Navigation – Plan du site

AccueilRubriquesEspace, Société, Territoire1999Socio-spatial lifestyles and segr...


Socio-spatial lifestyles and segregation

Styles de vie socio-spatiaux et ségrégation
Schnell Izhak et Benjamini Yoav


Cet article propose un nouvel indicateur de la ségrégation socio-spatiale, qui met l’accent sur la ségrégation des agents dans leur vie quotidienne. Nous montrons que la ségrégation devrait être étudiée, pour les individus, en tenant compte de tous les espaces dans lesquels ceux-ci accomplissent leur vie quotidienne. Notre indice fait référence aux contextes à la fois spatiaux et socio-interactifs de sept aspects (foyer, groupes de foyers voisins, voisinage et ville dans le contexte spatial et amis, travail et activités de loisirs dans le contexte social) des activités de la vie quotidienne telles qu’elles s’accomplissent effectivement dans l’espace-temps quotidien. Cet article calcule l’indice de ségrégation de sept personnes qui représentent des modes de vie typiques. Chaque mode de vie diffère par ses préférences pour un lieu dans des espaces sociaux et par l’exécution des cycles d’activité de la vie quotidienne.
L’indice peut recevoir des valeurs négatives représentant la tendance des agents à s’exposer à des membres d’autres groupes et des valeurs positives représentant la tendance à la ségrégation. Dans la pratique, la répartition de l’indice de ségrégation varie entre +- 4.0. De plus, aucune corrélation entre la ségrégation spatiale et sociale n’a été établie. Plusieurs agents ont fait beaucoup plus l’objet d’une ségrégation socialement que spatialement, d’autres par contre plus spatialement que socialement.

Haut de page

Notes de l’auteur

The paper is based on a presentation given at the Conference of the Commission on Urban Development & Urban Life, Bucharest, 17-21.8.1998

Texte intégral


1Most segregation measurements are based on the dissimilarity index, which in practice measures social groups’ degree of uneven distribution in residential space. More sophisticated indexes improve the measurement by calculating the effort needed in order to return to an even distribution. In this paper we propose to develop an index that measures the degree of isolation and exclusion that social groups may experience in performing their daily life. Such an approach may be rooted in Buttimer’s interpretation of the French school of geographic thought (1981), Hagerstrands’ time geography (1975) and Giddens’ (1984) structuration theory. Segregation should represent not only a person’s location in respect to the identities of ones neighbours in residential space, but also in respect to the structure of a person’s daily activity spaces. This has become increasingly relevant in the era of telecommunication. Thus, everyday life spaces are treated as major arenas for the structuration of society and human agency with segregation defined as a territorial strategy that social groups may adopt in order to isolate themselves or others from the rest of society. Our proposed index measures degrees of isolation of individual members of a social group from or exposure to other groups’ populations at their sites, as a continuous function in space and in respect to the actors’ actual everyday life spaces.

2In the study we identify seven different lifestyles, each of them representing different styles of using everyday life spaces. We investigate the lifestyle of one particular person that was carefully chosen to be closest to an "ideal type" of one of the proposed lifestyles and we measure her degree of isolation from other social groups, or exposure to other groups. In the next section of the paper we summarise our reservations concerning conventional approaches to the measurement of segregation, leading to the construction of our proposed index in the third section. The last section presents the patterns of isolation and exposure to members of other groups in the seven proposed lifestyles.

Segregation indexes

3Defining segregation as a state of socio-spatial exclusion and isolation among social groups (Boal, 1987 ; Sibley, 1995 ; Jackson, 1994) raises the issue of how to measure segregation. Over the years five groups of measurements have been proposed : ‘unevenness’ or ‘dissimilarity’, ‘concentration’, ‘centralisation’, ‘clustering’ and ‘exposure’ or ‘isolation’ (Lieberson, 1981). Interestingly, most methodological advancements pay attention mostly to either geometric or topological characteristics of social groups’ distribution in residential spaces. The most widely used measurements are based on Duncan and Duncan’s dissimilarity index (1955) comparing distributions of two population groups over predefined discrete spatial units. New attempts to improve measurements focus on calculating the number of residents who need to relocate and the degree of effort needed by them to enable an even distribution among these groups (Morgan, 1983 ; Morrill, 1991 ; Waldorf, 1993, Wong, 1998).

4The only indicator that claims to relate directly to spatial segregation, rather than to population distribution, is the ‘exposure index’. It measures the probability that an average member of any social group will share the same discrete territorial unit with the average member of other social groups in her closed vicinity. This may be interpreted also as an indication for potential encounters between average members of any social group and members of other groups in their everyday life (Boal, 1987).

5These indexes that claim to measure degrees of segregation of groups’ residential distributions, suffer from significant shortcomings. They treat individuals in terms of homogeneous members of pre-defined and self-evident social groups, expressing their collective degree of segregation by using one mathematical value. They assume also that residents’ life conditions and worldviews are determined by the continuous space that is defined by a large proportion of the group members within relatively closed boundaries. As a consequence of their pre-assumption, conventional indicators do not take into account the possibility that different members of the group, who are located in different places or even in the same places, may experience varied degrees of isolation or exposure in respect to various spheres of their everyday life. For example, the possibility that two communities with high dissimilarity and isolation indexes may represent completely different situations of segregation was not considered. One may represent a highly isolated group, which is blocked of any communication routes with other social groups in space. The other may represent a group of people who prefer to share a common neighbourhood but who maintain open routes of communication and socio-cultural dialogues with other groups. We suggest studying segregation for individuals in respect to their actual everyday life spaces. Only in a second stage of the analysis we suggest to aggregate agents with similar patterns of segregation and to identify their membership in larger social groups and territorial units.

6It seems that the reason for these indexes’ oversimplification of the geographical reality is rooted in their hidden assumption that the law of least effort spatially determines social relations. In accordance, it is believed that the world may be regionalised into continuous and discrete relatively closed territories, which have the power to determine their inhabitants’ lifestyles. We argue that these measurements oversimplify the spatial patterns of post-modern daily life in which spaces may be fragmented into configurations of discontinuous territories and places. These spaces may receive particularistic meanings by different social groups and may be relatively exposed to global information and networks. Current indexes has some advantage in measuring groups degrees of dissimilar distributions by using simple and accessible indexes, but we suggest supplementing them with the proposed index which may shed light on human beings segregation in their everyday life spaces which increasingly transcend local networks.

Socio-spatial lifestyles and segregation

7The basic unit of analysis is an agent who is located at her home site. We measure the agent’s level of segregation in respect to her everyday life activity spaces. Everyday life space is defined in this context as the spaces of human daily activity circulation as it is structured and restructured in society and space. Everyday life spaces are defined in terms of networks of interactions starting from the zero point of the home and reaching out to a person's horizons of involvement in her daily life (Buttimer, 1981). This orbit of everyday life activity is a person’s arena for the constitution of her lifeworld (Buttimer 1977 ; Giddens, 1984 ; Pred, 1989). The association of social lifestyle with location in social space and in constructing daily activity orbits, is based on the assumption that any social project requires the meeting with others and movement in space. Thus, social "projects" and spatial "paths" are dialectically associated as well as human lifeworlds and their outer surroundings (Hagerstrand, 1975 ; Pred, 1989).

8In our model we assume that any activity has a social dimension designated by the social identity of meaningful others in major activities performed by a person and a spatial dimension designated by the population structure in concentric territorial bases (Figure 1). The identity of meaningful others as members of a person’s group or other groups represents ones actual isolation or exposure, while the composition of the population in the aforementioned territories represents a person's probability to encounter others, either from her group or from alternative groups.

9Everyday life activities are frequently divided into several spheres of practices, with most literature distinguishing between work, social contacts, shopping, personal arrangements, and leisure activities. The most significant distinction in capitalist society is the socio-spatial separation between home and work with an increasing salience devoted to leisure activities with friends in the late capitalist era. We add to these two social spheres the realm of

10telecommunication, which through cheap international telephone systems, Internet and other telecommunication systems has shrunken the fraction of global horizons into agents’ everyday life reaches. In each of these three spheres of daily activities we investigate the social composition of agents’ meaningful others, whether they belong to their social group or to alternative ones.

Figure 1 : A general model of agents’ everyday life space

11Territorial bases are hierarchical in character with four qualitatively concentric scales (Saarinen, 1976 ; Rowles, 1981). The most intimate scale is of personal space which surrounds individuals like bubbles (Hall, 1966 ; Sommer, 1969), but which we consider to be irrelevant to our study. In the second scale the zero point of an agent’s sense of territoriality is projected at the home as the representation of the agent’s identity, and as the focus of her everyday life activities (Cooper, 1984). At the same time, home territories are perceived to be essential for the formation of autonomous identities (Bachelard, 1969 ; Altman, 1975). The range of daily activities is the third territorial sphere in which agents reach out of their homes to fulfil their daily necessities such as work, shopping, leisure etc. This scale is frequently sub-divided into three different sub-scales. The most intimate one is of clusters of neighbouring homes along a common street in which people are recognised and social interactions with neighbours tend to take place (Cluster). The intermediate one is the neighbourhood as a territorial setting, which represents agents’ social identity and lifestyle. The third and widest scale is the neighbourhood unit, which supplies its residents with daily services. Recent research, performed in Tel Aviv, has confirmed the relevance of the two more intimate characteristics of neighbourhoods as centres of social interactions, and as representing groups’ identities. However, they were only indeterminate evidences of the relevance of the neighbourhood as a centre of daily services (Goldhaber, 1998). The highest scale is that of far more extensive spatial units such as city and state with which agents become acquainted through mass communication, socialisation etc.

12Geographers have identified several models that describe different forms of using everyday life spaces as part of a wider social lifestyle. Buttimer (1981) distinguished between localist and urbanite lifestyles. Localists are characterised by their concentric and narrow orbits of daily activity patterns, which are associated also with strong attachment to their home vicinity. Urbanites are characterised by their high mobility in the metropolitan area and by the fragmentation of their activity orbits to separate socio-spatial contexts. Hall (1984) characterised suburban lifestyle as a dual space of home and community in the homogeneous suburb on the one hand, and work, shopping and services and social activities performed in the wide metropolitan reaches on the other hand. The new middle class has introduced a new post-modern lifestyle, characterised by condense time-space practices, which spread through long hours within the dense and heterogeneous inner city milieux (Rose, 1984 ; Short, 1989 ; Schnell, 1996). In addition, studies on ghetto members’ activity patterns portray, dense and short distance movements, with intensive intra-ethnic networks, which are restricted to the boundaries of the ethnic enclave (Boal 1987 ; Waldinger, 1993).

13In the analysis we measure each agent's rate of isolation from, or exposure to members of other groups in both the social and the territorial dimensions. We represent an agent’s social dimension, by the social identities of meaningful others in three major spheres of daily activities : work ; friends ; and telecommunication. We represent the spatial dimension by the agent’s probability to be exposed to members of other groups in four major hierarchically ordered territorial bases : house ; cluster of neighbouring houses ; neighbourhood ; and city (Figure 2).

14The impact of each of the daily activity spheres and territorial bases, on the total rate of segregation, is weighed according to the proportion of awakened daily hours spent in each activity sphere and in the territorial base. This is multiplied by the relative salience that interviewees assign to each activity sphere and territorial base, respectively. An agent’s overall rate of segregation is, then, calculated as the sum of her rates of isolation or exposure in respect to both the social identities of meaningful others in the three activity spheres, and potential encounters with members of other groups in meaningful territorial bases.

Figure 2 : A general model of socio-spatial segregation

15In constructing the equation for segregation we define an initial rate for segregation and a weighing method. In order to enable the weighing on a meaningful scale, the initial rate which is originally measured by a proportion p is re-expressed into the logistic scale : s(p)=log(p/(1-p)). Averaging segregations at different activity spheres and territorial bases is done in much the same way as explanatory variables explain the probability of an outcome in a logistic regression. As p varies between 0 and 1, s(p) varies monotonically over the entire unbounded range. At p=1/2 s(p)=0, and it is further anti-symmetric for this value : s(p) = - s(1-p).

16The index of segregation at each of the lowest levels, is the logistically transformed number of the group’s members ni, out of the number of alternative groups’ members Ni in the relevant territorial bases or activity spheres. Thus for example, for the House territory with ni(H) and Ni(H) defined as above, the logistically transformed segregation is

17In practise we add a third to each of the counts. Statistically, this sets the transformed values, approximately at their expected values for a sample from the logistic distribution. Technically, this also enables the use of the logistic transformation when ni = Ni or when ni = 0, (Tukey 1977). We therefore define the house segregation index as

18The proposed index for socio-spatial segregation is calculated as the average value of the segregation in the social and the spatial dimensions, each of the two, in turn, being the average of three activity spheres and four territorial bases. We give below the specific formulation, which includes the seven aforementioned activity spheres and territorial bases. We suggest adopting this formulation as a standard structure, in order to make the index friendlier for comparative studies. Nevertheless, the index is open to the consideration of alternative social or spatial categories if they are relevant for the analysis of a particular group’s everyday life practices.

19The weighing process is essential to the characterisation of the overall segregation assessment and may involve two sets of weights : time related weights and salience related weights. One set of time weights measure the proportion of time the agent spends in each of her territorial bases for the spatial segregation index. A second set of time weights measure the proportion of time the agent spends in each of her daily activities for the social segregation index. Similarly, two other sets of weights express the relative salience they assign to each of these spatial and social settings. The information leading to the weights can be based on actual observations, on registered self-reporting (diaries), or the respondents’ answers to a questionnaire.

20If Ti(r) is the time spent at home territory r, and Vi(r) is its perceived salience, we form their product W*i(r)= Ti(r) Vi(r), and normalise it by the sum of these weights over all territorial bases,

21to acquire a set of weights that equal 1. Then, we define

22Similarly, if Ti(a) is the time spent at activity a, and Vi(a) is its perceived salience, we form their product W*i(a)= Ti(a) Vi(a), and normalise it by the sum of these weights over all activity spheres,

23to obtain a set of weights whose sum is1. Then, we define

24To obtain the overall socio-spatial segregation index we combine SPi and SOi using two fixed weights W(SP) and W(SO) whose sum is 1, to obtain the weighed average

25SPSi = [W(SP) SPi + W(SO) SOi].

26In conclusion, socio-spatial segregation is calculated as an agent’s weighed average rate of isolation expressed on a logistic scale, in a set of social and territorial contexts within agents’ everyday life spaces. The individual weights reflect the relative weight of each of these social and territorial dimensions of the agent everyday life, both in terms of time spent and in their perceived importance.

27The SPS index of segregation enjoys the same properties as the initial segregation indexes it averages. It varies over the entire range of numbers, positive values indicating high degrees of socio-spatial isolation of the individual from members of alternative groups, and negative values indicating high degrees of exposure to members of the other groups. Values close to 0, (e.g. -1< SPS < +1) express inter-group mixture of about equal probability for socio-spatial interactions with members of the same group or of other groups. It may add some helpful insight to recall that the re-expression eSPS/(1+ eSPS) takes the value of the combined SPS back into the probability range 0 to 1.

28It is important to emphasise that the proposed segregation index SPSi is evaluated for an individual residing at her site. A different question now is how to study the spatial pattern of those who live in segregation. The most common solution would be to create an isoline map based on the triangulation of the values of the agents’ segregation rates. But several other common mapping solutions may be implemented as well. Nonetheless, some difficulties may arise from the fact that two agents residing in the same place, while sharing the same territorial related basic initial segregation values, may end up with different SPSi, either because of different weights or because of different social segregation levels SOi. A second cause of difficulty is those sites with no agents from the studied group. Using computerised mapping methods like homopletic maps or response surface analysis may solve the first problem. A remedy for the second problem is to use only the spatial component in the definition, i.e. setting SPSi = SPi. A third remark may be useful at this point. By adding one third to the initial proportion of the group members relative to others, enables the calculation of the spatial indexes even in places where no individual from the investigated group resides. Furthermore, in such a case, if one house has 10 individuals and the other has 40, the SP at the location of the first house will be higher then at the other, correctly reflecting the spatial segregation an individual from the investigated group has if he joins either houses. We now demonstrate the applicability of our proposed index in the case of African migrant workers in Tel Aviv, including the application of solutions to some of the aforementioned problems.

Lifestyles and segregation

29In this section we describe the daily activity patterns of seven individuals who represent typical urban lifestyles. We deliberately choose individuals whom we expect to be closest to one of the lifestyles that have been established in the geographic research and whom we expect to characterize typical lifestyles in Tel-Aviv. We have identified seven such models, which form different daily activity patterns. For each of the respondents we calculate the metrics of segregation in order to demonstrate how differences in lifestyle are expressed in the proposed segregation index. Each respondent was asked to describe the identity that characterise her in respect to her neighbourhood using it as a criteria to distinguish between members of the group and members of alternative groups. We standardise the results to a hypothetical milieu of 100,000 inhabitants in which the members of the group are ten percent of the total population. This procedure enables us to construct a general model for comparative analysis of socio-spatial lifestyles, based on our careful selection of respondents who represent typical geographical lifestyles. In our hypothetical model, the respondent’s neighbourhood inhabits six thousand members, of which two hundred share with the proposed dweller the same block (Cluster), and twenty-four people share the same house. The types of dwellers differ from each other in choosing their neighbours and their meaningful others in performing their daily life.

30We exemplify the model of the typical member of the ethnic ghetto with an Israeli Palestinian who was born in the Arab neighbourhood of Jaffa. Most Arabs in Jaffa are much more open to the Jewish majority, but we chose him in order to demonstrate the case of extreme segregation in ethnic ghettoes (Schnell, 1995 ; Gonen and Hadas, 1994 ; Omer, 1997). He knows Hebrew but he prefers to speak Arabic in which he expresses himself much easier. He takes advantage of the "ethnic enclave" in finding a job, shopping and other activities. He owns a grocery store, which serves the Palestinian community, and in the afternoons he is active in the community center. He is deeply involved in local ethnic organisations in a way that many feel obliged to buy from his store. In accordance, most of his activities are fulfilled within the ghetto’s territory with other members of the ethnic group. Trips outside the ghetto are minimal and short, because he tends to feel a stranger in none ethnic spaces and he is afraid to be exposed to negative stereotypes. Almost all his meaningful others are members of the Palestinian community in Jaffa. His strong sense of attachment to his close surrounding is in the neighbourhood of Agami and his sense of strangeness in outer spaces leads him to assign high priority to the ghetto’s territory, and to spend most of his time there. Thus, his close surrounding, highly dominated by Arabs, house and cluster are highly weighted by him in contrast to the rest of the city that receives only a low weight (Table 1). An interesting attitude is presented toward the neighbourhood level. Like most Palestinians in Jaffa, he assigns relatively high salience to the whole neighbourhood of Agami because it is associated in the Palestinian collective memory with the Palestinian identity despite the fact that it is today a mixed neighbourhood in which he feels alienated.

Table 1 : The Pattern of a member of the ghetto segregation

Note : The nominator describes the number of members of the group.

31The denominator describes the total number in the studied unit.

32The resulting segregation index is SPS=+3.7 which represents the respondent’s deep tendency toward isolation from non-Palestinian communities both in the spatial dimension (SP=+4.2) and the social dimension (SO=+3.3). Since, large aspects of his daily life are performed within the cluster and the house, they supply the highest contribution to the overall segregation (values of 2.2 and 1.6, respectively). The other spheres of activities that contribute to the overall degree of segregation are friends and work, the first mainly due to their relative importance and the second mainly due to the long hours spent at work. The city beyond the neighbourhood and telecommunication, which were assigned only marginal relevance almost, do not influence the overall value of segregation.

33A member of the neighbourhood of Ezra, located in the eastern quarter of Ramat Gan, represents the typical localist from oriental Jewish and working class origin. She centers her daily life in narrow circles around her house. She is attached to her neighbourhood, mainly through her children. She spends at least three hours a day with her children walking to school, playing in the garden etc. In addition, she performs a large part of her daily activities like shopping and social activities within the neighbourhood. Unlike the Palestinian from the ghetto, she works outside the neighbourhood, some of her friends are distributed around the city and even beyond, and she enjoys travelling from time to time outside her neighbourhood for shopping, leisure activities and visiting friends. Since she identifies with the neighbourhood she assigns relatively higher weights to her cluster and neighbourhood but not to her standard flat in the public project (Table 2). Her social life and her status as a good friend receive by her higher salience than her job as a manual worker in the factory. In accordance, she assigns high weight to her social life.

34The resulting segregation index for the aforementioned person shows a relatively low level of isolation from non-ethnic milieux with SPS=+1.8. The localist shows a relatively significant degree of spatial segregation (SP=+3.3) but her social networks are already mixed (SO=+0.3), free of ethnic discrimination. The tendency to choose friends among oriental Jews is more the result of localist lifestyle in a neighbourhood that inhabits mainly Oriental Jews. The homogeneous character of the Oriental Jewish cluster supplies the highest contribution to the person’s socio-spatial segregation. These concentrations of Oriental Jews in marginalised neighbourhoods is the result of public housing policies in the 1950’s, in which many immigrants were settled in the fringe of existing towns, usually on Palestinian refugees’ Lands (Gonen, 1997).

Table 2 : The Pattern of a localist’s segregation

Note : The nominator describes the number of members of the group. The denominator describes the total number in the studied unit.

35A seventy-five year old lady from the southern neighbourhood of Florentine represents the typical isolated elderly. Yuppies, Migrant workers, or businesses replaced most of the veteran members of the neighbourhood. Only a few of her friends remained in her vicinity. Her two children live far away from her - one in the U.S.A. and one in the northern periphery of the country. In accordance, she is the only member of her social group in the house, and one of the last ones to remain in the cluster and in the neighbourhood. She assigns extremely high weight to her house both because she spends most of her day at home and because she assigns high importance to her home in which she has accumulated many memories. Most of her contacts are with her family and few good friends that still remained with her. From time to time she either visit or host them, but much more frequently she talks to them on the phone. In accordance she assigns high weights to friends and telecommunication (Table 3).

36The resulting pattern of segregation is a mixed lifestyle (SPS=-0.3) in which spatial dispersion (SP=-2.7) is coupled with social isolation (SO=+2.0). But the negative result for the spatial dimension may be misleading to some extent. It does not represent the elderly choice to locate herself in a mixed neighbourhood, but rather a result of her isolation from her friends and relatives who were replaced by younger generations and some other new groups. As a response, the studied lady closes herself in her home on the one hand and develops telecommunication networks with her relatives and friends from whom she is spatially isolated.

Table 3 : The Pattern of an isolated elderly segregation

Note : The nominator describes the number of members of the group. The denominator describes the total number in the studied unit.

37A person who moved out of the lower class oriental Jewish "Ezra" neighbourhood represents the typical exposed person. He feels that by his move, he has escaped the marginality and stigma as an "Oriental Jewish failure", joining the milieu of assimilated middle class Israelites. In accordance, he chose a neighbourhood that is largely inhabited by Israeli and European born householders, and he mentions almost only Israeli and European born meaningful others in describing his social networks (Table 4). He moves freely around the city, avoiding any discrimination among spaces and he assigns proportional salience to each of the spheres of his activity pattern. The resulting pattern of segregation is a situation of exposure both in the spatial dimension (SP=-2.1) and in the social dimension (SO=-2.1). In accordance, the overall segregation of the investigated exposed person reaches the level of SPS=-2.1 with all the spheres of activity contributing their share to his sense of exposure to non-ethnic groups.

Table 4 : The Pattern of an exposed person’s segregation

Note : The nominator describes the number of members of the group. The denominator describes the total number in the studied unit.

38The typical "yuppie" (Young urban professional) or "Dinky" (double income no kids) tends to locate herself in the heterogeneous and crowded sections of the inner city, where she may perform her intensive lifestyle in a condensed time-space context (Schnell, 1996). A thirty year old, non-married woman from Sheinkin Street in the inner city of Tel-Aviv may represent this group. She does not know her neighbours or others in her cluster. Instead, she chooses her meaningful others from around the inner city surrounding, as well as from the rest of the world. She is always connected to a wide network of meaningful others by portable telephone and she spends about three hours a day on Internet communication. Except for a few close friends, she tends to mix business with social and leisure activities. She is hardly attached to her home and she spends most of the day outside it in intensive activity patterns, which include a mixture of long hours of work business meetings, and sophisticated consumption of leisure and culture. In accordance, she devotes high weights to work and telecommunication, while her home receive only marginal importance (Table 5). Corporealy, she spends most of her time in her inner city surrounding (neighbourhood), although long distance networks are highly essential for her. The resulting pattern of segregation represent an example of a lifestyle that takes place in a mixed milieu both in the social and the spatial dimensions (SP=-0.9 and SO= +0.9).

Table 5 : The Pattern of a yuppie’s segregation

Note : The nominator describes the number of members of the group. The denominator describes the total number in the studied unit.

39The typical member of the middle class suburban lifestyle tends to share his time between work in the city and his suburban neighbourhood. His private house becomes a salient source of identification. The house is defined as a castle that supplies refuge from the stresses of the competitive world of business. He tends to spend about eight hours a day in the city at work, and before travelling home for about one hour each direction, he may spend another two hours either shopping or busy with leisure activities. He becomes involved in his community life only in crisis situations when his lifestyle or the value of his property is under risk. Therefore, he may assign high weight to the house on the one hand and to the rest of the city on the other (Table 6). Similarly he assigns high weights to his career and his household. The result is a pattern of mixture in which the investigated persons is exposed both to members of his group and to others in large reaches of the metropolitan area and beyond.

Table 6 : The Pattern of a cosmopolitan’s segregation

Note : The nominator describes the number of members of the group. The denominator describes the total number in the studied unit.

40An illegal African migrant worker represents a typical model of extreme social isolation that is not necessarily followed by a high degree of residential segregation. He settled in the neighbourhood of Nave-Sha’anan in which about two-thirds of the residents are migrant workers from different origins and the rest are Israeli working class households and yuppies. Despite the heterogeneity of the area, he is almost completely isolated from socio-spatial networks with Israelis, and he very rarely develops socio-spatial networks with migrant workers from different origins (Table 7). As an illegal and temporary resident he avoids unnecessary movements out of home as much as possible, and he spends long hours at work. He cleans two Israeli homes a day, five days a week, but he refuses to perceive his employers as meaningful others expressing the transparency of his relations with Israelis. Accordingly, the house receives the highest weight due to its high salience while the rest of the city as the work place, receives also a relatively high weight. Friends from his country of origin became his ultimate support group while shame to expose his difficulties to his family in the home country, and the high cost of telecommunication prevent him from maintaining extensive telecommunication networks. The resulting pattern of segregation for the studied migrant worker is high isolation in the social dimension (SO=+3.1) while he lives in a mixed surrounding with SP=-0.8. The overall segregation in this case is therefore, SPS=+1.2. The major factor of the migrant worker’s isolation is his lack of friends and other networks with Israelis, including at work when relations with Israelis remain transparent.

Table 7 : The Pattern of a migrant worker’s segregation

Note : The nominator describes the number of members of the group.The denominator describes the total number in the studied unit.


41We have proposed an alternative index of segregation, which measures socio-spatial isolation from mixture, or exposure to other groups as a multi-dimensional measurement. The two main dimensions measure either the social or the spatial context of seven aspects of agents’ everyday life actions as they are actually performed in their everyday time-spaces. We believe that our more complex approach, which relates to the contribution of different aspects of agents’ everyday life practices in actual everyday life time-spaces, better represent current segregated life conditions, than the traditional model, which assumes that exclusive groups are contained within exclusive residential territories defined by relatively closed boundaries.

42The seven different lifestyles presented in the article represent different forms of using everyday life spaces. These differences are translated to weighing coefficients, which determine the impact of each spatial context on the index of segregation calculated for each dweller. Thus, we transform the index of segregation from measuring a person's proximity to other members of her group in residential Newtonian space, to measuring a person's probability for actual encounters with either members of her group or others in the actual spaces of ones everyday life. Everyday life spaces may be discontinuous, invested with various human meanings and of different size and structure for each agent. Segregation is measured in each person’s relevant space. This means that segregation is determined not only by ones choices for location in social space, but also by defining ones orbit of activity spaces.

43Figure 3 enables a comparison among the representatives of the seven lifestyles described in the paper. The graphs express the initial proportions among the groups’ populations and alternative groups by the length of the concentric lines. The longer the line the more the group is socio-spatially exposed to others, the shorter the line the more the group is isolated from alternative groups. The width of the flower at the centre of the graph represents the weights assigned to each of the spheres of activities-at the upper part of the graph-and territorial bases-at the lower part of the graph. The circle describes the overall degree of segregation in comparison to the dotted circle, which expresses the segregation index of zero. The seven examples are ordered according to their overall segregation indexes, with the ghetto representing the most segregated model both in the spatial and the social dimensions and the exposed person representing a high degree of exposure in both dimensions.

44A comparison among the seven lifestyles, assuming that they present significant lifestyles with wider validity, shows that information from either the social or the spatial dimensions is not redundant to each other. Certain empirical evidences may support our argument. Empirical investigations on migrant workers show that migrant workers who are spatially concentrated, may evoke antagonism by the majority and may reduce the migrant need for inter-ethnic relations (Esser, 1986). Other researches show that spatial aggregation has no impact on the intensity of inter-group encounters (Nauck, 1988 ; Alpheis, 1990).

45Furthermore, variances among the lifestyles are significantly higher in the spatial dimension. These results emphasise the limitation of the use of one number as an expression of segregation and the added value of a description that is based on a vector that describes the components of the segregation on the seven spatial and the social dimensions. On the one hand, the isolated elderly is highly exposed to alternative groups in the spatial dimension, but from the social dimension we learn that she rarely communicates with members of alternative groups who surround her in the spatial dimension. On the other hand, the localist is segregated in the spatial dimension but this spatial isolation does not block him from developing social networks with members of alternative groups out of his home territory. Furthermore, if these seven ideal types have larger validity, it seems that agents in Western societies tend to be significantly more segregated socially than spatially. In the two lifestyles of isolated elderly and migrant workers, a clear tendency toward spatial exposure was not followed by social exposure to members of alternative groups (values lower than -2.0). On the contrary, both cases show a high tendency towards isolation from alternative groups (values higher than +2.0).

Figure 3 : Segregation graphs

Figure 3 : Segregation graphs

Figure 4 : Relations between social and spatial segregation

46The graphs may also distinguish between the more localist lifestyles of the ghetto and the localist, which are performed around the home and the cluster. In contrast, the rest represent much more mobile lifestyles in space, practising large parts of the metropolitan area as their everyday life action space. The cosmopolitan and the migrant worker exemplify the discontinuity of their everyday life space by assigning extremely low relevance to the cluster and the neighbourhood, while emphasising the salience of the house on the one hand and the rest of the city on the other. A further challenge will be to test the possibility to identify groups with similar graphs or lifestyles and to distinguish among groups with different lifestyles. In addition a further challenge is to create segregation maps for each group.

Haut de page


Alpies, H. 1990 "Erschwert die etische kontzentration die eingleiderung ?" in Esser, H., Friedriech, J. (Eds.) Generations und identitat, Opladen : Westdeutscher Verlag, Pp. 147-184.

Altman, I. 1975, The Environment and Social Behavior : Privacy, Personal Space and Crowding, Brooks/Cole : Monteray, Cal.

Bachelard, G. 1969, The poetics of Space, Beacon Press : Boston.

Boal, F. W. 1987, "Segregation", Peach, C. (Ed.), Social Geography, Progress and Prospect, Croom Helm, Beckenham : New York.

Buttimer, A. 1981, "Social Space and the Planning of Residential Areas", in Buttimer, A. and Seamon, D. (Eds.) The Experience of Space and Place, St. Martin’s Press : New York.

Cooper, C. 1975, Easter Hill Village, Free Press : New York.

Duncen, D. and Duncan, B. 1955. "A Methodological Analysis of Segregation Indexes", American Sociological Review, 20, 210-217.

Esser, H. 1986 "Ethnische Kolonien : Binnenintegration oder Gesellschaftliche Isolation ?" in Hoffmeyer-Zlotnik (Eds.) Segregation und Integration, Berlin Pp. 106-117.

Giddens, A. 1984, The Constitution of Society : Outlines of the Theory of Structuration, Polity Press : Cambridge.

Goldhaber, R. ‏‏‏1998, The Relevance of the Neighbourhood in Post- Modern Tel-Aviv-Jaffa, M.A. Thesis, Geography Department, Tel Aviv University (Hebrew).

Gonen, A. 1995 Between City and Suburb, Avebury :Aldershot.

Gonen, A., and Hadas, E. 1994 Jews and Arabs in a Mixrd Neighbourhood, The Floersheimer Institute for Policy Studies : Jerusalem,

Hall, E. T. 1966, The Hidden Dimension, Doubledy and Company : New York.

Jackson, 1994, Construction of Place, Race and Nation, University of Minnesota Press : Minneapolis.

Liberson, S. 1981, "An Asymmetrical Approach to Segregation", in Peach, C., Robinson, V. and Smith, S. (eds.), Ethnic Segregation in Cities, Croom Helm : New York, 61-82.

Morgan, B. S. 1983, "A Distance-Decay Interaction Index to measure Residential Segregation, Area, 12, 211-216.

Morrill, R. L. 1991, "On the Measure of Geographical Segregation", Geography Research Forum, 11, 25-36.

Nauck, B. 1988 "Sozial-okologischer knotex und auserfamiliare beziechungen. Ein intercultureller und interkontextueller vergleich am beispiel von Deutschen und Turkischen familien", Kolner Zeitschrift fur soziologie und sozialpsychologie Sonderheft, 29, 310-327.

Omer, I. 1997, Ethnic Groups’ Residential Segregation in Urban Space as a Structuration Process, Ph.D. Dissertation, Tel Aviv University : Tel Aviv.

Racine, J. B. and Mager, C. 1997, The Foreigner and the City, Institute Geographie, (Unpublished Paper).

Rose, D. 1984 "Rethinking Gentrification beyond uneven development in Marxist urban theory", Society and Space, 2, 47-74.

Rowles, G. D. 1981, "Toward Geography of Growing Old", in Buttimer, A. and Seamon, D. (Eds.), The Experience of Space and Place, St. Martin Press : New York.

Saarinen, T. F. 1976, Environmental Planning, Miflin Co. Boston.

Schnell, I., Graicer, I. 1994 Back to Tel-Aviv, The Center for Urban and Regional Studies, Technion : Haifa.

Schnell, I. 1995 "Arab Ghetto in Jaffa", Unpublished manuscript : Tel Aviv university.

Schnell, I. 1998, Migrant Workers in Tel Aviv, Floersheimer Institute for Policy Studies : Jerusalem (Hebrew).

Short, J.R. 1989 "Yuppies, Yuffies, and the New Urban Order", Trans. Br. Geogr. 14, Pp. 173-188.

Sibley, D. 1995 Geographies of Exclusion, Routledge : London.

Sommer, R.1969, Personal Space, Prentice-Hall : Englewood, Cliff. New Jersey.

Tukey, J. W. 1977, Exploratory Data Analysis, Addison-Wesley : Massachusetts.

Waldinger, R. 1993 "The Ethnic Enclave Debate Revisited", Events and Debates, 444-452.

Waldorf, B. S. 1993, "Segregation in Urban Space : A New Measurment Approach", Urban Studies, 30, 7, 1170-1192.

Wong, D. W. 1998, "Measuring Multiethnic Spatial Segregation", Urban Geography, 19, 1, 77-87.

Haut de page

Table des illustrations

Fichier image/png, 9,4k
Fichier image/png, 9,9k
Fichier image/png, 790 octets
Fichier image/png, 501 octets
Fichier image/png, 440 octets
Fichier image/png, 669 octets
Fichier image/png, 571 octets
Fichier image/jpeg, 40k
Fichier image/jpeg, 36k
Fichier image/jpeg, 36k
Fichier image/jpeg, 40k
Fichier image/jpeg, 36k
Fichier image/jpeg, 36k
Fichier image/jpeg, 40k
Fichier image/jpeg, 48k
Titre Figure 3 : Segregation graphs
Fichier image/png, 3,6k
Haut de page

Pour citer cet article

Référence électronique

Schnell Izhak et Benjamini Yoav, « Socio-spatial lifestyles and segregation  », Cybergeo : European Journal of Geography [En ligne], Espace, Société, Territoire, document 94, mis en ligne le 27 avril 1999, consulté le 29 novembre 2020. URL : ; DOI :

Haut de page


Schnell Izhak

Geography Department, Tel Aviv University, Tel Aviv, P.O.B 39040, Israel
Fax : 972 3 6406243

Benjamini Yoav

Department of Statistics and OP, Tel Aviv University, Tel Aviv, P.O.B 39040, Israel
Fax : 972 3 6406243

Haut de page

Droits d’auteur

Licence Creative Commons
La revue Cybergeo est mise à disposition selon les termes de la Licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 3.0 non transposé.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search