# RECORD 6c554bbc30f053c6546ae8df71b70b126d3f4178
# TITLE: Evaluating the Impact of Urban Blue Space Accessibility on Housing Price: A Spatial Quantile Regression Approach Applied in Changsha, China
# YEAR: 2021.0  VENUE: Frontiers in Environmental Science  DOI: 10.3389/fenvs.2021.696626
# AUTHORS: Tuo-Fu Huang; Qing-Yun He; Dongxiao Yang; Xiao Ouyang
# PAGES: 15  CANDIDATE TABLE PAGES (1-based): [8, 10]
# ABSTRACT: Urban nature spaces are increasingly recognized as essential urban features providing crucial amenities to the residents’ health and well-being. While many studies have been conducted focusing on the influence of green spaces on house prices, very few have explored the impact of urban blue spaces. In this study, we analyzed the proximity effects of different types and sizes of urban blue spaces on property value in Changsha metropolis, China, and examined the spatial quantile effect across different housing prices. A two-stage instrumental method (2SLS) hedonic model was employed to evaluate the impact of different types of urban blue space: river (mainstream and tributary), wetland park, and lakes (large, medium, and small). Spatial quantile regression (SQR) was then used to measure the spatial effect of accessibility on various house price ranges. The 2SLS results show that, except for small-sized lakes, proximity to blue spaces significantly increases property value. Analysis of the SQR model reveals that proximity to major blue spaces increases the marginal willingness-to-pay among homebuyers of high-priced properties, while ordinary blue spaces are more attractive to buyers of low- and medium-priced houses. This may be broadly related to the level of education, utility, and sensitivity to ecosystem services across income groups. Based on these findings, we recommend that urban planners adopt different strategies to promote space utilization efficiency. This is one of the few studies that analyze the capitalization of blue space accessibility in house prices across different quantiles. By categorizing blue spaces and employing the SQR model, this study found the effect of blue spaces on housing prices to be heterogeneous, providing new perspectives to the existing literature.


===== PAGE 1 (score 14, numeric-paren 0, table-words 0, var-words 14) =====

Evaluating the Impact of Urban Blue
Space Accessibility on Housing Price:
A Spatial Quantile Regression
Approach Applied in Changsha, China
Huang Tuofu 1, He Qingyun 1*, Yang Dongxiao 2 and Ouyang Xiao 3
1College of Geographic Science, Hunan Normal University, Changsha, China, 2School of Economics and Trade, Hunan
University, Changsha, China, 3Hunan Institute of Economic Geography, Hunan University of Finance and Economics, Changsha,
China
Urban nature spaces are increasingly recognized as essential urban features providing
crucial amenities to the residents’ health and well-being. While many studies have been
conducted focusing on the inﬂuence of green spaces on house prices, very few have
explored the impact of urban blue spaces. In this study, we analyzed the proximity effects
of different types and sizes of urban blue spaces on property value in Changsha
metropolis, China, and examined the spatial quantile effect across different housing
prices. A two-stage instrumental method (2SLS) hedonic model was employed to
evaluate the impact of different types of urban blue space: river (mainstream and
tributary), wetland park, and lakes (large, medium, and small). Spatial quantile
regression (SQR) was then used to measure the spatial effect of accessibility on
various house price ranges. The 2SLS results show that, except for small-sized lakes,
proximity to blue spaces signiﬁcantly increases property value. Analysis of the SQR model
reveals that proximity to major blue spaces increases the marginal willingness-to-pay
among homebuyers of high-priced properties, while ordinary blue spaces are more
attractive to buyers of low- and medium-priced houses. This may be broadly related to
the level of education, utility, and sensitivity to ecosystem services across income groups.
Based on these ﬁndings, we recommend that urban planners adopt different strategies to
promote space utilization efﬁciency. This is one of the few studies that analyze the
capitalization of blue space accessibility in house prices across different quantiles. By
categorizing blue spaces and employing the SQR model, this study found the effect of blue
spaces on housing prices to be heterogeneous, providing new perspectives to the existing
literature.
Keywords: blue space, housing price, spatial quantile regression, accessibility, urban China
INTRODUCTION
In the process of intensiﬁed urbanization across modern China, the well-being of residents faces
numerous challenges. Deteriorating ecological environment, unhealthy lifestyle, and highly
competitive employment market (Wu Q. et al., 2020, Wu et al., 2021), have put signiﬁcant
pressure and risks on the urban population’s health and well-being. In response to these social
Edited by:
Yanfei Li,
Hunan University of Technology and
Business, China
Reviewed by:
Han Li,
University of Miami, United States
He Nie,
National University of Singapore,
Singapore
Chen Ziyue,
Norwegian University of Science and
Technology, Norway
*Correspondence:
He Qingyun
heqingyun1848@gmail.com
Specialty section:
This article was submitted to
Environmental Economics and
Management,
a section of the journal
Frontiers in Environmental Science
Received: 17 April 2021
Accepted: 03 May 2021
Published: 20 May 2021
Citation:
Tuofu H, Qingyun H, Dongxiao Y and
Xiao O (2021) Evaluating the Impact of
Urban Blue Space Accessibility on
Housing Price: A Spatial Quantile
Regression Approach Applied in
Changsha, China.
Front. Environ. Sci. 9:696626.
doi: 10.3389/fenvs.2021.696626
Frontiers in Environmental Science | www.frontiersin.org
May 2021 | Volume 9 | Article 696626
1
ORIGINAL RESEARCH
published: 20 May 2021
doi: 10.3389/fenvs.2021.696626


===== PAGE 8 (score 186, numeric-paren 132, table-words 5, var-words 7) =====

been corrected to a certain extent in the 2sls model and that the
use of instrumental variables in the spatial 2sls model is needed to
circumvent endogeneity.
Based on the regression coefﬁcients of the spatial lag term, we
found that the 30–90 quantile is signiﬁcant at the 10% level, and
the spatial dependence coefﬁcient increases signiﬁcantly with the
increase of the quantile (from 0.013 to 0.087). This indicates a
more pronounced spatial dependence in high-priced housing.
The conﬁdence interval above the 50th quantile exceeds the 2SLS
conditional mean of 0.022, indicating that the 2SLS results may
overestimate the spatial dependence of low-priced houses and
underestimate the high-priced houses.
Blue Space of Changsha
The results in Table 4 show that the Xiangjiang River has a
signiﬁcant impact on property value, with distance negatively
correlated with price. Each kilometer increase in the distance to
Xiangjiang River correlated with a decline of about 1.72% in
property value (i.e., 210 RMB per m2). The capitalization effect
of the Xiangjiang River (less than Yanghu wetland) was not
consistent with its status as the center of the nature landscape in
Changsha, which may due to increased ﬂood risk. In 2017, the
Xiangjiang River experienced ﬂooding, resulting in economic
losses for residents. Recent ﬂood events can signiﬁcantly alter
people’s perception and behavior towards ﬂood risks (Bin and
Polasky, 2004). With respect to the different housing prices,
proximity to the Xiangjiang only has a signiﬁcant impact above
the 70th quartile, peaking at 90th (from 2.36% to 3.14%, each
km). This suggests that nearness to the Xiangjiang river affects
mainly
house
prices
for
high-value
properties.
This
heterogeneity in proximity effects may be caused by better
river scenarios and increased safety found mainly in high-
value properties, similar to the conclusions of Wu Q. et al.
(2020).
The Liuyang River is a tributary of the Xiangjiang River,
ﬂowing through the entire eastern part of the city, and is an
TABLE 3 | OLS and QR model results.
Variable
OLS (1)
0.1 (2)
0.2 (3)
0.3 (4)
0.4 (5)
0.5 (6)
0.6 (7)
0.7 (8)
0.8 (9)
0.9 (10)
Floor area
–0.015
0.0126
0.00578
–0.0059
–0.0127
–0.0228*
–0.0292**
–0.0239**
–0.0390***
–0.0469***
[0.0107]
[0.0170]
[0.0122]
[0.0128]
[0.0124]
[0.0122]
[0.0123]
[0.0117]
[0.0133]
[0.0160]
Plot ratio
–0.0056**
–0.0042*
–0.0051***
–0.0066***
–0.0072**
–0.0078**
–0.0072*
–0.0013
–0.0035
–0.0056***
[0.0023]
[0.0019]
[0.0019]
[0.0019]
[0.0026]
[0.0028]
[0.0043]
[0.0045]
[0.0042]
[0.0033]
PM
0.0129***
0.0207***
0.0205***
0.0198***
0.0158***
0.0138***
0.0148***
0.00982*
0.00575
0.00663
[0.0036]
[0.0069]
[0.0060]
[0.0036]
[0.0027]
[0.0038]
[0.0040]
[0.0050]
[0.0068]
[0.0060]
Building age
–0.0016***
–0.0027***
–0.0026***
–0.0028***
–0.0025***
–0.0022***
–0.0018***
–0.0016***
–0.00080***
–0.00057***
[0.0001]
[0.0002]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
QMS
0.0834***
0.0495***
0.0510***
0.0644***
0.0750***
0.0739***
0.0781***
0.0827***
0.0845***
0.0979***
[0.0109]
[0.0189]
[0.0167]
[0.0150]
[0.0171]
[0.0171]
[0.0131]
[0.0188]
[0.0175]
[0.0167]
QPS
0.0294**
0.0427**
0.0222
0.0235*
0.0177
0.0157
0.0274***
0.0389***
0.0468***
0.0299*
[0.0118]
[0.0192]
[0.0232]
[0.0164]
[0.0187]
[0.0170]
[0.0167]
[0.0158]
[0.0152]
[0.0174]
Subway
0.0481***
0.0606***
0.0417***
0.0451***
0.0500**
0.0612***
0.0553***
0.0690***
0.0480**
0.00169
[0.0129]
[0.0264]
[0.0211]
[0.0184]
[0.0197]
[0.0226]
[0.0180]
[0.0178]
[0.0216]
[0.0256]
LF
0.0081***
0.0097***
0.0031***
0.0035***
0.0041***
0.0045***
0.0034***
0.0039***
0.0083***
0.0159***
[0.0003]
[0.0005]
[0.0006]
[0.0005]
[0.0004]
[0.0004]
[0.0003]
[0.0003]
[0.0004]
[0.0005]
Dcbd
–0.0625**
–0.103***
–0.0675***
–0.060**
–0.069*
–0.0705**
–0.0615***
–0.0489**
–0.03295
–0.054
[0.0290]
[0.0331]
[0.0395]
[0.0302]
[0.0297]
[0.0333]
[0.0323]
[0.0195]
[0.0505]
[0.0813]
D.Xriver
–0.0168***
–0.0125
–0.0085
0.00265
0.00066
–0.0064
–0.0087
–0.013*
–0.0252**
–0.0280**
[0.0056]
[0.0111]
[0.0083]
[0.0072]
[0.0061]
[0.0078]
[0.0072]
[0.0071]
[0.0097]
[0.0083]
D.Lriver
–0.0245**
–0.0357
–0.0318**
–0.0315***
–0.0263***
–0.0223***
–0.01385
–0.0075
–0.0025
–0.0171
[0.0120]
[0.0224]
[0.0158]
[0.0119]
[0.0083]
[0.0061]
[0.0371]
[0.0435]
[0.0446]
[0.0537]
D.YHwetland
–0.0251***
–0.00795
–0.0183**
–0.0166*
–0.0235**
–0.0273***
–0.0299***
–0.0312***
–0.0405***
–0.0420***
[0.0049]
[0.0097]
[0.0101]
[0.0092]
[0.0095]
[0.0087]
[0.0086]
[0.0069]
[0.0067]
[0.0084]
D.LL
–0.0173***
0.00134
–0.0156
–0.0164
–0.0205*
–0.0207**
–0.0223**
–0.0243***
–0.0209**
–0.0319***
[0.0066]
[0.0181]
[0.0131]
[0.0121]
[0.0089]
[0.0069]
[0.0052]
[0.0068]
[0.0077]
[0.0105]
D.ML
–0.0143***
–0.0344***
–0.0156
–0.0039
–0.0014
–0.004
–0.0082
–0.0066
–0.0053
–0.0019
[0.0045]
[0.0108]
[0.0098]
[0.0048]
[0.0043]
[0.0045]
[0.0055]
[0.0052]
[0.0048]
[0.0058]
D.SL
–0.0124
–0.0214
0.00524
–0.0001
0.0044
0.00533
–0.0113
–0.0118
–0.0098
–0.0421
[0.0167]
[0.0260]
[0.0288]
[0.0212]
[0.0241]
[0.0216]
[0.0139]
[0.0183]
[0.0293]
[0.0378]
D.YL
–0.0234***
–0.0277***
–0.0323***
–0.0349***
–0.0311***
–0.0281***
–0.0251***
–0.0190**
–0.0182**
–0.0212**
[0.0045]
[0.0068]
[0.0068]
[0.0056]
[0.0037]
[0.0052]
[0.0044]
[0.0078]
[0.0061]
[0.0073]
D.NWG
–0.0124***
–0.0082***
–0.0108***
–0.0861***
0.0089***
–0.0138***
–0.0135***
–0.0131***
–0.0095***
–0.0054
[0.0047]
[0.0021]
[0.0022]
[0.0015]
[0.0022]
[0.0024]
[0.0022]
[0.0026]
[0.0026]
[0.0038]
Cons
9.603***
9.136***
9.503***
9.502***
9.596***
9.734***
9.738***
9.701***
9.836***
10.14***
[0.1021]
[0.2057]
[0.2139]
[0.1399]
[0.1168]
[0.1379]
[0.1226]
[0.1027]
[0.1305]
[0.1737]
Adj.R2
0.519
Pseudo-R2
0.321
0.307
0.313
0.346
0.331
0.358
0.329
0.343
0.341
***p < 0.01; **p < 0.05; *p < 0.10
Frontiers in Environmental Science | www.frontiersin.org
May 2021 | Volume 9 | Article 696626
8
Tuofu et al.
Evaluate the Impact of Urban Blue


===== PAGE 9 (score 14, numeric-paren 0, table-words 2, var-words 9) =====

important blue space in Changsha. As shown in Table 4,
proximity to the Liuyang River has a signiﬁcant impact on
housing prices (at the 10% level), with a 2SLS conditional
mean regression coefﬁcient of 2.36% value decline per km.
The SQR results show that only low and medium-priced
houses (i.e., 30th–50th quantile) are affected by proximity to
this river.
Proximity to the Yanghu National Wetland Park was also
found to signiﬁcantly affect housing prices, similar to the ﬁndings
of Mei et al. (2018), with every kilometer distance correlating with
a 2.27% decline in housing prices (equivalent to 277 RMB per
m2). Similar to the effects of the Xiangjiang River, the SQR shows
that the price effect caused by proximity to the Yanghu wetland
increases as the quantile rises and is most pronounced for high-
value properties (From 1.53% to 2.85%). Yanghu National
Wetland Park was found to have a strong effect on housing
prices among the different landscape factors in Changsha and has
an estimated coefﬁcient almost higher than any blue and green
space. This may suggest wetlands are increasingly becoming more
valued in cities.
While urban lake is the most common blue space in the city,
our results found noticeable differences between varying sizes.
The conditional mean coefﬁcient of large lakes (LL) using 2sls was
1.66% value loss per km (i.e., 202 RMB per m2), signiﬁcant at the
1% level, and is higher than most landscape coefﬁcients. LL has a
signiﬁcant impact on house prices above the 40th quantile. The
estimated LL coefﬁcients increased signiﬁcantly, especially for
mid-to-high-priced housing above the 60–70 range (2.31%,
2.35%, respectively), similar to the impact of wetland parks
and the Xiangjiang River landscape. The impact of medium
lakes (ML) was signiﬁcant at the 10% level, resulting in a
1.45% value decline per km. The ML coefﬁcient value was
smaller than for large lakes and had a nonlinear trend
across quantiles. According to the SQR, ML affected house
prices below the 10th quantile but had no signiﬁcant impact
on those above the 10th quantile. Small lake (SL) is the
commonest blue space in Changsha city. Although SL has a
negative coefﬁcient, the variable was not statistically signiﬁcant.
Housing prices seem to be unaffected by small water bodies,
which is in sharp contrast to the impact on price of other kinds of
blue space.
Green Spaces in Changsha
For comparison, we used the non-water body green spaces as the
control variables. As shown in Table 4, the conditional mean loss
of property value for the Yuelu Forest Park was 2.03% per km,
relatively similar to some previous studies of forest parks
(Czembrowski and Kronenberg, 2016). The SQR results show
that proximity to the Yuelu forest does not signiﬁcantly inﬂuence
house prices above the 80th quantile. For the other common
green spaces, their impact was found to have a noticeable steady
trend. The 2SLS results were signiﬁcant at the 1% level, with a
value loss of 1.21%. Most of the quantile coefﬁcients for non-
water green spaces were higher. These ﬁndings are consistent
with existing studies (Liao and Wang, 2012), and the estimated
accessibility coefﬁcient is also comparatively similar.
FIGURE 2 | Coefﬁcients of blue spaces estimated by QR.
Frontiers in Environmental Science | www.frontiersin.org
May 2021 | Volume 9 | Article 696626
9
Tuofu et al.
Evaluate the Impact of Urban Blue


===== PAGE 10 (score 190, numeric-paren 129, table-words 6, var-words 10) =====

Other Key Variables
The study also found that the housing price gradient can diversify
with the distance from the CBD (Wen et al., 2014). By 2SLS, each
kilometer away from the CBD leads to a 5.75% decay of house
prices. The SQR results showed that the price for accessibility to
the CBD decreases for higher quantiles (from 13.65% to 3.81%,
each km). Low- and medium-priced properties were inﬂuenced
more strongly by CBD than high-priced houses. Neighborhood
variables, such as proximity to quality primary and middle
education facilities (2.80% and 8.17%, respectively) and nearness
to a subway station (4.72%), were also found to have signiﬁcant
impact on housing prices. These results suggest education-related
factors are becoming important factors affecting property house
prices in China.
Density Test of Urban Blue Accessibility
As suggested by Coulson and McMillen (2007), the distribution
of house prices varying to the explanatory variables could exam
the heterogeneity across dependent variable quantiles. If a change
in the value of the independent variable (distance) causes a
change in the dependent variable’s distribution, this would
mean that different quantiles would have varying responses to
the independent variable and that heteroscedasticity exists across
different quantiles. To investigate this effect, we drew the kernel
density curve to compare the distribution of prices among
varying accessibility to various urban blue parameters. The
results are presented in Figure 4. For the Xiangjiang River, the
kernel curve changes its kurtosis as the distance increases.
Similarly, the parameters Liuyang river, Yanghu wetland, large
lakes, and middle lakes have a signiﬁcant change in the
distribution of housing prices when the value of accessibility
independent variable changing. Most have changes in the kurtosis
(right tail of the distribution), with some increase in variance
distribution, coincident with the SQR results. In contrast, the
kernel density curve for small lakes has little distributional
change, which means no heteroscedasticity in this variable.
TABLE 4 | 2Stage Least Squares and SQR model results.
Variable
2SLS (1)
0.1 (2)
0.2 (3)
0.3 (4)
0.4 (5)
0.5 (6)
0.6 (7)
0.7 (8)
0.8 (9)
0.9 (10)
Floor area
–0.0153
0.0099
0.00319
–0.00587
–0.0131
–0.0224*
–0.0273**
–0.0278**
–0.0436**
–0.0566**
[0.0147]
[0.0138]
[0.0147]
[0.0120]
[0.0094]
[0.0098]
[0.0080]
[0.0125]
[0.0160]
[0.0219]
Plot ratio
–0.00549**
–0.00525**
–0.005**
–0.00677***
–0.00704*
–0.00759
–0.00684
–0.00135
–0.00263
–0.0065
[0.0023]
[0.0021]
[0.0024]
[0.0027]
[0.0023]
[0.0041]
[0.0045]
[0.0048]
[0.0037]
[0.0040]
PM
0.0129***
0.0212***
0.0211***
0.0199***
0.0159***
0.0143***
0.0143***
0.0103**
0.00461
–0.000734
[0.0036]
[0.0050]
[0.0045]
[0.0037]
[0.0044]
[0.0056]
[0.0053]
[0.0045]
[0.0072]
[0.0046]
Building age
–0.0015***
–0.0026***
–0.0024***
–0.0026***
–0.0024***
–0.0021***
–0.0020***
–0.0017***
–0.0009***
–0.0006***
[0.0001]
[0.0002]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
[0.0001]
QMS
0.0817***
0.0623***
0.0478**
0.0653***
0.0705***
0.0669***
0.0743***
0.0797***
0.0773***
0.0981***
[0.0111]
[0.0210]
[0.0145]
[0.0115]
[0.0136]
[0.0099]
[0.0127]
[0.0134]
[0.0124]
[0.0164]
QPS
0.0280**
0.0503***
0.0208
0.0244
0.0182
0.0104
0.0236*
0.0348**
0.0462**
0.0342
[0.0119]
[0.0136]
[0.0140]
[0.0149]
[0.0134]
[0.0178]
[0.0138]
[0.0161]
[0.0156]
[0.0306]
Subway
0.0472***
0.0534***
0.0470***
0.0417**
0.0475***
0.0573***
0.0553***
0.0707***
0.0444*
0.0165
[0.0129]
[0.0220]
[0.0178]
[0.0142]
[0.0196]
[0.0216]
[0.0183]
[0.0200]
[0.0205]
[0.0230]
LF
0.0079***
0.0113***
0.0033***
0.0031***
0.0045***
0.0056***
0.0023***
0.0045***
0.0073***
0.0125***
[0.0003]
[0.0006]
[0.0007]
[0.0006]
[0.0005]
[0.0004]
[0.0003]
[0.0003]
[0.0004]
[0.0007]
D.CBD
–0.0575**
–0.1365***
–0.079***
–0.0595***
–0.0672***
–0.0557*
–0.0555**
–0.0381**
–0.00635
–0.0417
[0.0267]
[0.0450]
[0.0271]
[0.0202]
[0.0266]
[0.0291]
[0.0274]
[0.0224]
[0.0507]
[0.0984]
D.Xriver
–0.0172***
–0.0128
–0.00706
0.0004
–0.0007
–0.0073
–0.0066
–0.0115
–0.0236**
–0.0314***
[0.0057]
[0.0106]
[0.0084]
[0.0071]
[0.0086]
[0.0091]
[0.0097]
[0.0135]
[0.0132]
[0.0097]
D.Lriver
–0.0236**
–0.0312
–0.0325
–0.0221**
–0.0252***
–0.0214***
–0.0112
–0.0062
–0.0015
–0.0083
[0.0118]
[0.0512]
[0.0529]
[0.0111]
[0.0104]
[0.0081]
[0.0261]
[0.0263]
[0.0363]
[0.0543]
D.YHwetland
–0.0227***
–0.0125
–0.0153*
–0.0171**
–0.0222***
–0.0237***
–0.0235***
–0.0215*
–0.0260**
–0.0285***
[0.0056]
[0.0099]
[0.0081]
[0.0069]
[0.0067]
[0.0071]
[0.0095]
[0.0086]
[0.0071]
[0.0073]
D.LL
–0.0166**
–0.0074
–0.0189
–0.0152
–0.0210**
–0.0179**
–0.0231***
–0.0235***
–0.0195***
–0.0143*
[0.0008]
[0.0011]
[0.0184]
[0.0181]
[0.0103]
[0.0087]
[0.0095]
[0.0093]
[0.0077]
[0.0081]
D.ML
–0.0145*
–0.0332**
–0.0159
–0.0051
–0.0013
–0.0038
–0.0071
–0.0051
–0.0049
–0.0050
[0.0085]
[0.0137]
[0.0122]
[0.0066]
[0.0066]
[0.0047]
[0.0050]
[0.0048]
[0.0055]
[0.0041]
D.SL
–0.0128
–0.0207
0.00136
–0.0065
0.00368
–0.0008
–0.0107
–0.0137
–0.0218
–0.029
[0.0167]
[0.0267]
[0.0216]
[0.0231]
[0.0220]
[0.0228]
[0.0227]
[0.0284]
[0.0284]
[0.0345]
D.YL
–0.0203***
–0.0424***
–0.0373***
–0.0320***
–0.0285***
–0.0211***
–0.0214***
–0.0176*
–0.0066
–0.00744
[0.0058]
[0.0084]
[0.0085]
[0.0073]
[0.0077]
[0.0072]
[0.0066]
[0.0103]
[0.0118]
[0.0111]
D.NWG
–0.0121***
0.0043
–0.005
–0.0082**
-0.0091**
–0.0155***
–0.0164***
–0.0183***
–0.0114***
–0.00907
[0.0017]
[0.0032]
[0.0034]
[0.0025]
[0.0014]
[0.0018]
[0.0021]
[0.0022]
[0.0030]
[0.0301]
WlnP
0.0221**
–0.0028
–0.0072
0.0136*
0.0152*
0.0443**
0.0468**
0.0447***
0.0809***
0.0874***
[0.0121]
[0.0076]
[0.0084]
[0.0081]
[0.0091]
[0.0223]
[0.0232]
[0.0192]
[0.0231]
[0.0157]
Cons
9.364***
10.18***
9.799***
9.377***
9.444***
9.236***
9.273***
9.214***
8.895***
9.047***
[0.2961]
[0.3726]
[0.4683]
[0.3900]
[0.3734]
[0.3357]
[0.3232]
[0.3933]
[0.6152]
[0.6469]
Adj.R2
0.531
Pseudo-R2
0.335
0.311
0.329
0.352
0.341
0.367
0.339
0.361
0.353
***p < 0.01; **p < 0.05; *p < 0.10.
Frontiers in Environmental Science | www.frontiersin.org
May 2021 | Volume 9 | Article 696626
10
Tuofu et al.
Evaluate the Impact of Urban Blue


===== PAGE 11 (score 0, numeric-paren 0, table-words 0, var-words 0) =====

FIGURE 3 | Coefﬁcients of blue spaces estimated by 2SQR.
FIGURE 4 | Shifts in the distribution of house prices.
Frontiers in Environmental Science | www.frontiersin.org
May 2021 | Volume 9 | Article 696626
11
Tuofu et al.
Evaluate the Impact of Urban Blue