Study on the Impact of Green Finance on Regional Total Factor
Carbon Productivity: Analysis of Spatial and Temporal Heterogeneity
Based on Beijing-Tianjin-Hebei, Yangtze River Delta and Pearl River
Delta Regions
Xiangran Cheng
a
, Zhu Kai
b
, Nie Yan
c
, Yao Rui
d
, Yidong Liu
e
and Yanan Zheng
*
College of Economics, North China University of Technology, Tangshan, Hebei, China
*
1135233112@qq.com
Keywords: Green Finance, Green Total Factor, Carbon Productivity, Group Regression, SBM-DDL, GML Index.
Abstract: Looking at his Beijing-Tianjin-Hebei, Yangtze River Delta to Pearl River Delta type regarding teacup as every
prey. The total factor carbon productivity and the effect of green finance are measured using the GML index
model with SBM directional distance function from 2011 to 2020 is analyzed by constructing indicators
regarding the level of advancement in green financing and using group regression model. As a result of the
total regression results, the results are displayed, green securities have a strong contribution to TFCP, and
green securities and carbon finance have the strongest significant effect on TFCP in the region; green
insurance has a more significant effect on TFCP in the Pearl River Delta region; green insurance and carbon
finance have a stronger significant effect in the Yangtze River Delta region.
1 INTRODUCTION
The stage of China's economic development is
shifting from one of rapid expansion to one of high-
quality development. Promote China's economy to
green, low-carbon, environmentally friendly direction
to achieve benign development, to enhance the scale
of economic growth efficiency has been the general
trend. In the process of forming a strategic shift in the
economic development model, green finance can
build a bridge and link between economic
development and environmental regulation; green
finance promotes the allocation of financial resources
toward environmental protection and enhances the
ability of society to resist risks, while enhancing
economic vitality for green industries. The total factor
carbon productivity (TFCP) index includes carbon
emissions as an input variable, which can reveal the
impact of a country's (region's) resource endowment
on carbon productivity. Under the current national
goal of "carbon neutrality" and "carbon peaking", it is
a
https://orcid.org/0000-0002-9177-177X
b
https://orcid.org/0000-0003-3864-148X
c
https://orcid.org/0000-0002-8039-5019
significant to investigate the TFCP. As a result,
researching the interactions between green finance
and TFCP is critical to the high-quality development
of the green economy. As the three major urban
agglomerations in China, Beijing-Tianjin-Hebei
region (Region I), Yangtze River Delta (Region II),
and Pearl River Delta (Region III), it is of exemplary
significance to investigate the effect of green finance
on TFCP and the spatial variability over time.
Considering how green finance is developing at
the moment, five products are selected as explanatory
variables, and relevant data are collected for 10 years
from 2011 to 2020 for three regions with intensive
economic development for measurement and
evaluation; meanwhile, the non-radial and non-angle
SBM-DDL model is used to measure and analyze the
TFCP of the three regions. Finally, a group regression
model is constructed using the measured TFCP to
examine spatial and temporal heterogeneity of green
financial development on TFCP among the three
regions.
d
https://orcid.org/0000-0002-3582-7652
e
https://orcid.org/0000-0002-0524-5635
140
Cheng, X., Kai, Z., Yan, N., Rui, Y., Liu, Y. and Zheng, Y.
Study on the Impact of Green Finance on Regional Total Factor Carbon Productivity: Analysis of Spatial and Temporal Heterogeneity Based on Beijing-Tianjin-Hebei, Yangtze River Delta and
Pearl River Delta Regions.
DOI: 10.5220/0012027000003620
In Proceedings of the 4th International Conference on Economic Management and Model Engineering (ICEMME 2022), pages 140-148
ISBN: 978-989-758-636-1
Copyright
c
2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
2 LITERATURE REVIEW
2.1 Research on Green Finance on
Total Factor Productivity
At present, academics mainly study the drivers of
TFCP. Different scholars have derived the influence
of various factors on TFCP from environmental
regulation (Chen Dongjing, Liu Kun 2022; Guo
Weixiang, Sun Hui 2020), industrial structure (Hu
Biqing 2019), foreign direct investment (Jin
Shucheng 2022), and Internet development (Bai
Xuejie; Sun Xianzhen 2021).
In addition, some researchers also focus on the
impact of developing green finance on TFCP and
GTFP: Shu Taiyi (2022) analysis of the impact of
green funding on TFP from 2011 to 2019 using a
fixed-effects model, and green finance had a
significant contribution to GTFP under the fixed-
effects model; Zhang Yuan(2022) empirically
analyzed by building a panel model and using the
level of economic development to determine how
green finance affects GTFP for heterogeneity
analysis, the mechanism of green finance on green
total factor productivity is examined. The findings
reveal that there is a significant regional
heterogeneity between green finance and GTFP.
Pengfei Ge (2018) et al found that there are different
non-linear relationships between financial scale,
structure, efficiency and deepening and GTFP; Ziju
Yin (2021) et al discovered that the geographical
structure of green finance and GTFP development in
China was "high in the east," "flat in the center," and
"poor in the west." "Cheng Gui et al (2022) found that
there is a significant non-linear relationship between
green financial development and GTFP. There are
differences in curve shape, inflection point and the
moment when different provinces cross the inflection
point between groups, and green finance cannot
effectively encourage the development of GTFP
enhancement early on, and the relationship curve
shows a significant U-shaped characteristic.
2.2 On the Measurement of TFCP
TFCP covers both desired and undesired outputs, and
its comprehensive consideration of economic and
environmental benefits seeks the organic
coordination of economy and ecology. In the face of
the increasingly serious environmental problems and
urgent economic development transformation in
recent years, how to accelerate the green
development is a hot topic of discussion in academic
circles.
Regarding the measurement of TFCP, most of the
domestic and international studies focus on the
optimization of measurement indicators and
improvement of measurement methods. In terms of
the selection of measurement indicators, energy,
labor, and capital inputs are mostly used as input
variables, while outputs are mostly expressed in
terms of gross national product. For carbon dioxide
emissions, some scholars include them as input
variables (Zhang Lifeng 2013; Li Yinrong 2020).
Some scholars include it in non-desired outputs into
the measurement system to reduce the bias of
production efficiency caused by undesirable outputs
(Li Bo et al.2016 ; Chen Dongjing, Liu Kun 2022)
include CO
2
emissions in non-desired indicators for
TFCP measurement; there are differences between
the two in the perspective of TFCP: the former will
emit a certain amount of CO
2
to produce a certain
amount of output and benefits on The former
analyzes carbon dioxide emissions to produce output
and benefits; the latter analyzes input factors to
produce carbon dioxide emissions; although the
research perspectives are different, there are
achievements in theoretical research. In terms of
TFCP measurement method, Zhang Lifeng (2013) for
the first time used the DEA method to measure the
Malmquist index. Subsequent scholars have
continuously improved and refined this method:
Dongjing Chen and Kun Liu (2022); Yifei Liu and
Kai Wang (2022) used the Super-SBM model with
non-expected output for measurement; Xuejie Bai
and Xianzhen Sun (2021); Wenjing Gao et al (2018)
used the Global Malmquist-Luenberger index method
based on the SBM directional distance function for
measurement.
3 MATERIALS AND METHODS
3.1 TFCP Measurement Method
For the measurement of TFCP, academics usually use
the DEA method to measure productivity change
using Malmquist index (Zhang Lifeng 2013), but this
method cannot include non-desired output into the
model; CHUNG Y, FÄRE R (1997) further extended
this method to form the Directional distance function
(DDF), which can include both desired and non-
desired output in a certain period.
Directional distance Function (DDF), which can
include both desired and non-desired outputs and be
widely used to measure total factor productivity for a
certain period of time, but this method has certain
shortcomings and cannot effectively overcome the
Study on the Impact of Green Finance on Regional Total Factor Carbon Productivity: Analysis of Spatial and Temporal Heterogeneity Based
on Beijing-Tianjin-Hebei, Yangtze River Delta and Pearl River Delta Regions
141
problem of linear programming nonsolutions and
measurement bias caused by the different selection of
radial or angular directions. The GML index was
introduced by OH (2010) to address this issue;
subsequent scholars further optimized the non-radial,
non-angle function model by combining the SBM
directional distance function with the GML index
(Fukuyama et al. 1997); some scholars in the
measurement of TFCP and GTFP also Some scholars
also followed this method in measuring TFCP and
GTFP (Xiang Yang 2015, Teng Zewei 2017, Liu
Zhang-sheng 2017). Therefore, in this paper, the
GML index is chosen to measure TFCP.
Considering each province as a decision making
unit (DMU) separately, each province uses N inputs
x = (x
1
, ... , x
N
) ∈ R
N
to produce M desired outputs y
= (y
1
, ... , y
M
) ∈ R
+
M
and I non-desired outputs b =
(b
1
, ... , b
I
) ∈ R
M
at each period t = 1, ... , T the kth k
= 1, ... , T the kth k = 1, ... , the inputs and outputs of
the K provinces are (x
k, t
, y
k, t
, b
k, t
). Construct the set
of production possibilities containing desired and
undesired outputs.
() {( , ): , ;
1
,; , ; 1, 0, }
11 1
K
tttttt
Px yb zy y m
kkm km
k
KK K
ttt tttt t t
zbbi zxxxn z z k
k ki ki k kn kn kn k k
kk k
(1)
Where Zrepresents the weight of each cross-
sectional observation. In some cases,the production
of P(x) will regress. Based on this problem, Oh
(2010) created a list of potential worldwide
productions, which has improved the consistency and
comparability of the production frontier.
() {( , ): , ; ,
11 11
;,;1,0,}
11 11
TK TK
Gtt ttt ttt
Px yb zy y m zb b
k km km k ki ki
tk tk
TK TK
tt t t t
izxxnzzk
kkn kn k k
tk tk
(2)
Directional distance function for the SBM. based
on Fukuyama's et al (1997) construction of the SBM
directional distance function considering the resource
environment, the global SBM directional distance
function is obtained from the ideas of Yang Xiang
(2015) in carbon productivity measurement as
'''
'
''
11
,,,
,,
11 11
11
1
11
()
1
(, ,,,,)max
2
.. , ;
,; + ,;
0, ; 0, ; 0,
xyb
xb
NM
ni
x
y
nm
Gtk tk tk x y b
nm
V
sss
TK TK
tt x t t
kkn n k
k
tk tk
TK
tyt ttbt
km m k ki i
km ki
tk
K
txy
knm
k
ss
NgM g
Sx y b ggg
St z x s x n z
ysy m zbsbi
zksns
;0,
b
i
ms i
(3)
where (x
t, k′
, y
t, k
, b
t, k′
) denotes the input and output
vectors for province k′ and (s
x
, s
y
, s
b
) denotes the
vectors of input and output slack. (g
x
, g
y
, g
b
) denotes
the directional vectors of desired output expansion,
undesired output and input compression taking
positive values.
The GML index, since the Malmquist⁃Luenberger
index is often not cyclic and there is an unsolvable
linear programming nonsolution problem, constructs
the GML index built with SBM directional distance
function according to Oh (2010) and Yang Xiang
(2015)
with the following equation.
ICEMME 2022 - The International Conference on Economic Management and Model Engineering
142
1(,,;,,)
1
111
1(, ,;,,)
yGttt x b
Sxybggg
t
V
GML
t
yGt t t x b
Sx y b ggg
V
(4)
The change is represented by the GML index. in
period t + 1 relative to period t. Whenever the index
exceeds 1, it means that TFCP has an upward trend;
conversely, if it is less than 1, it means that TFCP has
a downward state; if it is equal to 1, it means that
TFCP is in a largely steady condition. The GML
index, however, is not similar because it measures
how time t + 1 changed from period t, therefore,
before the analysis, it is converted into a cumulative
index by drawing on the methods of Qiu Bin et al
(2008) and Li Bin (2013), if the TFCP in period 1 is
1, then the TFCP in period t + 1 is:
1
1
t
ttt
GTFP GML GTFP
(5)
3.2 Input and Output Indicators
Selection and Data Description
(1) Input indicators: labor input, fixed asset input and
energy input are selected as input indicators. For
labor input, most of the existing studies use the year-
end employment number of each province to
characterize, and this paper also continues this idea.
For fixed asset inputs, the permanent inventory
method (PIM) proposed by Goldsmith. The following
formula is used to calculate each province's
productive capital stock:
1
(1 )
tt t t
KK I
－
(6)
(2) Output indicators: for desired output, the
nominal GDP of each province is used to
characterize; for non-desired output, carbon dioxide
emissions are used to characterize, and carbon
emissions are measured using the IPCC method with
the following model equation.
=
ii
i
i
EC
CE
EE
(7)
3.3 Analysis of TFCP Measurement
Results
The results of TFCP measurement in three regions
from 2011 to 2020 are obtained by using MATLAB
software and expressed by GML cumulative index, as
shown in Figure 1. In the long term, the TFCP of the
three largest urban areas in China is steadily
increasing and the growth rate has been increasing,
which indicates that the green economic development
of the three regions in China has achieved significant
results. However, the GML index decreased in the
period of 2012-2015 compared with that of 2011 and
2012. The reason for this phenomenon may be that
the control on the use of fossil energy such as coal
and the supervision and management of heavy
industrial industries were effective at the early stage
of policy promulgation, but there were problems of
lax supervision and lack of systematic and effective
management in the subsequent development.
Therefore, in the subsequent development the
government and other business sectors should
increase the supervision of environmental protection
to achieve a balance between the improvement of
production and living standards and carbon
emissions.
From the comparison of the three regions, the
GML accumulation index of the three major urban
clusters shows a pattern of "Region III > Region II >
Region I". The GML accumulation index of Beijing-
Tianjin-Hebei region is fluctuating in the past ten
years and has been declining since 2017, while the
GML index of Tianjin is the smallest in most years,
contributing the least to the GML accumulation index
and lagging behind in relative terms; the GML
accumulation index of Yangtze River Delta and Pearl
River Delta region are both on a stable upward trend,
and the development speed of Pearl River Delta is
faster than that of Yangtze River Delta. It may be due
to the rapid development of high-tech in Guangdong
Province, especially in Shenzhen and Guangzhou,
which provides technical support and financial
guarantee for industrial green development, coupled
with the inclined support of national policies, so the
PRD region has been in the front line of green
economic development.
4 EMPIRICAL ANALYSIS OF
THE IMPACT OF GREEN
FINANCE ON REGIONAL
TOTAL FACTOR CARBON
PRODUCTIVITY
4.1 Model Setting and Variable
Selection
In order to investigate how regional TFCP is affected
by green funding, a mixed OLS model (8) is used as
the benchmark regression equation for establishing
panel data.
11
it it it
TFCP digitfin
(8)
Study on the Impact of Green Finance on Regional Total Factor Carbon Productivity: Analysis of Spatial and Temporal Heterogeneity Based
on Beijing-Tianjin-Hebei, Yangtze River Delta and Pearl River Delta Regions
143
In this model, the explanatory variable TFCP
it
is
the regional TFCP, the core explanatory variable
digitfin
it
is the degree of regional green finance
development, and εit is a random disturbance term.
To investigate the spatial differences of green
finance on TFCP, i.e., the effects of green finance in
three regions on TFCP among the three regions,
models (9), (10), and (11) are developed to conduct
heterogeneity analysis of the structure of the core
explanatory variables, respectively.
22
it it it
TFCP digitfin
(9)
33
it it it
TFCP digitfin
(10)
44
it it it
TFCP digitfin
(11)
The explanatory variables and control variables of
models (9)-(11) are the same as the model, and the
subscripts i and t of the variables in the above model
denote province and time
Table 1: Descriptive statistics for variables.
Var ia ble t yp e
Explained
variable
Explanatory variable
Variable name
Total Factor
Carbon
Productivity
Green
credit
Green
securities
Green
investment
Green
insurance
Carbon
finance
Sample size 70 70 70 70 70 70
Maximum value 1.448 0.615 0.3 0.003 9.171 0.003
Minimum value 0.758 0.192 0.012 0 0.067 0.002
Mean 1.03 0.392 0.088 0.001 1.841 0.009
Standard
deviation
0.153 0.111 0.057 0.001 2.111 0.006
Median 1.007 0.348 0.084 0.001 0.665 0.008
4.2 Data Sources
In this paper, the panel data of region I,region II,
region III from 2011 to 2020 are selected as the
research samples. The relevant data are mainly
obtained from the statistical yearbooks of each
province. Missing data are supplemented by
interpolation method for completeness. respectively,
and εit is a random disturbance term.
Baseline regression
The regression results of model (8) are shown in
Table 2
Table 2: Linear regression analysis results.
Standardized
coefficient
t P VIF R²
Adjustment
R²
F
Beta
Constants - 20.527 0.000*** -
0.488 0.448
F=12.186
P=0.000***
Green credit -0.111 -0.899 0.372 1.904
Green securities 0.384 3.042 0.003*** 1.996
Green investment -0.257 -2.415 0.019** 1.418
Green insurance -0.715 -6.736 0.000*** 1.408
Carbon finance -0.502 -2.769 0.007*** 4.106
The analysis of the results of the F-test can be
obtained that the significance P-value is 0.000***,
Therefore, the model basically meets the
requirements. The t-statistics of four indicators,
namely, green securities, green investment, green
insurance and carbon finance, are significant,
indicating that their effects on TFCP are more
significant.
The standardized coefficient of green securities is
0.384, which indicates that green securities have a
positive effect on TFCP; meanwhile, the standardized
coefficients of green investment, green insurance and
carbon finance are all negative, which indicates that
these three green financial products have a negative
effect on TFCP.
2. Heterogeneity analysis
The grouped regression results of models (9) to
(11) are shown in Table 3.
ICEMME 2022 - The International Conference on Economic Management and Model Engineering
144
Table 3: Regression results for subgroups.
Overall
Re
g
ion I Re
g
ion II Re
g
ion III
Constants
1.267***
(20.527)
1.013***
(15.998)
0.772
(1.474)
1.488***
(3.953)
Green
Credit
-0.153
(
-0.899
)
-0.079
(
-0.791
)
-0.744
(
-0.318
)
0.783
(
1.073
)
Green
Securit
y
1.035***
(
3.042
)
0.809***
(
5.003
)
3.568
(
0.672
)
-0.122
(
-0.062
)
Green
Investment
-67.556**
(-2.415)
-7.923
(-0.509)
-242.515
(-1.108)
-67.649
(-1.286)
Green
Insurance
-0.052***
(
-6.736
)
-0.007
(
-0.925
)
1.043*
(
2.652
)
-0.108***
(
-5.769
)
Carbon
Finance
-12.377***
(
-2.769
)
-4.695**
(
-2.089
)
33.121
(
0.538
)
-55.301***
(
-3.884
)
Sample size 70 30 10 30
R
2
0.488 0.548 0.940 0.687
Adjustment R
2
0.448 0.454 0.866 0.621
F
F
(
5,64
)
=12.186,
p
=0.000
F (5,24)=5.816,p=0.001 F (5,4)=12.644,p=0.015 F (5,24)=10.514,p=0.000
The p-values of the three regions were compared
by F-test and were significant at the degree of
confidence of 95%, indicating that green finance has
a significant contribution to TFCP in three regions.
Further analysis shows that green securities and
carbon finance have the strongest significant effect on
TFCP in Beijing-Tianjin-Hebei region; green
insurance has a more significant effect on TFCP in the
Pearl River Delta region; green insurance and carbon
finance have a stronger significant effect in the
Yangtze River Delta region. and the standardized
coefficients of each indicator in Beijing-Tianjin-
Hebei and Yangtze River Delta regions are consistent
with the overall model, while the standardized
coefficients in the PRD region differ slightly, which
may be due to too little data.
Table 4: Regression coefficient variance test.
Name Item 1 Item 2
b
1
b
2 Difference t
p
Green
Credit
Re
g
ion I
Region III
-0.079 -0.744 0.665 2.117 0.041**
Region I Region II
-0.079 0.783 -0.862 -2.289 0.028**
Region III Region II
-0.744 0.783 -1.527 1.857 0.072*
Green
Security
Region I Region III
0.809 3.568 -2.759 -11.588 0.000***
Region I Region II
0.809 -0.122 0.931 -3.896 0.000**
Region III Region II
3.568 -0.122 3.690 -1.953 0.056
Green
Investment
Region I Region III
-7.923 -242.515 234.592 1.089 0.281
Region I Region II
-7.923 -67.649 59.726 1.460 0.150
Region III Region II
-242.515 -67.649 -174.866 10.268 0.000***
Green
Insurance
Region I Region III
-0.007 1.043 -1.050 7.692 0.000***
Region I Region II
-0.007 -0.108 0.101 -1.742 0.090
Region III Region III
1.043 -0.108 1.151 1.241 0.222
Study on the Impact of Green Finance on Regional Total Factor Carbon Productivity: Analysis of Spatial and Temporal Heterogeneity Based
on Beijing-Tianjin-Hebei, Yangtze River Delta and Pearl River Delta Regions
145
Name Item 1 Item 2
b
1
b
2 Difference t
p
Carbon
Finance
Region I Region III
-4.695 33.121 -37.816 -0.589 0.560
Region I Region II
-4.695 -55.301 50.605 5.439 0.000***
Region III Region II
33.121 -55.301 88.421 3.721 0.001***
Table 5: Results of robustness test.
Name
Re
g
ion I Re
g
ion II Re
g
ion III
Constants 0.958**(12.563) 1.368**(3.217) 0.889(1.645)
Credit
-0.059
(-0.587)
1.055(1.396) -3.402(-1.646)
Security 0.651**(3.086) 0.030(0.014) 6.899(1.382)
Investment 0.803(1.082) -0.516(-0.094) -3.289(-0.844)
Insurance -0.004(-0.590) -0.111**(-4.553) 1.371*(2.850)
Carbon Finance -3.446(-1.349) -58.136**(-3.651) 115.943(1.515)
Sample size 30 30 10
R
2
0.564 0.665 0.934
Adjustment R
2
0.473 0.595 0.852
F
F (5,24)=6.215,
p=0.001
F (5,24)=9.532,
p=0.000
F (5,4)=11.322,
p=0.018
By comparing the differences in regression
coefficients among the three regions, the regions and
variables that are significant in each region are
bolded, as shown in Table 4. According to the chart,
It is possible to draw the following conclusions: the
effect of green credit and green investment on the
TFCP of the three regions is not significant; green
securities bring a greater impact effect in Region I,
and their p-values are more significant compared with
the Region II and Region III, with greater spatial
heterogeneity; green insurance has a strong
significance in the Region III and theRegion II; the p-
values of the Region III compared with the Region I
are heterogeneous at the p-value in the PRD region is
heterogeneous at 90% confidence level compared to
the Region I, while the heterogeneity is not
significant compared to the Region II, indicating that
the effect level of green insurance scale on TFCP is
not much different between the two regions. Carbon
finance has a strong heterogeneity in Region I,Region
II, Region III.
3.Robustness test
By referring to the test method of grouped
regression by Chang-Rong Wu et al (2022), this paper
performs robustness test by replacing the indicators
of explanatory variables.
The amount of money invested in reducing
environmental pollution and the amount of money
spent on energy efficiency and environmental
protection are some of the development indicators of
green investment; the development indicators of
green credit include the proportion of agricultural
insurance scale and agricultural insurance payout
rate
. And the indicators of agricultural insurance
payout rate were removed from the stepwise
regression after the strong covariance problem of the
variables. Therefore, the core explanatory variables
in the baseline regression models (9) to (11) were
replaced with the share of fiscal spending on
environmental preservation, energy efficiency, and a
new regression with three regional groupings was
conducted, and the obtained results are shown in
Table 5. The F-test shows that the overall regression
effect of each region is still significant, and a
comparison with Table 2 reveals that this is generally
agreeable to the outcomes of the benchmark
regression model (9) in terms of the core explanatory
variables; indicating that the econometric model and
regression results of this study have high reliability
and robustness.
ICEMME 2022 - The International Conference on Economic Management and Model Engineering
146
5 RESEARCH CONCLUSIONS
AND COUNTERMEASURE
SUGGESTIONS
5.1 Research Conclusions
First, at the overall level, the four indicators of green
securities, green investment, green insurance and
carbon finance have a significant effect on TFCP in
the three regions, and green credit has no significant
effect on its impact. Among them, green securities
have a positive promoting effect on TFCP, while the
other three indicators have a negative hindering effect
on TFCP. It may be because green finance, in
promoting the flow of capital in the market, both
increases carbon emissions in the production process
and generates economic benefits at the production
end. The economic benefits generated by green
securities are the strongest in the capital flow, and the
capital flows more to the high-end production end,
and the economic benefits generated are greater than
the impact brought by carbon emissions; while the
production end to which the capital flows such as
green insurance, green investment and carbon finance
are more to the primary production end, and the
reasons for the capital flows are mostly policy-driven
factors, and the economic benefits brought are less
than the impact of carbon emissions.
Second, at the regional level, the positive
contribution of green securities to TFCP in Beijing-
Tianjin-Hebei region is more obvious than other two
regions, probably because in recent years, Beijing-
Tianjin-Hebei region has been promoting green
transformation in various industries, increasing green
financial expenditures, and taking a number of
practical actions to implement the national energy
conservation and emission reduction policies, in
addition to the joint hosting of the 2022 Winter
Olympics. In addition, the joint hosting of the 2022
Winter Olympics by Beijing and Zhangjiakou has
also contributed to the green transformation and
upgrading of the region's industries. Green insurance
has a more obvious and significant effect on TFCP in
both the PRD and the Yangtze River Delta, while the
impact of carbon finance on TFCP varies greatly
among the three regions, and the possible reason for
the variability is the imbalance of economic strength
and technological innovation capacity.
5.2 Suggestions for Countermeasures
The following recommendations are provided for the
growth of green finance and the enhancement of
TFCP based on the research findings mentioned
above.
First, the three major city clusters, especially the
Yangtze River Delta, should take the initiative and
fully utilize the benefits of industrial agglomeration
to take the lead in achieving a sustainable and steady
growth of total factor carbon productivity, and at the
same time, they can also play a radiating effect to
drive other regions to make a rapid change to high-
quality economic development, while paying
attention to the balance between the speed of
economic development and the utilization of
environmental resources.
Secondly, strengthen top-level design, enhance
regional synergy and linkage across provinces and
cities, strengthen information sharing between
regions, promote the construction of green financial
standard system, increase the scale share of
agricultural insurance, improve the innovation
capacity of financial industry, and promote TFCP.
Finally, improve and perfect the relevant rules and
regulations, strengthen the supervision and
management of high energy consumption, high
pollution, low value-added industries, and promote
the green transformation and modernization of
industry; can further encourage collaboration
between educational institutions and businesses.,
increase the financial and policy support for research
institutes and universities as well as the
environmental research enterprise sector, train
innovative and complex talents, and promote
technological innovation to help the development of
green economy.
REFERENCES
Bai Xuejie, Sun Xianzhen.(2021).Internet development
affects TFCP: cost, innovation or demand induced[J].
China Population-Resources and Environment, 31(10):
105-117.
Chen, Dongjing, Liu, Kun. (2022).A study on the impact of
heterogeneous environmental regulations on total
factor carbon productivity[J]. Forestry Economics,
44(06): 35-49.DOI:10.13843/j.cnki.lyjj.20220725.001.
Chen Jingquan, Lian Xinyan, Ma Xiaojun, Mi Jun. (2022).
Total factor energy efficiency measurement and its
driving factors in China [J]. China Environmental
Science, 42(05):2453-2463.
DOI:10.19674/j.cnki.issn1000-6923.0120.
Cheng Gui,Zeng Lesong.(2022).Heterogeneous impact of
regional green financial development on green total
factor productivity[J]. Gansu Finance, (03):4-14.
CHUNG Y, FÄRE R.(1997).Productivity and undesirable
outputs: a directional distance function approach [J].
Journal of environmental management, 51(3):229-240.
Study on the Impact of Green Finance on Regional Total Factor Carbon Productivity: Analysis of Spatial and Temporal Heterogeneity Based
on Beijing-Tianjin-Hebei, Yangtze River Delta and Pearl River Delta Regions
147
FUKUYAMA H, WEBER W L.(2009).A directional
slacks-based measure of technical inefficiency [J].
Socioeconomic planning sciences, 43(4) : 274 - 287.
Gao WJ, Ren Xuedi, Kang XH, Zhao GH.(2018).Analysis
of resource optimization allocation path for industrial
carbon productivity improvement[J]. Macroeconomic
Research,2018(05):166-175.DOI:10.16304/j.cnki.11-
3952/f.2018.05.016.
Hu Biqing. (2019).Research on the impact of industrial
structure on carbon productivity in China's service
industry [D]. Zhongnan University of Economics and
Law.
Jin Shucheng.(2022).Research on the impact of foreign
direct investment on carbon productivity in China[D].
Liaoning University.
DOI:10.27209/d.cnki.glniu.2022.000467.
Li B, Zhang WZ, Yu JH.(2016).Service industry
development, informationization level and TFCP
growth-an empirical study based on threshold effect[J].
Geography Research,35(05):953-965.
Liu Yifei, Wang Kai. (2022).TFCP of tourism in Yangtze
River Economic Zone and its interactive response with
industrial agglomeration[J]. Journal of Hubei Academy
of Arts and Sciences,43(08):48-56.
Liu Huiqun, Peng Chuanli. (2022)OFDI, reverse
technological spillover and total factor energy
efficiency--analysis based on PVAR model[J].
Ecological Economics,38(04):68-76.
Liu Zhangsheng, Song Deyong, Gong Yuanyuan.(2017).A
study on the spatio-temporal divergence and
convergence of green innovation capability in China
[J]. Journal of Management, 14(10) : 1475 - 1483.
Qiu B, Yang SH, Xin Pei J.(2008).A study on FDI
technology spillover channel and manufacturing
productivity growth in China: an analysis based on
panel data [J]. World Economy, 31 (08) :20 - 31.
Shu, Taiyi, Zhao, Tiantian, Zhang, Xiaolei. (2022).
Research on the impact of green finance on GTFP--an
empirical test based on fixed effects model [J]. Green
Technology,24(15):255-258.
DOI:10.16663/j.cnki.lskj.2022.15.035.
Wu C. R., Chen C.Meng X. W., Dong J. Yuan. (2022).
Differential analysis of the impact of digital inclusive
finance on residents' consumption in the South and
North--empirical evidence based on provincial panel
data [J]. Times Business and Economics, 19(05):22-30.
DOI:10.19463/j.cnki.sdjm.2022.05.036.
Zhang Yuan. (2022).Research on the impact of green
finance on GTFP[D]. Shanxi University of Finance and
Economics, DOI:10.27283/d.cnki.gsxcc.2022.000044.
ICEMME 2022 - The International Conference on Economic Management and Model Engineering
148
