# RECORD 5bed224cb10ce261eda55857ae903598b9f96697
# TITLE: Investigation of the impact of railways on housing values; the case of Istanbul, Turkey
# YEAR: 2019.0  VENUE: International Journal of Transport Development and Integration  DOI: 10.2495/tdi-v3-n4-295-305
# AUTHORS: E. Cengiz; H. Çelik
# PAGES: 11  CANDIDATE TABLE PAGES (1-based): [6]
# ABSTRACT: Housing units located near transportation facilities tend to have a higher market value than those that are at a greater distance to transport facilities but have similar characteristics. Provision of public infrastructure has a great impact on urban areas, especially on urban development patterns, spatial distribution of urban land use and real estate values. Mass transit systems are one of the most important public infrastructures in cities. Mass transit operations have a significant effect on property values, since they increase accessibility in urban areas. Railway projects are long-lasting investments and by their nature are extremely expensive when compared to other land transportation modes. It is not always possible or reasonable for many countries to fund these high-price projects only with their national budget generated by taxes collected from all citizens. Literature suggests that property values tend to rise in parallel to land values after any kind of public infrastructure investment. Nowadays transportation, especially railway system investments, is regarded as the most important public infrastructure in urban areas impacting land values maximally. Railway investments are becoming widespread in Turkey, especially in Istanbul as Istanbul is the most important and most populated city in country. As due to their high operation costs urban public transport systems, especially railway systems, rarely make a profit, private investors are not willing to get involved in these projects. Furthermore, it is not possible to finance all rail system investments from public funds. This article aims to prove the impact of railway systems on housing values and investigate the possibility of internalizing this increased value for financing railway investments. The main objective of this article is to calculate the impact of distance to a railway station on the value of properties by using the hedonic price model on an existing railway corridor and applying this anticipated impact on a proposed railway corridor. Finally, the calculated value is used as a preliminary basis as a funding model for railway projects in developing countries like Turkey.


===== PAGE 1 (score 13, numeric-paren 0, table-words 0, var-words 13) =====

© 2019 WIT Press, www.witpress.com
ISSN: 2058-8305 (paper format), ISSN: 2058-8313 (online), http://www.witpress.com/journals
DOI: 10.2495/TDI-V3-N4-295-305
	 Elif Can Cengiz & Hüseyin Murat Çelik, Int. J. Transp. Dev. Integr., Vol. 3, No. 4 (2019), 295–305
INVESTIGATION OF THE IMPACT OF RAILWAYS ON 
HOUSING VALUES; THE CASE OF ISTANBUL, TURKEY
ELİF CAN CENGİZ1, & HÜSEYİN MURAT ÇELİK1
1Istanbul Technical University, Turkey
Abstract
Housing units located near transportation facilities tend to have a higher market value than those that 
are at a greater distance to transport facilities but have similar characteristics. Provision of public infra-
structure has a great impact on urban areas, especially on urban development patterns, spatial distribu-
tion of urban land use and real estate values. Mass transit systems are one of the most important public 
infrastructures in cities. Mass transit operations have a significant effect on property values, since they 
increase accessibility in urban areas. Railway projects are long-lasting investments and by their nature 
are extremely expensive when compared to other land transportation modes. It is not always possible or 
reasonable for many countries to fund these high-price projects only with their national budget gener-
ated by taxes collected from all citizens. Literature suggests that property values tend to rise in parallel 
to land values after any kind of public infrastructure investment. Nowadays transportation, especially 
railway system investments, is regarded as the most important public infrastructure in urban areas 
impacting land values maximally. Railway investments are becoming widespread in Turkey, especially 
in Istanbul as Istanbul is the most important and most populated city in country. As due to their high 
operation costs urban public transport systems, especially railway systems, rarely make a profit, private 
investors are not willing to get involved in these projects. Furthermore, it is not possible to finance all 
rail system investments from public funds. This article aims to prove the impact of railway systems on 
housing values and investigate the possibility of internalizing this increased value for financing railway 
investments. The main objective of this article is to calculate the impact of distance to a railway station 
on the value of properties by using the hedonic price model on an existing railway corridor and applying 
this anticipated impact on a proposed railway corridor. Finally, the calculated value is used as a prelimi-
nary basis as a funding model for railway projects in developing countries like Turkey.
Keywords: hedonic price model, housing values, metro investment, public transport, railway funding. 
1  Introduction
The concern of economists about urban rent values dates back to 18th century. Today’s urban 
land value theories are derived from Von Thünen’s location differential rent theory [1]. As a 
successor of Von Thünen; Alonso developed the bid rent theory which states that rents tend 
to increase close to the city center as individuals wish to minimize daily transportation costs 
[2]. Nowadays, unlike their historical counterparts, cities have more than one center. These 
centers are formed by transportation networks. In the past, cities had one market place where 
all economic activities took place. Thus being close to the central business district (CBD) was 
an important determinant for housing prices. After transportation became easily accessible 
throughout the city, transportation access points which connect the CBD with locations at 
greater distances became new attraction points. Thus, property prices increased around such 
transportation access points as these meant less mobility time and cost [3].
Housing units at closer distance to public transportation networks mostly tend to demand 
higher market values than those with similar characteristics but at a further distance. Such 
differentiation is due to the savings in transportation costs [4]. 
Provision of public infrastructure has a deep influence on urban areas, especially on urban 
development patterns and spatial distribution of urban land use and real estate values [5]. 
Mass transit systems are one of the most important public infrastructures in cities. Mass 
transit operations have a significant impact on property values as they increase accessibility 


===== PAGE 6 (score 30, numeric-paren 0, table-words 8, var-words 7) =====

300	
Elif Can Cengiz & Hüseyin Murat Çelik, Int. J. Transp. Dev. Integr., Vol. 3, No. 4 (2019)
In order to calculate the value impact of proximity to railway stations, a questionnaire 
survey is conducted with rail estate agents working with 4–5 students on the site around the 
stations. On-site survey was started on 1 April 2017 and lasted for 1 week. Detailed informa-
tion on housing prices and their attributes are collected in the control case which has five 
stations. A number of 665 questionnaires are collected in total around these five stations. 
After the model is developed, the impact of being 1 m closer to a station is calculated, and the 
obtained coefficient is used for the main case study [16]. 
In this part of the study, the hedonic price model is run for the control case. The impact 
of each and every attribute on the housing price is calculated using the data gathered via 
questionnaires. The study continues with detailed discussions on the first model. The Google 
Earth program is used to obtain the distance between housing units and stations. The distance 
used in the model is air distance; the reason for using air distance instead of walking distance 
is the calculation methodology for the anticipated value increase on the main case.
3  MODEL RESULTS
After questionnaires digitized to excel sheet, model is run with R-Studio program by defining 
every variable in order to obtain an equation. Price of the apartment is the dependent variable 
to the other features of it. Table 2 shows the model results obtained from the related program.
Table 2:  Model results.
Price_new
Coef.
Std. Err.
t
P>ItI
[95% Conf. Interval]
air distance
–53.84791
19.80832 –2.72
0.007
–92.74359
–14.95222
age
–3,856.095
241.4267 –15.97
0.000
–4,330.161
–3,382.029
size
612.0502
158.6117
3.86
0.000
8,559.484
923.5007
rooms
18,641.78
5,134.587
3.63
0.000
8,559.4840
28,724.07
baths
23,672.19
6,665.66
3.55
0.000
10,583.47
36,760.9
alaturka
14,625.76
4,799.835
3.05
0.002
5,200.786
24,050.73
corner
–25,102.66
10,815.1
–2.32
0.021
–46,339.24
–3,866.091
south
–22,210.63
10,416.01
–2.13
0.033
–42,663.55
–1,757.709
west
–20,810.5
11,364.87
–1.83
0.068
–43,126.59
1,505.596
north
–19,716.2
14,065.37
–1.40
0.161
–47,335.01
7,902.614
elevator
22,008.68
4,815.631
4.57
0.000
12,552.69
31,464.67
_cons
167,938
14,697
11.43
0.000
139,078.8
196,797.1
X1: air distance, X2: age, X3: size, X4: rooms X5: bathrooms, X6: alaturka, X7: corner, X8: south, 
X9: west, X10: north, X11: elevator.
Total
3.2397e+12
664
4.8791e+09
Root MSE
=
46008
Adj R-squared
=
0.5662
Residual
1.3822e+12
653
2.1168e+09
R-squared
=
0.5733
Model
1.8575e+12
11
1.6886e+11
Prob > F
=
0.0000
F(11, 653)
=
79.77
Source
SS
df
MS
Number of obs
=
665
.regress price_new airdistance age size rooms baths alaturka corner south west north elevator


===== PAGE 7 (score 15, numeric-paren 1, table-words 3, var-words 8) =====

	
Elif Can Cengiz & Hüseyin Murat Çelik, Int. J. Transp. Dev. Integr., Vol. 3, No. 4 (2019) 301
To begin with, according to the calculated t-test values asserted above, almost all the 
exploratory variables exhibit statistically significant coefficients, except for X9 (west or not) 
and X10 (north or not). Second, according to the calculated F-test, the model is statistically 
significant. In other words, the theoretical investigation which refers to the hedonic pricing 
model of the related housing units seems statistically meaningful. Third, the calculation of 
the coefficient of determination (R^2), which is equal to 0.5733 and is a kind of acceptable 
value for such an exploratory investigation, asserts that the stated independent variables in 
modeling housing prices are able to statistically explain the related price changes around 
56.62 percentage level [16]. 
Besides the t-test values, the coefficients for variables seem significant. As distance and 
housing value have an inverse ratio, the coefficient for air distance carries a negative value. 
That means, as the distance from the station increases, the price of apartments decrease. Sim-
ilar to distance, age also carries negative value. As apartments get older their price decreases. 
As the size of the apartments increases, the price also increases; in other words, bigger apart-
ments are more expensive. Number of bathrooms, existence of an alaturka toilette and eleva-
tor also increase the price. But it is surprising that the facade of the apartment does not have 
a positive impact on the price, especially as according to literature corner apartments are 
always more expensive than others since they have a dual facade. But in this data set, it has 
a negative impact. This can only be explained by the cultural characteristics of the neigh-
borhood. Most of the buildings in the area are old and very close to each other. This may 
influence users’ preferences as for privacy reasons they may not wish to be too close to other 
buildings. Finally, the equation for an apartment depending on all its attributes is as follows; 
P=167,938–53.84×X1-3,856.09×X2+612.05×X3+18,641×X4+23,672×X5+14,625.7
6×X6–25,102.66×X7–22,210.63×X8–-20,810.5×X9–19,716.2*X10+22,008.68×X11
In the second part of the study, the calculated coefficient for distance in the control case is 
used to calculate the total value increase in the main case. The methodology for this is obtain-
ing the total number of single housing units served by each station.
4  DISCUSSION
The positive impact of railway investments on the housing market is demonstrated with the 
help of literature and the use of two case studies in Istanbul. As a further research, a new 
model will be developed to collect some part of this increased value from property owners 
for the financing of the railway. Since his kinds of urban infrastructure projects are mainly 
financed by governmental funds, some parts should turn to national budget in order to main-
tain the further investment funds. In the “Conclusion” section, the betterment tax collection 
methodology will be discussed as a new financial instrument. 
5  CONCLUSION
Based on the model, being located 1 m closer to the railway station increases the property 
value by 53.84 TL. The model is developed for the properties within a 500-m radius from 
the stations. Thus, this area is divided into zones of 100-m range see Fig. 5. The number of 
individual housing units is established within these zones. The average distance for each zone 
is accepted as follows: 
1–100 m----------50 m
100–200 m------150 m
200–300 m------250 m
300–400 m------350 m
400–500 m------450 m