Mobile Edition
Print
Get the Mag
Weekly UpdatesIn the last few years, the property market has enjoyed high appreciation in values in the Istanbul housing market. This has occurred along with urban growth, and especially the changing economic structure and a new regulatory system in housing finance.
3.2. The data
The database employed in this study was generated using two data sets. The first dataset is gathered from two major real estate agent's websites and this data set contains 2,175 transactions of single-family homes sold in Istanbul in November 2006 and in April. This dataset compiles observations from 348 submarkets constructed from 946 neighbourhoods in 32 districts. The second dataset is derived from a survey that was undertaken by Istanbul Greater Municipality and provides information about the socio-economic structure of the neighbourhoods and the satisfaction of inhabitants of the city. [Deleting observations with missing values reduces the sample size to 1,517].
Table 1 presents the descriptive statistics for the transactions data provided from two major real estate agencies, Remax and Turyap in Turkey. This data set provides the property characteristics of the hedonic model. The database comprises information on key variables such as, location, price, age, floor area, construction type, number of storeys of the building and the housing unit, elevators, garage, garden, balcony, security unit and swimming pool.
The average transaction price for the 2,175 properties is $251,082, ranging from $34,000 to $8,000,000 (Figure 1). The average property area has 170[m.sup.2] of living space with 3.2 rooms and was 12 years old at the time of the sale.
Approximately half of the transactions (52.2%) on the sale range from 0 to 8 years old, and this correlates with the Marmara Earthquake in 1999. Although there are buildings up to 150 years old in the range, the percentage of the 61-150 year old building age group is 1.1% (Figure 2). The earthquake and also the increase in housing prices together with the trend of investing in the property market caused rapid construction process in the last years.
The living space ranges from 45 [m.sup.2] to 1920 [m.sup.2] whereas the number of the rooms ranges from 1 to 15 (Figure 3). The number of rooms in the housing unit vary from 1 to 15 and the avarage number of rooms is 3 (Figure 4). The average number of storeys of the buildings where the housing units exist is 6 (Figure 5) and 64% of the buildings has elevators. 90% of the housing units are flat (Figure 6) whereas 92% has a balcony, 78% has a garage and 79% has a garden. 46% of properties have a security system and 19% of them have swimming-pool.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
In most of the housing studies, neighbourhoods are defined as areas with homogeneous housing characteristics, property values, socio-economic property characteristics, political jurisdictions, and school districts (Clapp and Wang, 2006). Therefore, like the studies by Watkins (2001) for Glasgow, Goodman and Thibodeau (2003) for Dallas and Kauko (2004) for Amsterdam, administrative boundaries are taken on board as submarkets boundaries in this research. Housing submarkets are constructed using the administrative boundaries of the Istanbul Greater Municipality. This assumption also allows for the identification of the socio-economic structure, neighbourhood quality and housing prices segmentation in Istanbul.
In this research, each transaction is associated with its neighbourhood administrative boundary. The survey was not held in each of the neighbourhoods and therefore the adjacent neighbourhood to the submarket where the housing unit exists, is taken as the representative neighbourhood. In order to display the socio-economic and neighbourhood quality characteristics of the neighbourhoods, the dataset is composed from the survey held in 2005 by the Istanbul Greater Municipality. The data of this survey were collected according to a systematic sampling method with a sample size of 3,863, and by taking the density and land values into consideration in some of the 946 neighbourhoods. This data set provides the socio-economic and neighbourhood quality characteristics of the hedonic model The variables for socio-economic characteristics such as income, travel time to jobs and schools, travel time for shopping, the length of time the inhabitants have lived in Istanbul, the length of time the inhabitants have lived in the neighbourhood, household size and the variables for neighbourhood quality characteristics such as satisfaction from schools, transportation, municipality, health service, cultural facilities, playground facilities, security, neighbours, home, neighbourhood quality are provided by this survey.
[FIGURE 6 OMITTED]
This is shown in Table 2, which presents the descriptive statistics for the neighbourhood characteristics provided from the survey of Istanbul Greater Municipality. The database comprises information for a range of key variables and these can be seen in Table 2.
The average household income is $1,072, ranging from $333 to a maximum of $4444. The average household size is 3.5 and ranges from 1 to 6.5. The length of time the inhabitants have lived in Istanbul is 29.5 years whereas the length of time the inhabitants have lived in the neighbourhood is 13.5 years. Travel time for shopping is 17 minutes on average, whereas for jobs and schools it increases to approximately half an hour. In order to measure the satisfaction from different kinds of facilities, the respondents were asked to score these facilities on a scale from 1 to 7, with 1 being unsatisfactory, and 7 being satisfactory. According to the results, the less satisfaction rates belong to security, playground and cultural facilities. On the other hand, health service, school, transportation, and municipality satisfaction rates are valued as average. The higher satisfaction scores belong to neighbourhood quality, neighbour quality and home satisfaction. The factors affecting the housing prices are measured according to the variables listed in Table 3.
3.3. Methodology
Housing prices can be modelled using hedonic price functions. The hedonic approach is based on the assumption that a residential unit is composed of a bundle of individual components, where each one has an implicit price. The theory of hedonic price is formulated as a problem in which the entire set of implicit prices guides both consumer and producer locational decisions in characteristics space (Rosen, 1974). The hedonic price model is a method by which the price of the housing unit is delineated by structural, locational, and environmental attributes. This technique is based on a statistical analysis that characterises the price of housing unit as a dependent variable, and the structural, locational, and environmental factors are employed as independent variables in order to explain the dependent variable that is housing prices. Housing prices are affected not only by the structural characteristics of the housing units, but also by the socio-economic, behavioural environment, neighbourhood quality, and locational factors like amenities and disamenities. It is possible to interpret the implicit price of each attribute from the coefficients estimated from the hedonic function. This also allows comparisons between the prices paid for different qualities of the commodity, by examining individual attribute prices and the aggregate prices paid for heterogeneous housing units.
The hedonic price model is based on an assumption that the market contains a heterogeneous housing stock and heterogeneous consumers. Heterogeneity causes variation in house prices within a location, providing housing consumers with a range of housing unit options. In addition, housing consumers differ according to socio-economic and behavioural characteristics. Different households with different socio-economic composition have different tastes for housing structures that vary with respect to a range of components like size, number of rooms, and construction type. The heterogeneity of the housing stock and housing buyers denotes that the urban housing system is composed of submarkets, each of which will have a different market price for property attributes.
Hedonic price estimation is often used in housing submarket studies. The most significant implication of heterogeneity in housing market modelling studies is segmentation in the housing market. The urban housing market is most accurately represented as a collection of diverse yet interrelated submarkets (Rothenberg et al., 1991). In many studies, urban housing markets were investigated by taking submarkets as bases (Goodman and Thibodeau, 1998; Fletcher et al., 2000; Bourassa et al., 2007). In this study, housing price determinants are examined by employing a hedonic pricing model that incorporates neighbourhood administrative boundaries which can reflect the heterogeneous physical and socio-economical configuration. The variables included in the hedonic function can be grouped in four categories: property characteristics, socio-economic characteristics, neighbourhood quality characteristics, and locational factors.
Property characteristics include price, age, living area, number of rooms and total storeys of the building. Other property characteristics are represented with dummy variables, such as the type of the property (flat, detached), the existence of an elevator, balcony and/or garden. In addition, the characteristic such as "site" represents the dummy variable if the housing unit location is in a secured site with swimming pool and garage. The other characteristic "low storey" embodies if the storey of the building is lower than 5. "Site" and "low storey" variables were taken into account with respect to the preferences of the house buyers in Istanbul. After the 1999 Marmara Earthquake, house consumers preferred to live in the lower storey buildings at the highly secured low density sites that have swimming pools and facilities.
FEED