Abstract
Increasing ethnic and social diversity in cities does not translate into diverse networks of urbanites. Particularly for white middle-classes in gentrified neighbourhoods, there is evidence on boundary drawing to ‘unwanted groups’ such as ethnic minorities and lower-classes. Rarely have these studies focused on the networks of ethnic minorities, the actual diversity-bringers. I will contribute to the understanding of why and under what circumstances diversity in neighbourhoods gets translated into people’s daily practices, hence also networks, by analysing those of middle-class Turkish-Germans in Berlin. Based on interviews and network analysis, I will show that a neighbourhood’s ethnic diversity, fellow residents’ attitudes towards diversity and the built environment play an important role in building category-crossing ties. Owing to a lack of reciprocity in establishing ties, Turkish-Germans in neighbourhoods with a high share of native-Germans actually have more ties to Turks than those in a more diverse neighbourhood.
Background: Networks, Homophily and Social Closure
In 2010, German Chancellor Angela Merkel declared that the system of ‘multikulti’ had utterly failed in Germany. She argued that the state alone cannot support immigrants, but that they also have to show more effort to integrate. In that same year, then member of the Social Democrats Thilo Sarrazin published the highly debated book Deutschland schafft sich ab (‘Germany Abolishes Itself’), in which he describes lower-class immigrants coming to Germany, who pass on their low IQ to their children, or their inability to integrate, particularly for Muslims (cf. Schuler, 2010). In an interview with Lettre International, preceding the book publication, Sarrazin claimed that he would not recognize anyone as part of German society who constantly ‘produces new little girls with headscarves’ (Berberich, 2009: 200; my own translation) – the ultimate sign to him indicating the existence of a parallel society and the subordination of women to men (Foroutan, 2011). Two years later, Heinz Buschkowsky, the then mayor of Neukölln – the Berlin neighbourhood often fancied by the media as an example of an immigrant ghetto – published his book, likewise describing failed integration by Turkish and Arab immigrants, largely making them responsible for their situation (cf. Schmitz, 2012).
Indicators of failed integration or the existence of parallel societies mentioned by these popular authors are poor education, insufficient language skills, the lack of interethnic friendships, a low number of mixed marriages, or being stuck in the ethnic economy (cf. Foroutan, 2011).
These issues have also been taken up in scholarly research. Although less loaded, the broad question is still why some ethnic minority groups are not well integrated into majority society, hence adopting a deficit perspective. Closed, strong ethnic networks are presented as a mobility trap and a sign of lacking integration (Gestring et al., 2004; Haug, 2010; Kalter and Kogan, 2014). Ethnic segregation in neighbourhoods or networks is often explained by homophily, a preference to live and interact with (ethnically) similar people.
Lazarsfeld and Merton (1954) were two of the first scholars studying homophily. They observed that people who are categorically similar tend to form friendships with each other. This tendency is, however, only an observed correlation, not an explanatory concept – an important difference they only mention in a footnote. Rather than assuming an a priori preference for similar people, studying homophily is a ‘complex problem of determining the degree to which such selectivity varies for different kinds of social attributes, how it varies within different kinds of social structure, and how such selective patterns come about’ (Lazarsfeld and Merton, 1954: 22). This complexity has often been neglected, and homophily served as the explanation for homogeneous networks, not the observation that had to be explained.
Having a preference for socializing with co-ethnics is sometimes interpreted as a sign of lacking networking skills or ‘abilities to integrate’ (Haug, 2010: 254; my own translation
Networks of ethnic minorities have been studied for several reasons. Within social capital research, the main question is whether ethnic minorities develop bridging ties, or whether they continuously have bonding ties to co-ethnics. The former ones are interpreted as a sign of successful integration, since they help with getting information on jobs and other important issues (Briggs, 2005; Coleman, 1988; Granovetter, 1973). The prevalence of bonding ties, in contrast, is seen as a hindrance to integration (Bankston, 2014; Friedrichs and Blasius, 2001; Gestring et al., 2004; Ryan et al., 2008). Urban policy is likewise increasingly interested in bridging contacts. Many urban policy programmes forwarding social and ethnic mixing in neighbourhoods expect that lower-class residents build ties to middle-class residents and thus profit from their social capital. However, these category-crossing ties rarely emerge (Blokland, 2008; Kleit, 2005; Watt, 2009).
Network analysts focus on the geographical scope of personal networks and have shown that we nowadays live in liberated communities. Networks are no longer locally based, but dispersed across the city, and for migrants often across countries. Who we socialize with is no longer a question of spatial proximity, but of our own choosing (Blokland and Rae, 2008; Wellman and Leighton, 1979). There has, however, often been a dichotomy, with (lower-class) migrants’ predominantly local and homogeneous networks on the one end and highly skilled migrants’ locally dispersed and transnational ties on the other end (Ryan and Mulholland, 2013; Smith and Favell, 2006). Rather neglected is the important ‘middling’ group: the skilled and educated among the globally mobile …: students, nurses, mid-level technical and clerical employees, ambitious or adventurous upwardly mobile middle-classes, migrants from a range of intermediate developing states, and many more it would be hard to describe as ‘elites’. (Smith and Favell, 2006: 2)
This is a research gap I address with the current study on middle-class Turkish-Germans.
Taking seriously the assertion that a preference to socialize with co-ethnics is not an explanation, but the observation that has to be explained, the question arises of why networks are more or less homogeneous. Spatial proximity and group size are important mechanisms influencing network diversity (Blau, 1977; Briggs, 2007; Wimmer, 2013). For group size and exposure, Briggs (2007) showed that a high share of same-race friends among white individuals in the US results from the area’s low share of other racial groups. In an ethnically homogeneous neighbourhood, a high presence of co-ethnics in the network is thus not an expression of preference, but of a lack of availability. Homophily based on class will be high in a college or university. Thus, if baseline or ‘induced’ (McPherson and Smith-Lovin, 1987) homophily is high, emerging ties also tend to be homogeneous.
In addition to exposure and availability, there have to be places where people can meet one another, so-called foci (Feld, 1981: 1016): a ‘social, psychological, legal, or physical entity around which joint activities are organized’. A neighbourhood can be a foci, but it can also contain other foci such as cafés, bars, schools, or playgrounds. People organize their daily lives around such foci and, as Small (2009: 87) has shown with the example of a children’s day care centre, ‘organizations can institutionally perform much of the “work” required to sustain strong friendships’.
Wimmer (2013: 142) furthermore points to the important influence of balancing mechanisms, which include ‘the tendency of a friendship to be returned (reciprocity) and of friends of friends to befriend one another (triadic closure)’. Studying the Facebook friendships of more than 1600 students attending a private college in the US, Wimmer finds that ‘rates of same-race reciprocity and closure are not higher but lower than those of racially heterogeneous dyads and triangles’ (Wimmer, 2013: 162). Ethnic or racial homogeneity in networks does not occur ‘automatically’, and its existence has to be explained.
Nevertheless, even in favourable contexts, category-crossing ties do not automatically form. The taste for diversity expressed by white middle-classes in various countries and cities rarely translates into diverse networks (Blokland and van Eijk, 2010; Butler, 2003; Butler and Robson, 2001; Kleit, 2005; May, 1996; Tissot, 2011). For a London neighbourhood, Butler and Robson (2001: 2156) found ‘a gulf between a widely circulated rhetorical preference for multicultural experience and people’s actual social networks and connections’. Blokland and van Eijk (2010) assessed networks of people in a multicultural area in Rotterdam and found that even ‘diversity-seekers’ – native Dutch who consciously chose to move to an ethnically heterogeneous area – do not have ties to ethnic minorities, even though they use the local infrastructure. Thus, ‘even for those who, whether middle-class or not, come into a mixed neighbourhood with openness to diversity, this openness does not translate in more diverse networks’ (Blokland and van Eijk, 2010: 327).
These studies, however, deal only with native white middle-classes. There are no data on ethnic minorities’ networks. As we will see, middle-class Turkish-Germans are also diversity-seekers, and some also display diverse networks. The neighbourhood, its ethnic makeup and built environment play a crucial part in explaining why some networks are more or less diverse.
Method and Data
This article is based on a qualitative study of 41 upwardly mobile Turkish-Germans 1 in Berlin. In-depth interviews, lasting around 1.5 hours, were conducted by the author in 2012, inquiring on neighbourhood choice, ethnic and social identification, and networks. I compared people living in socioeconomically disadvantaged neighbourhoods of Berlin to those who have moved out of such areas into more advantaged ones. For this article, I will only consider the latter group of movers − 23 in number. The event of moving presents a unique opportunity to study how and to whom new contacts are established. For forms of support, local contacts continuously play an important role (Wellman and Leighton, 1979). The movers have been living in their respective neighbourhood for at least one year, but mostly five to seven years. The time dimension is crucial since making new contacts and developing a local support network takes time (Ryan et al., 2008).
The respondents were largely selected using the Berlin Turkish yellow pages. The participants, mostly between 35 and 50 years, were 1.5 and 2nd generation immigrants – they immigrated as children or were born in Berlin. The interviews were conducted in German – all respondents spoke the language fluently. This was an important requirement, particularly regarding networks, since insufficient language skills can be a barrier to making contacts with the ethnic majority. Moreover, linguistic differences can trigger mistrust and keep native-Germans from forming category-crossing ties (Schaeffer, 2014: 92).
Among the respondents were many professionals, working as lawyers, tax or insurance consultants, or in the media sector. They all have experienced upward mobility and have higher economic and cultural capital than their parents, who were – as most guest workers – employed as manual labourers, and only had basic schooling (in Turkey). Among the movers, all but two were married or in a partnership, and 18 out of 23 had children.
Part of the interview was a network analysis to assess the Turkish-Germans’ ego-centred network (Burt, 1984; Hennig, 2008; McCallister and Fischer, 1978). Using name generators, I asked who the respondent would turn to for a variety of emotional and instrumental support, such as discussing personally relevant issues, getting help with small repairs in the house, or help with taking care of the children. For all named contacts, I elicited information regarding relationship type (e.g. partner, kin, friend, neighbour), place of residence, and ethnic and educational backgrounds. The respondents named on average 12 contacts. Of all contacts (290), around 30 per cent had a native-German background, almost 60 per cent a Turkish background, and 10 per cent another ethnic background. For education, 11 per cent of the contacts had no degree, 27 per cent and 20 per cent lower or higher secondary schooling, respectively, and a third had a university degree. These will be assessed in more detail in the following part.
It is important to keep in mind that the network data are based on 23 cases, which have not been randomly sampled. Hence, the numbers presented here cannot be generalized to other Turkish-Germans. Consequently, I will look at interesting network patterns and develop some hypothesis on possible explanations, supplemented by interview material and a closer assessment of individual respondent’s networks.
Results
Turkish-German Diversity-Seekers
Thinking about networks and tie formation, particularly in the wake of a move, the question arises regarding what kind of contacts the respondents actually look for. The Turkish-Germans strongly stress their hybrid identity (Foroutan and Schäfer, 2009), pointing out that they socialize with Turks and Germans alike. In their daily life, as well as in a neighbourhood, they value a strong sense of community, which they see as an important characteristic of their Turkish background. Lawyer Behçet, married and father of three children, sums up how he believes people with a Turkish family background are more community oriented: Community life is much more important [for Turks]. There is a pleasant being together, a feeling of belonging, which is stronger; taking care of each other is stronger, communication is stronger. And we’re more affectionate, we know far better how to party [laughs], we’re more open, hospitality is much more important. So if you don’t have bread, we get bread so that we can share.
This sense of community is a feature of neighbourhood life that the respondents highly value. Real estate agent Selami lives with his wife and two children in a multicultural neighbourhood and expresses its positive characteristics: Now, I live close to the park. Well, there are many shops and bars; there is always life, urban life. In the summer, life takes place outside, in the streets, and that is great.
In contrast is, for example, Özcan’s place of residence, which lacks such street life. Residents only come together for official events, such as a spring feast. The Turkish-Germans thus value contact to fellow residents, and express a distaste towards ‘anonymous’ (Varol) areas. They seek neighbourly relations and see this as a characteristic of ‘urban life’. Part of this urban life is diversity. The Turkish-Germans value diversity in a neighbourhood in two respects. First, they value ethnic diversity, hence neighbourhoods with an ethnically diverse population. Second, diversity refers to the built environment. The respondents prefer areas that offer parks, public and ‘third places’, such as cafés (cf. Oldenburg, 1997), where residents can meet. Having places where people can run into each other on a regular basis can create public familiarity, which is the basis for developing more durable ties (Blokland and Nast, 2014). In the following, we will see that not all movers find neighbourhoods with a community life and diversity. This gets mirrored in their networks.
Network Diversity
Overall, nine of the 23 Turkish-German respondents living in the heterogeneous neighbourhood of Westgrove 2 named 105 contacts and the 14 respondents living in the homogeneous Southfield named 185; thus, 12 and 13 per respondent on average, respectively. The respondents in Westgrove have more than twice as many native-Germans and twice as many people with another ethnic background in their networks than those living in the homogeneous Southfield (Figure 1). The latter, in contrast, have a network made up of about 71 per cent of people with a Turkish background, compared to only around 42 per cent of the Turkish-Germans living in Westgrove.
Movers: Ethnic network diversity based on neighbourhood diversity, percentage and numbers.
This pattern includes all contacts, but it holds when considering local ties only. Based on spatial proximity and exposure, the Turkish-Germans in Westgrove should have more native-Germans in their networks than those in Southfield. The opposite is the case. Of the Southfield residents’ named 63 contacts, only 23 (36.5%) have a native-German background and 33 (52.4%) a Turkish one. The local networks in Westgrove include 18 (66.7%) native-Germans and only six (22.2%) Turks. The higher raw numbers for those in Southfield reflect the slightly bigger size of that group.
Assessing the networks’ diversity regarding social class – the respondent’s education serves as a proxy – a similar, unexpected pattern emerges (Table 1). Since the respondents live in socioeconomically similar neighbourhoods, there should not be any difference in the social make-up of the networks. However, there are clear differences. For the movers living in a diverse neighbourhood, 25.3 per cent of all contacts have lower or no education. In the networks of the Turkish-Germans living in a homogeneous area, more than half of all contacts (54%) fall into this category. In contrast, for the Turkish-Germans in Westgrove, almost half of all contacts have a college degree, in contrast to just over 30 per cent for those in homogeneous Southfield.
Educational level of contacts, percentage and numbers.
This might indicate the overlap between ethnic background and social class: people with a Turkish family background fare worse than native-Germans and other ethnic minority groups regarding educational attainment (Hartmann, 2014; Sauer, 2009). If, thus, the respondents do not succeed in establishing ties with native-Germans, they might keep ties to their old peer group, to people with a Turkish family background, who often only have a lower education.
Recalling that the respondents are similar in socioeconomic background, age, family status, and that they are all diversity-seekers, the difference in network diversity cannot be attributed to differences in individual characteristics or stated preferences to socialize only with co-ethnics. In the following, I will therefore assess in more detail the networks of Turkish-Germans and supplement them with interview material, presenting possible reasons as to why particular networks are more diverse than others.
The Lack of Reciprocity and Foci as Driving Forces Behind Homogeneous Networks
A reason for the low share of native-Germans in some movers’ networks is that Turkish-Germans in homogeneous areas often face hostility from their ethnic majority neighbours. Their wish to develop ties is not reciprocated, so their networks remain ethnically homogeneous. The Turkish-Germans living in diverse neighbourhoods, in contrast, are able to translate this diversity into their networks.
The Turkish-Germans who moved to a homogeneous neighbourhood with a high share of native-Germans did not have much knowledge about that new area. They mainly moved out of their old, socioeconomically disadvantaged neighbourhood to find a better educational environment for their children. Thus, whereas knowledge about local schools was good, knowledge on neighbourhood life was limited. After some time, they realized that it did not match their idea of a good neighbourhood life.
Lawyer Varol, 42 years old, and his wife moved to Southfield because of the area’s good schools. He describes the neighbourhood as very residential, dominated by conservative native-Germans who are not exactly fond of having Turkish neighbours. Even after 13 years of living there, he does not feel he belongs there, but instead misses his old neighbourhood. Varol particularly complains about lacking neighbourly relations: That [good relations] is missing in Southfield. There, you are, how can I put it, more anonymous. But do you know any families who live there? Yes, we know some. But we don’t visit each other. You did that maybe once or twice, but not anymore. They all want to seal themselves off.
When he and his wife moved to Southfield, they tried to establish ties with their fellow residents, by inviting them over, but their attempts were not reciprocated. Serkan had similar experiences when moving to Southfield with his wife and two children: At the beginning it was hard. It was … Conversely, for the people who live in Southfield it was hard to accept foreigners. For those who live there, many elderly and Germans, it was hard to accept us. They thought ‘Oh, foreigners, what are they doing here? If one is here they will all come’. But did you expect that when you moved there? No.
Serkan, Varol and others were pioneers: they were among the first Turkish-German families moving to a neighbourhood with a high share of native-Germans. Being a pioneer may be a difficult position, particularly for members of a stigmatized ethnic group such as Turkish-Germans. Research in the US has shown that white individuals prefer to live in neighbourhoods with only a small percentage of black or Hispanic residents. After a tipping point of 5–20 per cent is reached, they start moving out of these neighbourhoods (Lewis et al., 2011; St John and Bates, 1990). In Europe, and Germany in particular, with lower levels of segregation, such preferences are not that pronounced. In Germany, only 23.3 per cent of the population would prefer to live in an ethnically homogeneous neighbourhood and 37.9 per cent actually do – these are among the lowest numbers in Europe (Semyonov et al., 2007). Moreover, education and income positively influence the attitude towards living in an ethnically diverse neighbourhood. Thus, middle-class natives should be rather tolerant towards ethnic minorities, at least when they share the same social status.
Even if ethnic minority pioneers might evoke fears about declining school quality and property values or increasing crime rates among white neighbourhoods, more time spent in the neighbourhood and a higher familiarity with the ethnic other should decrease natives’ negative attitudes (Pettigrew, 1998; Semyonov et al., 2007). This, however, does not seem to be the case. Even after living in the neighbourhood for many years, the Turkish-Germans did not succeed in building longer-lasting ties, but feel that they are still perceived as the outsider by their native-German neighbours. Sometimes, even the benefit of ‘friendly recognition’, implying a recognition of the person as such, which is ‘the normative, minimum principle of interaction among people who consider each other neighbours, and the foundation for the development of deeper neighbourly relationships that eventually form networks and communities’ (Kusenbach, 2006: 291) is denied. Enginalp, a lawyer and politician, who has been living in Southfield with his wife and son for five years, recounts just that: So, if you are out on the streets, there is not much small talk? I would say so, yes. And what about just nodding, saying ‘hi’? Yes, a very afflicted nod, if you directly look at each other.
There might be a nod between the Turkish-German and the native-German residents, but only if eye contact makes it unavoidable. Otherwise, even a simple nod is withheld.
In extreme cases, there is not only disregard, but open racism. When he walks around in Southfield, Varol still sees people looking out of their windows when he – the Turkish neighbour – passes by. Enginalp has been confronted with racism most severely. Interestingly enough, he calls his fellow residents decent, ‘except that they have a problem when a Turkish family lives among them’ – which clearly shapes his experience in the neighbourhood. He had neighbours who spat in his garden, or let the dog defecate in front of his door. One time he even had a swastika drawn on the front door. His wife, who also has a Turkish family background, had a neighbour ranting to her because she was hanging up laundry in the garden – on a Sunday. Enginalp’s experiences are very extreme, but other Turkish-Germans living in Southfield also know what it means not to be greeted by the (native-German) neighbours, or have the door shut when walking by.
Tightly connected to the missing relations between residents is the built environment of the neighbourhood. Southfield is a rather residential area, not offering many foci to meet other locals. Ferda, asked what she thinks is typical for Southfield, says that: ‘It’s quiet, everyone for himself. I don’t know whether that’s typical for here or for these areas with single family homes.’ Sealing off and keeping to oneself can be a characteristic of more residential areas, as Jacobs (1992: 65) has described: ‘In areas that lack a natural and casual public life, it is common for residents to isolate themselves from each other to a fantastic degree.’ Ferda points out that a more anonymous neighbourhood life can be due to the area’s residential character, and the predominance of single family homes. Alkalin mentions the lack of public places. His reaction when asked to describe his ‘Kiez’ – the popularly used word in Berlin for one’s micro-neighbourhood – is telling: Well, it’s not a Kiez. If I think of Kiez, I think of a place where people know each other a bit, that there is a bar, a bakery, something like that. So it’s more, more anonymous.
While research has shown that children can work as brokers for interethnic contact (cf. Kleit, 2005; Schaeffer, 2014), this does not seem to be the case in Southfield. Schaeffer’s (2014) research indicated that children can facilitate interethnic contact if parents use the local parks and playgrounds. The parents who rarely use these foci also have few category-crossing ties. A possible explanation is that native-German parents select homogeneous neighbourhoods because they ‘fear the potential disadvantage for their children’s development that they see as being associated with ethnic diversity’ (Schaeffer, 2014: 128; cf. also Butler and Robson, 2001). If the native-Germans in Southfield selected the neighbourhood because of its ethnic homogeneity, boundary drawing towards ethnic others becomes more likely than the development of category-crossing ties. The brokering role of children is thus context-dependent. Southfield seems to be rather unfavourable as there are few meeting opportunities in the first place, and the native-German parents are seemingly reluctant to develop interethnic ties.
Connected with the dissatisfaction with the missing community life in Southfield is a nostalgic view of neighbourhood relations in the old neighbourhood. Blokland (2003: 191) has shown that ‘people’s recollections about the past reveal at least as much about their interpretations of the current social reality as about the way life used to be’. This is why the community-like relations from the previous neighbourhood figure prominently in these nostalgic accounts. Asked about whether he misses anything from his previous neighbourhood Northend, Özcan refers to the public life in the streets, which he perceives to be typical for more ethnically diverse neighbourhoods with many public places: So, in Northend a lot is going on in the streets. If you go out, just to have a walk, you would have met all the grocers, you would have been on a first-name basis with everyone, and you would have … more this street life. Here [Southfield] it is only event-life. You have summer fest, garden fest, arbour fest, those are the fests where you meet … there is just no street life.
Özcan contrasts the spontaneous encounters typical for street life with the regulated, clearly planned event-life that characterizes neighbourhood life in Southfield. People do not just go out, strolling around, knowing that they will always meet someone who is up for some small talk. In a similar manner, Ferda nostalgically talks about the good neighbourly relations in Northend: I hated it earlier, when we went grocery shopping with my mom. It was terrible for me, when we went shopping and I always had to stand there half an hour because she always met some friends. And now I miss it. When I am out I would also like to see familiar faces, and stop to have a chat. Now I know what that means. And those are things I miss from Northend. And when I am there now and see a familiar face, I really enjoy it.
Consequently, she enjoys going back to Northend, and spends a lot of time there. The same is true for others such as Varol or Alkalin, who likewise go to their old neighbourhood to socialize, but rarely invite friends over to their place. Northend is their ‘second neighbourhood’ (Briggs et al., 2010: 109), and Ferda and Alkalin even intend to move back there with their families.
The missing foci in the neighbourhood, plus the negative attitudes of the German residents towards their Turkish-German neighbours, explain why they do not figure highly in the networks. As Enginalp puts it: ‘And then you feel that you stand out less among certain kinds of people, and then you start looking for these people to socialize with.’ These people are often family, kin and friends from school, mostly with a Turkish background. The network then looks rather closed, but it does not arise out of a preference to socialize with co-ethnics, or a lack in networking skills. The respondents have both the skill (such as language command) and the will (shown by attempts to establish ties) to build category-crossing ties. The homophilous networks result from failed, not reciprocated, attempts to establish contact with native-Germans.
This can have significant consequences. The lack of a diverse local network might exclude the Turkish-Germans in Southfield from important information, particularly ‘child-related social capital’ (Nast and Blokland, 2014: 495). Moreover, not having a local support network implied that Varol and the other respondents would rather not ask their neighbours for small favours, but seek other ways, which are often more time-consuming or expensive.
However, while some respondents are dissatisfied with their neighbourhood and have homogeneous networks, other respondents in Southfield – likewise dissatisfied – still display some ethnic network diversity. Enginalp and Selim, for example, have native-Germans and ethnic others in their networks. These are not on the local level and result mostly from work or free time. Selim works in a café in a gentrified, multi-ethnic neighbourhood and became friends with many of his customers, which enabled him to ask them for small favours. Enginalp works as a lawyer and politician and thus has various contacts to native-Germans and people with another ethnic background. Therefore, ethnically diverse networks do not exclusively result out of living in an ethnically diverse neighbourhood. Nevertheless, neighbourhood satisfaction is linked to the existence of local contacts, which is why Enginalp and Selim do not evaluate Southfield as their ideal neighbourhood.
Notwithstanding these negative examples, neighbourly relations can be more positive, and get translated into diverse networks. Dalim is a 47-year-old owner of an agency for the brokerage of bi-cultural media hosts. He is married, has two daughters of school age, and has been living in Westgrove since 2008, an area he is very fond of: It’s a nicely bourgeois, regenerating area. And we fit in well and we like living there. It’s central. The schools are good. And it is a Kiez. We have Bismarckplatz on the one side, and a couple of hundred metres further is Lindenplatz, which is a very nice square, with the playground, the church, the café, bistro. Beautiful.
Dalim specifically calls his neighbourhood a ‘Kiez’ – referring to the two squares and places where locals meet. Bedia, who is living in Westgrove with her partner, likewise refers to the local places that facilitate regular encounters, particularly stressing the relations between fellow residents and local shopkeepers: Well, in the shops, I basically have some regular shops, the library, a café where I am often, the bakery, there you always talk to people, I like that. But it’s a Kiez, it’s normal.
Bedia considers the good neighbourly relations as ‘nothing special’. The expectation is to have interactions with fellow residents or local shopkeepers. As seen in the previous part, however, such ‘normal’ interactions do not exist in every neighbourhood. Bedia’s quote shows that good relations are the standard against which different (absent) relations are assessed.
Behçet, a lawyer who is well known in the neighbourhood due to his twin daughters, describes how neighbourly relations, through the shared use of local places: … grow organically. You only go to Lindenplatz, just to have some ice-cream, you meet five, six families, kids, you can gather with them, even if it’s just a hello. There is a quiet, pleasant, familial atmosphere, it’s nice.
In contrast to Southfield, children actually function as brokers of interethnic contact in Westgrove. In contrast to Southfield, the native-Germans in Westgrove likely self-selected themselves into that particular neighbourhood at least partly because of its ethnic diversity. In line with what has been found for white middle-classes in other cities, they value diversity and choose their neighbourhood accordingly. In addition, the built environment of Westgrove facilitates interaction between families, as parents share daily rounds, or stop at the same cafés. Dalim recounts the round he used to take for a long time: In the kindergarten it was like this, I picked up the kids. Then we went our way, passing along this square and you just can’t cross it. You know everyone, all the kids know each other, so you sit down in a café … We have coffee, the kids play and you’re in the Kiez. And that’s really the great thing about this Kiez, and I once called it, like an extended living room.
Repeated interactions in the kindergarten or school, regulated by the hours to drop off and pick up the children, facilitate the work entailed in building and keeping ties (Small, 2009). If the parents use even more foci together, such as for after-school activities, the opportunities to form longer-lasting friendships increase. The importance of foci within a neighbourhood where people can meet was also confirmed by Nast and Blokland (2014: 494) who stress that ‘mixture is more interactive in institutional settings than in neighbourhoods’. Here, the mixture occurs in neighbourhood institutions, but is also practiced in daily life.
While respondents in Southfield describe their neighbours as conservative, the Turkish-Germans in Westgrove find their native-German fellow residents rather open. Dalim thinks of himself as progressive – his wife has the steady, well-paying job, whereas he is self-employed and more flexible so that he mainly takes care of the children. Although he believes that this progressiveness makes him different from other Turkish-Germans, he still adheres to some practices that he attributes to his Turkish background. He is fond of those practices and tries to share them with his native-German friends in Westgrove: There is this typical German garden-fence mentality. People really say ‘this is mine’. And I only experience it with Germans that if you knock on somebody’s door, when you want to borrow something, the door remains the barrier. That is unthinkable for me. I would say ‘come in for a bit’, and then you see what the person needs and you get it. But I think that I made it pretty clear in my inner circle of friends, they don’t do that anymore when I am at the door.
Dalim’s ties to native-Germans still allow him to act out parts of his Turkish background. There are mutual learning processes and his neighbours, some of whom he became friends with, also adopt some of Dalim’s practices. It is the prospect of developing actual friendships, not just superficial ties, that makes people cross boundaries: ‘constructive contact relates more closely to long-term close relationships than to initial acquaintanceship’ (Pettigrew, 1998: 76). When people think that contact remains superficial, they may not engage in the burden of crossing boundaries, which may include insecurities about how the other person might react and behave. Hence, ‘“friendship potential” is an essential … condition for positive intergroup contact effects that generalize’ (Pettigrew, 1998: 76) beyond the one person, and that can be encouraged by repeated encounters with the same people in the neighbourhood.
All those reporting good relations between Westgrove residents also display high neighbourhood satisfaction. Good neighbourly relations, thus, seem to be crucial for being satisfied with one’s place of residence. That good relations actually translate into a support network is not obvious, though. Public familiarity – a form of ‘Vergemeinschaftung’ that follows from people regularly meeting each other in a locality (Blokland, 2003: 93) – does not imply neighbours actually helping out. Selbi, a part-time university lecturer in her early thirties, has two German neighbours she did not know before who she would and did already ask for small favours, such as taking a package from the postman when she is not home. Besides, she already went to her German neighbour to have a cup of tea, and also gave the son a birthday present. Likewise, Lacin, who would not hesitate to ask the neighbours in the same building for small favours. Since moving to Westgrove, moreover, he made friends in the area, whom he would ask for various forms of support.
Contrasting the networks of Dalim (Figure 2) and Varol (Figure 3) clearly illustrates the difference regarding network diversity of a person living in a homogeneous versus a diverse neighbourhood.
Dalim’s network.
Varol’s network.
Based on proximity and exposure, given the high share of native-Germans in Southfield, we would expect Varol to have more native-Germans in his network than Dalim, particularly on the local level. However, the opposite is the case. The only German contact in Varol’s network is a neighbour. Contact with her is rare, and he does not know much about her, such as her educational background. His parents, sister and two friends have no or only lower secondary education. Based on the networks, we could be quick to claim that Varol has a preference for co-ethnics. Going further we could even argue that Varol is not well integrated into majority society, and lacks networking skills. Dalim, on the other hand, seems to be well integrated: the only Turkish contacts in his networks are his parents; all ‘chosen’ contacts are with native-Germans and they have higher secondary education or a university degree. Only his parents have no formal education.
This interpretation would be seriously flawed, though. Varol has the same taste for diversity as Dalim, but did not succeed in translating this into actual practice; hence his network is dominated by co-ethnics. Homophily has complex reasons, and assessing the selective mechanisms has shown that it is not based on a sheer preference to socialize with co-ethnics. What it indicates is social closure from the ethnic majority.
For homophily, the neighbourhood, its built environment, but especially native-Germans’ attitudes towards diversity, make a difference. Owing to a lack of reciprocity in the attempt to establish ties, Varol’s support network is closed and ethnically homogeneous – although he lives in spatial proximity to native-Germans. A neighbourhood with public space, such as Dalim’s, where residents are open to diversity, not only rhetorically but also in reality, can facilitate interethnic contacts and lead to a diverse network. The differences cannot be explained by lacking social networking skills and an a priori preference to socialize with co-ethnics. An explanation for homogeneous networks is reciprocity. Owing to self-selection in residential choice, the willingness to reciprocate contact across categories may differ between residents of different neighbourhoods. If the willingness to build cross-categorical ties exists, the neighbourhood can facilitate this by providing places where people can meet and build such ties.
Summary and Discussion
The analysis started out with the intriguing pattern that Turkish-Germans in neighbourhoods with a high share of native-Germans actually have more homogeneous networks with a high share of Turks than those living in ethnically diverse neighbourhoods. Since all respondents highly value positive neighbourly relations and a sense of community, this pattern cannot be explained by a preference for socializing with co-ethnics. A likely explanation is that Turkish-Germans in neighbourhoods with a high share of native-Germans are perceived with suspicion, if not open racism. Attempts to establish ties with neighbours are not reciprocated. In addition, these neighbourhoods have a residential character and do not provide many spaces for locals to regularly encounter each other. Thus, residents do not become familiar with each other, and the Turkish-Germans are not able to develop a local support network.
The opposite experience exists as well. In ethnically more diverse, but equally socioeconomically advantaged neighbourhoods, Turkish-Germans build category-crossing ties to fellow residents. In these neighbourhoods, encounters and interaction between residents are encouraged through the availability of places such as squares with playgrounds and cafés around. Such familiarity develops into actual forms of support.
For the first group of respondents, the networks of the native-Germans seem to be closed. The Turkish-Germans do not lack networking skills, but their networking attempts are not reciprocated. This has important implications for social capital, since the Turkish-Germans may thus be excluded from networks that might otherwise benefit them and their children. Particularly networks formed in schools may be very important, since they provide information on secondary schools, but also on doctors or other child-related subjects. Since the respondents had the same social status as the other residents and parents, this kind of exclusion is based on ethnic background. This is a difference to prevailing research that has focused on differences in social capital between the lower and the middle-classes, including (actively blocked) streams of information between people of different social backgrounds (Blokland, 2003, 2008). Even when people have the same social status, however, information might be withheld by the ethnic majority group, which is often the more powerful one and thus engaging in opportunity hoarding processes (Tilly, 1998).
These results also imply that the relationship between the number of interethnic ties and the degree and willingness of integration is not as straightforward as still too often assumed. Instead, ‘missing contacts do not always result from migrants’ preferences, but may result from missing opportunities or from a refusal by Germans’ (Sauer, 2009: 239, my own translation). According to the Turkish-Germans’ account, this explanation seems much more likely for network homogeneity than the one claiming a lack of networking skills. The respondents were educationally and professionally successful, spoke the German language and thus the ‘exemplary immigrant’, as asked for in the speech by Angela Merkel or the books by Sarrazin and Buschkowsky mentioned previously. This, however, is not a guarantor for integration, due to the missing reciprocity from native-Germans, of whom some seem to be reluctant to build ties with Turkish-Germans.
Nevertheless, we should not forget that the analysis has also shown the potential of ethnically diverse, family-friendly neighbourhoods, with a built environment that facilitates contact (cf. also Schaeffer, 2014). If the residents in such neighbourhoods embrace diversity, the Turkish-Germans are able to build category-crossing ties. There is reciprocity in the attempt to establish new ties, and this reciprocity extends to mutual learning processes.
Footnotes
Acknowledgements
I want to thank Talja Blokland, Patrick Le Galès and particularly Tommaso Vitale for very helpful comments on earlier versions of this article. Thank you also to the two anonymous referees who provided very valuable comments which greatly helped to improve the article.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.