880.
Sardosky, P., 2010. The impact of financial development on energy consumption in Wu, H., Ren, S., Yan, G., Hao, Y., 2020b. Does China’s outward direct investment
emerging economies. Energy Policy 38, 2528–2535. https://doi.org/10.1016/j. improve green total factor productivity in the “Belt and Road” countries? Evidence
enpol.2009.12.048. from dynamic threshold panel model analysis. J. Environ. Manag. 275, 111295.
Sardosky, P., 2011. Financial development and energy consumption in central and https://doi.org/10.1016/j.jenvman.2020.111295.
eastern European frontier economies. Energy Policy 39, 999–1006. https://doi.org/ Xia, F., Xu, J., 2020. Green total factor productivity: a re-examination of quality of
10.1016/j.enpol.2010.11.034. growth for provinces in China. China Econ. Rev. 62, 101454. https://doi.org/
Scheel, H., 2001. Undesirable outputs in efficiency valuations. Eur. J. Oper. Res. 132 (2), 10.1016/j.chieco.2020.101454.
400–410. https://doi.org/10.1016/S0377-2217(00)00160-0. Xie, H., Ouyang, Z., Choi, Y., 2020. Characteristics and influencing factors of green
Seiford, L.M., Zhu, J., 2002. Modeling undesirable factors in efficiency evaluation. Eur. J. finance development in the Yangtze River Delta of China: analysis based on the
Oper. Res. 142 (1), 16–20. https://doi.org/10.1016/S0377-2217(01)00293-4. spatial durbin model. Sustainability 12 (22), 9753. https://doi.org/10.3390/
Shen, C.H., Lee, C.C., Lee, C.C., 2010. What makes international capital flows promote su12229753.
economic growth? An international cross-country analysis. Scott. J. Polit. Econ. 57 Xu, X., Li, J., 2020. Asymmetric impacts of the policy and development of green credit on
(5), 515–546. https://doi.org/10.1111/j.1467-9485.2010.00529.x. the debt financing cost and maturity of different types of enterprises in China.
Shen, N., Liao, H., Deng, R., Wang, Q., 2019. Different types of environmental J. Clean. Prod. 264 (10), 1–13. https://doi.org/10.1016/j.jclepro.2020.121574.
regulations and the heterogeneous influence on the environmental total factor Yan, Z., Zou, B., Du, K., Li, K., 2020. Do renewable energy technology innovations
productivity: empirical analysis of China's industry. J. Clean. Prod. 211, 171–184. promote China’s green productivity growth? Fresh evidence from partially linear
https://doi.org/10.1016/j.jclepro.2018.11.170. functional-coefficient models. Energy Econ. 90, 104842. https://doi.org/10.1016/j.
Song, M., An, Q., Zhang, W., Wang, Z., Wu, J., 2012. Environmental efficiency evaluation eneco.2020.104842.
based on data envelopment analysis: a review. Renew. Sust. Energ. Rev. 16 (7), Yu, J., Zhou, K., Yang, S., 2019. Regional heterogeneity of China’s energy efficiency in
4465–4469. https://doi.org/10.1016/j.rser.2012.04.052. “new normal”: a meta-frontier super-SBM analysis. Energy Policy 134, 110941.
Tian, P., Lin, B., 2017. Promoting green productivity growth for China's industrial https://doi.org/10.1016/j.enpol.2019.110941.
exports: evidence from a hybrid input-output model. Energy Policy 111, 394–402. Yu, C.H., Wu, X., Zhang, D., Chen, S., Zhao, J., 2021. Demand for green finance:
https://doi.org/10.1016/j.enpol.2017.09.033. resolving financing constraints on green innovation in China. Energy Policy 153,
Tone, K.A., 2001. Slacks-based measure of efficiency in data envelopment analysis. Eur. 112255. https://doi.org/10.1016/j.enpol.2021.112255.
J. Oper. Res. 130 (3), 498–509. https://doi.org/10.1016/s0377-2217(99)00407-5. Yuan, H., Feng, Y., Lee, C.C., Cen, Y., 2020. How does manufacturing agglomeration
Tone, K.A., 2002. Slacks⁃based measure of super⁃efficiency in data envelopment analysis. affect green economic efficiency? Energy Econ. 92, 104944. https://doi.org/
Eur. J. Oper. Res. 143 (1), 32–41. https://doi.org/10.1016/S0377-2217(01)00324- 10.1016/j.eneco.2020.104944.
1. Zeng, X.-W., Liu, Y.-Q., Man, M.-J., et al., 2014. Measurement and analysis of China’s
Walker, N.L., Williams, A.P., Styles, D., 2020. Key performance indicators to explain green financial development degree. J. China Exec. Leader. Acad. Yan’an 7 (6),
energy & economic efficiency across water utilities, and identifying suitable proxies. 112–121.
J. Environ. Manag. 269, 110810. https://doi.org/10.1016/j.jenvman.2020.110810. Zhang, D., 2021. Green credit regulation, induced R&D and green productivity: revisiting
Wang, E.Z., Lee, C.C., 2022. The impact of clean energy consumption on economic the Porter Hypothesis. Int. Rev. Financ. Anal. 75, 101723. https://doi.org/10.1016/
growth in China: is environmental regulation a curse or a blessing? Int. Rev. Econ. j.irfa.2021.101723.
Financ. 77, 39–58. https://doi.org/10.1016/j.iref.2021.09.008. Zhao, P.J., Zheng, L.E., Lu, H.Y., Zhou, Y., Hu, H.Y., Wei, X.Y., 2020. Green economic
Wang, Y., Sun, X., Guo, X., 2019. Environmental regulation and green productivity efficiency and its influencing factors in China from 2008 to 2017: based on the super-
growth: empirical evidence on the Porter Hypothesis from OECD industrial sectors. SBM model with undesirable outputs and spatial Dubin model. Sci. Total Environ.
Energy Policy 132, 611–619. https://doi.org/10.1016/j.enpol.2019.06.016. 741, 140026. https://doi.org/10.1016/j.scitotenv.2020.140026.
Wang, R., Wang, Q., Yao, S., 2021. Evaluation and difference analysis of regional energy Zheng, M., Feng, G.F., Jang, C.L., Chang, C.P., 2021. Terrorism and green innovation in
efficiency in China under the carbon neutrality targets: insights from DEA and Theil renewable energy. Energy Econ. 104, 105695. https://doi.org/10.1016/j.
models. Aust. J. Environ. Manag. 293, 112958. https://doi.org/10.1016/j. eneco.2021.105695.
jenvman.2021.112958. Zhou, X., Tang, X., Zhang, R., 2020. Impact of green finance on economic development
Wang, E.Z., Lee, C.C., Li, Y., 2022. Assessing the impact of industrial robots on and environmental quality: a study based on provincial panel data from China.
manufacturing energy intensity in 38 countries. Energy Econ. 105, 105748. https:// Environ. Sci. Pollut. Res. 27, 19915–19932. https://doi.org/10.1007/s11356-020-
doi.org/10.1016/j.eneco.2021.105748. 08383-2.
15

1 s2.0 S0140988322000469 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S0140988322000469 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0140988322000469 Main

Uploaded by

Copyright:

Available Formats

Energy Economics 107 (2022) 105863

Contents lists available at ScienceDirect

1. Introduction models postulate that a sustainable economy only depends on technol­

constraints. consumer and investment behavior guidance, the implementation of

the financial constraints of environmentally friendly enterprises and

Fig. 4. Spatial distribution of China's GTFP over 2006–2018.

POLS RE FE Two-way GMM

GF 0.9200 1.3438*** 1.6632*** 1.1682** 1.8141***

LGDP HGDP LFDI HFDI LFS HFS

GF 0.7589 2.7525*** 0.1187 3.0207*** 0.9020** 2.0104***

Table 4 has proclaimed a series of energy conservation, emission reduction

Lwater Hwater Lgas Hgas Lsolid Hsolid

GF 0.2486 4.0426*** 0.6828* 3.0383*** 0.0625 3.5022***

GTFP GTFP GTFP GTFP GTFP GTFP

GF 1.6104*** 0.7696 1.7308*** 1.9681*** 1.0077** 1.1842***

See Table A1 to Table A3

Type Variable Definition

Input Labor force Number of employees

Year GTFP GTC GEC

2006 1.0310 0.9974 1.0402

GTFP GTC GEC

Beijing 1.0842 1.0012 1.0856

Guangdong 1.0833 0.9991 1.0840

See Table B1 to Table B3

1st-class index 2nd-class index

Green credit Total amount of green credit

1st-class index 2nd-class index

You might also like

1. Introduction models postulate that a sustainable economy only depends on technol

GF 0.9200 1.3438* 1.6632* 1.1682 1.8141*

GF 0.7589 2.7525* 0.1187 3.0207* 0.9020 2.0104*

GF 0.2486 4.0426*** 0.6828* 3.0383* 0.0625 3.5022*

GF 1.6104* 0.7696 1.7308* 1.9681* 1.0077 1.1842***