Neighbourhoods and Municipalities as Contextual Opportunities for Interethnic Contact

Abstract

This article investigates consequences of spatial contexts on interethnic contact. Despite the acknowledged integrative effects of pleasant interethnic relationships, several unresolved issues remain in this research field: investigations at two contextual levels simultaneously—i.e. neighbourhood and municipality levels; investigations of several contextual characteristics simultaneously, e.g. ethnic concentration, physical contact opportunities, population size; investigations on different kinds of interethnic contact, for example, contact with neighbours, with friends or in general. The present study contributes to these issues by analysing interethnic contact from a native’s perspective using a German nation-wide dataset. A considerably high proportion of Germans (72 per cent) have contact with foreigners in at least one out of four measured types. Ethnic concentration is the strongest contextual predictor for all kinds of interethnic contact. Physical contact opportunities in the immediate neighbourhood foster interethnic contact in the neighbourhood only, while municipality size mostly diminishes interethnic contact.

1. Introduction

This article investigates consequences of different levels of spatial context on interethnic contact. Interethnic contact is not only a key aspect of integrating immigrants in the host society (Esser, 1986; Gijsberts and Dagevos, 2007) but an important asset for societal integration in general, as bridging ties can create valuable social capital (Briggs, 2007; Woolcock, 1998). Interethnic relationships are valuable for social life where they foster tolerant attitudes and mutual acceptance, expand social identities and resolve intergroup conflicts (Farwick, 2009; Hewstone, 2009; Pettigrew and Tropp, 2006).

Given the largely acknowledged positive impacts of interethnic contact, opportunities promoting this type of social interactions should be investigated. In social theory, contextual opportunities are substantial ‘supply-side’ factors for meaningful social contact (Blau, 1994; Marsden, 1990; Verbrugge, 1977). Studies that investigate contextual opportunity are mainly concerned with effects of ethnic concentration. However, the empirical evidence from studies especially on Germany is mixed: while some show a proportional relation between ethnic concentration and interethnic contact at the neighbourhood level (Oberwittler, 2007; Wagner et al., 2003), others found no significant effect (Esser, 1986; Drever, 2008). Additionally, there are contradictory effects on interethnic contact at neighbourhood and municipality levels (Briggs, 2007). Contradictions between strong theoretical statements and mixed empirical evidence of previous research indicate several unresolved issues in this research field.

First, even if recent research involves contextual opportunities, studies that include sub-structures and macro structures simultaneously are rare. Secondly, although recent studies include ethnic concentration as a primary and important contextual characteristic, contact opportunities are not restricted to ethnic concentration but should relate to other socio-spatial characteristics as well. Thirdly, interethnic contact is often measured as exposure in friendships (Briggs, 2007) but there is little knowledge on different kinds or different levels of contact—i.e. to friends, family members, neighbours or colleagues.

The study presented here contributes to the literature by analysing the interethnic contact of Germans with foreigners differentiating contact in general, in the neighbourhood and with friends and acquaintances, by testing hypotheses at the neighbourhood and municipality levels derived from macro-sociological and socio-spatial theories. While the contact of immigrants or ethnic minorities with native populations is well researched (Babka von Gostomski and Stichs, 2008; Drever, 2004; Fong and Isajiw, 2000; Vervoort et al., 2011), little is known about members of host societies. This study focuses on the natives’ perspective because interethnic contact is determined by the willingness to integrate and the behaviour of the host population (Esser, 1980). Analyses in this study are done with a nation-wide dataset. This enables the comparison of many different localities with widely varying contextual characteristics. Hence, more general conclusions on contextual opportunities on interethnic contact can be drawn.

This article is organised in five sections. The theoretical background of contact processes from which several context hypotheses are deduced is outlined in section 2. In section 3, the dataset is briefly introduced: a German general social survey conducted in 2006. Analyses and hypotheses tests are described in section 4. Finally, results are discussed and some conclusions are drawn in the last section.

2. Theoretical Background and Hypotheses

2.1 Contextual Opportunities for contact

Social contacts are complex products of contextual conditions, individual characteristics and personal preferences (Mardsen, 1990; Verbrugge, 1977). Verbrugge (1977) suggests a two-stage process of ‘meeting and mating’ assuming that contacts are more likely under conditions of spatial proximity and status similarity (meeting) and that some contacts progress to stable personal ties if spatial proximity and frequent interaction continue and are accompanied by mutual attraction (mating).

Meeting processes are related to Feld’s focus theory (1981). He argues that contact is often an unintended consequence of everyday activities within joined foci like living in the same neighbourhood, encountering in public spaces or being at the same site (workplaces, sites of recreation, theatre etc.). It is not only that foci bring previous strangers together into contact but rather create “positive sentiments indirectly through the generation of positively valued interaction” (Feld, 1981, p. 1017). On these grounds, mating processes base on homophily. Lazarsfeld and Merton (1954) found that value and attitude homophily and mutual perceptions determine friendship choices. More recent research indicates that behaviour and activity similarities are more relevant (McPherson et al., 2001).

It can be concluded that people who see each other regularly, because their paths cross in their respective daily routines, becoming acquainted is opportunity-driven, while actively sought friendly contacts as well as the avoidance of unwanted contacts are preference-driven. Obviously, both components are needed but it is claimed here that opportunity structures gain relative weight for initial encounters (Gans, 1961).

It can be followed from such processes that interethnic contact is easily established if opportunities for contact and no obstructions like social distance or value divergences exist (Esser, 1986). Interethnic contact in the broadest sense means any kind of interaction between people from different groups that are defined according to ethnicity, nationality or migration background.

Macro-structures and sub-structures of opportunity are important contextual conditions for interethnic contact, as theorised in the structural theory by Blau (1994) that extends Simmel’s concept of intersecting social circles. Opportunity structures can be related to different levels of socio-spatial context—for example, neighbourhoods as sub-structures are nested within municipalities as macro structures. Investigations in neighbourhoods are particularly interesting because neighbourhoods have the necessary proximity out of which interethnic contacts evolve. Context effects on interethnic contact might be most pronounced at the neighbourhood level because it is the small-area geographical unit where people presumably spend most of their leisure time (Tolsma et al., 2009).

However, higher-level contexts like municipalities offer contact opportunities as well. To be sure, social life is not restricted to one’s neighbourhood. Frequently visited sites—for example, workplaces, friends’ homes, churches, recreational areas, clubs—may scatter across a city. Moreover, it is assumed that neighbourhoods are especially important for fleeting encounters and casual interethnic contacts but municipalities are in particular significant for meaningful relationships. Finally, according to Blau (1994, p. 5), sub-structures promote intergroup contact but higher-level contexts may entail segregation that limits this contact. Therefore, I investigate contextual effects at the neighbourhood and municipality levels.

2.2 Hypotheses on the Neighbourhood Level

In accordance with structural theories, neighbourhoods embody in a way an important ‘market of possibilities’ to join in common activities (Galster, 2008, p. 12). A neighbourhood is not only a relevant site for a number of activities but also one of the most important starting-points for social intergroup contact. In the words of Verbrugge (1977, p. 577), “people whose daily rounds intersect are more likely to become acquaintances than others”. The neighbourhood determines interethnic contacts in two ways: a ‘quarter effect’ derived from the physical structure and a ‘concentration effect’ derived from the population structure (Friedrichs and Blasius, 2000).

The quarter effect primarily takes various behavioural opportunities and restrictions into account that come from the material-physical and architectonic characteristics of neighbourhoods. In this respect, public spaces and transitional zones between the private and public spheres are of particular interest. Especially, transitional spaces function as links between the protected private space and the uncertain public spaces of the wider urban area. These transitional zones can help to bridge the boundaries between the self and the other (Ellsworth, 2005). Examples are stairways of residential buildings, gardens, porches, courtyards and the small stretch of road in front of a house (Gehl, 1987; Hoogland, 2000).

Dines and Cattell (2006) identified preconditions for social contact in public spaces, including familiarity, regular use and available facilities. The mechanism behind the effect of transitional spaces on interethnic contact is that familiarity, regular use and available facilities give purpose to a space and enhance its social vitality. Social vitality breeds contact or as Goffman (1963, p. 17) puts it, encounters with others happen in public spaces when there is a ‘gathering’ and people conceive that they are close enough to see and be seen. It can be assumed that familiarity and regular use are given for transitional spaces of people’s residential buildings and their immediate neighbourhood, but social vitality can only be found in densely populated neighbourhoods of predominantly multiple family dwellings. Hanhörster and Mölder (2000, p. 392) point to the fact that opportunities for contact in transitional spaces of residential buildings and nearby green spaces stimulate interethnic contact. Based on this quarter effect of neighbourhoods, hypothesis (H1) follows

H1: The more contact opportunities transitional spaces in the immediate neighbourhood have the more interethnic contact.

The concentration effect points to the relevance of the population structure. Opportunities and restrictions for interethnic contact arise from different proportions of ethnic populations in spatial contexts and in neighbourhoods in particular. Ethnic concentration affects interethnic contact by at least two mechanisms. First, higher proportions of the ethnically defined out-group let opportunities to contact out-groupers grow. If opportunities are taken up by residents of the neighbourhood, interethnic contact will increase. This is not a sociological hypotheses but just a ‘mathematical truism’ (Blau, 1994, p. 30). The higher the concentration of ethnic others in the neighbourhood, the more interethnic contact can be expected.

Secondly, people adjust their identifications and orientations according to the context (Esser, 1999, p. 452; Galster, 2008, p. 11). People are aware of the others’ behaviour just by sharing the same context (Lofland, 1973). Socialisation and learning are a second mechanism that mediates between ethnic concentration and interethnic contact. Surrounded by visible and audible ethnic diversity, locals experience the everyday behaviours, manners and habits of strangers. Even if somebody does not interact directly, ethnic concentration helps orientating oneself by observing ethnic others. Social norms of mutual dealings can develop from these experiences. Furthermore, everyday observations and experiences of interethnic contacts in an ethnically diverse environment enhance adapting of behavioural roles. If these orientations, norms and roles contribute to a positive evaluation of peaceful co-existence of diverse groups in the neighbourhood, interethnic contact will be stimulated.

However, co-ordination problems such as language barriers might restrict interethnic contact. Furthermore, everyday observations and experiences may result in negative evaluations resulting in withdrawal from interethnic contact. In addition to it, conflict theories presume that competition between ethnic groups over resources and representation leads to negative attitudes towards other groups and reduces mutual trustworthiness (Blalock, 1967; Quillian, 1995; Stephan and Renfro, 2002).

According to Putnam’s constrict theory (2007) it is ethnic diversity rather than ethnic concentration that causes residents to ‘hunker down’. The mix of foreign groups evokes feelings of insecurity and perceived threat that result in less social contact. Unfortunately, testing this diversity argument is not possible with the data at hand.

Own interethnic contact will be more likely if interethnic encounters are experienced as something positive, or at least as nothing negative (Kecskes, 2003; Farwick, 2009). If people from different socio-cultural backgrounds live together in close proximity, they will begin to know and appreciate each other as individuals, thus contributing to the elimination of stereotypes and prejudices. In sum, ethnically diverse neighbourhoods raise the chances for interethnic contact. A pronounced ethnic heterogeneity and a positive and friendly relationship between in- and out-groupers in the neighbourhood are opportunities for interethnic contacts. This leads to hypothesis (H2)

H2: The higher the ethnic concentration of a neighbourhood is and the more positive the relationship between in- and out-group members in the neighbourhood is perceived, the more interethnic contact.

Some authors argue that processes of social change like individualisation as well as technological innovations weaken the relationship between space and social interactions (Zelinsky and Lee, 1998; Castells, 2000). It should not be expected that people share a large part of their social life with neighbours and locals. Social interactions are not bound to the neighbourhood but are geographically more dispersed (Wellman, 1996).

Hence, interethnic contact in the neighbourhood is directly influenced by contextual opportunities of transitional spaces and ethnic concentration in the neighbourhood, but more general forms of interethnic contact that mostly take place outside the neighbourhood are only indirect and weakly affected by neighbourhood characteristics. Moreover, interethnic contact in the neighbourhood is only a minor part of all interethnic interactions among family, friends, colleagues and acquaintances. It can be expected that contextual opportunities at the neighbourhood level (as stated in H1 and H2) are rather influential to neighbourly interethnic contact than to more general forms of interethnic contact.

2.3 Hypotheses on the Municipality Level

Because of the mathematical truism, the effect of ethnic concentration can be expected for municipalities too. Higher concentrations of out-group members in municipalities provide higher chances for interethnic contacts as well as higher chances to observe and experience interethnic contact by others. Hypothesis (H3) follows from this

H3: The higher the ethnic concentration in a municipality, the more interethnic contact.

According to Blau (1994), neighbourhoods and municipalities develop separate and divergent effects on interethnic contact. Because municipalities vary largely in size ranging from the smallest villages to the biggest metropolitan cities, one of these divergent effects is related to the population size of municipalities. There are two arguments for negative impacts of municipal population size on interethnic contact: more freedom of choice for certain contacts and segregation effects.

First, it is obvious that options for contacts increase with population size: there are more potential interaction partners in larger than smaller municipalities. Weber already highlights that trivial characteristic of cities: “the reciprocal personal acquaintance of the inhabitants, elsewhere characteristic of the neighbourhood, is lacking” (Weber, 2008, p. 1212). Residents of larger municipalities have more options to choose their contacts—i.e. contact processes shift from opportunity-driven to preference-driven (Fischer, 1976, p. 118)—and people prefer intraethnic over interethnic contact (McPherson et al., 2001); hence, interethnic contact will decrease in larger municipalities.

Secondly, macro structures entail segregation that limits interethnic contact (Blau, 1994, p. 5). Both, ethnic concentration and ethnic segregation are macro-structural concepts that result from socio-spatial migration processes but affect interethnic contact in divergent ways. If in- and out-group members live segregated in a municipality—i.e. in different neighbourhoods—then fewer opportunities for intergroup contact exist (Friedrichs, 2008, p. 384; Wilson, 1987). Hence, interethnic contact will decrease in segregated municipalities. It is empirically proven that segregation is higher in municipalities with larger populations (Janßen and Schroedter, 2007, p. 468).

The two arguments will not follow a linear relationship. The freedom of choice argument should affect all municipalities larger than villages. The segregeation argument should only affect bigger cities. Both mechanisms together will result in a more linear relationship between municipal population size and interethnic contact. The final hypothesis (H4) is formulated on the basis of these arguments

H4: The more inhabitants live in a municipality, the less interethnic contact.

3. Data

The German General Social Survey 2006 (ALLBUS) is used to empirically test the hypotheses. The sample is based on the adult population in Germany (Wasmer et al., 2007). Data are generated by face-to-face interviewing. The following analyses are based on the sub-sample of respondents with German citizenship. The disproportionality of East and West German sub-samples is compensated for by a design weight. In the following, operationalisations of the variables (descriptive statistics in Table 1) and the analysis strategy are presented.

Table 1.

Descriptive statistics

Variables^a	Minimum	Maximum	Mean	S.D.	n
Contact with foreigners (dependent variables)
Contact in the neighbourhood	0	1	0.381		3151
Contact with friends and acquaintances	0	1	0.488		3154
Any contact with foreigners	0	1	0.719		3102
Neighbourhood context (independent variables)
Type of residential building	1	7	3.325	1.538	3115
Perceived foreigners in the home environment	1	4	1.740	0.737	3142
Relation between Germans and foreigners in the home environment
Be on good termsNeutral or no foreignersBe on bad terms	000	111	0.1750.7730.052		305930593059
Municipal context (independent variables)
Share of foreigners at county level, grouped	1	23	7.862	5.039	3156
Municipal population in 1000s	0.775	957.015	152.148	299.370	3156
Grouped municipal population
0–4999 inhabitants (villages)5000–19,999 inhabitants (small towns)20,000–99,999 inhabitants (mid-sized towns)100,000 and more inhabitants (big cities)	0000	1111	0.1930.2860.2670.254		3156315631563156
Socio-structural background (control variables)
Migration background (1 = with)	0	1	0.182		3156
Educational achievement
Lower educational levelMedium educational levelHigher educational level	000	111	0.4090.3340.257		311531153115
Professional qualification
Lower professional qualificationMedium professional qualificationHigher professional qualification	000	111	0.1870.6020.210		314531453145
Employment status
Pupils and students(Self-)employed, full-time(Self-)employed, part-timeUnemployedNon-workingRetired/pensioned	000000	111111	0.0400.3990.0860.0630.1160.295		315131513151315131513151
Income in 1000 Euro	0	12	2.185	1.291	2500
Gender (1 = female)	0	1	0.520		3156
Age	18	94	50.056	17.319	3148
Household size	1	9	2.469	1.244	3129

The sample includes people aged 18 or older holding German citizenship. Data are weighted according to the disproportional sample for East and West Germany.

3.1 Dependent Variables

Interethnic contact is operationalised as exposure to contact of Germans with foreigners. The ALLBUS 2006 distinguishes four types of contact that can be interpreted as foci (Feld, 1981). The reported personal contact ranges from 26 per cent in the focus of family or close kinship, to 38 per cent in the neighbourhood and 43 per cent at the workplace, to 49 per cent in the circle of friends and acquaintances. The proportion of respondents with interethnic contact in at least one focus substantially rises to almost 72 per cent.

Note that ALLBUS asks about exposure to interethnic contact. No information is available on aspects like frequency (certain temporal continuity), quality (positive, neutral or negative evaluations) or bonding (strong or weak). Because previous research showed that interethnic contact varies greatly along those lines (Petermann, 2011; Reuband, 1989; Smith, 1999), the contact measure can have different interpretations. Given the reported high proportion of cross-group contact, it is reasonable that the measure sets a low and hence more inclusive threshold on what respondents report as interethnic contact. The contact measure mainly involves fleeting encounters, sporadic contacts or weak network ties. Moreover, these should be predominantly friendly contacts since Petermann and Schönwälder (2012) showed that interethnic contact is overwhelmingly positively evaluated—only a tiny minority of 1 per cent perceive these conversations as unpleasant.

Identifying foreigners as contact partners is another issue. Given the nature of fleeting contact or weak ties, respondents might classify people according to perceivable characteristics—for example, appearance, dress code, language—rather than nationality. This practice identifies people with migration background rather than foreigners and may lead to a more precise description of contact between different ethnic groups.

3.2 Independent Variables

The number of households in the residential building of the respondent is used as indicator for the contact opportunities in transitional spaces to test hypothesis H1. This measure ranges from a farm building to a tower block with more than nine floors. The seven categories are scaled such that higher values represent more contact opportunities. The average of this scale is 3.3, meaning that average respondents live in terraced houses or residential dwellings with up to four homes.

Hypothesis H2 is tested with two variables. The subjective perception of foreign residents in the home environment is taken as a measure of ethnic concentration. The scale ranges from 1 (hardly any foreigners in the home environment) to 4 (mostly foreigners in the home environment) and reaches an average of 1.7, meaning that a vast majority perceive only a few foreigners in their neighbourhoods. The subjective perception might not measure the neighbourhood effect of ethnic concentration, but rather individual knowledge of it. However, I am confident that even the subjective perceptions are good enough to measure ethnic concentration since evidence from a survey of 50 German neighbourhoods (Petermann et al., 2012) shows that objective ethnic concentration and the aggregated subjective perception of it are highly related to each other (r = 0.81). Another issue of a subjective measure of ethnic concentration is that it may be influenced by contact with foreigners, since the fact that if someone has contact with foreign co-residents it might make this person more aware of foreigners in the neighbourhood. If so, people with contact with foreigners should overestimate the share of foreigners in their neighbourhoods compared with people without contact with foreign co-residents of the same neighbourhood. Yet in 76 per cent of the neighbourhoods under study, residents acquainted with foreigners do not perceive significantly higher shares of foreigners than co-residents without such contacts. Hence, it is reasonable to assume that perceived contact opportunity fosters contact and not vice versa. If at least some foreigners live in the home environment, respondents should evaluate the local relationship between Germans and foreigners. On a three-point scale with positive, neutral (including respondents with hardly any foreigners in the home environment) and negative categories, a vast majority (77 per cent) is assigned to the middle position. As hypothesis H2 states that ethnic concentration is moderated by the evaluative variable, both measures form interaction terms in the analyses. However, I did not include the main effects of relations between Germans and foreigners in the home environment because I am interested in moderating or conditional effects on the association between perceived foreigners and interethnic contact and not in the main effects of relations between Germans and foreigners on interethnic contact. In this simpler model, interpretations of main and conditional effects of perceived foreigners are much easier. While the main effect is in this case the effect for people evaluating the relation between Germans and foreigners in their home environment as neutral, interaction effects indicate differences in slopes between neutral and positive/negative evaluated neighbourhoods.

To test hypothesis H3, ethnic concentration is measured as the percentage of foreigners in the county to which a municipality belongs.¹ The data were available from public statistics and have been matched with the survey data. Percentage of foreigners varies between 1 per cent and 23 per cent around an average of just under 8 per cent.²

The number of inhabitants is used to test hypothesis H4. In order to estimate a linear effect of the population size, the available indicator with seven categories is weighted by the average population size per category using official statistics (Statistisches Bundesamt, 2007, p. 40). The average municipality has 152,148 inhabitants and a standard deviation that is almost twice as high as the average.

Individual-level explanations should be taken into account to control for compositional effects as well as to avoid the risk of overestimating contextual deteminants. I employ socioeconomic (highest achieved graduation of both education and occupational training, employment status, household income) as well as socio-demographic (migration background, gender, age, household size) variables reflecting individual living conditions.

3.3 Data Analysis

In order to test the hypotheses, a stepwise strategy was employed. First, the impact of locality characteristics on contact to foreigners in the neighbourhood was investigated because this type of interethnic contact should most obviously be influenced by spatial contexts at least by neighbourhood traits. Analysis begins with a base model that includes control variables only. Next follows a context model by adding locality characteristics to the base model. The total impact of spatial contexts can therefore be estimated. Secondly, context models of contact with foreigners among friends and acquaintances as well as any contact with foreigners were analysed. Assuming that these types of interethnic contact are spatially more spread, neighbourhood characteristics should be less influential. Finally, some refinements on the measurement of municipality size are made accounting for non-linear effects on interethnic contact.

I employ multilevel analyses to take into account the nesting data structure (Snijders and Bosker, 1999). The survey design has a hierarchical two-level structure with respondents at the first level nested within 161 municipalities at the second level. The intraclass correlation ρ is 0.24 for contact with foreigners in the neighbourhood indicating substantial variance between municipalities that makes multilevel modelling necessary. Variables at the first level are all characteristics that vary between respondents. These are the dependent variables of interethnic contact, independent individual control variables as well as independent contextual variables of the neighbourhood. Note that respondents are randomly selected within municipalities, hence respondents are not clustered at the neighbourhood level but belong to different neighbourhoods. Variables at the second level are the independent contextual variables of the municipalities. Due to the dichotomous dependent variables, multilevel logit regression models were estimated.

4. Results

The base model on contact with foreigners in the neighbourhood shows that control variables reflect just a minor part of the variance, although several variables have strong effects (base model in Table 2).

Table 2.

Multilevel logit regressions on different kinds of contact with foreigners

variables	Contact in the neighbourhood(base model)		Contact in the neighbourhood (contextmodel 1)			Contact with friends (context model 2)		Contact in general (contextmodel 3)
	Coefficient	S.E.	Coefficient	S.E.	Predicted contact chance^b(percentages)	Coefficient	S.E.	Coefficient	S.E.
Fixed part
Intercept	−0.384	0.141	−0.368	0.130	40.9	0.181	0.122	1.888	0.160
Social background (individual variables at the 1st level)
Migration background (1 = with)	0.352	0.115	0.102	0.123	43.4	0.572	0.118	0.384	0.149
Education (ref. = medium education)
Lower educationHigher education	0.050−0.049	0.1120.139	−0.083−0.028	0.1190.145	38.940.2	−0.1740.318	0.1100.138	0.0170.059	0.1320.176
Professional qualification (ref. = medium prof. qual.)
Lower professional qualificationHigher professional qualification	−0.0530.188	0.1300.139	−0.290 0.167	0.1400.145	34.145.0	−0.2520.023	0.1330.137	−0.428 0.380	0.1510.173
Employment status (ref. = self-employed, full-time)
Pupils and students(Self-)employed, part-timeUnemployedNon-workingRetired/pensioned	0.188 −0.360 −0.161−0.129 −0.539	0.2880.1700.1980.1560.158	0.272 −0.451 −0.393−0.240 −0.472	0.3010.1790.2130.1670.167	47.630.631.835.230.1	0.348−0.143−0.261−0.181−0.255	0.3280.1660.1910.1560.154	0.452−0.271 −1.081 −0.880 −0.815	0.4820.2240.2210.1910.182
Income in 1000 Euro^a	0.018	0.044	0.093	0.047	43.8	0.073	0.046	0.128	0.062
Income not known	−0.015	0.117	0.026	0.121	41.5	−0.222	0.115	−0.167	0.138
Gender (1 = female)	0.029	0.099	0.074	0.104	42.7	−0.224	0.098	−0.085	0.116
Age^a	−0.008	0.004	−0.006	0.004	38.6	−0.038	0.004	−0.036	0.005
Household size^a	0.071	0.041	0.133	0.044	44.9	0.002	0.042	0.083	0.055
Neighbourhood (context variables at the 1st level)
Type of residential building^a			0.120	0.038	43.8	−0.052	0.036	0.002	0.043
Perceived foreigners in the home environment^a			1.333	0.092	64.9	0.450	0.082	0.898	0.111
Interaction of perceived foreigners x relations between Germans and foreigners in the home environment (ref. = neutral or no foreigners)
Be on good termsBe on bad terms			0.231 −0.725	0.2170.195	66.460.7	0.121 −0.723	0.1840.196	0.286 −1.056	0.3030.237
Municipality (context variables at the 2nd level)
Share of foreigners at county level^a			0.094	0.017	52.7	0.110	0.016	0.145	0.022
Municipal population in 1000 and grouped^a			−0.001	0.000	36.4	−0.001	0.000	−0.001	0.000
Random part
S.D. of intercept	1.040	0.085	0.566	0.073		0.557	0.065	0.769	0.083
Model statistics
N (people/municipalities)	2,988 / 161		2,988 / 161			2,989 /161		2,943 / 161
LR test (d.f.)	87.48 (15)		540.79 (21)			496.77 (21)		562.49 (21)
McFadden pseudo R²	0.024		0.149			0.127		0,180
BIC	3,670.150		3,264.854			3,588.140		2,742.921

Variables are centred on its mean values.

This is the estimated probability of contact with foreigners in the neighbourhood if dichotomous variables change from 0 to 1 and if metrical variables (marked with ^a) change by a standard deviation.

Note: Bold printed coefficients are significant at the 0.05 level.

In the next step, the base model was extended by characteristics of spatial contexts to test the hypotheses (context model 1 in Table 2). The context model substantially improves the base model as McFadden pseudo R² rises from 0.02 to 0.15. The estimated contact probabilities are additionally computed to compare the effects of the independent variables.

There is a proportional relationship between contact opportunities in transitional zones of one’s neighbourhood and contact with foreigners in the neighbourhood that confirms hypothesis H1. The analysis indicates that the contact probability increases the more apartment units a residential building has. The material-physical and architectonic characteristics of neighbourhoods shape interethnic contact in the neighbourhood such that densely inhabited neighbourhoods breed more interethnic contact (see Figure 1a). This relationship exists even under control of the ethnic make-up of the neighbourhood—i.e. it cannot be due to the fact that shares of foreigners are lower in areas with predominantly single-family houses.

Figure 1.

Neighbourhood-level effects on contact with foreigners in the neighbourhood (context model 1 in Table 2).

Hypothesis H2 on the effect of ethnic concentration is tested by the perception of foreigners in the neighbourhood and interaction terms of this perception with the relationship between Germans and foreigners in the neighbourhood. The proportional effect is the strongest in the context model leading to a contact probability of almost 65 per cent by a rise of the independent variable’s standard deviation. Figure 1 (solid lines) demonstrates this strong effect by comparing both neighbourhood effects. The distinct impact of ethnic concentration in the neighbourhood is also confirmed by the interaction effects of ethnic concentration and the relationship between Germans and foreigners. If this relationship is on good terms, the effect of ethnic concentration is even larger but not significantly distinct from people with a neutral evaluation. On the other hand, if people assess a conflictual relationship between Germans and foreigners that has a diminishing effect of ethnic concentration. Yet it is interesting to see that even under bad circumstances the probability of interethnic contact increases with higher levels of ethnic concentration but just to a smaller extent (dotted line of Figure 1b). That means that, even if people are aware of frictions between co-residents of national versus immigrant origin, they do not withdraw from neighbourly intergroup interactions. Perceived ethnic concentration and its interaction effects strikingly confirm hypothesis H2.

The strong correlation may result from a self-selection bias—i.e. people deliberately move into neighbourhoods with high ethnic concentrations in order to live according to preferred lifestyles, needs and interests or with like-minded people (Friedrichs, 2008; Horr, 2008). If such a preference was crucial for moving, then interethnic contact will be presumably high. However, there are strong arguments against self-selection bias. First, the choice of a neighbourhood is primarily determined by the structural constraints of the housing market, by the economic resources available for housing and by the infrastructural and environmental characteristics of the area. Only if these criteria are satisfied, does the preference for the social make-up of the neighbourhood come into play. Secondly, people overwhelmingly have preferences for homogeneous and not diverse areas. A recent German study about motivations to move shows little empirical evidence for a self-selection bias. While 62 per cent of West Germans said that characteristics of the home (such as price, space, fittings, tenancy termination, ownership) are the main reason for moving tso another neighbourhood, only 16 per cent stated ‘other reasons’ which might include but does not exclusively mean the ethnic concentration of neighbourhoods (Kemper, 2008). Moreover, people who prefer ethnic diversity do not differ from people without this preference in the ways in which they practice diversity in their daily lives (Blokland and van Eijk, 2010). Thus, it is more plausible to assume that causality goes from context characteristics to interethnic contact.

Hypothesis H3 postulates that ethnic concentration at the municipal level affects contact with foreigners. The percentage of the foreign population shows a significant proportional effect. While a share of foreigners of 3 per cent, a usual proportion of East German municipalities, predicts a probability of 29 per cent for interethnic contact in the neighbourhood, a share of foreigners of 13 per cent which is usual on average for West German municipalities, predicts a probability of 50 per cent. This substantial effect on contact with foreigners in the neighbourhood is depicted in Figure 2a. Hypothesis H3 is distinctly confirmed.

Figure 2.

Municipality-level effects on contact with foreigners in the neighbourhood (context model 1 in Table 2).

Finally, there is a disproportional significant effect of population size on contact with foreigners as hypothesised in H4. The larger the population of a municipality the less contact—i.e. the probability of contact with foreigners decreases to 36 per cent if the population size increases by a standard deviation. This effect is depicted in Figure 2b. Regression coefficient, contact probability and curve shape indicate that the strength of municipal population size is rather moderate compared with other contextual characteristics. However, hypothesis H4 is confirmed too.

To sum up, the neighbourhood as well as the municipality are meaningful spatial contexts for contact with foreigners in the neighbourhood. Neighbourhood characteristics and municipal conditions make significant and relevant contributions to explain this type of contact even under control of individual traits. However, it is questionable if this holds for other types of contact. Contacts in the neighbourhood are only a minor subset of all contacts a person holds and they lean towards casual rather than socially significant relationships, not least because of new communication and transport technologies (Sampson, 1988; Wellman, 1996, 2001). Hence, further context models with different kinds of interethnic contact can secure and verify the hypotheses. Subsequently, analyses are performed examining contextual effects on contact with foreigners among one’s friends and acquaintances as well as on contact with foreigners in general.

The findings related to context conditions are similar for both kinds of contact with foreigners (context models 2 and 3 in Table 2). The effect for type of residential building is now insignificant, meaning that transitional zones around the house are irrelevant for contacts with foreign friends and acquaintances and foreigners at large. Hypothesis H1 has to be refuted for these types of contact with foreigners. The contact opportunities in transitional zones of the immediate neighbourhood are only appreciable for interethnic contact in the neighbourhood.

Effects of perceived foreigners in the neighbourhood moderated by the relationship between Germans and foreigners in the neighbourhood remain significant and important. Perceptions of many foreigners in the neighbourhood and positive relations between Germans and foreigners seem to be suitable conditions for interethnic contact, whereas few foreigners and a conflictual relationship hinder interethnic contact. Hence, hypothesis H2 can be largely confirmed.

The results of both context models show that share of foreigners has the hypothesised proportional effect. There is more contact with foreigners among friends and acquaintances and more contact with foreigners at large if municipalities have high shares of foreigners. More meeting opportunities at the macro structural level lead to more interethnic contact. Further, the significant effects emphasise the relevance of municipal contexts for interethnic contact. The findings of all three context models are in line with hypothesis H3.

Finally, contact with foreigners among friends and acquaintances and any contact with foreigners are less likely in larger municipalities. These significant relationships between population size and different kinds of interethnic contact largely support hypothesis H4. However, it is not possible to determine the extent to which the effects found are the result of segregation tendencies or of homophilous preferences.

The relationship is rather weak compared with other contextual characteristics. One reason might be that there are certain thresholds rather than a linear relationship. This goes along with the ideas of critical mass in forming sub-cultures (Fischer, 1982). To test whether the relationship between municipal population size and interethnic contact is linear or not, municipalities are classified into four groups: villages (0–4999 inhabitants), small towns (5000–19,999 inhabitants), mid-sized towns (20,000–99,999 inhabitants) and big cities (100,000 and more inhabitants). In context models 4 to 6 (Table 3), the population size measure is replaced by this classification.

Table 3.

Multilevel logit regressions on different kinds of contact with foreigners

Variables	Contact in the neighbourhood (context model 4)		Contact with friends (context model 5)		Contact in general (context model 6)
Variables	Coefficient	S.E.	Coefficient	S.E.	Coefficient	S.E.
Fixed part
Intercept	−0.109	0.163	0.438	0.157	2.162	0.210
Social background (individual variables at the 1st level)
Migration background (1 = with)	0.100	0.123	0.576	0.118	0.384	0.148
Education (ref. = medium education)
Lower educationHigher education	−0.076−0.037	0.1190.145	−0.164 0.310	0.1100.138	0.0110.053	0.1320.176
Professional qualification (ref. = medium prof. qual.)
Lower professional qualificationHigher professional qualification	−0.296 0.172	0.1400.145	−0.2550.026	0.1330.137	−0.438 0.389	0.1510.173
Employment status (ref. = self-employed, full-time)
Pupils and students(Self-)employed, part-timeUnemployedNon-workingRetired/pensioned	0.304 −0.455 −0.375−0.230 −0.466	0.3010.1780.2130.1670.167	0.374−0.134−0.249−0.163−0.248	0.3280.1660.1910.1560.154	0.501−0.283 −1.070 −0.876 −0.812	0.4820.2240.2210.1910.182
Income in 1000 Euro^a	0.091	0.046	0.074	0.046	0.128	0.062
Income not known	0.031	0.121	−0.215	0.115	−0.166	0.138
Gender (1 = female)	0.071	0.104	−0.228	0.098	−0.091	0.116
Age^a	−0.006	0.004	−0.038	0.004	−0.036	0.005
Household size^a	0.131	0.044	0.001	0.042	0.081	0.055
Neighbourhood (context variables at the 1st level)
Type of residential building^a	0.115	0.039	−0.059	0.037	0.014	0.044
Perceived foreigners in the home environment^a	1.312	0.093	0.444	0.083	0.899	0.112
Interaction of perceived foreigners x relations between Germans and foreigners in the home environment (ref. = neutral or no foreigners)
Be on good termsBe on bad terms	0.250 −0.713	0.2170.195	0.130 −0.728	0.1840.196	0.279 −1.063	0.3030.238
Municipality (context variables at the 2nd level)
Share of foreigners at county level^a	0.084	0.017	0.095	0.017	0.157	0.023
Municipal population (ref. = 20,000–99,999 inhabitants[mid-sized towns])
0–4999 inhabitants (villages)5000–19,999 inhabitants (small towns)100,000 and more inhabitants (big cities)	−0.458 −0.147 −0.557	0.224 0.178 0.189	−0.294−0.292−0.519	0.2090.1760.187	−0.049−0.050−0.941	0.2690.2320.246
Random part
S.D. of the intercept	0.545	0.073	0.553	0.065	0.744	0.081
Model statistics
N (people / municipalities)	2988 / 161		2989 / 161		2943 / 161
LR test (d.f.)	546.97 (23)		498.16 (23)		570.99 (23)
McFadden pseudo R²	0.151		0.128		0,183
BIC	3274.682		3602.751		2750.388

Variables are centred on its mean values.

Note: Bold printed coefficients are significant at the 0.05 level.

The findings of all three kinds of contact with foreigners are similar in that the relationship is curvilinear with the highest likelihood of interethnic contact in mid-sized towns. On closer inspection, differences between mid-sized towns and smaller municipalities are only gradual and mostly non-significant, but interethnic contact is far less likely in big cities compared with mid-sized towns after controlling for share of foreigners—i.e. the higher meeting opportunities that are mostly in larger municipalities are taken into account. The findings suggest that negative influences on interethnic contact caused by segregation tendencies and opportunities for intraethnic contact preferences evolve in big cities only.

There is one exception from the general picture—a significantly negative relationship between villagers and contact with foreigners in the neighbourhood. This effect needs an alternative explanation because it is implausible to assume segregation and opportunities at large for preferred contact in villages. A possible explanation is that neighbourhood is quite narrowly defined by villagers as just the immediate neighbours. Then, interethnic contact might be almost impossible given the already low number of foreigners in German villages.

5. Discussion

The aim of the present study was to investigate three research issues on the relationship between socio-spatial context and interethnic contact. First, while most of the research in this field has indicated that both neighbourhood and municipality levels are influential spatial contexts, many studies investigate only a single level. The present study has tried to bridge this gap by including both localities at the same time. The analysis shows that this is an important step forward because ethnic concentration at both levels independently fosters interethnic contact. Moreover, and in line with Blau (1994, p. 5), sub-structures promote intergroup contact but macro structures entail segregation that limits this contact.

The second issue deals with different contextual conditions. Most studies in the field are concerned with the extent to which the ethnic or multinational make-up shapes interethnic contact. It is hypothesised here that other factors like neighbourly contact opportunities of transitional zones and municipal population size are influential as well. However, the analysis shows that ethnic concentration has the strongest impact. Especially, perceived foreigners in the neighbourhood moderated by the perceived relationship between Germans and foreigners strongly contribute to the interethnic contacts of Germans. Other contextual conditions are influential as well, but municipalities’ population size and the physical semi-public space of neighbourhoods are not as powerful as the ethnic make-up of neighbourhoods and municipalities. At least, they are significant even under control of ethnic concentration. The huge impact of ethnic concentration as well as effects of neighbourly contact opportunities of transitional zones and municipal population size proves structural contact opportunities true. Moreover, the results on the municipal context together imply that the inconsistency of previous research on effects of ethnic concentration may be due to different characteristics that have been disentangled in this paper.

The third issue focuses on different kinds of interethnic contact. Three kinds were distingished and analysed: interethnic contact in the neighbourhood, freely chosen interethnic contact with friends and acquaintances, and interethnic contact in general. Effects of ethnic concentration and municipal population size remain relatively stable and relevant across these different forms of contact. This is insofar surprising as neighbourhood context opportunities should affect interethnic contact in the neighbourhood only. However, a significant relationship of physical contact opportunities in transitional zones was found for interethnic contact in the neighbourhood, but not for the two other kinds of interethnic contact. It can be concluded that the material-physical characteristics of neighbourhoods open up opportunities for contact that are restricted to this sub-structure and that such conditions affect neighbourly encounters only. A shortcoming is the vague measure of interethnic contact that is open to various interpretations. Future research should investigate social interactions from fleeting encounters to strong ties.

Finally, the negative effect of municipal population size does not provide secure evidence of more disintegration in big cities. There is also an alternative interpretation of more specialised interethnic contact. There are three reasons that support this view. First, there is not per se less interethnic contact in big cities, but just under control of ethnic concentration. Empirically, there is a high correlation between ethnic concentration and municipality size but the effects on interethnic contact of the two characteristics are opposed. A rerun of the analyses without share of foreigners reveals that interethnic contact is lower in villages and small towns but not in big cities. Secondly, ethnic segregation is much less prevalent in Germany than in the US. Schönwälder and Söhn (2009) showed for selected German big cities that inner-urban ethnic concentration is relatively low. Thirdly, social life is organised in much more complex ways in big cities than in smaller municipalities. For example, sub-cultures need critical masses to develop and this precondition exists in big cities in particular. If urban societies have complex organisation into sub-cultural networks, then a few interethnic contacts that bridge these networks might have more impact than many interethnic contacts within a small municipality. However, the research question of whether or not the societal organisation of interethnic interactions is more influential than the sheer existence of interethnic contact should be investigated in future reseach.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Notes

References

Babka von Gostomski

Stichs

. (2008) Der Einfluss von Gelegenheitsstrukturen auf die Häufigkeit des Kontakts von Zuwanderern mit Deutschen, in: Hillmann

Winzio

(Eds) Migration und städtischer Raum: Chancen und Risiken der Segregation und Integration, pp. 279–296. Opladen: Budrich UniPress.

Blalock

(1967) Toward a Theory of Minority-group Relations. New York: John Wiley.

Blau

P. M.

(1994) Structural Contexts of Opportunities. Chicago, IL: University of Chicago Press.

Blokland

Eijk

G. van

(2010) Do people who like diversity practice diversity in neighbourhood life? Neighbourhood use and the social networks of ‘diversity-seekers’ in a mixed neighbourhood in the Netherlands, Journal of Ethnic and Migration Studies, 36(2), pp. 313–332.

Briggs

(2007) ‘Some of my best friends are …’: interracial friendships, class, and segregation in America, City and Community, 6(4), pp. 263–290.

Castells

(2000) The Rise of the Network Society. Oxford: Blackwell.

Dines

Cattell

(2006) Public Spaces, Social Relations and Well-being. Bristol: The Policy Press.

Drever

A. I.

(2004) Separate spaces, separate outcomes? Neighbourhood impacts on minorities in Germany, Urban Studies, 41(8), pp. 1423–1439.

Drever

A. I.

(2008) Germans in Germany’s ethnic neighborhoods, Schmollers Jahrbuch, 128, pp. 175–190.

10.

Ellsworth

(2005) Places of Learning: Media, Architecture, and Pedagogy. New York: Routledge.

11.

Esser

(1980) Aspekte der Wanderungssoziologie: Assimilation und Integration von Wanderern, ethnischen Gruppen und Minderheiten: Eine handlungstheoretische Analyse. Darmstadt/Neuwied: Luchterhand.

12.

Esser

(1986) Social context and inter-ethnic relations: the case of migrant workers in West German urban areas, European Sociological Review, 2, pp. 30–51.

13.

Esser

(1999) Soziologie: Spezielle Grundlagen. Band 1: Situationslogik und Handeln. Frankfurt am Main: Campus.

14.

Farwick

(2009) Segregation und Eingliederung: Zum Einfluss der räumlichen Konzentration von Zuwanderern auf den Eingliederungsprozess. Wiesbaden: VS Verlag.

15.

Feld

S. L.

(1981) The focused organization of social ties, American Journal of Sociology, 86, pp. 1015–1035.

16.

Fischer

C. S.

(1976) The Urban Experience. New York: Harcourt Brace Jovanovich.

17.

Fischer

C. S.

(1982) To Dwell among Friends: Personal Networks in Town and City. Chicago, IL: University of Chicago Press.

18.

Fong

Isajiw

W. W.

(2000) Determinants of friendship choices in multiethnic society, Sociological Forum, 15(2), pp. 249–271.

19.

Forrest

Kearns

(2001) Social cohesion, social capital and the neighbourhood, Urban Studies, 38(12), pp. 2125–2143.

20.

Friedrichs

(2008) Ethnische Segregation, in: Kalter

(Ed.) Migration und Integration, pp. 380–411. Wiesbaden: Westdeutscher Verlag.

21.

Friedrichs

Blasius

(2000) Leben in benachteiligten Wohngebieten. Opladen: Leske und Budrich.

22.

Galster

G. C.

(2008) Quantifying the effect of neighbourhood on individuals: challenges, alternative approaches, and promising directions, Schmollers Jahrbuch, 128, pp. 7–48.

23.

Gans

(1961) Planning and social life, Journal of the American Institute of Planners, 27(2), pp. 134–140.

24.

Gehl

(1987) Life between Buildings: Using Public Space. New York: Van Nostrand Reinhold.

25.

Gijsberts

Dagevos

(2007) The socio-cultural integration of ethnic minorities in the Netherlands: identifying neighbourhood effects on multiple outcomes, Housing Studies, 22, pp. 805–831.

26.

Goffman

(1963) Behavior in Public Places: Notes on Social Organization of Gatherings. New York: The Free Press.

27.

Hanhörster

Mölder

(2000) Konflikt- und Integrationsräume im Wohnbereich, in: Heitmeyer

Anhut

(Eds) Bedrohte Stadtgesellschaft: Soziale Desintegrationsprozesse und ethnisch-kulturelle Konfliktkonstellationen, pp. 347–400. Weinheim: Juventa.

28.

Hewstone

(2009) Living apart together? The role of intergroup contact in social integration. Working Paper No. 09–12, Max Planck Institute, Göttingen.

29.

Hoogland

(2000) Semi-private Zones as a Facilitator of Social Cohesion. Nijmegen: Katholieke Universiteit Nijmegen.

30.

Horr

(2008) Ethnische und soziale Unterschiede der Wohnungssuche und Wohnortwahl, in: Hillmann

Windzio

(Eds) Migration und städtischer Raum: Chancen und Risiken der Segregation und Integration, pp. 175–192. Opladen: Budrich UniPress.

31.

Janßen

Schroedter

J. H.

(2007) Kleinräumliche Segregation der ausländischen Bevölkerung in Deutschland: Eine Analyse auf der Basis des Mikrozensus, Zeitschrift für Soziologie, 36(6), pp. 453–472.

32.

Kecskes

(2003) Ethnische Homogenität in sozialen Netzwerken türkischer Jugendlicher, Zeitschrift für Soziologie der Erziehung und Sozialisation, 23(2), pp. 68–84.

33.

Kemper

F.-J.

(2008) Residential mobility in East and West Germany: mobility rates, mobility reasons, reurbanization, Zeitschrift für Bevölkerungswissenschaft, 33(3/4), pp. 293–314.

34.

Lazarsfeld

P. F.

Merton

R. K.

(1954) Friendship as social process: a substantive and methodological analysis, in: Berger

Abel

Page

C. H.

(Eds) Freedom and Control in Modern Society, pp. 18–66. Toronto: van Nostrand.

35.

Lofland

(1973) A World of Strangers: Order and Action in Urban Public Space. New York: Basic Books.

36.

Logan

J. R.

Spitze

G. D.

(1994) Family neighbors, American Journal of Sociology, 100, pp. 453–476.

37.

Marsden

P. V.

(1990) Network diversity, substructures, and opportunities for contact, in: Calhoun

Meyer

Scott

R. S.

(Eds) Structures of Power and Constraint: Papers in Honor of Peter Blau, pp. 397–410. New York: Cambridge University Press.

38.

McPherson

Smith-Lovin

Cook

J. M.

(2001) Birds of a feather: homophily in social networks, Annual Review of Sociology, 27, pp. 415–444.

39.

Oberwittler

(2007) The effects of ethnic and social segregation on children and adolescents: recent research and results from a German multilevel study. WZB [Social Research Centre], Berlin.

40.

Petermann

(2011) Räumlicher Kontext, migrationsbezogene Vielfalt und Kontakte zu Ausländern in der Nachbarschaft. MMG Working Paper No. WP 11–06, Max Planck Institut, Göttingen.

41.

Petermann

Schönwälder

(2012) Gefährdet Multikulturalität tatsächlich Vertrauen und Solidarität? Eine Replik, Leviathan, 40, pp. 482–490.

42.

Petermann

Heywood

Hewstone

Hüttermann

. (2012) The ‘diversity and contact’ survey 2010 (wave 1): technical report. MMG Working Paper No. WP 12–21, Max Planck Institute, Göttingen.

43.

Pettigrew

T. F.

Tropp

L. R.

(2006) A meta-analytic test of intergroup contact theory, Journal of Personality and Social Psychology, 90, pp. 751–783.

44.

Putnam

R. D.

(2007) E Pluribus Unum: diversity and community in the twenty-first century. The 2006 Johan Skytte Prize Lecture, Scandinavian Political Studies, 30(2), pp. 137–174.

45.

Quillian

(1995) Prejudice as a response to perceived group threat: population composition and anti-immigrant and racial prejudice in Europe, American Sociological Review, 60, pp. 586–611.

46.

Reuband

K. H.

(1989) Kontakte zwischen Deutschen und Gastarbeitern: Ein Indikatorenvergleich, ZA-Information, 24, pp. 72–83.

47.

Sampson

R. J.

(1988) Local friendship ties and community attachment in mass society: a multilevel systemic model, American Sociological Review, 53, pp. 766–779.

48.

Schönwälder

Söhn

(2009) Immigrant settlement structures in Germany: general patterns and urban levels of concentration of major groups, Urban Studies, 46 (7), pp. 1439–1460.

49.

Smith

T. W.

(1999) Measuring inter-racial friendships: experimental comparisons. GSS Methodological Report No. 91, National Opinion Research Center, University of Chicago.

50.

Snijders

T. A. B.

Bosker

R. J.

(1999) Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modelling. London: Sage.

51.

Statistisches Bundesamt (2007) Statistisches Jahrbuch 2007: Für die Bundesrepublik Deutschland. Statistisches Bundesamt, Wiesbaden.

52.

Stephan

W. G.

Renfro

L. C.

(2002) The role of threats in intergroup relations, in: Mackie

Smith

(Eds) From Prejudice to Intergroup Emotions, pp. 191–207. New York: Psychology Press.

53.

Tolsma

Meer

T. van der

Gesthuizen

(2009) The impact of neighbourhood and municipality characteristics on social cohesion in the Netherlands, Acta Politica, 44(3), pp. 286–313.

54.

Verbrugge

L. M.

(1977) The structure of adult friendship choices, Social Forces, 56(2), pp. 576–597.

55.

Vervoort

Flap

Dagevos

(2011) The ethnic composition of the neighbourhood and ethnic minorities’ social contacts: three unresolved issues, European Sociological Review, 27(5), pp. 586–605.

56.

Wagner

Dick

van, Pettigrew

T. F.

Christ

(2003) Ethnic prejudice in east and west Germany: the explanatory power of intergroup contact, Group Processes & Intergroup Relations, 6(1), pp. 22–36.

57.

Wasmer

Scholz

Blohm

(2007) Konzeption und Durchführung der ‘Allgemeinen Bevölkerungsumfrage der Sozialwissenschaften’ (ALLBUS) 2006. Methodenbericht 2007/09, ZUMA.

58.

Weber

(2008) Economy and Society: An Outline of Interpretative Sociology. Berkeley, CA: University of California Press.

59.

Wellman

(1996) Are personal communities local? A dumptarian reconsideration, Social Networks, 18, pp. 347–354.

60.

Wellman

(2001) Physical place and cyberplace: the rise of personalized networking, International Journal of Urban and Regional Research, 25(2), pp. 227–252.

61.

Wilson

W. J.

(1987) The Truly Disadvantaged. Chicago, IL: University of Chicago Press.

62.

Woolcock

(1998) Social capital and economic development, Theory and Society, 27(2), pp. 151–208.

63.

Zelinsky

Lee

B. A.

(1998) Heterolocalism: an alternative model of the sociospatial behaviour of immigrant ethnic communities, International Journal of Population Geography, 4, pp. 281–298.