1 Tackling the Dilemma of Local Variations in Ethnic Conflict and Integration

This is a book about social change in European cities as brought about by international migration. Among the many aspects of such long and complex waves of transformations what interests us in particular is how relations are structured and how they evolve in different and increasingly diverse local societies. The main research questions addressed in this book are thus the following ones: why do origin-based categories emerge much more clearly and powerfully as practical and symbolical boundaries in certain local contexts rather than in others within the same national and regional space or even within the same city? How much do urban contexts count in shaping inter-group relations and specifically in making ethnic categories more or less salient?

Given the fact that central aspects of our research questions, such as relations between the majority population, immigrants and their descendants and the salience of ethnic boundaries are generally considered to be key dimensions of integration processes though with a different emphasis in American and European literature (Gans 1992; Portes and Zhou 1993; Massey 1995; Esser 2001; Entzinger and Biezeveld 2003; Alba and Nee 2003; Zincone 2009), this is also meant to be a volume on immigrant integration in European cities.

Since ethnic tensions and conflict have been a major concern in the urban policies of several European cities over the last two decades, we pay special attention to these dynamics in our understanding of integration. Specifically, though we agree that positive relations and loss in salience of ethnic boundaries are crucial aspects of successful integration, we understand the latter not as a rigid state of peaceful coexistence conceptually opposed to conflict but as a dynamic achievement, a process that implies ongoing negotiations on the idea of ‘us’, whereby re-negotiation is possible. Accordingly the research project upon which this book is based is titled ‘Concordia Discors’,Footnote 1 an expression originating in the epistles of the Latin poet Horace, now paradigmatic of a dynamic state of ‘discordant harmony’. Actually, the idea that conflict is part of the dynamic process of social change that integration entails has belonged to the sociology of migration since the very beginning. The school of urban ecology of the University of Chicago, which began to deal with immigration in the early 1920s as part of an analysis of urban transformations, maintained that conflict does not necessarily have negative implications but, on the contrary, is an important step of the integration process through which groups become aware of their identity and their specific needs, and are able to make claims in terms of access to resources and rights (Park and Burgess 1921). However, in distinction to the Chicago school, here the function of conflict in the process of integration is not taken for granted or located within a staged path. We started from the assumption that conflict does not always progressively disappear and sometimes breaks out, even suddenly. Conflict can sometimes be an opportunity to know each other, to cross and eventually change group boundaries thereby enlarging the concept of ‘us’. At other times it makes such boundaries neater and more impermeable.

Like all societal phenomena, integration varies through time and space. The degree of integration between individuals or groups is unquestionably affected by the course of time and by the succession of generations. But it is also deeply influenced by the particularities of place,Footnote 2 as clearly integration is powerfully shaped by the specificities of the economic, political and spatial context in which the encounter occurs. However, problems arise when one tries to be more specific in determining the respective influence of different geographic scales on integration processes.

Throughout the 1990s, immigration scholars in Europe focused on the nation state as the key level for understanding processes and policies of immigrant integration (Brubaker 1992; Castles and Miller 2003; Schnapper 1992; Soysal 1994; Zincone 1991; Böhning 1984; Heckmann and Schnapper 2003). In the mid-1990s this focus on the national level started to shift, following a growing scientific and political awareness of the fact that most immigrants live in cities and their integration takes place primarily at the local level.

Since then, empirical migration studies have increasingly been focusing on the city level. This is in line with broader trends which have been making cities ever more crucial to understanding social and political phenomena, especially since the crisis of Fordism at the end of the 1970s opened the way to economic restructuring and new forms of governance (Kazepov 2005). The increasing complexity of society and social demands has indeed made it more difficult for central governments to impose rules from the top, thus fostering a ‘hyperlocalisation of the social’, i.e. the relegation of the treatment of complex social problems – including the integration of minorities – to the local sphere (Body-Gendrot and Martiniello 2000).

However, not only states but also contemporary cities have proved to be too internally fragmented and heterogeneous to be investigated as undifferentiated units: suburban nineteenth century working-class neighbourhoods are very different from inner gentrified districts, social housing areas built in the 1960s are hardly comparable to more recent business city centres and so on.

From our standpoint, this means that the integration outlook may vary significantly from one place to another, even within short distances. As Wallman (2003, 2005) highlights, in any city there are areas where diversity leads to private and public benefits and to sustainable development, while in others it may bring about social tensions, segregation and economic stagnation. Neighbourhoods do indeed play a decisive role within each city as high-visibility testing grounds of integration, i.e. as concrete contexts where integration ‘succeeds’ or ‘fails’, ‘microcosms at the epicentre of larger problems’ (Body-Gendrot and Martiniello 2000, p. 2). In positive terms, despite globalization, neighbourhoods as a social arena perform an important however specialized role as, for instance, the site of routinized practice and ‘repair work’ for everyday life and a source of social identity in addition to other functions (Forrest and Kearns 2001; Guest and Wierzbicki 1999; Henning and Lieberg 1996).

Given these assumptions, we believe that it is time to go beyond not only methodological nationalism (Wimmer and Glick Schiller 2002), but also what could be called methodological ‘city-ism’ (meant as a rigid and sometimes exclusive focus on the city level as the context for immigrant integration), and to regard neighbourhoods and even specific sites of interaction within neighbourhoods as a primary and crucially important level for integration. Some steps in this direction have already been taken. In European migration research the neighbourhood level initially gained attention in the field of segregation studies. Transferring frames and interests from North American literature, with its long history of investigating the formation of urban ghettos, that strand of research clearly showed that European cities, generally more heterogeneous than those in the US, experience lower levels of segregation (Musterd and Ostendorf 1998; Fortuijn et al. 1998; Peach 1996, 1998; Phillips 1998; Barbagli and Pisati 2012; Fainstein et al. 1992; Mollenkopf and Castells 1991).Footnote 3

Nevertheless, spatial segregation has become and remains an important issue on the political agenda of many European cities and it is increasingly accompanied by attention to immigration issues since in disadvantaged neighbourhoods the share of migrant population is often higher than the city average. Moreover, both segregation and diversity are usually framed as ‘problems’ because of tensions between people of different backgrounds that can sometimes turn into riots, so that policies are often aimed at reducing social and ethnic homogeneity rather than social inequality (Musterd et al. 2000).

Actually, as Forrest and Kearns (2001) pointed out, not only the policies but also the literature on neighbourhoods is still much focused on disadvantaged areas reflecting the policy agenda and providing a rather partial view. In this work we have adopted a partially different perspective using the level of diversity rather than the degree of disadvantage as the key criterion for selecting areas to investigate, and including both deprived and affluent neighbourhoods, with high and low shares of foreign residents. Given the above-mentioned focus on ethnic tensions, we have also made an effort to investigate neighbourhoods with differences in the salience of immigration and ethnic minorities issues, at least in public debate and collective representations, in order to make some empirically-grounded hypotheses on how urban and social dynamics influence the salience of ethnic boundaries and foster conflict or cooperation within the context of everyday life.

In analyzing ‘neighbourhood effects’, we look at the impact of the neighbourhood context on relations rather than on the aspirations, opportunities and actual achievements of individual residents as the neighbourhood effects literature does. As a matter of fact, neighbourhoods may be regarded as a series of overlapping social networks but little is known about differences in local patterns of social interaction within different types of neighbourhoods (Forrest and Kearns 2001).Footnote 4

Nevertheless, far from disregarding the literature on neighbourhood effects, we have paid specific attention to the main factors that this literature considers, such as the distribution of population per housing regime, education and employment status, on which data were collected and summarized in the Tables in Annex I. These data have been mainly used to sketch the contexts of interaction and were integrated with qualitative analysis of spatial dimension, social infrastructures, modes of economic development and integration, political and mass media discourse, history and patterns of migration of investigated neighbourhoods.

Neighbourhoods are here conceived of not just as containers, portions of urbanized land hosting certain amounts and types of population, but also as places with social, institutional, symbolical and physical dimensions (Amin 2002; Galster 2001; Simon 2000; Wessendorf 2010; Tyler and Jensen 2009; Ray et al. 2008; Robinson and Reeve 2006). Partly in opposition to segregation studies, this kind of approach starts from the observation, well made by Wallman (2005), that similar amounts of population mix can lead to very different kinds of diversity. As a matter of fact, the city chapters of this book clearly show that among the several factors impacting on ethnic tensions, migration-generated ethnic diversity itself is far from being the most significant. What matters more is how diversity is framed and played so that stereotypes and xenophobia may be higher in neighbourhoods with a very low proportion of migrants, as in the Hungarian neighbourhoods, or may decrease while the share of ethnic minorities increases as in British neighbourhoods.

Finally, it is worth saying that the focus on the neighbourhood level has also been crucial for the boundary-making approach adopted in this book and explained in the next section. In fact, as Wimmer suggests (2013), using territorial and spatial units such as neighbourhoods – rather than ethnic units and social groups, in general – as units of observation, allows us to observe patterns of everyday boundary-making and group formations without presuming ex-ante that these necessarily cluster along ethnic lines.

2 Eleven European Neighbourhoods: On the Diversity of Diversity

The majority of this book’s chapters present in-depth analyses on specific neighbourhoods located in five European cities: Bermondsey and Camberwell in London (United Kingdom), Poble Sec and Sagrada Família in Barcelona (Spain), Barriera di Milano and San Paolo-Cenisia in Turin (Italy), Werderau, Langwasser and Gostenhof in Nuremberg (Germany), and Józsefváros and Kőbánya in Budapest (Hungary). Not all of them are administrative units and they are part of different institutional settings. In Nuremberg, for instance, the lowest local institutional level is the City – although district coordinators have recently been introduced in some areas. In the case of Turin, too, neighbourhoods (Quartieri) as such are not administrative units as they were merged into ten larger institutional entities (Circoscrizioni) in the 1980s. In contrast, Hungarian, Spanish and British target areas have a more or less significant level of administrative autonomy: Józsefváros and Kőbánya are two Districts of Budapest, i.e. the lowest administrative level below the City; Poble Sec and Sagrada Família are Barris, i.e. the administrative units into which the Districts are divided; Camberwell and Bermondsey are Community Council Areas, i.e. the administrative level below the Borough.

Based on key features of their socio-economic history, our target neighbourhoods can be classified in two broad groups: the first and larger one is made up of former industrial neighbourhoods while the other consists of central neighbourhoods where the service sector (trade, leisure, museums and tourist destinations, etc,) prevails. The latter category includes Camberwell (London), Sagrada Família (Barcelona), Gostenhof (Nuremberg), and Józsefváros (Budapest). The first category encompasses Bermondsey (London), Poble Sec (Barcelona), Barriera di Milano and San Paolo (Turin), Werderau and Langwasser (Nuremberg), and Kőbánya (Budapest).Footnote 5 These neighbourhoods are reacting to the dismantlement of industrial sites differently: some of them seem to manage this change quite successfully, developing alternative social and economic organizational settings while others are trapped in the Fordist economic and social model, and are strongly affected by its decline. We will examine the significance of this more thoroughly in the concluding chapter. Here we will discuss the broader bi-type categorization and its relevance for inter-group relations since it seems to hold significance despite the ongoing transformations.

Indeed, the dissimilarities between these two categories of neighbourhoods produce different settings for inter-group relations. First, in former industrial neighbourhoods people you meet in public places are almost always residents since these areas do not offer special attractions. Furthermore, they are generally rather isolated and disconnected from the rest of the city due to urban barriers mostly related to the industrial past such as abandoned factories, railways which once served plants or huge and heavily congested roads which can hamper relations between people residing in different areas of the neighbourhood and even produce fragmented perceptions of the latter with the subsequent development of strong sub-neighbourhood identities. In contrast, central service-oriented neighbourhoods are characterized by substantial daily inflows of people coming from outside to use services and entertainment facilities. As a consequence, migrants tend to be just one category of outsiders among others, not always the most significant in neighbourhood narratives on the Other and in everyday practices. Furthermore, the centrality of these neighbourhoods usually produces a higher turnover not only in neighbourhood users but also in residents, making them used to settlement of migrant populations from other regions and from abroad.

Secondly, neighbourhoods’ economic and urban functions in addition to their degree of isolation have an impact on their social composition. Former industrial neighbourhoods are usually more socially homogenous both in terms of class and ethnicity, whereas central service-oriented neighbourhoods are more socially mixed as a consequence of differentiated economic activity and a higher turnover of population. The latter are used to diversity in its various forms and have developed socio-economic strategies to cope with it. In former working-class neighbourhoods, on the other hand, social cohesion may constitute an asset for migrant integration, but the nostalgic memory of a cohesive community with a shared history can also foster hostility towards outsiders, who are regarded as a threat.

As we will discuss further in the concluding chapter, these differences in the economic and social history of the neighbourhoods, as well as in their urban texture and the permeability of their borders greatly affect the everyday experience of diversity.

The 11 neighbourhoods analyzed in this book also display differences and similarities in demographic terms.Footnote 6 As we can see in Fig. 1, the demographic balance of all target neighbourhoods is essentially positive, as negative figures generally appear as 1-year exceptions.Footnote 7 The most steady and consistent growth is reported in the Italian and British neighbourhoods and, to a lesser degree, in the Hungarian ones, whereas the Spanish and German neighbourhoods show more swinging trends which alternate positive and negative values.

Fig. 1
figure 1figure 1

Demographic balance rate (%) (2006–2010)

It is worth adding that the migratory balance (see Fig. 2) is positive in all target neighbourhoods except for Gostenhof, whereas the natural balance (Fig. 3) is positive only in Barriera di Milano, Werderau, Gostenhof and marginally in Poble Sec. This means that the population increases shown in Fig. 1 are more often the result of new resident inflows rather than births. Consistently with the broader demographic landscape of an ageing Europe, these neighbourhoods are producing fewer children but since they are still able to attract residents from outside they are avoiding population decline.

Fig. 2
figure 2figure 2

Migratory balance rate (%) (2006–2010)

Fig. 3
figure 3figure 3

Natural balance rate (%) (2006–2010)

If we narrow the focus to foreign residents, we see that their share of the overall population (see Fig. 4) varies significantly even between neighbourhoods of the same city, being highest in Barriera di Milano, Gostenhof and Poble Sec, and very low in other German neighbourhoods and in Hungarian ones. As a matter of fact, compared to the city they belong to, neighbourhoods may be concerned by very different migration dynamics and similarities may be greater between neighbourhoods located in different cities than within the same city, as we said in the previous sections.

Fig. 4
figure 4figure 4

Proportion of people with foreign citizenship in total population – year 2010 (%) (Note: The figure is not available for the two London neighbourhoods)

Nevertheless, we should consider that these data also reflect very different international migration histories of the cities and countries where our target neighbourhoods are located: the foreign population is less numerous both in mature immigration countries, such as Germany, where the bulk of the population with a migratory background has now acquired the nationality of the country of settlement, and in countries such as Hungary, which are not – or at least, not yet – attracting very substantial migratory inflows.

Generally, whereas in most recent immigration countries data on nationality actually mirror the level of origin-based diversity, this is not the case in older immigration countries such as the UK and Germany. Indeed, in these latter countries, data on the residents’ migration background are systematically collected and show a rather different picture from the one revealed by data on nationality. Whereas in our German neighbourhoods the share of foreign residents ranges from 3 % to 11 %, the proportion of people with a migratory background as defined by the Statistik Nürnberg Fürth is 37 % in Langwasser, 48 % in Werderau and 54 % in Gostenhof; similarly the share of residents with an ethnic minority background as defined by the UK Census ranges from 37 % to 44 % in different parts of Bermondsey (Grange, Riverside and South Bermondsey) and from 48 % to 60 % in Camberwell’s wards (Brunswick Park, Camberwell Green, South Camberwell).

Not only the number of foreigners, but also the number of countries of origin is significant. Figure 5 aims at measuring and comparing the degree of ‘statistical ethnic diversity’ in target neighbourhoods by aggregating the first five nationalities in terms of number of residents and calculating their incidence in the total foreign population. Differences are evident: in Hungarian neighbourhoods foreigners are few but come from many different countries since the first five nationalities do not even account for 10 % of the foreign population whereas in Italian neighbourhoods foreigners account for a larger share of the residents but the level of diversity is rather low since the first five nationalities amount to more than 70 % of the total foreign population, primarily as a consequence of the very substantial presence of Romanian citizens. We can assume that these differences between neighbourhoods impact on collective representations of diversity and inter-group relations as well as on the policies dealing with these issues. On the other hand, as briefly discussed above in Sect. 1 and illustrated in greater detail in the following chapters, ‘statistical ethnic diversity’ is often less crucial than expected since its perception is affected by many other factors capable of emphasizing or disguising it in collective perceptions, so to present it in more positive or negative terms.

Fig. 5
figure 5figure 5

First five nationalities (as proportion of total foreign population)* (%) (2006–2010)*Data on foreign residents are not available for British neighbourhoods while in German neighbourhoods are grouped by geographical areas instead of by nationality

Finally, it is worth remembering that these differences in statistical categories are also specifically significant to our analysis of boundary-making, the key tenets of which are illustrated in the next section. As a matter of fact, statistical categories and classification criteria, as part and expression of a given institutional setting, are as much a means of boundary-making as are informal everyday relations (Wimmer 2013). In this sense, public statistics can be seen as macro-frames exactly like the political and media narratives which represent the specific objects of our research. In the same way, statistics reflect the unequal distribution of power among groups who struggle over the boundaries of belonging. As a consequence, they are usually native-biased since the majority population is usually legally, politically and symbolically advantaged compared to immigrant minorities thanks to its privileged relationship with institutions. In fact, categories employed to gather and elaborate data for official statistics have usually been defined by the majority population and they are usually taken for granted with little or no concern about the degree to which immigrant minorities self-identify with those categories. In Great Britain, for instance, this is evident in the use of the category BME (Black and Minority Ethnic). In Italian statistics, those born abroad have traditionally been distinguished into two main groups, i.e. people from developed countries (Western Europe, North America, Oceania, Israel and Japan) and people from ‘high-migration pressure countries’, a totally inconsistent and strongly Western-centric category which basically coincides with the rest of the world. People from Serbia are classified as Serbians in Budapest, but as former Yugoslavs in Nuremberg. It is then clear that who is counted in the diversity statistics, and how, is a matter of definition and boundary-making and ultimately of power. Bourdieu (1985) would have used the concept of symbolic capital and efficacy, i.e. the ‘action that is performed through signs capable of producing social things and, in particular, groups’ (Bourdieu 1985, p. 741). Statistics taken for granted as objective measurements are certainly a clear example of the use of effective symbolic power to build groups.

3 Beyond ‘Groupism’

In most of the empirical literature on ethnicity and migration, ethnic and origin-related boundaries are not treated as outcomes to be explained, but rather, as an independent variable, i.e. something used to explain. In this regard, Brubaker (2002, 2006) speaks about ‘groupism’ to refer to the tendency to take discrete and bounded groups for granted in the study of ethnicity and in social analysis in general. Here we have instead adopted the boundary-making perspective originally proposed by Fredrik Barth (1969) and more recently further developed and applied to empirical research by Brubaker (2004, 2009) and Wimmer (2007, 2008, 2013). According to this approach ethnic distinctions have a relational nature and they may (or may not) coincide with objective cultural differences. We look at ethnicity as a ‘cognitive phenomenon, a way of seeing and interpreting the world’ (Brubaker 2002, p. 184) rather than a substantial entity or feature.

In this perspective, ethnicity can then be regarded as the result of actors defining social relations and constructing social networks and institutions through references to ethnic elements (nationality, culture, language) rather than through other elements such as profession, class or place of residence (Taboada Leonetti 1989). In other words, it is one of the possible ways to apprehend social situations (Esser 2004). Therefore, we did not pre-cluster individuals into ethnic groups since the existence and the configuration of such groups is itself part of our empirical research findings.

In this line of thought a fundamental distinction is the one between categories and groups. While groups can be defined as bounded collectivities with a sense of solidarity, shared identity and sense of belonging, and capacity of collective action, categories are cognitive frames (Wimmer 2013; Brubaker 2002, 2006; Jenkins 1994, 1997). Therefore, as Brubaker (2002, 2006) points out, ethnic categories can exist without groups. In any case, ethnic categories can have real consequences since they influence – in some cases quite strongly – ways of seeing, thinking, talking and behaving.

Considering this assumption, the analyses presented in the following chapters have not started with ‘groups’, but with ‘categories’ asking how, why and in what contexts ethnic categories are used to make sense of everyday life, events, issues addressed in public debate, and so on. In this sense, ours has been a category-centred as opposed to group-centred approach to the study of ethnicity. This approach obviously does not exclude the presence of groups but starts from the idea that ‘bounded and solidary groups are one modality of ethnicity’ (Brubaker 2004, p. 3). Actually, we found many categories and few groups.

As social constructions, groups’ boundaries are situationally defined, and inter-group representations and behaviours do not develop following general rules but vary according to the context (Jenkins 1994; Lamont 2000; Wimmer 2004; Galaty 1982). Therefore, the boundary-making perspective is also consistent with the approach explained above which gives particular relevance to the specific characteristics of places – in our case neighbourhoods, each with its own socio-economic profile, urban texture, social history and identities which all concur in providing a specific setting and specific stakes for social interaction.

Given this perspective, we have chosen not to focus on groups legally or statistically defined on the basis of country of origin or citizenship; nor have we always used the same categories in the city chapters since ‘boundarying’ and labelling activities are themselves objects of analysis, and to import cognitive frameworks from one context to another would have been in contradiction with our initial assumptions. Categories such as ‘minorities’, ‘black’ and ‘white’ belong to everyday frameworks and speech in London’s neighbourhoods but not in Turin’s neighbourhoods where ‘migrants’, ‘Muslim’, ‘Senegalese’, ‘Peruvians’ or other national identities are more usual terms while ‘white’ is regarded as an expression of racist views by a large part of the population. As a result, the following chapters refer to different ‘ethnic categories’ reflecting the most salient boundaries in the investigated urban contexts.

Furthermore, we have tried to develop a broader perspective than the one usually adopted by migration studies, looking at migration as just one facet, however crucial, of the social and urban transformations of contemporary European cities. In keeping with our ambition to go beyond a narrow definition of migration studies we have looked not only at migration-generated ethnic cleavages but also at other cleavages which have emerged from the fieldwork as relevant in structuring groups and reinforcing, weakening or blurring the ones produced by migration. This approach is consistent with the super-diversity paradigm (Vertovec 2006) according to which the interplay between different ‘profiles’ of diversity is a crucial aspect of contemporary societies. As the following city chapters illustrate, the main cleavages that emerged from the city chapters run along age, socio-economic status and length of stay in the neighbourhood. These were not identified a priori, but were the result of empirical fieldwork starting from the above-mentioned assumption that social situations can be defined on the basis of many different criteria of which ethnicity is just one. As a matter of fact, we found not only that the ethnic groupings emerging from the field are different from those we would have expected but even that ethnic belonging is not always particularly relevant in structuring the social life of the researched neighbourhoods.

4 A Multilevel and Multi-focused Analysis of Inter-group Relations

It is commonly acknowledged that the processes through which social cleavages and neighbourhood representations are constructed are influenced not only by everyday experiences and practices but also by wider political and media discourses (Petersson and Tyler 2008; Ray et al. 2008; Simon 2000, 2005; Taboada Leonetti 1989; Wessendorf 2010). We have therefore tried to provide an insight into the relationship between these different levels by investigating macro-frames as generated by local media and policy communities; representations of groups and narratives; social practices and relations among groups as observable in everyday situations.

These three levels are evidently related to one another: everyday behaviours and performances depend on the identity and status that we attribute to the interlocutor on the basis of more general factors (Goffman 1959) including media and policy macro-frames.Footnote 8 However, the relation between these levels should not be seen as linear, made up of dependent and independent variables, but as a cycle which integrates both the macro and micro level (Wimmer 2008). In the meantime, the analysis is made more challenging by possible mismatches between these levels of analysis. First of all, everyday practices and narratives may not coincide with the general picture of inter-group relations as provided by macro-frames. Secondly, representations, i.e. ‘what people have in mind’ about groups, and practices, i.e. ‘what people do’, may or may not overlap and converge. Positive perceptions of other groups do not in fact always trigger positive interactions, just as negative perceptions may not prevent cooperation. All these are questions that will be specifically explored in the next chapters.

These theoretical assumptions required mixed methodologies derived from different disciplines and fields of study, as well as the use of different sources. Starting from the assumption that the urban context constrains the available options for identity-building and the structuring of social relations (Wallman 2005; Lamont 2009), we reconstructed the urban and social context of the investigated neighbourhoods. This empirical background is provided in Sect. 2 of this Introduction, partially integrated in Annex I, and in the single city chapters.

We then investigated the macro-frames looking at ‘neighbourhood policy communities’ and local mass media. ‘Neighbourhood policy communities’ are meant here as the sum of all actors involved in policymaking regardless of their legal status, i.e. public, non-profit and profit actors (Jordan 1990; Rhodes 1990; Marsh and Rhodes 1992). Therefore, we considered the neighbourhood-level representative institutions (where existing) as well as housing associations, spontaneous groups of residents, NGOs, etc., paying special attention to policy frames because, as already emphasized, ethnic categories may imbue, to a varying extent, institutional cognition and recognition processes and organizational routines (Brubaker 2002; Tilly 1998). The analysis of mass media was focused on local newspapers or local pages of nationwide newspapers and it covered the decade 2001–2010. The main aims were to see to what extent the investigated neighbourhoods were associated with immigration and to identify the dominant representations of inter-group relations there. Given that the contribution of these analyses is strongly valued by comparison, the findings of this strand of research are jointly presented in the chapter on media and further elaborated and discussed in the final chapter.

Finally, local ethnographies which constitute the core of the city chapters, were the main ways to explore the actual working of society and to understand how diversity is experienced and negotiated on the basis of the everyday situation by adopting the approach of what has been called ‘everyday multiculturalism’ (Berthoud and Gershuny 2000; Amin 2002; Watson 2006; Wise and Velayutham 2009; Semi et al. 2009). Specifically, interviewsFootnote 9 with residents and ‘neighbourhood-users’ and direct observation in specific sites of interaction allowed us to catch residents’ representations of differences as well as everyday practices and daily encounters.

A systematic and more detailed illustration of the timing and articulation of our fieldwork methodology can be found in the Methodological Annex (Annex II) at the end of this volume.

Though based on the same overarching questions and on the same research methodology, each chapter develops a distinct thematic core consistent with the multi-focused approach inspiring this work.Footnote 10 The chapter by O. Jensen and B. Gidley, comparing two neighbourhoods in the London borough of Southwark, has a special focus on housing stock composition and its management in shaping social and ethnic diversity and therefore inter-group-relations. This issue is further developed in the chapter by C. Köhler on three neighbourhoods of Nuremberg characterized by different locations within the city and various degrees of ‘permeability’ and housing mix which turn into different patterns of ethnic diversity, collective identities and inter-group relations. The chapter by C. Köhler also pays particular attention to processes which lead to framing material conflicts as ethnic conflicts. The chapter on Budapest allows us to see how inter-group relations work in a context marked by low immigration rates and strong stigmatization against an internal minority (Hungarian Roma) rather than against foreigners. Moreover, this chapter offers deep insight into inter-group dynamics within a small business environment, showing how immigrant communities largely employed in this sector, such as Arabs and Chinese, might be represented in very different ways by the Hungarian majority. Albeit analyzed throughout the volume, everyday interactions in micro-sites, such as public gardens and open markets, together with the relevance of identity-building processes and embedded social capital in shaping the transition from mono-ethnic industrial districts with considerable doses of diversity generated by internal migration to multi-ethnic and economically diverse neighbourhoods, emerge as central themes, especially in the chapters on Barcelona and Turin. The chapter by A. Pogliano carries out an analysis of local media narratives focusing on the neighbourhood dimension, something which is usually disregarded by media studies and which shows how local policy communities can shape media narratives on neighbourhoods. The final chapter proposes some interpretative hypotheses on inter-group relations and the role played by urban context and policy communities based on a comparative analysis of the empirical contents of the previous chapters.