Landscape and Urban Planning 48 (2000) 45±55
Residential-location preferences according to demographic characteristics in Istanbul Vedia DoÈkmeci, Lale BerkoÈz* Istanbul Technical University, Faculty of Architecture, Department of Urban and Regional Planning, Tas,kis,la, Taksim, 80191, Istanbul, Turkey Received 13 May 1999; received in revised form 23 November 1999; accepted 3 December 1999
Abstract Istanbul is a rapidly growing city whose urban structure and socio-economic characteristics are constantly evolving. Residential preferences of the population are continually adapting to its changing conditions. According to many studies, the life cycle is the main explanatory factor in intra-urban migration. This study investigates residential-location preferences with respect to different age groups, household sizes and income groups. The results, which show a strong desire for mobility in middle and older age groups, are in contrast with the ®ndings of Western studies. At the same time, while young people's preferences are concentrated in the periphery, a large percentage of middle and older age groups prefer to move to the intermediate area between the core and the periphery, now the most easily accessible zone in the city. # 2000 Elsevier Science B.V. All rights reserved. Keywords: Demography; Residential; Preferences
1. Introduction Istanbul is a rapidly growing city wherein lifestyles and housing prices are continually in a ¯ux. Thus, residential preferences are constantly adapting to these changing conditions. In general, research on residential preferences deals with housing characteristics Ð cost, size and location Ð (Timmersmans, 1984) as well as demographic characteristics, such as
*
Corresponding author. Fax: 90-212-251-4895. E-mail addresses:
[email protected] (V. DoÈkmeci),
[email protected] (L. BerkoÈz) 0169-2046/00/$20.00 # 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 9 - 2 0 4 6 ( 9 9 ) 0 0 0 8 0 - 8
age, sex and income (Michaelson, 1977). This paper concentrates on residential-location preferences according to age groups and family sizes. It has been observed by many researchers (Clark and Van Lierop, 1987) and (Nijkamp et al., 1993) that the life cycle Ð rather than economic motives Ð is the predominant factor in residential relocation decisions. The temporal context of residential preferences has been conceptualised in terms of age or life cycle stage. Values change over the course of life, and these changes are presumably re¯ected in change of residential preferences (Michaelson, 1977; Rossi, 1980). In general, immobility increases with increasing age. Middle-age households (35±64 years old) are less likely to express a desire to move than younger house-
46
V. DoÈkmeci, L. BerkoÈz / Landscape and Urban Planning 48 (2000) 45±55
holds, and the elderly are the least potentially mobile (Lu, 1998). In some studies, age or life cycle stage has simply been measured and found to explain (Lamanna, 1964) or not to explain variation in preferences (Fuguitt and Zuiches, 1975). Other authors (Lindberg et al., 1988) showed that age affects residential preference on account of changes in the perceived importance of different values. Two studies have shed light on the relationship between reasons for moving and age (Speare et al., 1974; Clark and Onaka, 1983). According to these studies, housing unit adjustment, which means that young couples with small households need smaller houses than older couples with more children, and that the elderly with shrunken households may return to smaller houses again, is the most frequent explanation for mobility over all age groups. Changes in life cycle and neighbourhood adjustment are the second and third most frequent reasons. For young people married or not, adjustments in housing cost, tenure and structure type are the important motivating forces for relocation. For those at mid-life points (that is, those where the age of household head is between 35±45, Ð generally couples with young children or with teenage children Ð adjustment in tenure, housing unit size, and housing quality are important reasons. A recent study by Lindberg et al. (1992) examined the hypothesis that preference for residential location is based on housing-attribute evaluation derived from beliefs about value ful®lment, and that changes in these beliefs account for variation of residential-location preferences across the life span. The results of this study showed that life span changes in residentiallocation preferences depended on how much the preferences were in¯uenced by the value of comfort relative to freedom, well-being, and togetherness. A review of the literature shows that most studies about preferences for residential location have been done for developed countries. However, there are a few studies which investigate the intra-urban migration in developing countries, such as (Gilbert and Gugler, 1982; van Lindert, 1991; Conway and Brown, 1980; Vaughan and Feindt, 1983) in Latin American countries, Ozo (1986) in Nijeria and Ahmad (1992) in Pakistan. The present study investigates the residential-location preferences in Istanbul according to different age
groups and family sizes. It was expected that these various groups would differ in their residential-location preferences. The organisation of the paper is as follows: Section 2 gives background information about the distribution of age groups and household sizes in Istanbul; Section 3 investigates residential-location preferences according to age groups, household sizes and income groups. Section 4 presents the conclusion and suggestions for further research. 2. General characteristics of Istanbul Istanbul, with a population 7.3 million, is Turkey's largest city as well as its ®nancial and cultural centre. At the same time, it is a city world-famous for its natural beauty and historical monuments, re¯ecting its role as the capital of three separate empires. It enjoys the unique amenities of shorelines on the Black Sea, the Marmara Sea and the Bosphorus Strait. The rapid growth of the city since the 1950s, due to rural migration, has affected the quality of life in various sections of the city. While some of the modern districts have become comparatively more attractive, the historic districts have lost population due to the deterioration of their neighbourhood. In addition, the construction of bridges on the Bosphorus and the Golden Horn have changed accessibility of various areas measurably, and have thus caused a transformation in the pattern of land-use. These changes have created locational advantages and disadvantages re¯ected in the real estate market and intra-urban migration. Consequently, the Istanbul metropolitan region, with its dynamic social, economic, demographic and urban structure, makes an interesting case study for research. In this study, the distribution of population characteristics is de®ned according to concentric rings around the centre of the city. The core area is up to 3 km, which corresponds to the old CBD. The ®rst ring is from 3 to 12 km from the centre, which covers the area occupied by the city in the 1950s; the second ring is taken as the peripheral area beyond the ®rst ring (see Fig. 1). The city's population distribution according to the rings illustrates the following pattern: between 1980 and 1990, the share of population in the second ring increased from 39 to 54%, while that of the ®rst
V. DoÈkmeci, L. BerkoÈz / Landscape and Urban Planning 48 (2000) 45±55
47
Fig. 1. Location of districts in Istanbul.
ring decreased from 54 to 41% and that of the core area from 7 to 4.6%. Meanwhile, the actual population grew in both, the ®rst and the second rings, but decreased in the core area (DoÈkmeci and BerkoÈz, 1994). The geographic distribution of employment followed the population growth pattern in Istanbul. Before 1970, the main employment area was in the core of the city. However, by 1985, the percentage of the city's jobs in the ®rst ring (53%) had exceeded that in the combined core (27%) and the second ring (21%). This shift in job location and employment opportunities, in turn, resulted in changes in age distribution among the rings. After construction of the bridges over the Bosphorus and the Golden Horn and the connecting peripheral highways, upper and
middle income groups vacated the downtown residential areas and leapfrogged to newly constructed neighbourhoods into modern housing projects in the periphery (Figs. 2 and 3). Rural migrants moved into these deserted old houses (Figs. 4 and 5). Filtering processes, together with suburbanization, have been observed as a result of centrifugal population movement. The reversal of this process, witnessed during the past two decades in Boston, New York, Chicago and San Francisco, has yet to make itself felt in Istanbul. The success of the American cities in this respect was the result of government policies that encouraged housing construction, deindustrialization and recapturing of disamenity areas and a maturing urban aesthetic involving an appreciation of views, architecture and human contact (Ford, 1991).
48
V. DoÈkmeci, L. BerkoÈz / Landscape and Urban Planning 48 (2000) 45±55
Fig. 2. AtakoÈy Ð modern housing project from the II. Ring.
In order to provide a demographic picture of the city, age groups according to concentric rings are illustrated in Table 1. The core area of Istanbul has a small, but balanced distribution of age groups which is similar to that of the city as a whole. The most populous group in the core area is the low-income younger age group (DoÈkmeci et al., 1993, 1996). In other words, the centre of the city neither has the large numbers of old people Ð as in many developed countries Ð nor does it attract large numbers of middle-class young people who might participate in its regeneration, as in some inner cities of developed countries (Smith and Williams, 1986; Champion and Congdon, 1988). The distribution of age groups in the ®rst ring shows that while the percentage of the younger age group is lower than the average of this group in Istanbul, the Table 1 Distribution of age groups across rings (according to 1990 census ®gures) Age groups
Fig. 3. Atas,ehir Ð modern housing project from the II. Ring.
0±29 30±59 60 Total percent
Ring (%)
Total percent
Core
I
II
2.87 1.55 0.31 4.73
21.69 13.32 3.27 38.28
36.49 17.72 2.78 56.99
61.05 32.59 6.36 100.00
V. DoÈkmeci, L. BerkoÈz / Landscape and Urban Planning 48 (2000) 45±55
49
Fig. 4. Old housing from the Core (EminoÈnuÈ District).
middle and older age groups are higher. The higher percentage of the middle-age group can be explained by the higher percentage of job opportunities located
in this ring and higher residential prices. The higher percentage of the older age group may indicate a preference to stay near their middle-aged children
Fig. 5. Old houses from the Core (BeyogÆlu District).
V. DoÈkmeci, L. BerkoÈz / Landscape and Urban Planning 48 (2000) 45±55
50
Table 2 Distribution of household size according to urban zones (According to 1990 census ®gures) Household size
1±2 3±4 5±6 7 Total percent
Ring (%)
Total percent
Core
I
II
0.8 2.1 2.6 1.5 7.0
5.7 16.3 10.7 3.3 36.0
7.9 22.9 19.4 6.8 57.0
14.4 41.3 32.7 11.6 100.0
who are already located in the ®rst ring. There is a traditional family culture in Turkey as mentioned earlier, a fact borne out as well by international studies. The distribution of age groups in the second ring, however, illustrates the reverse situation. While the percentage of the younger age group is more than the average of this group in Istanbul, the percentages of the middle and older age groups are lower than the averages of these groups in the city. The higher percentage of the younger age group can be explained by a large number of migrant families in the periphery
of the city due to the development of squatter settlements (DoÈkmeci et al., 1993, 1996). The distribution of household size across rings has also been investigated. Results are given in Table 2. While the percentage of small households is lower than that of Istanbul's average, the percentage of large households is higher than that of Istanbul's average, both in the core area and in the periphery of the city. This can be explained by the higher number of migrant families in these zones, supporting John Turner's theory and the further expansion of this theory by BaÈhr (1994). According to this theory, at the earliest growth stages of a city, migrants are located in the core, because of low-paid jobs and the availability of cheap housing at the edges of the city centre. After their adjustment to city life, they move to the periphery in order to work at better-paid jobs and to live in better housing. After the growth of the city and development of industry in the periphery, it is possible for new migrants to ®rst establish themselves in the old squatter settlements in the periphery, where there are more job alternatives, and then to move on to better housing as their socio-economic situation improves (Figs. 6 and 7).
Fig. 6. The view of the District of Bes,iktas, from the I. Ring.
V. DoÈkmeci, L. BerkoÈz / Landscape and Urban Planning 48 (2000) 45±55
51
Fig. 7. New housing from the I. Ring (ErenkoÈy, KadikoÈy, District).
3. Distribution of residential-location preferences according to age groups and family sizes The spatial distribution of residential preferences of individuals is investigated here demographically. Respondents were posed questions in two sections. The ®rst section consisted of questions about the demographic characteristics of the subject. The second section consisted of questions asking the subject if they wanted to move to a particular district of Istanbul and why. The questions were asked in a multiple choice format in which the respondent was asked to rank his preferences from among many options. The sample size was 1105 households, proportionally taken from different districts. The results of the survey are evaluated by cross-tabulation. According to the results, 71.2% of the sample want to change their location. Broken down by age groups, 76.4% of the young people, 72.3% of the middle-age group and 62.9% of the older age group want to change their residential location. Among these, the 25±30 years old are of particular importance, because that is the age at which most people have settled down to family life and moved to apartments or singlefamily housing in the suburbs according to their
income level (Alonso, 1980). It is interesting to note that the high degree of expressed desire to move among the middle and older age groups is in contrast with the ®nding of studies in western countries (McHugh et al., 1990; Lu, 1998 ). Another uncommon result is that 70.2% of the older age group and 31.9% of the middle-age group who want to change their residential location have lived in their houses more than 10 years. This contradicts the general belief that as the number of years in the same location increases, people's stability increases also. Thus, these results illustrate that people in all age groups in Istanbul are under the in¯uence of some form of dynamic force which makes them open to physical and socio-economic transformations. The reasons for people's spatial preferences vary according to their stage in their life-cycle. For instance, proximity to job location is the most important factor for young people's residential preferences. The following reasons rank subsequently: proximity to relatives, clean environment, social environment, and amenities. The most important reason for the middle-age group is proximity to relatives, then followed by job location, a clean and quiet environment, social environment and amenities. Aesthetic is more
V. DoÈkmeci, L. BerkoÈz / Landscape and Urban Planning 48 (2000) 45±55
52
important for the middle-age group who can afford to pay for it, than the older and the younger groups. With respect to proximity to relatives, Ahmad (1992) found similar results in Karachi. The most important factors for the older group are proximity to relatives as well as a clean and quiet environment. Other factors rank as follows: social environment, job location, and amenities. The older people do care more for green areas and view than the other groups; probably, they have more time to enjoy the natural beauty of Istanbul. It is interesting to note that the middle-age group considers proximity to relatives more important than the younger group. Older people also place a higher value on proximity to their relatives as well as on a clean and quiet environment, such as they were accustomed to in the `old days' of Istanbul, before rural migration caused overcrowding. Older people's desire to be close to their relatives is similar to the international pattern: the results of several studies disclose that elderly parents typically have at least one child living within commutable distance of their place of residence (Warnes, 1986; Harper, 1987; Lin and Rogerson, 1995; Smith, 1998). Furthermore, subjects in different life-cycle groups were asked their preference for living at various distances from the city centre (Table 3). According to the results, 36.7% of the young people (under 30 years of age) prefer to live in the ®rst ring, while the remainder (63.3%) prefer the second ring. They do not want to live in the dilapidated neighbourhoods of the core (Figs. 4 and 5). These results correspond to the results from western countries where young people with small children prefer to live in the periphery of the cities. Of the young people who prefer to be in the ®rst ring, the main reason for their preference is proximity to their jobs (36.4%). However, the main reason of those who want to be in the second ring is to be close to relatives, and secondarily to be close to jobs and a clean environment (Table 4). Table 3 Spatial distribution of location preference according to agea
0±29 30±59 60 a
Core
I Ring
II Ring
Total (%)
± 1.1 2.4
36.7 43.6 54.9
63.3 55.3 42.7
100.0 100.0 100.0
Sample size: 1105.
While the largest percentage of the middle-age group prefers to live in the second ring which has better quality housing (Figs. 2 and 3), the largest percentage of the older group prefers to live in the ®rst ring (Figs. 6 and 7) (Table 3). The main reason for wanting to move in the case of both groups is proximity to relatives (Table 4). Only a small percentage of the middle-age and older groups prefer to live in the core. The main reasons given by the middle-age group are proximity to jobs and to relatives. These results correspond closely to the existing distribution of these groups within the city. None of those surveyed, under 30 years of age, wants to live in the core area. Thus, the return of young people to the inner part of the city that has been observed in many old western cities as a result of the gentri®cation movement has not yet started in Istanbul. The impact of view and aesthetic did not have an important impact on mobility, because only a small percentage of people can afford to live in such places. If the residential preferences of individuals are analysed with respect to the existing location of households, the following results are obtained. A larger group of the young people who are located in the periphery want to continue to live in the same ring than young people who are located in the ®rst ring (Table 5). None of the younger people want to live in the core area, because it has neither promising jobs to offer to young people nor decent housing. Although there are some small-scale revitalization projects that were started in the 1990s, the limited budget of the city government coupled with a growing democratic base have imposed severe limitations on the extent of largescale urban redevelopment and reinvestment. Cultural and aesthetic in¯uences also mitigate any wish from the middle-income and upper-income groups to gentrify the inner city. Besides, the `rent-gap' is insuf®cient to stimulate private sector supply of new or restored homes as a likely option in the city centre. Larger shares of middle aged and older people than younger people prefer to stay where they currently are. This trend is similar to the results obtained in western countries (Weesep and Kempen, 1992). If residential preferences are investigated according to life-cycle and income, the following results are obtained. A large share of lower and middle-income young people want to live in the ®rst and second rings (Table 6).
V. DoÈkmeci, L. BerkoÈz / Landscape and Urban Planning 48 (2000) 45±55
53
Table 4 Reasons for aspiring to spatial mobility of householdsa Age
Zones
Green
Clean
>30
Core I Ring II Ring
± 9.1 5.3
± 4.5 21.1
30±59
Core I Ring II Ring
± 2.6 4.9
60
Core I ring II Ring
± 6.7 8.6
a
Relatives
Jobs
View
± 18.2 31.6
± 36.4 23.7
± ± 2.6
± 13.0 16.5
40.0 34.4 27.2
60.0 18.2 10.3
± 20.0 20.0
100.0 31.1 28.6
± 13.3 8.6
Age
Current location
