How does green finance affect the low-carbon economy? Capital allocation, green technology innovation and industry structure perspectives

Figures & data

Figure 1. China’s low-carbon economy level in 2005 and 2019. Source: Authors.

Figure 2. China’s green finance level in 2005 and 2019. Source: Authors.

Figure 3. The average value of green finance and low-carbon economy during 2005–2019. Source: Authors.

Table 1. Comprehensive evaluation system of green finance.

Index	Index description	Index attribute
Green credit	Interest expense of six high energy consuming industries/total industrial interest expense	–
Green insurance	Agricultural insurance income/gross agricultural output value	+
Green investment	Investment in environmental pollution control/GDP	+
Government support	Financial environmental protection expenditure/ financial general budget expenditure	–

Source: Authors.

Table 2. The definitions of variables.

Variables	Acronym	Definition	Unit	Obs	Mean	Std. dev.	Min	Max
Low-carbon economy	cp	GDP/carbon emissions	10^8yuan/ ton	450	1.977	1.058	0.360	7.067
Green finance	gf	Entropy Model	–	450	0.155	0.098	0.044	0.793
Capital misallocation	cmi	STATA was used for calculation	–	450	0.262	0.294	0.000	2.461
Green technology innovation	gti	Number of green invention patent applications	piece	450	6.683	1.676	1.946	10.354
Industrial structure change	chg	Output value of service industry/industry	%	450	1.174	0.650	0.527	5.234
Human capital	hc	EXCEL was used for calculation	year/person	450	8.928	0.976	6.558	12.701
Openness	open	Foreign direct investment/GDP	%	450	0.023	0.017	0.000	0.082
Infrastructure level	fund	Proportion of investment in fixed assets to GDP	%	450	0.702	0.248	0.208	1.516
Economic development	eco	Per capita GDP	10^4 yuan	450	10.411	0.647	8.560	12.011
Environmental regulation	er	Entropy Model	–	450	0.534	0.529	0.000	2.585

Source: Authors.

Table 3. Correlation coefficient of each variable and multicollinearity test.

variables	cp	gf	cmi	chg	gti	hc	open	fund	eco	er	VIF
cp	1
gf	0.830**	1									3.29	5.98	3.66
cmi	0.402**	0.497**	1								1.45
chg	0.530**	0.753**	0.475**	1								3.82
gti	0.793**	0.714**	0.280**	0.292**	1								3.93
hc	0.624**	0.758**	0.452**	0.624**	0.628**	1					3.80	4.35	3.85
open	0.256**	0.236**	0.112*	0.102*	0.156**	0.323**	1				1.25	1.30	1.24
fund	−0.103*	−0.266**	−0.185**	−0.177**	−0.024	−0.195**	−0.332**	1			1.68	1.65	1.43
eco	0.763**	0.748**	0.427**	0.444**	0.814**	0.796**	0.196**	0.058	1		4.51	5.22	5.95
er	−0.029	−0.02	−0.181**	−0.338**	0.260**	−0.072	−0.067	−0.133**	0.061	1	1.21	1.41	1.34
Mean VIF											2.46	3.39	3.09

Note: *p < 0.1, **p < 0.05, ***p < 0.01.

Source: Authors.

Table 4. Regression results of spatial Durbin model for the whole sample.

Variables	(5) Basic model		(7) Scale effect		(8) Technique effect		(9) Structural effect
	Direct	Indirect	Direct	Indirect	Direct	Indirect	Direct	Indirect
gf	6.874***	−0.997	5.550***	3.177	−0.047	5.458	5.483***	2.143
	(18.132)	(−0.421)	(10.212)	(0.988)	(−0.04)	(0.868)	(7.751)	(0.486)
cmi/gti/chg			−0.310**	−0.211	0.139***	0.113	0.073	−0.643*
			(−2.402)	(−0.331)	(3.882)	(0.692)	(1.04)	(−1.794)
gf*cmi			2.090***	−3.933
			(3.583)	(−1.122)
gf*gti					1.501***	−1.458
					(6.124)	(−1.076)
gf*chg							0.300*	0.049
							(1.742)	(0.046)
hc	−0.001	0.376	0.003	0.478	0.112*	0.203	−0.033	0.560*
	(−0.021)	(1.284)	(0.056)	(1.588)	(1.845)	(0.694)	(−0.537)	(1.82)
open	−0.343	16.264**	0.292	18.000**	−0.655	17.054**	−0.352	14.284*
	(−0.213)	(2.161)	(0.179)	(2.3)	(−0.411)	(2.366)	(−0.215)	(1.812)
fund	0.114	0.137	0.195*	−0.173	0.143	0.162	0.086	0.164
	(1.007)	(0.205)	(1.72)	(−0.259)	(1.339)	(0.265)	(0.767)	(0.242)
eco	0.454***	−2.217***	0.619***	−2.844***	0.614***	−2.456***	0.650***	−2.811***
	(4.502)	(−4.121)	(5.776)	(−4.485)	(5.615)	(−4.173)	(5.404)	(−4.456)
er	−0.125**	0.608*	−0.120*	0.351	−0.131**	0.330	−0.117*	0.561*
	(−2.06)	(1.842)	(−1.757)	(0.991)	(−2.224)	(1.118)	(−1.858)	(1.692)
$\lambda$	0.681***		0.691***		0.674***		0.682***
	(15.731)		(16.42)		(15.465)		(15.851)
R^2	0.943		0.946		0.949		0.945
Log L	−39.000		−30.775		−16.352		−34.171
Moran’s I	0.617***		0.618***		0.560***		0.582***

Note: *p < 0.1, **p < 0.05, ***p < 0.01; the t value is in parentheses.

Source: Authors.

Table 5. Regional heterogeneity results of spatial Durbin model.

Variables	(5) Eastern basic model		(7) Scale effect		(8) Technique effect		(9) Structural effect
	Direct	Indirect	Direct	Indirect	Direct	Indirect	Direct	Indirect
gf	8.712***	0.730	7.914***	1.568*	8.370***	2.662	8.487***	1.369
	(36.744)	(0.978)	(25.052)	(1.715)	(9.623)	(1.162)	(22.658)	(1.302)
cmi/gti/chg			0.131	0.203	−0.014	−0.05	0.149	−0.217
			(1.37)	(0.791)	(−0.395)	(−0.541)	(2.708)	(−1.508)
gf*cmi			0.879**	−1.229
			(2.301)	(−1.024)
gf*gti					0.112	−0.479
					(0.574)	(−0.892)
gf*chg							−0.118	0.033
							(−1.22)	(0.106)
	(5) Central basic model		(7) Scale effect		(8) Technique effect		(9) Structural effect
Gf	15.519***	−26.498***	18.681***	−40.722***	10.765***	7.139	15.764***	−24.384***
	(18.729)	(−4.514)	(19.824)	(−5.182)	(7.915)	(0.766)	(18.141)	(−3.637)
cmi/gti/chg			1.844***	−5.557*	0.250***	−1.842***	0.051	−0.138
			(4.525)	(−2.016)	(5.09)	(−5.703)	(0.411)	(−0.143)
gf*cmi			36.624***	−105.13**
			(5.599)	(−2.183)
gf*gti					1.158**	−11.476***
					(2.685)	(−3.745)
gf*chg							2.242	9.499
							(0.985)	(0.527)
	(5) Western basic model		(7) Scale effect		(8) Technique effect		(9) Structural effect
Gf	16.888***	−0.108	14.695***	−0.931	9.068***	0.914	16.462***	2.335
	(26.435)	(−0.03)	(19.282)	(−0.228)	(9.403)	(0.223)	(23.375)	(0.66)
cmi/gti/chg			−1.393***	−0.428	0.371***	−0.201	0.559***	−0.659
			(−5.662)	(−0.359)	(11.293)	(−1.433)	(4.953)	(−0.987)
gf*cmi			−13.296***	2.204
			(−4.377)	(0.142)
gf*gti					2.931***	−5.039***
					(11.754)	(−4.241)
gf*chg							8.300***	−0.847
							(5.064)	(−0.077)
control	YES		YES		YES		YES

Note: *p < 0.1, **p < 0.05, ***p < 0.01; the t value is in parentheses.

Source: Authors.

Table 6. Dynamic threshold regression results.

Variables	Model 10	Model 11	Model 12
L.cp	0.751***	0.815***	0.843***
	(10.76)	(17.57)	(17.94)
$\widehat{\gamma }$	0.148	8.296	1.322
95% confidence interval	[ 0.114–0.459]	[ 4.771–8.878]	[0.744–1.575]
Effect of gf*cmi/gti/chg
Estimated threshold value	cmi	gti	chg
${\widehat{\beta }}_1$	−20.550***	1.658***	4.309***
	(−3.06)	(4.62)	(2.87)
${\widehat{\beta }}_2$	3.814***	1.196***	0.145
	(2.88)	(3.51)	(0.95)
_cons	−3.628***	−6.516***	−3.683***
	(−5.47)	(−9.22)	(−8.86)
control	YES
AR(1)	−2.952	−3.482	−3.164
	[0.003]	[0.001]	[0.002]
AR(2)	1.162	0.111	0.279
	[0.245]	[0.911]	[0.780]
Sargan test	22.889	24.951	23.457
	[1.000]	[1.000]	[1.000]
Wald test	21,721.22	203,994.93	89,115.07
	[0.000]	[0.000]	[0.000]
Obs.	420	420	420

Note: *p < 0.1, **p < 0.05, ***p < 0.01; [ ] means P value, ( ) means Z value.

Source: Authors.