# RECORD d956659c01d343fed17ef54a0b1e248a34a32995
# TITLE: THE EFFECT OF URBAN GREEN SPACES ON HOUSE PRICES IN WARSAW
# YEAR: 2018.0  VENUE: International Journal of Strategic Property Management  DOI: 10.3846/ijspm.2018.5220
# AUTHORS: R. Trojanek; Michał Głuszak; Justyna Tanas
# PAGES: 14  CANDIDATE TABLE PAGES (1-based): [7, 9, 10, 11]
# ABSTRACT: In the paper, we analysed the impact of proximity to urban green areas on apartment prices in Warsaw. The data-set contained in 43 075 geo-coded apartment transactions for the years 2010 to 2015. In this research, the hedonic method was used in Ordinary Least Squares (OLS), Weighted Least Squares (WLS) and Median Quantile Regression (Median QR) models. We found substantial evidence that proximity to an urban green area is positively linked with apartment prices. On an average presence of a green area within 100 meters from an apartment increases the price of a dwelling by 2,8% to 3,1%. The effect of park/forest proximity on house prices is more significant for newer apartments than those built before 1989. We found that proximity to a park or a forest is particularly important (and has a higher implicit price as a result) in the case of buildings constructed after 1989. The impact of an urban green was particularly high in the case of a post-transformation housing estate. Close vicinity (less than 100 m distance) to an urban green increased the sales prices of apartments in new residential buildings by 8,0–8,6%, depending on a model.


===== PAGE 1 (score 16, numeric-paren 0, table-words 0, var-words 16) =====

Copyright © 2018 The Author(s). Published by VGTU Press
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unre-
stricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
*Corresponding author. E-mail: r.trojanek@ue.poznan.pl
International Journal of Strategic Property Management
ISSN: 1648-715X / eISSN: 1648-9179
2018 Volume 22 Issue 5: 358–371
https://doi.org/10.3846/ijspm.2018.5220 
THE EFFECT OF URBAN GREEN SPACES ON HOUSE PRICES IN WARSAW
Radoslaw TROJANEK1,*, Michal GLUSZAK2, Justyna TANAS3
1 Department of Microeconomics, Poznan University of Economics and Business, Al. Niepodleglosci 10,  
61-875 Poznan, Poland
2 Department of Real Estate and Investment Economics, Cracow University of Economics,  
ul. Rakowicka 27 bld. B, 31-510 Krakow, Poland
3 Department of Investment and Real Estate, University of Gdansk, ul. Bazynskiego 8, 80-309 Gdansk, Poland
Received 17 July 2017; accepted 18 November 2017
Abstract. In the paper, we analysed the impact of proximity to urban green areas on apartment prices in Warsaw. The data-
set contained in 43 075 geo-coded apartment transactions for the years 2010 to 2015. In this research, the hedonic method 
was used in Ordinary Least Squares (OLS), Weighted Least Squares (WLS) and Median Quantile Regression (Median QR) 
models. We found substantial evidence that proximity to an urban green area is positively linked with apartment prices. 
On an average presence of a green area within 100 meters from an apartment increases the price of a dwelling by 2,8% to 
3,1%. The effect of park/forest proximity on house prices is more significant for newer apartments than those built before 
1989. We found that proximity to a park or a forest is particularly important (and has a higher implicit price as a result) 
in the case of buildings constructed after 1989. The impact of an urban green was particularly high in the case of a post-
transformation housing estate. Close vicinity (less than 100 m distance) to an urban green increased the sales prices of 
apartments in new residential buildings by 8,0–8,6%, depending on a model.
Keywords: hedonic methods, parks, urban green, property prices.
Introduction
The market value of a property depends mainly on its 
physical characteristics, out of which the most important 
is location. The location choices depend mostly on indi-
vidual preferences but generally can be related to the dis-
tance from work, schools, and hospitals, an accessibility 
of public transport, positive or negative neighbourhood 
effects both about built and natural environment. Among, 
various factors affecting housing choices, and thus house 
prices, environmental attributes rank very high in the 
hierarchy of importance (Źróbek, M. Trojanek, Źróbek-
Sokolnik, & R. Trojanek, 2015; Gluszak & Marona, 2017).
Green spaces not only provide a pleasant and natural 
environment but also improve the quality of life in urban 
areas and undertake essential environmental functions 
(B. Zhang, Xie, C. Zhang, & J. Zhang, 2012; Streimikiene, 
2014). A considerable range of benefits provided by urban 
green areas has been studied and reported in the literature. 
The list includes, but is not limited to (Konijnendijk, Anner-
stedt, Maruthaveeran, & Nielsen, 2013): human health and 
wellbeing, social cohesion, tourism, biodiversity, air qual-
ity and carbon sequestration, water management and the 
role of parks in the cooling of urban areas. Moreover, green 
spaces have a significant economic effect, most notably an 
impact on neighbouring property values. The latter effect 
was reported in a great number of research papers (for ex-
ample Espey & Owusu-Edusei, 2001; Jim & Chen, 2010).
The topic of urban green in CEE (Central and Eastern 
Europe) cities has been neglected in the discussion on the 
post-1989 transformation of the major metropolitan areas. 
Like many cities in the region, Warsaw experienced dis-
continuity of urban development (Staniszkis, 2012), and 
as a metropolitan area entity consists of two different sys-
tems. First of them is a heritage of a socialist period, and 
it is key characteristics are relatively low housing quality, 
but low development densities (building coverage ratios), 
greenness, and well-planned accessible public infrastruc-
ture. The second urban system is an effect of spontaneous, 
chaotic and dynamic development (Kusiak, 2012), that 
was not adequately regulated, due to malfunctions in ur-
ban planning. The higher quality of housing often comes 
at a price of high development density, lack of urban green 


===== PAGE 7 (score 38, numeric-paren 0, table-words 3, var-words 22) =====

364
R. Trojanek et al. The effect of urban green spaces on house prices in Warsaw
construct models that will not be biased because of a low-
quality dataset). The research is based on transaction data. 
The sample is based on residential property sales from the 
1st quarter of 2010 to the 4th quarter of 2015. The data was 
obtained from Board of Geodesy and Municipal Cadastre 
in Warsaw. All non-market transactions (e.g. debt enforce-
ment sales) were removed from the final dataset. Notarial 
contracts contain following information: transaction date, 
sales price, area of a dwelling, the floor on which a dwell-
ing is located, and the area of any auxiliary premises (e.g. 
a garage/parking spot in an indoor parking lot or a cellar/
storage). On the other hand, notarial contracts do not in-
clude information on various price determinants such as 
construction technology, architecture, building quality. 
Using cadastre data, additional information was gathered 
on the height of buildings and the year of construction. Fi-
nally, using Street View application on maps.google.com, 
the missing data concerning the height, and most impor-
tantly building technology was added. Locations of trans-
actions (exact addresses) were geocoded with google maps 
API. Based on WMS servers of the Warsaw City Hall, the 
vector layer of parks was prepared. In this research, we ex-
cluded apartments built before 1939. The reason for which 
was the fact that in the case of such apartments, the techni-
cal condition of a building is a significant determinant of 
their value (our base does not include this factor), which 
might burden the obtained results. Moreover, during the 
Second World War Warsaw was almost completely demol-
ished, so the share of such transactions is very small (should 
not burden the results). The final dataset contained 43 075 
geo-coded apartments sold between 1q 2010 and 4q 2015 
(Figure 1). Taking into account Dubin (1988) and Malpezzi 
(2008) suggestions three major categories of characteristics 
of dwellings may be distinguished: (1) structural attributes 
S, (2) neighbourhood related services and features N, (3) 
location and accessibility L. Additionally it is important to 
control for time (T).
The choice of qualitative and quantitative data was 
limited by the availability of information gathered. Table 2 
presents variables used in the study.
To conclude we controlled for various characteristics 
typically used in hedonic pricing models for residential 
market: location (administrative district), construction 
technology, floor, time of construction, area of an apart-
ment, presence of a garage, presence of a basement, height 
of building, and distance to a city centre, a kindergarten, a 
primary school, a subway and an urban green areas. The 
summary statistics of variables are presented in Table 3.
The results of the hedonic models’ estimation are pre-
sented in Section 3.
Table 3. Summary statistics of variables used in the research (source: own elaboration)
Category
Variable
Mean
Standard 
deviation
Category
Variable
Mean
Standard 
deviation
T
2010
0.16
0.37
S
area
52.86
24.23
2011
0.16
0.37
area2
3381.06
4089.24
2012
0.14
0.35
basement
0.43
0.50
2013
0.17
0.38
cons1950
0.18
0.38
2014
0.17
0.38
cons1960
0.16
0.36
2015
0.19
0.39
cons1970
0.11
0.31
L
d1
0.06
0.24
cons1980
0.04
0.20
d2
0.10
0.30
cons1990
0.06
0.23
d3
0.06
0.23
cons2000
0.40
0.49
d4
0.15
0.36
cons2010
0.06
0.24
d5
0.05
0.23
floor1
0.13
0.34
d6
0.10
0.30
floor2
0.51
0.50
d7
0.03
0.17
floor3
0.36
0.48
d8
0.01
0.09
garage
0.42
0.66
d9
0.09
0.28
height
1.67
0.67
d10
0.05
0.22
technology
1.73
0.44
d11
0.04
0.20
N
kindergarten
0.28
0.19
d12
0.07
0.25
school
0.45
0.29
d13
0.01
0.11
subway
2.78
2.48
d14
0.01
0.09
duga100
0.06
0.24
d15
0.03
0.18
duga200
0.11
0.31
d16
0.02
0.13
duga300
0.12
0.32
d17
0.10
0.30
duga400
0.11
0.32
d18
0.02
0.14
duga500
0.10
0.30
dcc
6.31
3.37
duga501
0.50
0.50


===== PAGE 8 (score 14, numeric-paren 0, table-words 3, var-words 8) =====

International Journal of Strategic Property Management, 2018, 22(5): 358–371
365
2.4. Hedonic price models
In a great number of researchers, the aim of which is to es-
timate the implicit price for each characteristic of a good, 
the hedonic method is used. Since Lancaster’s (1966) 
and Rosen’s (1974) seminal work, the hedonic method is 
widely employed in housing studies. Hedonic price mod-
els focus on the utility derived from individual character-
istics of a good. The core of the hedonic methodology is 
the assumption that the price of any heterogeneous good 
(apartment in this study) is the function of its attributes 
(localisation, area, quality of an apartment). As housing is 
heterogeneous good, it is difficult to even indicate a full 
list of crucial attributes. In our study, we examine the im-
plicit value of one of neighbourhood related environmen-
tal features, namely the proximity to the urban green area 
(duga). We hypothesise that transaction price is a function 
of a distance to the urban green area (duga), while con-
trolling for other relevant structural (S), neighbourhood 
(N), and location attributes (L), as well as time (T). In 
general, the price is a hedonic function:
 
 (
, , , , )
price
f
duga S N L T
=
.	
(1)
In our research, we use a typical log-linear hedonic 
model specification, commonly used in the literature 
(Malpezzi, 2008). Regress apartment’s price (natural loga-
rithm of sales price) on a set of independent variables. 
Our baseline model is given by:
0
1
100
2
200
3
300
4
400
5
500
ln
.
k k
m
m
i i
j
j
price
duga
duga
duga
duga
duga
S
N
L
T
= β +β
+β
+β
+
β
+β
+ ∑θ
+ ∑ω
+
∑γ
+ ∑τ
+ ε
	
(2)
In this research, we used several variants of hedon-
ic regression, namely standard Ordinary Least Squares 
(OLS), robust Weighted Least Squares (WLS) and Median 
Quantile Regression (Median QR) models. We use differ-
ent estimation techniques to ensure the results are robust 
and reliable. The housing literature is quite consistent on 
treating housing as heterogeneous in several dimensions. 
From econometric (or data analysis) perspective, this het-
erogeneity can create heteroscedasticity in the residuals, 
while estimating the price function using standard OLS. 
Therefore, we decided to address the problem using differ-
ent analytical approaches. Firstly, a robust model, employ-
ing OLS with heteroscedasticity-correction (WLS) was es-
timated. Secondly, we decided to employ median quantile 
regression. The quantile regression relies on minimisation 
of weighted absolute deviations, and during the process, 
conditional quantiles (percentile) functions are estimated 
(Koenker & Bassett, 1978; Koenker & Hallock, 2001). The 
quantile regression involves weighing, both symmetric (for 
the median, quantile = 0.5), and asymmetric (for all other 
quantiles). In this research we decided to use symmetric 
weighting (quantile-0.5). Within the quantile regression 
approach, there is no limitation imposed on explaining the 
mean of the dependent variable. The quantile regression is 
more flexible, as it can be employed to explain the implicit 
prices of housing attributes at any point of the distribution 
of the dependent variable, thus for low-priced, medium-
priced, and high-priced properties (J. Zietz, E. Zietz, & 
Sirmans, 2008).
The mechanism to perform the quantile regression 
is similar to ordinary regression. The difference in the 
mechanism to perform the quantile regression to ordinary 
regression lies in the way of searching for the margin of 
sums of squared residuals; the quantile regression looks for 
the margin of weighted sums of absolute residuals. There 
are several theoretical advantages of a hedonic quantile 
regression reported in the literature. The technique can be 
particularly useful in case of heteroscedasticity, outliers, 
and unobserved heterogeneity found in the empirical data 
on housing transactions (Liao & Wang, 2012).
Due to the high number of independent variables 
available, multicollinearity may be a serious concern. 
Multicollinearity leads to unstable coefficients and inflated 
standard errors. The Variance Inflation Factors (VIFs) was 
used to detect it. The VIF values in models do not exceed 
6.8 which is in line with the most conservative rules of 
thumb that the mean of the VIFs should not be consider-
ably larger than 10.
3. Results
We estimated two groups of regression models. The first 
group of models, later referred to as a general model, is 
based on the whole sample of housing transactions con-
ducted in Warsaw (apartments built after 1950). The sec-
ond group of models was based on the same sample. How-
ever, new variables were introduced to track the influence 
of a construction period (different housing estate) on 
proximity to the urban green area. The estimation results 
for the general model is presented in Table 4.
As a robustness check, alongside baseline OLS model 
we used WLS and quantile regression. The results are rela-
tively similar, albeit we found several differences regarding 
statistical significance for selected parameters. The esti-
mates discussed below refer to quantile regression model.
In our study nominal prices were used (we did not 
correct for inflation) and we observed that within the pe-
riod under study (2010–2015), the time had a significant 
impact on transaction prices. It is worth mentioning that 
housing prices in the biggest cities in Poland increased by 
about 100% between 2006 and 2007 (Trojanek, 2012). At 
the end of 2007, the subsequent decreasing phase in the 
house price cycle began resulting from this abnormal price 
increase and the beginning of the financial crisis.
The regression coefficients of the locational variables 
(districts) can be interpreted as district values. In this 
research, the base variable was d9 corresponding to the 
Downtown district. Statistically significant negative coef-
ficients of regression of other districts confirmed the fact 
that consumers for apartments in the city centre are will-
ing to pay more (Trojanek, 2015).
In general, the relations between physical character-
istics of apartments and sales prices were consistent with 
expectations. Additionally, we found that the increase 


===== PAGE 9 (score 45, numeric-paren 0, table-words 6, var-words 6) =====

366
R. Trojanek et al. The effect of urban green spaces on house prices in Warsaw
Table 4. Estimation results – general mode (dependent variable is natural logarithm of sale price) (source: own elaboration)
OLS
WLS
Median QR
Variable
Coefficient
Probability
Coefficient
Probability
Coefficient
Probability
const
11.9593
0.0001
11.8682
0.0001
11.9052
0.0001
Y2011
–0.0071
0.0130
–0.0071
0.0027
–0.0088
0.0011
Y2012
–0.0805
0.0001
–0.0809
0.0001
–0.0788
0.0001
Y2013
–0.1290
0.0001
–0.1282
0.0001
–0.1261
0.0001
Y2014
–0.0987
0.0001
–0.0995
0.0001
–0.0981
0.0001
Y2015
–0.0907
0.0001
–0.0951
0.0001
–0.0922
0.0001
d1
–0.1572
0.0001
–0.1490
0.0001
–0.1358
0.0001
d2
–0.2174
0.0001
–0.2347
0.0001
–0.2115
0.0001
d3
–0.1479
0.0001
–0.1553
0.0001
–0.1393
0.0001
d4
–0.0746
0.0001
–0.0802
0.0001
–0.0687
0.0001
d5
–0.1553
0.0001
–0.1396
0.0001
–0.1291
0.0001
d6
–0.1708
0.0001
–0.1690
0.0001
–0.1578
0.0001
d7
–0.2860
0.0001
–0.2837
0.0001
–0.2705
0.0001
d8
–0.1126
0.0001
–0.1154
0.0001
–0.0987
0.0001
d10
–0.2220
0.0001
–0.2142
0.0001
–0.2046
0.0001
d11
–0.1309
0.0001
–0.1258
0.0001
–0.1021
0.0001
d12
–0.0102
0.2429
–0.0244
0.0018
0.0013
0.8743
d13
–0.1069
0.0001
–0.0993
0.0001
–0.0870
0.0001
d14
0.0177
0.2303
0.0182
0.1767
0.0478
0.0006
d15
–0.1504
0.0001
–0.1186
0.0001
–0.1221
0.0001
d16
–0.2111
0.0001
–0.1911
0.0001
–0.1733
0.0001
d17
–0.1953
0.0001
–0.1859
0.0001
–0.1758
0.0001
d18
–0.0651
0.0001
–0.0797
0.0001
–0.0685
0.0001
area
0.0227
0.0001
0.0256
0.0001
0.0242
0.0001
area2
–0.0001
0.0001
–0.0001
0.0001
–0.0001
0.0001
basement
0.0294
0.0001
0.0204
0.0001
0.0216
0.0001
cons1960
0.0147
0.0001
0.0154
0.0001
0.0180
0.0001
cons1970
0.0341
0.0001
0.0245
0.0001
0.0307
0.0001
cons1980
0.0608
0.0001
0.0371
0.0001
0.0415
0.0001
cons1990
0.1354
0.0001
0.1310
0.0001
0.1366
0.0001
cons2000
0.2273
0.0001
0.2079
0.0001
0.2139
0.0001
cons2010
0.2213
0.0001
0.2210
0.0001
0.2059
0.0001
floor2
0.0269
0.0001
0.0256
0.0001
0.0269
0.0001
floor3
0.0284
0.0001
0.0288
0.0001
0.0302
0.0001
garage
0.0291
0.0001
0.0250
0.0001
0.0267
0.0001
height
–0.0061
0.0001
–0.0037
0.0063
–0.0026
0.0757
technology
0.0497
0.0001
0.0482
0.0001
0.0484
0.0001
dcc
–0.0268
0.0001
–0.0217
0.0001
–0.0233
0.0001
kindergarten
–0.0119
0.0131
–0.0029
0.5049
–0.0081
0.0718
school
–0.0352
0.0001
–0.0288
0.0001
–0.0307
0.0001
subway
–0.0160
0.0001
–0.0188
0.0001
–0.0186
0.0001
duga100
0.0452
0.0001
0.0276
0.0001
0.0304
0.0001
duga200
0.0145
0.0001
0.0138
0.0001
0.0077
0.0042
duga300
0.0120
0.0001
0.0069
0.0044
0.0065
0.0126
duga400
0.0074
0.0074
0.0021
0.3788
0.0028
0.2807
duga500
0.0056
0.0510
0.0015
0.5372
0.0029
0.2778
R-squared
0.8702
0.8495
–
N
43075
43075
43075


===== PAGE 10 (score 39, numeric-paren 0, table-words 6, var-words 11) =====

International Journal of Strategic Property Management, 2018, 22(5): 358–371
367
in distance from the city centre negatively affects the 
value of the apartment. The negative relation between 
the distance and a property price was also observed in 
case of a kindergarten, a primary school and a subway 
station (Trojanek & Gluszak, 2017). What is important, 
estimation results confirm the findings from earlier stud-
ies in Poland which suggest that the proximity to an ur-
ban green area has a positive impact on property values 
(Zygmunt & Gluszak, 2015; Czembrowski & Kronenberg, 
2016; Trojanek, 2016).
The results suggest that proximity to urban green has 
a significant positive impact on apartments’ prices up 
within 400 metres distance, and the reported price pre-
miums were the highest within 100 meters distance band. 
The effect is statistically insignificant beyond this distance 
threshold. These findings are also in line with the existing 
body of evidence, which suggests that urban green affects 
property process within 500–600 meters radius, and the 
values of adjacent properties are affected the most (Lut-
zenhiser & Netusil, 2001; McCord et al., 2014).
On average the distance to urban green areas has a 
significant nonlinear impact on residential prices. Direct 
proximity to a park or an urban forest (up to 100-meter 
distance) increased apartment sale prices by 2.8%–3.1% 
(depending on a model) compared with apartments lo-
cated outside 500 distance band. The premium declined 
with distance – apartments located in the second distance 
band (100–200 meters, thus relatively close, but not in di-
rect proximity to the urban green) were still sold for a 
higher price than comparable apartments situated further 
from the green area, but the difference was only by about 
1.0%. The premium had somewhat steep distance decay 
function – proximity to the urban green did not affect 
significantly sales prices of otherwise comparable apart-
ments located outside 400 meters distance band from ur-
ban green. The result can be intuitively explained – it is 
worth to pay for proximity to the park/forest only if it is 
easily accessible (a short walk from the building entrance, 
plus sometimes a relaxing view from the apartment). If 
the trip to the park/forest demands longer walk/drive the 
direct benefits from urban green will diminish.
To discuss the impact of an urban green on a housing 
value of different housing estates, we estimated an alterna-
tive model with interaction terms. We interacted distance 
to an urban green (duga) with period1 and period2 dum-
my variables. Variable period1 takes value 1 if the dwelling 
was built during the years 1950–1989, otherwise it takes 0. 
Analogously, period2 is equal to 1 if the dwelling was built 
after 1989, and 0 otherwise. The models were estimated 
again, and the results were presented in Table 5.
Table 5. Hedonic regression estimation results (dependent variable is natural logarithm of sale price) (source: own elaboration)
OLS
WLS
Median QR
Variable
Coefficient
Probability
Coefficient
Probability
Coefficient
Probability
const
11.9685
0.0001
11.8746
0.0001
11.9074
0.0001
Y2011
–0.0068
0.0174
–0.0070
0.0031
–0.0079
0.0024
Y2012
–0.0805
0.0001
–0.0810
0.0001
–0.0796
0.0001
Y2013
–0.1291
0.0001
–0.1283
0.0001
–0.1274
0.0001
Y2014
–0.0989
0.0001
–0.0996
0.0001
–0.0984
0.0001
Y2015
–0.0905
0.0001
–0.0951
0.0001
–0.0913
0.0001
d1
–0.1602
0.0001
–0.1521
0.0001
–0.1387
0.0001
d2
–0.2198
0.0001
–0.2363
0.0001
–0.2118
0.0001
d3
–0.1451
0.0001
–0.1533
0.0001
–0.1354
0.0001
d4
–0.0747
0.0001
–0.0801
0.0001
–0.0673
0.0001
d5
–0.1547
0.0001
–0.1396
0.0001
–0.1284
0.0001
d6
–0.1753
0.0001
–0.1712
0.0001
–0.1599
0.0001
d7
–0.2900
0.0001
–0.2869
0.0001
–0.2738
0.0001
d8
–0.1253
0.0001
–0.1230
0.0001
–0.1083
0.0001
d10
–0.2259
0.0001
–0.2180
0.0001
–0.2056
0.0001
d11
–0.1345
0.0001
–0.1295
0.0001
–0.1051
0.0001
d12
–0.0041
0.6345
–0.0188
0.0161
0.0069
0.3836
d13
–0.1136
0.0001
–0.1062
0.0001
–0.0902
0.0001
d14
0.0008
0.9589
0.0067
0.6314
0.0368
0.0063
d15
–0.1449
0.0001
–0.1153
0.0001
–0.1161
0.0001
d16
–0.2159
0.0001
–0.1961
0.0001
–0.1771
0.0001
d17
–0.1972
0.0001
–0.1881
0.0001
–0.1763
0.0001
d18
–0.0620
0.0001
–0.0749
0.0001
–0.0631
0.0001
area
0.0227
0.0001
0.0257
0.0001
0.0243
0.0001
area2
–0.0001
0.0001
–0.0001
0.0001
–0.0001
0.0001
basement
0.0287
0.0001
0.0198
0.0001
0.0215
0.0001
cons1960
0.0144
0.0002
0.0143
0.0001
0.0181
0.0001


===== PAGE 11 (score 35, numeric-paren 0, table-words 5, var-words 10) =====

368
R. Trojanek et al. The effect of urban green spaces on house prices in Warsaw
The results of the second group models estimation 
(with variables which should distinguish the influence of 
proximity to an urban green area on different housing es-
tate) provide interesting results. On an average proximity 
to an urban green areas did not have a significant impact 
on sale prices (neither in OLS, not WLS and Median QR 
model) in case of apartments built in years 1950–1989. 
Even immediate proximity (an urban green area located 
within 100 meters from the apartment) did not influence 
house prices. No significant (at alpha = 0.05) difference 
was observed for other distance bands.
The results differed significantly in case of building 
constructed after 1989. The impact of an urban green was 
particularly strong in case of post-transformation housing 
stock. Close vicinity (less than 100 m distance) to urban 
green increased the sales prices of apartments in new resi-
dential buildings by 8.0–8.6%, depending on a model. It 
did not have a significant impact in case of apartments 
located in socialist housing estates. The empirical results 
provide arguments in support of the research hypothesis. 
Apparently, relative lack of green areas around buildings 
and dense development typical for post-transformation 
residential development increases the hedonic value of 
public urban green areas. In the Figure 3 duga’s regression 
coefficients of quantile models are presented.
OLS
WLS
Median QR
Variable
Coefficient
Probability
Coefficient
Probability
Coefficient
Probability
cons1970
0.0342
0.0001
0.0238
0.0001
0.0297
0.0001
cons1980
0.0564
0.0001
0.0328
0.0001
0.0408
0.0001
cons1990
0.1197
0.0001
0.1190
0.0001
0.1237
0.0001
cons2000
0.2116
0.0001
0.1960
0.0001
0.2026
0.0001
cons2010
0.2074
0.0001
0.2098
0.0001
0.1949
0.0001
floor2
0.0268
0.0001
0.0254
0.0001
0.0267
0.0001
floor3
0.0281
0.0001
0.0288
0.0001
0.0305
0.0001
garage
0.0293
0.0001
0.0256
0.0001
0.0265
0.0001
height
–0.0059
0.0002
–0.0038
0.0046
–0.0023
0.1029
technology
0.0495
0.0001
0.0476
0.0001
0.0492
0.0001
dcc
–0.0270
0.0001
–0.0219
0.0001
–0.0237
0.0001
kindergarten
–0.0117
0.0145
–0.0024
0.5690
–0.0082
0.0596
school
–0.0354
0.0001
–0.0290
0.0001
–0.0294
0.0001
subway
–0.0146
0.0001
–0.0176
0.0001
–0.0175
0.0001
period1 x duga100
–0.0037
0.4587
–0.0057
0.1814
0.0022
0.6303
period1 x duga200
–0.0015
0.6965
0.0020
0.5545
–0.0034
0.3284
period1 x duga300
0.0009
0.8070
–0.0001
0.9685
–0.0017
0.6198
period1 x duga400
–0.0009
0.8015
–0.0046
0.1672
–0.0043
0.1940
period1 x duga500
0.0042
0.2963
0.0015
0.6614
0.0002
0.9484
period2 x duga100
0.0993
0.0001
0.0768
0.0001
0.0826
0.0001
period2 x duga200
0.0310
0.0001
0.0273
0.0001
0.0208
0.0001
period2 x duga300
0.0220
0.0001
0.0136
0.0002
0.0147
0.0001
period2 x duga400
0.0139
0.0008
0.0084
0.0191
0.0111
0.0032
period2 x duga500
0.0040
0.3152
0.0003
0.9339
0.0051
0.1665
R-squared
0.8709
0.8498
–
N
43075
43075
43075
End of Table 5
–0.02
0.00
0.02
0.04
0.06
0.08
0.10
0
100
200
300
400
500
Coef.
Distance to urban green area
All
1950‐1989
1990 and afer
Figure 3. Regression coefficients of distance to urban green 
areas (quantile models) (source: own elaboration)
We believe that this finding can have interesting policy 
implications. It is easy to dismiss positive effects of urban 
green areas when the city is relatively green. Increasing de-
mand for residential land and chaotic housing development 
resulted in a decrease of green areas at the urban fringe, 
where most of the new housing projects have been locat-
ed. The relative shortage of green areas has increased the 
implicit price of proximity to parks and urban forests in 


===== PAGE 12 (score 24, numeric-paren 0, table-words 0, var-words 23) =====

International Journal of Strategic Property Management, 2018, 22(5): 358–371
369
Warsaw. The residential value increase around parks and 
value can be captured in property taxes, but currently, this 
option is not exploited (in Poland nowadays property taxes 
are based on the size of a dwelling, not on its value – the 
location does not influence the tax amount). Currently, the 
tax system in Poland does not allow to link taxation with 
property value directly. Although the law allows for tax 
rate differentiation (based on location and other factors), 
in practice tax rates in most of the cities in Poland (Warsaw 
included) are set uniformly at a highest possible level for 
residential property (thus no differentiation is possible if we 
want to increase revenues). The increase in revenues from 
property tax revenues could be used to establish and main-
tain urban green and recreational areas, that would gener-
ate positive externalities and help to mitigate the negative 
effects of air pollution, typical to major cities in Poland.
Conclusions
We found strong evidence that proximity to a park is posi-
tively linked with apartment prices what is in line with 
previous studies. On the average presence of an urban 
green area within 100 meters from apartment increases 
the price of the dwelling by 2.8% to 3.1%. However, the 
effect of park/forest proximity on house prices is more 
significant for newer apartments than those built before 
1989. We found that proximity to a park is particularly 
important (and has a higher implicit price as a result) 
in the case of buildings constructed after 1989. Modern 
residential buildings are located smaller lots and built ac-
cordingly to the density maximisation principle (both re-
garding floor area ration and building coverage ratio). As 
a result, green areas within gated communities are often 
limited and can be treated like a club good, available only 
to residents. In that context, proximity to a publicly ac-
cessible green area outside the gates, only within a short 
walking distance is highly beneficial and translates into a 
significantly higher price of subject apartments.
We did not find a significant effect of urban green prox-
imity on prices of apartments located in housing estates 
built during the years 1950–1989. This somewhat surpris-
ing result may stem from the fact that contrary to modern 
residential buildings there is an abundance of green ar-
eas (children playgrounds, recreation facilities, conveni-
ent walking paths) within housing estates for public use. 
Presence of additional public park in proximity may not 
result in significantly higher residential satisfaction, and as 
a consequence will not be capitalised in apartment price.
The latter result may shed new light on previous find-
ings, notably Trojanek (2016) and Czembrowski and Kro-
nenberg (2016), that overlooked potential differences in 
hedonic prices of urban green areas for different types 
of housing estates. The results may be interesting for (i) 
housing developers, providing incentives for better loca-
tion choices regarding residential projects, as well as for 
(ii) urban planners, providing arguments to protect and 
maintain existing parks and green belts in the city.
References
Anderson, L. M., & Cordell, H. K. (1998). Influence of trees on 
residential property values in Athens, Gerogia (U.S.A.): a sur-
vey based on actual sales prices. Landscape and Urban Planning, 
15(1-2), 153-164. https://doi.org/10.1016/0169-2046(88)90023-0
Anderson, S. T. (2000). The effect of open space on single-family, 
residential home property values. Macalester Journal of Eco-
nomics, 10, 261-270.
Bark, R. H., Osgood, D. E., Colby, B. G., & Halper, E. B. (2011). 
How do homebuyers value different types of green space?. 
Journal of Agricultural and Resource Economics, 36(2), 395-
415. http://jareonline.org.
Bertaud, A., & Bertaud, M. A. (2000). The spatial development 
of Warsaw Metropolitan area. Comments on Warsaw Devel-
opment Strategy until the Year 2010 prepared for the World 
Bank, 11 June 2000.
Biao, Z., Gaodi, X., Bin, X., & Canqiang, Z. (2012). The effects of 
public green spaces on residential property value in Beijing. 
Journal of Resources and Ecology, 3(3), 243-252. 
https://doi.org/10.5814/j.issn.1674-764x.2012.03.007
Bolitzer, B., & Netusil, N. R. (2000). The impact of open spaces 
on property values in Portland, Oregon. Journal of Environ-
mental Management, 59(3), 185-193. 
https://doi.org/10.1006/jema.2000.0351
Conway, D., Li, C. Q., Wolch, J., Kahle, C., & Jerrett, M. (2010). 
A spatial autocorrelation approach for examining the effects 
of urban greenspace on residential property values. Journal of 
Real Estate Finance and Economics, 41(2), 150-169. 
https://doi.org/10.1007/s11146-008-9159-6
Crompton, J. L. (2001). The impact of parks on property values: a 
review of the empirical evidence. Journal of Leisure Research, 
33(1), 1-31. https://doi.org/10.1080/13606710500348060
Crompton, J. L. (2005). The impact of parks on property values: 
empirical evidence from the past two decades in the United 
States. Managing Leisure, 10(4), 203-218. 
https://doi.org/10.1080/13606710500348060
Czembrowski, P., & Kronenberg, J. (2016). Hedonic pricing and 
different urban green space types and sizes: insights into the 
discussion on valuing ecosystem services. Landscape and Ur-
ban Planning, 146, 11-19. 
https://doi.org/10.1016/j.landurbplan.2015.10.005
Des Rosiers, F., Thériault, M., Kestens, Y., & Villeneuve, P. (2002). 
Landscaping and house values: an empirical investigation. 
Journal of Real Estate Research, 23(1/2), 139.
Degórska, B. (2012). Spatial growth of urbanised land within the 
Warsaw Metropolitan Area in the first decade of the 21st cen-
tury. Geographia Polonica, 85(3), 77-95. 
https://doi.org/10.7163/GPol.2012.3.19
Do, A. Q., & Grudnitski, G. (1995). Golf-courses and residential 
house prices – an empirical-examination. Journal of Real Es-
tate Finance and Economics, 10(3), 261-270. 
https://doi.org/10.1007/BF01096941
Dombrow, J., Rodriguez, M., & Sirmans, C. F. (2000). The market 
value of mature trees in single-family housing markets. Ap-
praisal Journal, 68(1), 39-43.
Doss, C. R., & Taff, S. J. (1996). The influence of wetland type 
and wetland proximity on residential property values. Journal 
of Agricultural and Resource Economics, 21(1), 120-129.
Dubin, R. A. (1988). Estimation of regression coefficients in the 
presence of spatially autocorrelated error terms. The Review 
of Economics and Statistics, 70(3), 466. 
https://doi.org/10.2307/1926785