Abstract
This paper explores the extent to which people of different origins, natives and migrants, come together in everyday life in Europe. Instead of looking at overall ‘perceptions’ and ‘stances’, which are context-dependent and mediated through political-ideological currents and discourses as well as broader patterns of prejudice, we focus on sustained close contacts that suggest meaningful and organic relationships. Since it is most often people of migrant background who are blamed for leading ‘parallel lives’ and ‘not integrating’, we chose to focus on them and their interethnic friendships. Moreover, we seek to understand the relevance and role of the neighbourhood context in the development of those relationships. Despite the expressive fears in public discourses about the supposed negative impact of the presence of immigrants and ethnic minorities on social cohesion, our findings indicate that close interethnic relationships are not uncommon in diverse European cities. They further highlight that the neighbourhood context plays an important role in the first years of migrants’ settlement. Relationships in the neighbourhood develop in less formal social settings and are also less demanding in terms of host-country cultural skills on the part of the migrants, thus giving the opportunity to newcomers to develop close interethnic relationships with natives. Finally, the analysis supports the positive role of diversity at the neighbourhood level in the development of interethnic friendships and stresses the importance of the neighbourhood’s socio-spatial characteristics and its location in the wider urban net.
Keywords
Introduction
Ongoing immigration in European societies amidst the economic and political turmoil at times of globalisation, restructuring and crisis, have given rise to growing concerns over social cohesion. A shift in policy agendas and public debates during the past decade or so have signalled a retreat towards assimilatory views, and a concomitant backlash against multiculturalism and diversity (e.g. Vertovec and Wessendorf, 2010). Alarming talk of ‘inassimilable’ difference and of immigrants and minorities living ‘parallel lives’ are much concentrated in European cities and in specific neighbourhoods that are deemed as ‘problematic’. Such worries, widely reproduced by mainstream media, are often vocalised by politicians publicly announcing the failure of multiculturalism as a viable political project.
A series of mostly quantitative studies have come to confirm such fears, suggesting an overall negative relationship between ethnic diversity and social cohesion. 1 An influential paper by Putnam (2007) has proposed that, at least in the short and medium run, diversity negatively affects social solidarity and social capital. This gave way to a nascent subfield exploring different aspects of this relationship. Yet, findings remain to date contradictory and therefore inconclusive.
Most of these studies are based on attitudinal data collected through opinion surveys – data that may largely reflect local discourses on ethnic relations without providing an accurate picture of the experience of actual everyday interaction (see Jerolmack and Khan, 2014). One may expect that in negatively represented neighbourhoods or in socioeconomically deprived areas concentrating stigmatised social categories (broader) polarising discourses are more prevalent. The prevalence of such discourses, however, does not necessarily preclude the existence of positive interethnic contact locally (Noble, 2011; Wise, 2005). Neither does a more diversity-friendly neighbourhood discourse necessarily indicate sustained interethnic interaction (Simon, 2000; Wessendorf, 2013).
While conclusions are commonly drawn in relation to increasing anxieties surrounding ‘segregation’ and a supposed decline in social capital in diverse Western cities, it is rather surprising how little is known about how diversity is actually lived on the ground (Wise, 2009: 42). Research has shown that the development of interethnic relations in everyday life follows a different logic than that represented in national discourses and assumed by policy-makers (e.g. Pratsinakis, 2014; Simon, 2000; Wimmer, 2004). To be sure, power relations are always present, as are various degrees of intolerance and discomfort towards difference (Pratsinakis, 2014; Wise, 2009). However, despite the alarming talk about immigrants and minorities concentrating in ethnically diverse neighbourhoods and related worries about social cohesion, available evidence shows that people of diverse ethnic backgrounds do get along in shared urban spaces (Lobo, 2010; Noble, 2009; Wessendorf, 2013; Wise, 2009). As Wise (2009: 42) aptly argues ‘the closer one looks, the more it becomes obvious that “parallel lives” are not necessarily the prevailing norm’. In contrast, despite its apparently tenuous, apolitical invisibility, ‘everyday multiculturalism’ in many cases works as a cohesive force which resists and transcends fragmentation and division (Werbner, 2013: 416).
To assess the extent of and circumstances in which local experiences comply to this image, further research is necessary on everyday practices, actual interaction and the development of sustained relationships between people of different origins in multi-ethnic cities. As a step in that direction and drawing on a data set generated from a survey conducted in 18 neighbourhoods in six European cities (see Górny and Torunczyk-Ruiz, 2014), this paper explores the development of interethnic friendships and seeks to understand the relevance and role of the neighbourhood context in their development. Given that the patterns of interethnic contact may differ for people of native or migrant background (Górny and Torunczyk-Ruiz, 2014; Lancee and Dronkers, 2011), and considering that immigrants and their descendants are found at the epicentre of rising concerns over social cohesion, we restrict our attention to them.
Diversity, interethnic contact and social cohesion
Central to the theorising in the field of social cohesion and interethnic relations is a schematic and rather misleading juxtaposition of the so-called contact and conflict theories. 2 Accordingly, the former assumes that intergroup contact leads to the reduction of prejudice while the latter suggests that intergroup contact may actually increase it. Recent research on social cohesion and diversity provides evidence in both directions. Some scholars have concluded that these are negatively associated (Alessina and La Ferrara, 2002; Costa and Kahn, 2003), while others found that diversity improves perceptions of, and relations between, ethnic groups (Marschall and Stolle, 2004; Oliver and Wong, 2003). Along these lines, Putnam’s US-based study (2007) came not only at a time of high politicisation of migration and diversity issues, but also just when a theoretically sensitive critique of previous results started being formulated (e.g. Hooghe, 2007). Putnam interpreted his findings through a framework which he called constrict theory. Contrasting the contact hypothesis, which in his reading predicts that diversity erodes the in-group/out-group distinction and enhances out-group solidarity, and the conflict theory, which predicts the opposite, he proposed that diversity in fact reduces both in-group and out-group solidarity and social capital.
His argument has stimulated much debate, ranging from criticism (e.g. Portes and Vickstrom, 2011) to replications of his study in Europe. The latter largely refute Putnam’s thesis (Gijsberts et al., 2012; Vervoort et al., 2011) with the exception of the impact of (ethnic) diversity on trust, generally found to be negative (Lancee and Dronkers, 2010, 2011; Laurence, 2011). This growing body of work nevertheless provides contradictory evidence, partly deriving from different data sets, variables explored, analytical procedures and interpretations. For instance, with respect to what is actually analysed, whilst some studies examine ‘trust’ (variably measured) as their dependent variable (Lancee and Dronkers, 2010, 2011), others focus on composite indices of social capital (including trust) (Gundelach and Freitag, 2014; Letki, 2008), social cohesion (including categories of social capital) (Gijsberts et al., 2012; Laurence, 2011), interethnic attitudes (Havekes et al., 2014) or neighbourhood attachment (Górny and Torunczyk-Ruiz, 2014). Similarly, with respect to findings, the structure and direction of causality differ: rather than diversity, there may be socio-economic factors negatively affecting the above dimensions of social cohesion, such as deprivation (Letki, 2008), disadvantage (Laurence, 2011) or neighbourhood decline and disorder (Havekes et al., 2014).
Few recent studies on diversity and social cohesion take into account actual contact between people of different backgrounds. When they do, this comes into the analysis in the form of independent variables or items in composite indices of social capital, e.g. concerning the respondents’ extent, frequency and/or quality of (interethnic) contact (Gijsberts et al., 2012; Gundelach and Freitag, 2014; Lancee and Dronkers, 2011; Letki, 2008; Putnam, 2007). In other cases, interethnic contact is examined in its mediating role in the relationship between diversity and social trust, or between diversity and tolerance (Górny and Torunczyk-Ruiz, 2014; Laurence, 2011). Only few studies in this strand of research have directly examined actual interpersonal contact as a dependent variable (Kouvo and Lockmer, 2013; Vervoort, 2012; Vervoort et al., 2011). Even scarcer is a focus on close interethnic contacts, i.e. ‘strong’ ties such as friendships.
Regarding immigrants’ interethnic friendships in particular, evidence from the Netherlands and Canada suggests that these concern mostly relationships with the majority population, which are subject to time and relate to immigrants’ integration (Martinovic et al., 2009, 2011). The latter partly depends on the local context of settlement as a field of encounter and socialisation. As emerging from our own survey, the neighbourhood appears to be a focal point of immigrants’ social life (Schnell et al., 2012) and thus a key field where bridging ties may potentially develop. To the extent that the local context may determine interethnic contact, some of the above-mentioned studies tend to focus on neighbourhood population structures. Accordingly, Vervoort et al. (2011) and Vervoort (2012) found that ethnic concentration positively influences interethnic relations among different minority groups but not with natives (Vervoort, 2012; Vervoort et al., 2011), while Martinovic et al. (2011) concluded that living in a less concentrated area increases interethnic contact. In these studies however, the dependent variable was a measure of respondents’ interethnic social ties in general and not of those that developed in the neighbourhood itself. This makes the examination of the relationship between ethnic concentration and the development of interethnic relations problematic.
This paper aims at critically contributing to this debate by focusing on the development of actual interethnic contacts and the role of the neighbourhood context. More concretely, we focus on the close interethnic ties, i.e. friendships, of people of migrant background, which we approach as a positive situation reflecting pathways to integration, social interaction and participation. We further investigate the role of the local context, in our case the neighbourhood, as a site of meaningful encounter and contact, and examine which neighbourhood characteristics seem to foster interethnic friendships. Building on the same data set as Górny and Torunczyk-Ruiz (2014), which is unique in differentiating between immigrants’ interethnic ties that developed in their neighbourhoods of residence from those formed elsewhere, we go beyond conventional assumptions that the more interethnic relations the residents of a neighbourhood have, the more inducing the local characteristics are in the development of such relations.
To the best of our knowledge, this study is also the first on interethnic friendships to be based on an international data set comprising different neighbourhood types across six European cities, making our finding more generalisable and providing a measure of the development of interethnic friendships in European urban neighbourhoods. As such the paper further contributes to another strand of recent literature, that on everyday multiculturalism and interethnic belonging in the city (Lobo, 2010; Noble, 2009; Werbner, 2013; Wise, 2009; Wise and Velayutham, 2009), with which the current paper shares a common critical stance towards the problematisation of ethnic diversity.
The nascent literature studying diversity and social cohesion in the tradition of social capital research initiated by Putnam (2007) has been blind to qualitative accounts of everyday encounters, neglecting insights gained from empirical work on local sites of habitual contact. In so doing, it tends to reproduce mainstream views of multiculturalism discussed from a top-down perspective, as a set of policies concerned with the management of diversity by nation states. In this perspective, the problematisation of diversity implies specific policy remedies for ‘immigrants and minorities living parallel lives’, which are however based on a moralistic discourse of conditional tolerance promoting national cultural norms and whiteness as prerequisites for social cohesion (e.g. Amin, 2002; Fortier, 2008; Phillips, 2007). Such a discourse and related policies forces racialised ethnic minorities to ‘integrate’, rather than addressing society as a whole and setting the conditions that would foster meaningful intercultural interaction in shared urban spaces of multiethnic coexistence.
Recent literature on everyday multiculturalism has inquired into those conditions paying most attention to everyday encounters occurring in ordinary spaces and situations in the ebb and flow of daily life (Blokland and Nast, 2014; Noble, 2009; Wise, 2009). Although this literature shows that diversity is being experienced as a normal part of everyday social life in many places, it does not provide a measure of the degree to which everyday encounters translate to sustained close relations and the conditions under which this happens. Studying the development of interethnic friendship is important because it is through more regular and sustained relationships that meaningful intercultural dialogue can take place. Pettigrew and Tropp (2006) showed that contact in the form of cross-group friendships is more strongly associated with positive intergroup attitudes than other forms of contact. More recent experimental longitudinal studies support this finding evincing of a causal relationship between cross-group friendships and positive out-group evaluations (Davies et al., 2011; however, see Matejskova and Leitner, 2011). Still, although research evinces of the positive consequences of interethnic social ties and friendship for social cohesion, relatively little is known about the degree and determinants of such ties (Fong and Isajiw, 2000; Sigelman et al., 1996). Our study aims to contribute in this respect by exploring the development of interethnic friendships in diverse European neighbourhoods and understanding the relevance and role of the neighbourhood context in their development.
Data and methodology
Our analysis is based on data generated through the GEITONIES (meaning ‘neighbourhoods’ in Greek) international research project. A survey was conducted during 2009–2010 in 18 neighbourhoods in six European cities, namely, Lisbon, Bilbao, Thessalonica, Rotterdam, Vienna and Warsaw, offering examples from the different migration experiences of northern, southern and Central/Eastern Europe. Three neighbourhoods were selected in each city, all with a share of migrants higher than that of the whole city but with different characteristics in terms of their socio-demographic and urban structure (for an overview of neighbourhood profiles see Górny and Torunczyk-Ruiz, 2014: 6–7). The survey generated a randomly selected sample of approximately 200 respondents per neighbourhood, making up a total of 3668 individuals, split in about half between people of native and migrant background. The latter group includes the ‘second generation’, i.e. native-born respondents of immigrant background and those who arrived in the host country as minors up to six years of age, comprising 16.5% of the ‘migrant’ sample (though largely concentrated in Vienna and Rotterdam).
The survey gathered a wealth of information, including details on the respondents’ social networks. Concretely, the GEITONIES study looked at two types of interpersonal relationships: the respondents’‘overall social contacts’, i.e. their wider social circle, and their ‘most important’ ones, i.e. their network of ‘close’ relationships (including relatives) outside the household. In this paper, we focus on this latter type of networks. Respondents were asked to list up to eight people according to four broad categories of contacts: (a) people they spend their free time with, (b) people they would ask for advice in important decisions (and vice-versa), (c) people they give or receive help, and (d) other people close or important to them. The mean number of reported contacts for respondents of migrant background was as low as 3.1, suggesting that the respondents reported the core of their social networks. For each of these contacts, the survey recorded key information about their socio-demographic profile, the context and circumstances of meeting, and the patterns of contact in two periods: ‘then’ (at the time respondents moved into the neighbourhood) and ‘now’ (at the time of the survey). Respondents have not reported themselves whether their close contacts are of the same background or not; this was estimated by comparing one’s origin to that of her/his contact, using parents’ country of birth as a proxy for both while correcting for the second generation.
Strong interethnic ties are found to be much more frequent among respondents of immigrant background; 52.4% had at least one close ‘interethnic’ contact (compared to just 9.5% among natives), and for the majority (86.5%) these include relationships with natives, with 31% maintaining close ties with natives only. It thus appears that strong interethnic ties are not uncommon among migrants and their descendants in European cities. Given that immigrants’ social networks tend to be more concentrated spatially, the neighbourhood appears to be an important field for socialisation for them, yet only partly for the development of strong interethnic ties. Among immigrants’ total number of strong interethnic ties, 17% were originally met as neighbours, while the majority (38.5%) met as colleagues or fellow students.
Approximately one-third of the 815 immigrants with ‘interethnic’ friends have first met (at least one of) those contacts in their neighbourhood of residence. Restricting our attention to these contacts only, we find that their distribution along the 18 neighbourhoods does not follow the same pattern as the general distribution of close interethnic ties. In fact, in the areas with the highest shares of immigrants having ‘interethnic’ friends, the majority of such contacts were first met outside these neighbourhoods; hence the areas themselves do not seem to foster the development of neighbourly interethnic relations (but simply concentrate more immigrants with interethnic contacts). The reverse is the case for (some of) the areas with low shares of immigrants generally having interethnic ties, but most of these ties were actually first met in the neighbourhood. Yet, in this latter case, the neighbourhoods with the highest neighbourly interethnic relations share similar characteristics in terms of their location in the urban structure, built environment, economic functions and local migrant concentration; and so do those with the lowest shares of locally formed interethnic ties. Three hypotheses could be drawn out of these observations:
It is possible that the urban environment itself influences sociability at the neighbourhood level with the neighbourhoods having more dense social infrastructures and more socially inviting public spaces inducing the development of interethnic relations.
Location and hence distance from and access to the urban core may play a role in driving residents to use the neighbourhood for free time activities, thus providing more opportunities to meet each other.
When immigrants form negligible numerical minorities in the neighbourhood they may tend to stay within co-ethnic social circles.
Our subsequent analytical strategy is as follows. First, we identified the individual characteristics of migrants maintaining strong interethnic ties across the six cities. Then we performed a Factor (Principal Component) analysis, allowing us to classify our neighbourhoods by a number of contextual features (including those deriving from the above hypotheses), thus investigating the role of the neighbourhood in the development of strong interethnic ties. Finally, restricting our attention to those interethnic relations that developed in the neighbourhood, we combined immigrants ‘individual’ and ‘contextual’ neighbourhood characteristics in a logistic regression model, in order to shed light into the mechanisms influencing the development of intimate interethnic relations at the local level.
Individual characteristics
We first explored the extent of close interethnic ties among people of migrant background in our sample, by cross-tabulating their key characteristics with the existence or not of interethnic contacts. A number of statistically significant results allow for some initial observations, with respect to their socio-demographic and migratory profile:
Interethnic contacts are more frequent among the second generation. The share of second-generation immigrant respondents having at least one contact of different ethnic background is nearly 90%, while the respective share for the first generation drops to 44.8% (N = 1535, Pearson’s Chi square= 178,644, p < 0.05).
Among first generation immigrants, interethnic friendship is expectedly more common among those longer settled in the host country: more than half of those who had lived there for at least 20 years had some interethnic tie, compared with just one out of four of those who arrived three years prior to fieldwork (Pearson’s Chi square = 31,216, p < 0.05).
While we found no differences by gender and family status, certain family arrangements seem to matter. Having a partner of different ethnic background (about one-fourth of those married) appears to play a role: the share among them who have interethnic ties is 74.2%, nearly double than those whose partners are of the same origin (N = 1226, Pearson’s Chi square = 80,169, p < 0.05), while in most of these cases (66.1%) contacts with people of different ethnic background were actually formed after the relationship was initiated. On the other hand, the proportion of migrants with interethnic ties is far higher (63%) among those who do not have any children as compared with parents (47.4%) (N = 1016, Pearson’s Chi square= 69,372, p < 0.05).
Such evidence confirms that interethnic intimacy is a dynamic process associated with broader trends of migrants’ incorporation. This latter also includes factors pertaining to the institutional and socio-cultural environment and the migratory histories of the national and urban contexts of our study. Therefore, the extent of interethnic friendship was also explored in an additional set of immigrants’ characteristics that inevitably relate to such broader settings and arrangements. Although statistical significance was not the case here, a number of noteworthy patterns emerged.
We first assessed the impact of immigrants’ religious beliefs and ethnic background on interethnic relations. Neither religion nor ethnicity were taken as individual ‘attributes’; rather, they were treated relationally. Religion was examined in relation to the dominant religion in each city (weighting the relevant responses of natives). Accordingly, the frequency of interethnic contacts among immigrants belonging to a religion different to the dominant one is generally lower than among those who share the host country’s religion(s) or who are not religious at all. As a measure of ethnicity we grouped together countries of origin in two categories; those that have some sort of bond (historical, colonial, etc.) with the countries of residence and those that do not. Findings differed across cities, with interethnic contact being more frequent among immigrants from former Portuguese colonies in Lisbon, Soviet Greeks in Thessalonica and Germans in Vienna but not for immigrants from former Spanish or Dutch colonies in Bilbao and Rotterdam, respectively.
Next we examined interethnic contacts in relation to immigrants’ language skills, focusing on the first generation only. Accordingly, interethnic friendship is expectedly more frequent among people who command the host country’s language better, yet with the paradoxical exception of those who speak the host country language as their mother tongue, for which significant differences across counties were recorded. Language, therefore, is important but its significance on the development of interethnic relations is far from self-evident and may be seen as context-bound.
We also sought to examine the role of immigrants’ position in the host country polities, by looking at the extent of interethnic contact in broad categories of migrants’ legal status (namely citizenship, long term or permanent status, temporary status, and no permission). Results show that the more stable, safe and secure the migrants’ legal status is the more frequent close interethnic contact becomes, and this is a pattern in all cities despite differences in the distribution of different categories of migrants’ legal status.
Finally, we looked at key socio-economic characteristics, namely education and employment. With respect to the former, we observe an overall pattern of more frequent interethnic contact among immigrants with higher educational levels (ISCED categories), yet one that is not confirmed in all cities (the exceptions being Lisbon and Warsaw). By contrast, immigrants’ position in the labour market allows for some interesting insights (e.g. interethnic contacts are more frequent among students, retired persons, migrants employed as executives and professionals, but also skilled blue collar workers, and much scarcer among those involved in housework – mostly women – and the unemployed), but do not overall suggest any consistent pattern and there is a good deal of divergence across cities.
Neighbourhood characteristics
In order to specify which neighbourhood characteristics are to be considered, a factor analysis was performed to reveal underlying correlations between variables that could be used for a classification of our neighbourhoods. We tested 12 variables describing the urban, social, economic and demographic profile of our case study areas: four from secondary sources and eight from aggregated data derived from the GEITONIES survey. These latter were weighted according to the respective shares of native and immigrant residents in each neighbourhood, based on official statistics, so that our aggregated data are representative of the actual population. The selection of variables aimed at capturing different neighbourhood characteristics, such as socio-economic and demographic features, spatial attributes and residents’ perceptions, as illustrated in Table 1.
Variables describing the social and economic profile of case study areas.
Notes: aGiven the lack of GDP data at the neighbourhood or even urban level for all cities under study, we retreated to NUTS II, fully aware of the bias this may entail, yet confident that this should reflect the GDP per capita in the six metropolitan areas.
Length of residence was recalculated based on the respondents’ moment of entrance in each neighbourhood. For participants born in the neighbourhood, the respective length of residence equals their age.
The index was constructed by calculating a weighted average of responses to three questions on the extent and frequency of neighbourly relations: whether respondents personally know their neighbours, measured on a five-point Likert scale (1 = strongly agree, 5 = strongly disagree), and whether they engaged in small talk or visits with neighbours, both measured on a five-point scale indicating frequency (1 = more than 20, 5 = none).
A principal components analysis combined with a direct oblimin rotation method produced four components explaining 77.6% of the total variance in our data. 3 Given that the last component consists of one variable only (Educational similarity), it was excluded from our analysis. The retained three components according to which we have grouped our neighbourhoods are summarised in Table 2. Since a suitable rotated solution was reached, component scores were estimated for each neighbourhood, using the regression method. Given that these scores represent standardised values (mean equals 0 and standard deviation equals 1), the 18 Neighbourhoods were classified by receiving a ‘low’, ‘moderate’ and ‘high’ value corresponding to the respective components scores (< −1, −1 to 1, >1).
Rotated Components and description.
More specifically:
Three neighbourhoods scored exceptionally high on Component one, namely San Francisco (Bilbao), Afrikaandewijk (Rotterdam) and Nikopoli (Thessalonica). These are neighbourhoods with large shares of immigrants among the total population which are considered to be problematic in terms of reputation and safety (according to immigrant residents’ perceptions), while they are also characterised by high unemployment levels. On the other hand, all Warsaw neighbourhoods and Rekalde in Bilbao received low scores on this Component, given the exceptionally low shares of immigrants there and the respondents’ positive views in terms of safety and reputation. The remaining 11 neighbourhoods were classified as ‘moderate’ (values between −1 and 1).
All Viennese neighbourhoods and Schiemond in Rotterdam were assigned a ‘high value’ in Component two, denoting areas with high GDP/capita and long settlement histories. On the other extreme, Nikopoli, San Francisco and Szczesliwice (Warsaw) are areas with recent histories of immigrant settlement and belong to poorer regions in terms of GDP/capita; hence they received a ‘low’ value in this Component. The remaining 11 neighbourhoods were categorised as ‘moderate’.
Finally, Peraia (Thessalonica) Hoogvliet Noord (Rotterdam) and Costa da Caparica (Lisbon) were grouped together based on their high scores in Component three. These are neighbourhoods located far from the city centres, yet characterised by strong neighbourly relations among residents. San Francisco, Ludo-Hartmann-Platz (Vienna) and Szczesliwice received a ‘low’ value while the remaining 12 neighbourhoods were assigned a ‘moderate’ value.
The neighbourhood as micro-context of interethnic friendship
The derived component scores described above were used to create three new categorical variables (neighbourhood scores on Components 1, 2, 3), each coded with values ‘low’, ‘moderate’ or ‘high’, based on the respondents’ neighbourhood of residence. These new ‘exogenous’ variables were included together with the ‘individual’ immigrants’ characteristics found to affect interethnic relations, in addition to gender and socio-economic status, in a logistic regression predicting the variables which may explain (some of) the underlying mechanisms that foster the development of close interethnic ties locally. The model employed 12 independent variables altogether (see table 3) and the dependent variable is a dichotomous one (Yes/No) distinguishing between immigrants who met their close interethnic contacts in their neighbourhood of residence and those whose interethnic ties were established outside the neighbourhood.
Individual characteristics variables employed in the logit model.
The logit model was applied by requesting a forward stepwise (likelihood ratio – LR) method, since no reliable assumption was evident regarding the (accumulated) effect of the predictors (Menard, 1995). As a result of missing values in several of the employed variables, 528 out of 815 immigrant respondents with interethnic ties were ultimately included in this analysis. 4 Five steps were necessary to reach a reliable solution and five variables, namely ‘Language skills’, ‘Length of residence’ (in the host country) and ‘Neighbourhood scores’ on the three Components present significant values on Wald statistic, which suggests their importance as predictors in our model. The remaining seven variables were not retained, since their effect was not statistically significant. Table 4 provides a summary of these findings (obtained in the last step), showing the mean predicted chances (probabilities, expressed as percentages) for the ‘average’ immigrant resident of a neighbourhood to establish interethnic relations locally.
Summary of logit regression results.
Notes: Nagelkerke’s R2 = 0.243. x2(13) = 99.840. p < 0.01.
Figures in brackets compare the respective variable categories with the base category (0).
The model replicated the rather paradoxical earlier finding that being a native speaker in the host country’s language impacts negatively the development of interethnic relations also at the neighbourhood level. It reveals though an interesting local particularity: migrants whose language skills are ‘poor’ or ‘good’ have the same enhanced probability to develop interethnic ties at the neighbourhood, and they are 2.6 times more likely to make intimate interethnic contacts locally as compared with those speaking the host country’s language as mother tongue. Given that these two language skill categories are the most common among immigrants in our sample, it may be argued that linguistic skills do not constitute a significant barrier in the development of close interethnic contacts at the neighbourhood level, even if they do play a role in interethnic ties overall.
Equally contrasting to our descriptive analysis are the results on the migrants’ length of stay in the host country. While it was generally found that the more years immigrants have been in the country the more frequent their interethnic ties are, rather the opposite holds true for interethnic contacts met at the neighbourhood. Immigrants who have been in the country four to five years may be 4.3 times more likely to have neighbourhood-based interethnic contacts compared with those just arrived (base category), but also to those living in the country for six to ten years (3.2 times more possible compared with the base category). Moreover, those living in the country for 11–20 years are approximately as prone to interethnic friendship as those who are settled for more than 20 years (2.3–2.5 times more likely compared with the base group); yet both are considerably less likely to have neighbourhood-based contacts compared with those who are in the country for three to ten years. Excluding immigrants who have arrived very recently, the length of residence seems to be negatively affecting the probability of developing close interethnic contacts at the local level.
How are we to make sense of these results? The neighbourhood appears as a key place for the development of social relations with ‘others’ (i.e. mostly natives) during the first years. Through the passing of time, and as immigrants become more involved in the host society, engage in more social fields and are more mobile in their everyday lives, they face enhanced opportunities to meet people of different ethnic backgrounds and establish interethnic contact that might later develop to intimate relationships. Taking into account the results concerning language skills, it appears that the neighbourhood is a less demanding place of interethnic socialisation. Although poor language skills, generally a common barrier for newcomers, influence the development of interethnic contact with natives and other immigrants, this does not seem to be the case at the neighbourhood level. This may possibly be because relationships in the neighbourhood develop in less formal social settings which are also less demanding in terms of host-country cultural skills on the part of the immigrants.
Turning now to the impact of neighbourhood characteristics, immigrants who live in neighbourhoods that received a ‘moderate score’ on Component 1 (‘diverse and deprived neighbourhoods’) are 5.2 times more likely to develop interethnic relations locally, as compared with those living in neighbourhoods with ‘low’ scores in this Component, i.e. wealthier areas that are less diverse and thus do not offer opportunities for interethnic contact. Similarly, in neighbourhoods scoring ‘high’ (that is, very diverse but with social problems), residents are also more likely (3.5 times) to develop interethnic ties there. This is in line with the observation made earlier that when immigrants form negligible minorities and when diversity levels are low locally, they tend to stay within co-ethnic social circles.
The results on the effect of Component 3 (‘remote and sociable neighbourhoods’) support the hypothesis suggesting that immigrants settled in remote neighbourhoods (away from the city centre), with higher levels of neighbourly relations are (1.5 times) more likely to develop interethnic contacts within their residential areas, compared with those in more central districts where social interaction is less frequent (even though this effect is only statistically significant at the 0.1 level). A local neighbourhood ‘culture’ of socialisation can thus also foster interethnic relations. At the same time, neighbourly relations in this case are more intense in suburban areas rather secluded from the centre, where residents spend more time in the neighbourhood (e.g. in public spaces), highlighting the role of location and spatial characteristics.
Conclusion
Despite the expressive fears in public discourses about the supposed negative impact of the presence of immigrants and ethnic minorities on social cohesion and about immigrants living parallel lives, our findings indicate that close interethnic relationships are not uncommon in European urban neighbourhoods and that they mostly concern immigrants’ contacts with natives. In accordance to previous research (Martinovic et al., 2009, 2011), they further highlight that the most significant determinants are directly or indirectly related to time. This is not to say that there are no tensions related to cultural misunderstandings, growing socio-economic inequalities in cities – increasingly with ‘ethnic’ characteristics – the resurgence of racism and natives’ demands on migrants to comply with national norms. Yet, as our analysis highlights, early experiences of migrants’ settlement – especially in respect to developing contacts with natives – are subject to change in the short and medium run, hence far from indicative of their integration prospects, and thus should not be interpreted as signs of integration ‘failure’ or dismantling of social cohesion.
Migration destabilises and diversifies the structures of people’s affiliations and social networks, and devaluates much of the human and cultural capital they possess. Apart from their position in the class structure and institutional arrangements in their place of settlement, immigrants’ social integration is also typically constrained by native perceptions questioning the belongingness of immigrants in the ‘host society’ and placing them in an inferior position (Pratsinakis, 2014). However, despite those constrictions, over time immigrants do get intertwined with the wider society, and also develop close interethnic ties.
Concerning the role of neighbourhoods in particular, although they appear to be important fields for immigrants socialisation (Schnell et al., 2012), our findings indicate that their role in the development of strong interethnic ties is limited. As previous studies have shown, everyday cross-ethnic encounters in the neighbourhood do not easily translate to actual close relationships or friendships especially when the ethnic divisions are also coupled with class divisions (Blokland and Eijk, 2010). Thus policy-makers should not put too high hopes on government-sponsored mixing interventions, especially since such interventions reproduce stigmatising perceptions of immigrants’ concentration as a problem to be solved. Such mixing interventions sustain ideas of ‘immigrants and minorities living parallel lives and not wishing to integrate’ 6 while in practice, as Fortier argues (2008), they tend to be counterproductive calcifying cultures within boundaries across which one should mix.
However, policies could instead be directed towards facilitating the process of social integration by supporting both immigrants and natives, prioritising neighbourhoods that concentrate newcomers. Such measures may emphasise on local residents’ access to the labour market or welfare services, or on planning interventions targeting the urban and social infrastructure (which are often left to degrade, contributing to neighbourhood stigmatisation). Equally important are measures nurturing interethnic dialogue, easing feelings of distress among established residents while countering racist stereotypes. Such neighbourhoods may pose actual challenges for policy-makers but they also provide potentials for the fostering of immigrant–native relations: as our findings indicate, the neighbourhood appears to be an optimal setting for the development of interethnic relations among migrants in their early phases of settlement. Relationships in the neighbourhood develop in less formal social settings and are also less demanding in terms of host-country cultural skills on the part of the immigrants, thus giving the opportunity to newcomers to develop close interethnic relationships with natives. Over time the neighbourhood gradually loses its significance as a setting for interethnic contact, as immigrants become embedded in the host societies, acquiring all these skills and capabilities that may allow them to expand their social networks beyond their area of residence.
Finally in respect to the local context, our findings support that ethnic concentration and diversity plays a positive role. Notwithstanding the fact that high immigrant concentrations are often associated with negative reputation of those areas, our model revealed that this does not affect negatively the development of interethnic relations as such. We do see interethnic contact developing in highly mixed neighbourhoods rather than in districts with low migrant presence, which in our sample tended also to have a good reputation. Lastly, the analysis stresses the importance of the neighbourhood space itself and its location in the wider urban net (i.e. physical characteristics and position/function in the conurbation) – both of which are almost completely ignored in the relevant literature and necessitate further research.
Footnotes
Acknowledgements
We are grateful to our GEITONIES project partners who commented valuably on early versions of the analysis and arguments presented in this paper. We would also like to thank the Urban Studies Editors and three anonymous referees, for their valuable comments and suggestions that helped us improve the paper.
Funding
The GEITONIES project was financed under the Seventh Framework Programme of the European Community for Research, Grant Agreement number 216184.