Neighbourhood renewal, participation, and social capital in deprived areas: unintended consequences in a Nordic context Open Access

Community participation is widely reputed to bring about health and social benefits and has become a mandatory issue for many national and local governments in Western Europe. It can be located in the context of ‘a reworking of the state-citizen contract in a political project geared towards mobilising citizens to take greater responsibility of their own life outcomes’ (McKee 2015: 6). This political project manifests itself via certain phenomena; for example, within the last 20 years there has been a coalescence and alignment in housing, planning, and urban policy where the welfare state is considered a driver for social and moral decline through the creation of a dependency culture (McKee 2015). Austerity and welfare reforms are therefore being experienced by most Western European cities (McKee 2015: 2) and addressed through privatisation and physical and demographic reconfigurations of urban environments that aim to address urban deprivation (van Gent et al. 2009). As part of the Scandinavian context, Denmark, since the early 1970s, has fully developed what Esping-Andersen (1990) calls a universal welfare state, in which welfare provisions are institutional and apply to all citizens alike. An important argument regarding the welfare state is its impact on social cohesion and that there is a ‘more or less linear, income inequality-social cohesion/trust/esteem, etc., health status linkage’ (Coburn 2000: 136). Recently, however, the aforementioned development in Western Europe has been mirrored in Denmark and the other Scandinavian countries showing signs of norm and values fragmentation. In the Danish context, community processes based on concepts of social capital and expanding participation and responsibility by residents are increasingly identified as contributing to tackling neighbourhood deprivation, along with social and health inequalities. They share a number of common rationales; by providing and promoting social activities, they aim to build partnerships among the local municipality, the local housing organisation and residents to improve social cohesion and social inclusion, as well as living conditions in general (Munk 2002). Since personal levels of social support contribute to better self-reported health status (Grav et al. 2013; Lindström 2005; Poortinga 2006), lack of social capital is considered an important element in the decline in mental as well as physical health in areas deemed in need of intervention (Almedom 2005; Subramanian et al. 2002).

Social capital is therefore a widely used concept in community participation as a positive resource, which, in a non-specific way, leads to health gains (Gillies 1998; Rifkin et al. 2000; Wallerstein 2006). Many scholars stress the importance of geographical space in relation to social capital in understanding the effects on health (Carpiano 2006; Kawachi 1999; Wood and Giles-Corti 2008). However, despite the growing literature on community participation, social capital, and improved health, there remain many questions about implementation and measures of success. The ability of participation to foster empowerment, and more inclusive democracies, as classically defined by Rappaport (1984), can easily be undermined by the problem of ensuring representation from all groups. For example, Baum et al. (2000) find that participation in local groupings and networks is most likely to take place amongst the most privileged members of society. Thus, understanding who participates, why they participate, and the kind of resources gained from participating is essential if national and local policy is to treat community participation as a cornerstone in addressing social and health inequalities. This paper investigates these issues on the basis of a contemporary neighbourhood renewal project which aims to strengthen social networks and to increase social capital and health.

The origins of social capital lie in late nineteenth-century classics of sociology, where Durkheim (1897) emphasised the necessity of mutual trust to sustain a community. Durkheim defined social integration as attachment and regulation. While attachment is the extent to which an individual maintains ties with other members of society, regulation involves the extent to which an individual is held in the fabric of society by its norms, values, and beliefs. These perspectives form the basis for the later three dominant and also very diverse conceptualisations of social capital presented by Robert Putnam, James Coleman, and Pierre Bourdieu (Holt 2008: 228). Coleman (1988) worked within a functionalist sociology inspired by economic theory and was interested in the role of social capital in acquiring human capital, where human capital is considered to be the outcome of taking part in social relations whereas social capital is the by-product that connects people. He defined social capital as ‘a variety of entities with two elements in common: They all consist of some aspect of social structures, and they facilitate certain action of actors – whether persons or corporate actors – within the structure’ (Coleman 1988: S98).

However, the most influential contemporary manifestations of the concept of social capital are Bourdieu (1986) and Putnam (1995) (see Portes (1998) for a more comprehensive discussion of the dominant social capital theories). Both authors focus on properties of the social structure that facilitate (or hinder) social action. An important disagreement relates to whether social capital is an individual attribute or a property of collectives. Putnam's communitarian approach to social capital views the social network as the core element of social capital by conceptualising it in terms of ‘features of social organisation, such as networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit’ (1995: 66).

Building on Granovetter's (1973) categories of ‘strong’ and ‘weak’ ties, Putnam discusses two main components of the concept: bonding capital is the result of social connections in which people are similar in terms of social identity (e.g. race, ethnicity) while bridging capital is the result of social capital that connects different groups that are unlike each other, yet are more or less equal in terms of their status or power. Bridging ties connect people more loosely, which, according to Putnam, promotes greater social cohesion.

A common criticism of the communitarian approach to social capital is that it underplays the role of intra-community dynamics, the inequitable distribution of social capital, and the consequences for social and health equality (Moore et al. 2006). According to Rifkin (1996), power struggles in particular are critical to the long-term viability of participatory endeavours. Understanding social integration, participation, and social capital is therefore part of understanding the operations of power in daily life and how social structure, as practiced in daily life, creates and perpetuates inequality. According to Granovetter (1983: 209), ‘weak ties provide people with access to information and resources beyond those available in their own social circle’ and therefore function most effectively when they bridge a social distance. Thus, the value of the weak tie depends on the position of the member involved and whether that member is able to pass on useful information. Szreter and Woolcock (2004) therefore suggest incorporating a third form of social capital, ‘linking’ capital, which refers to the vertical ties that enable information and resources to flow along explicit, formal, or institutionalised power or authority gradients. Nonetheless, the communitarian approach to social capital as a panacea of social problems has legitimised its use by governments as a no-cost alternative for social welfare provision (Holt 2008: 229; Naughton 2014: 4).

On the other hand, Bourdieu's conceptualisation of social capital is part of his broader theory of fields, which addresses power and structural inequalities. In the work of both Putnam and Coleman, practice is seen as a product of rational choices. In an effort to understand the practice of agents, Bourdieu introduces the concepts of field, positions, and habitus, as well as the concepts of economic, cultural, and social capital. Capital, according to Bourdieu, is a resource that exists and functions in relation to fields (Bourdieu and Wacquant 1992: 101) that operate according to their own internal logics and dynamics. Economic capital and cultural capital are most developed in Bourdieu's work. For him, social capital is the sum of resources that an individual (or group) accrues by virtue of being enmeshed in ‘a durable network of more or less institutionalised relationships of mutual acquaintance and recognition – or, in other words, to membership in a group’ (Bourdieu 1986: 248). For Bourdieu (1986: 51),

An acquired disposition to gain and maintain knowledge of connections and skill at using them are therefore ‘integral parts of this capital’ (Bourdieu 1986: 52). Thus, Bourdieu's interest in social capital concerns primarily how resources lead to exchanges that are reflective of and affect the constant struggle for domination. He points primarily to the reproduction of the dominant class as the main explanation of social capital, where social capital is considered an investment of the dominant class engaged in mutual recognition as members of the group and where social capital is represented by aggregating (1) the size of the group and (2) the volume of capital possessed by the group members (Lin 2001: 25). However, as argued by Lin (1999: 34), the requirement of network density is only relevant when wanting to preserve or maintain resources, whereas weaker ties are more useful when searching and obtaining resources such as looking for a job. It all depends on the investment strategies of the members of the network and the type of return one expects. In the context of deprived areas, this implies that people who may be equal in economic and cultural capital, but who differ in social capital, can use social capital to dominate one another. Bourdieu's work therefore offers a useful starting point in focusing on the processes by which individuals in deprived neighbourhoods participate, that is, how they mobilise and invest in social capital; this perspective allows us to view social networks as embedded in and shaped by larger external social structures, and that participation as a strategy mediates the idea that individual practices are fundamentally interested. However, more conceptual work is needed to understand what the kinds of capital created in neighbourhood renewal projects are and what is actually exchanged through social ties and Bourdieu's concept of social capital does not lend itself either to a precise definition or a close empirical assessment. Instead, Bourdieu describes his ideas ‘like those of habitus, practice and so on, were intended, among other things, to point out that there is a practical knowledge that has its own logic, which cannot be reduced to that of theoretical knowledge’ (Bourdieu et al. 1991: 252). Thus, he suggests that the chief strength of his concepts lies in their empirical relevance when stating that they are ‘open concepts designed to guide empirical work’ (Bourdieu 1990: 107). This is where I suggest that social network analysis (SNA) might be of help. The primary tenet of SNA is that the structure of social relations (size and density or percentage of people who know each other) determines the content and behaviour of those relations (interactional characteristics such as frequency of interaction and degree of reciprocity) along with the function of the networks (practical aid, emotional support, etc.). Well established principles in sociology – the principles of homophily and heterophily – help to assess how likely it is that a set of relations carries rich or poor resources (Lin 2001). Bonding ties occur between people in similar situations (homophily) and are useful for the development of reciprocity and solidarity, whereas bridging ties (heterophily) connect people in different positions and are valuable resources of innovation. As remarked by Lin (2001: 27), Bourdieu's representation of social capital assumes that all members maintain strong and reciprocal relations. However, in his empirical studies of people who had recently changed employers, Granovetter (1973) found that many people discovered their new jobs through information from weak ties in terms of distant acquaintances rather than strong ties in terms of close friends. He therefore stressed the importance of bridges in networks in facilitating information and influence flows (Lin 2001: 27). In the context of area-based initiatives which aim to influence mobilise and invest resources by participating in activities, the following questions are relevant: Which individuals mobilise and invest resources by participating in activities? Which struggles and stakes are involved in these strategies? What kinds of resources are generated by being part of these activities?

I address these questions by drawing on Bourdieu's work and explore the nature and strength of social ties, distinguishing between the forms of bonding, bridging, and linking social capital. I examine these questions empirically in the context of data from a network study of community participation as this may enhance Bourdieu's concept of social capital when considering the different value that networks might have for accumulating resources.

Auburn (anonymised pseudonym), a neighbourhood 4 kilometres from the centre of a large provincial town in Denmark, has a population of 1800. The area is surrounded by woodland and sits on a ridge between valleys east and west of Auburn and is one of 29 designated social housing areas. Via area-based initiatives, and most recently with its ‘Ghetto Strategy’ (Danish Government 2010), the government aims to transform such areas into better neighbourhoods. Selection of the 29 ‘ghetto’ areas is based on three main criteria: (1) high proportion of unemployed residents (more than 40% of 18–64-year olds not in work or school), (2) high proportion of immigrants and descendants from non-western countries (more than 50%), and (3) high proportion of residents convicted of a crime (more than 2.7%). The strategy includes

• grants to physical renovation of housing estates to improve the living environment;

• co-operation between the municipality, housing company, residents’ board and voluntary organisations;

• mobilising and building of social networks;

• grants to initiate social, health-related, and cultural activities;

• integration of ethnic minorities;

• crime prevention.

Auburn has been targeted for neighbourhood renewal programmes since 2004. Many of the new residents moving into Auburn are placed there by the local authorities exercising their right to allocate 25% of the available apartments for social purposes. Therefore, a large number of the municipality's most disadvantaged families are concentrated in this particular area. According to the Danish Ministry of Housing, Urban and Rural Affairs (2014: 4), the rationale for the ghetto policy is that: ‘Housing estates which are physically and socially isolated from the surrounding community create good conditions for the emergence of parallel societies with deviant norms of behaviour and values, and constitute a democratic threat to social cohesion  … .’

As indicated, the problems in the so-called ghettos are attributable to the formation of parallel societies, understood as immigrants not integrating enough into Danish society, and thereby creating their own sets of rules and norms; it states that:

This study focuses on the part of the strategy aimed at building and strengthening social networks in the area ‘to integrate it into the Danish society’ and to increase ‘social cohesion’.

This study is part of a larger project which, on the basis of ethnographic fieldwork (Hindhede and Aagaard-Hansen, 2016), conducted an SNA of 17 identified groups totalling 133 members who, within a period of 10 months of fieldwork, were identified as participating in municipality-driven, community-based activities in Auburn (Table 1). I draw on the SNA to determine which individuals mobilise and invest resources by participating in activities, and I draw on qualitative interviews to determine which struggles and stakes are involved in these strategies as well as the kinds of resources generated by being part of these activities.

From the 75 individuals who participated, there was a varying degree of group membership, with a number of individuals being members of more than one group. Each of these 75 individuals was interviewed individually using a structured questionnaire that included questions about self-rated health, perception of the neighbourhood's physical and socio-cultural features, and economic, cultural, and social capital (Bourdieu 1986). Cultural capital was measured using highest education level. Social capital was measured as level of socialising with friends, family, group membership, and links to people in politically or financially influential positions (e.g. lawyers, politicians, doctors, and police officers). In relation to social capital, a network member's position is a source of opportunities to influence others and receive flows (including information, support, and material aid). Socialising (and a proxy for the strength of the social tie) was measured by regularity of face-to-face contact with family, friends, and group members on a four-point scale ranging from daily to monthly. This implies considering the type and nature of support (whether the interaction was material (as in lending tools), social (as in socialising together), or emotional (as in providing consolation)). Socialising was also measured by considering overlap of group membership. This allows differentiation between the ties between people in groups that do and do not result in the accrual of benefits. I also measured each centrality score. Centrality is the extent to which an individual inhabits a critical position in the network (Scott 2000: 82). The size and intensity of a person's network relationships in themselves say nothing about the resources the network might hold for that person; therefore, information on whether the network possesses resources relevant to the person's specific goals and needs was also included in the interviews. I also identified each network member's opportunity to act as a ‘broker’, who holds subnetworks together, which is a crucial determinant of influence (Gould and Fernandez 1989). Of the 75 participants, qualitative in-depth interviews were conducted with 22 ‘community influentials’; that is, individuals with the highest centrality and brokerage scores.

Interviews with community influentials (referred to using fictitious, randomly selected initials) were conducted at a meeting point in Auburn and lasted from 60 to 120 minutes. I presented myself as a researcher from a university and aimed to ask them about their social life in terms of daily activities, sense of identity in relation to place, the neighbourhood as a community, and what promoted or hindered personal involvement in this setting. Interviews were digitally recorded and transcribed verbatim.

Verbatim transcripts of interviews were analysed primarily in terms of Bourdieu's (1985) theoretical framework on social space, fields, and the genesis of groups. This goal entailed studying the data, systematically searching for items relating to determination to retain or improve social position, and how respondents aligned their goals with their assessments of their resources. The codes and categories are therefore a combination of codes which first capture the constitutive elements of bonding, bridging, and linking social capital as posed by Putnam, Szreter, and Woolcock, along with codes relating to Granovetter's notions of strong and weak ties. This was followed by a coding with the Bourdieusian concepts of position and position-taking, stakes, and struggles that denote how participation shapes community interaction in relation to residents’ choice of participation and what Bourdieu (1985: 732) would call strategies; that is, a means of understanding how people act on their environment, which is something neither wholly unconscious nor simply the result of rational calculation. Interpreting the data entailed systematically searching for items relating to participatory practices using a technique of constant comparison, with topics being grouped into theoretically based themes. These themes were chosen for their similarities and differences. For more details about the coding see Hindhede and Aagaard-Hansen (2016).

This project required approval by the Danish Data Protection Agency because it comprises person-identifiable data. In all cases, informed consent was gained prior to interviews, anonymity was guaranteed, and the participants were informed that they could withdraw from the study at any time. In presenting the data, individuals’ names, names of places, and other identifying details are omitted to ensure confidentiality. No formal ethical clearance was required.

When considering the outcomes possible as returns of social capital, the first part of the ethnographic fieldwork was to investigate the kind of activities on offer for the residents in order to identify the network characteristics. On the basis of this, the network locations were investigated in order to consider in more detail the resources embedded and accessed.

The fieldwork showed that among the activities in Auburn that aim to bring people closer together across age and cultural boundaries, there is an annual summer party, a flea market, and a festival. Also, 17 activities initiated by the community developers were identified. These activities included groups of 4–12 people from Auburn who met on a regular basis. In total, only 75 residents (4% of the population) were involved in these activities. Of the participants, 79% of participants were women and 21% were men, and 81% of all participants were unemployed. In general, participants were largely recruited through word of mouth or through knowing someone who was already involved, and they had all lived in Auburn for more than 10 years.

Based on the criteria used for labelling an area as a ‘ghetto’, low levels of economic and cultural capital are a general marker for all residents in Auburn, including the participants. Only a few of the 75 participants had linking ties connecting them to resourceful social networks outside the community. Data showed that most participants invested in dense bonding networks through regular contact with family and friends. Friends included both individuals from community activities and individuals living outside Auburn. Ties were primarily among friends similar to one another by age, gender, ethnicity, and religion. Thus, homophily (Rostila 2010) and the tendency for people to affiliate and associate with others similar to themselves were clearly visible.

All 22 influentials talked about the bad reputation of the area as exaggerated and stemming from inaccurate media reporting of a few occasional disturbances. For most respondents, one strategy was to maintain the established order by defending their neighbourhood and their own status against claims, which the quote below demonstrates. Despite visible cues of disintegration, including graffiti and vandalism, LB expresses pride in the area and its natural surroundings and felt that this neighbourhood was quiet and peaceful:

However, strategies of mutual distancing were also produced and operating in Auburn. This was particularly evident when I talked to the group of voluntary block ambassadors. Their official objectives are to welcome new residents, provide information about the area, and to serve as contact persons for the housing companies. From their narratives, there seemed to be a particular code of conduct in this position, having to do with not entering the apartment and to remain formal in contacts with the new resident, which is exemplified by the quote of BH (55, woman, unemployed, ethnic Dane): ‘We refuse when asked in – for safety reasons. It's better to just stay by the front door and hand over the information and the brochures. Otherwise it gets out of hand.’ Another member of this group, LB (73, woman, pensioner, ethnic Dane), explained her duties as a block ambassador:

LB seems to extend her power as a block ambassador as someone who can lay out rules of behaviour for newcomers and decide what the area around the blocks are to be used for. That the ‘foreigners’ had a different lifestyle created tension and struggles over space. Thus, the social capital created by being a voluntary block ambassador simultaneously creates boundaries that distinguish the ‘foreigners’ from the ‘natives’. Therefore, ‘insiders’ and ‘outsiders’ are mutually constituted in the same process (Blokland and Savage 2008: 12).

In considering who connects with whom, Bourdieu suggests that this process actually pertains to who does not get connected and how people are excluded. Online statistics on the neighbourhood indicate the presence of clusters of ethnic minorities: Greenlanders (4%), Iraqis (6%), Poles (4%), and Afghans (3%). However, not a single individual from these groups participated in community building. Many of the respondents talked about interethnic conflicts. For instance, people from Iraq were categorised uniformly as a group with particular attributes by several of the ethnic Danes. JA is an example of this: ‘their young boys are troublemakers’. People from Greenland were also labelled as a group and stereotyped by the ethnic Danes as exemplified by this quote: ‘they sit on the benches and drink all day long’ (DH, 30, woman, unemployed, ethnic Dane). MT (29, woman, hairdresser, Turk) explained to me how ethnic groups clustered and that Auburn differed from another area of the town that was also classified as a ghetto area: ‘I am the only Turk living in this area. It's a shame, actually because over there you find a lot of Turks so I go there very often to meet my friends.’ Bourdieu (1985: 743) states that in the space of ethnic groups, each group may be ‘characterised by the social positions of its members, by the rate of dispersion of these positions, and by its degree of social integration despite this dispersion’. From MT's quote, it seems that she actively turns to other groups of Turks, which means that for her, social relations with other Turks lead to constructive outcomes. Hence, the clustering of Greenlanders, Iraqis, Poles, Somalis, Turks, and Afghans may be based on consciously realised network structures that promote relative well-being for group members in a broader context of socioeconomic constraint and racial discrimination (see also Rostila 2010). Thus, whereas the Danish authorities aim to eliminate parallel societies, individual choices along with welfare state policies seem to influence residential segregation. At any rate, an identified group of stateless Palestinian women with a high degree of ethnic solidarity were among those LB talked about. They were densely connected in a cluster (see Figure 1) with strong practical, social, and emotional ties from a shared history of the consequences of statelessness and the ensuing discrimination, exclusion, and disadvantage. Bourdieu (1985: 737) argues that ties become established on the basis of recognition of the proximity of the participants, indicating that homophily functions to create and preserve social differentiation in Auburn. The Palestinian women explained that they gained resources from involvement in activities made available particularly for them. This seemed to give them bonding social capital advantages in helping one another, since these activities were explicitly prioritised. For these women, the level of reciprocal emotional ties through participation in the Koran club was described as increasing individual well-being,

Of the 12 members of this particular Koran club, 7 of them participated in three related groups: the Koran club, the running club, and the Arab tutoring club; however, only two of them were members of all three groups (see Figure 1). Three others, AA, LH, and GM, had high brokerage scores; thus their centrality is based on the fact that they lie between groups of individuals, rather than just individuals. By getting involved in groups with members of mixed ethnic backgrounds (e.g. the tutoring club), these three women seemed to profit from their investments in their networks and as brokers, successfully developing the skills required to navigate between cultures and to control the flow of information between groups:

According to Bourdieu, social capital is formed more or less consciously via integration into networks. Since it resides in social relations, it is characterised by a kind of indeterminacy and can therefore involve unspecified obligations, such as taking shifts in the tutoring club. For the three brokers, AA, LH, and GM, the power to organise and become engaged brought to light a horizontal kind of bonding social capital, expressed in terms of how this experience gave them a sense of connectedness resulting from networks of ‘solidarity’ through which they ‘could acquire the confidence’ needed to act as agents.

Bourdieu (1989: 20) points out that the meaning of ‘the social world can be perceived and expressed in a variety of ways, since they always include a degree of indeterminacy and vagueness’. Thus, ‘this plurality of visions  …  provides a base for symbolic struggles over the power to produce and to impose the legitimate vision of the world’. When considering the activities in Auburn five of the activities had an explicit health enhancing objective (Table 1) informed by national health priorities, such as obesity prevention, smoking cessation, and improving levels of physical activity. Levels of self-rated health were not higher in these groups. BH, who had poor experiences with these activities, explained: ‘They are too idealistic  …  young, skinny and well-trained instructors  …  they have to be kept at bay because they don't have the [health] problems themselves.’ Thus, these types of activities apparently did not translate well in this neighbourhood, in part because – as argued by BH – people did not conceive of themselves in the ways imagined by those who developed the programmes. Thus, the meaning and orientation of what should be offered in the neighbourhood was a space for interpretation and therefore contestation. ZC (51, woman, unemployed, Lebanese) explained about the gender aspects of participation and how socialising was the motivating factor for participation:

Another example of a clear basis for symbolic struggles over the definition of social reality was the committees of the two housing organisations. They were described by all respondents as groups difficult for newcomers to join and which perpetuated a dominant status. OH (49, male, disability pensioner, ethnic Dane) has invested in this struggle. He has lived in Auburn for 13 years and has been active in the neighbourhood since 2007. He explained how being a new member of the committee of the housing organisation had been tough and required skills and capacities to stand a chance of influencing decision-making:

As OH explains, the committees were very powerful groups; thus, the value of the connection to these groups was great and so were the obligations associated with membership. The social networks built were exclusive and give advantages to some while also downward-levelling norms that operate to keep outsiders at bay. These committee members initiated local social events which were seen by the renewal project as important vehicles to foster community engagement and to encourage social interaction. The activity group was another product of this strategy and arranged midsummer events with singing and campfires:

In OH's opinion, for many of the ethnic minorities, participation and engagement in the community has to do with trust and social confidence to go out into a public space or place. From his many years of doing voluntary work in Auburn, he has learned that mutual trust and commitment arises from group norms and repeated interaction. According to him, trust is therefore something that develops over time and that feeds back into the development of social networks.

Doing community volunteer work and participating in activities seemed to be a way to increase confidence and self-esteem. However, a challenge for many participants was that they could not recruit others for voluntary work – that their friends would act only if they received a private payoff from participation, as stated by AA: ‘they think I am kidding when I tell them that the job is unpaid’.

Detrimental to the development of social capital were the opposing opinions about how to reach the hard-to-reach in the neighbourhood. AN (69, male, pensioner, ethnic Dane) pointed out: ‘The way to the [minority] adults is through their kids. Kids are so important to them. So by doing things for the kids they will not be troublemakers when they are in their teens.’ Others like DH, however, pleaded for involvement of the adults instead of children for the activities to result in increased social cohesion, by noting:

OH spoke about participation fatigue being a problem, which accounted for residents’ deliberate self-exclusion from participation. He also pointed to the issue of potential volunteers being afraid that by participating they would be considered fit to work, and therefore participation could result in cancelation of benefits. The following quote illustrates OH's perception of the power of ordinary people to influence the course of events at the local community level:

As the person with highest overlap of group membership in the Community Renewal project, OH did not feel that the funders had much interest in residents’ views. Whereas OH, on the one hand, seems to have adopted what the community considers as recognition of collective values, on the other hand, he is not rewarded by his action strategies by gaining social position, ‘thereby helping to determine position in social space’ (Bourdieu 1985: 724).

In line with the arguments of Bourdieu (1977: 17), there is a competition for the symbolic power to determine the valuational standards of what are considered ‘innovative’ activities. When asked about community activities that would increase participation, GE explained: ‘Let us show what we are good at  …  let the youth help create positive changes  …  everybody would like to have their picture on the front of the local paper.’ This quote demonstrates a wish not to be a burden to the Danish welfare state, a wish to work toward greater acceptance, thereby engaging in the exercise of full citizenship. GE talked about increasing the influence of minorities in Auburn, giving them rewards and incentives for participating and to help create a positive culture to counter the alienation that many residents faced.

However, JA also explained that some husbands imposed restrictions on what could be done outside the household, and that most women were able to interact amongst themselves within their houses, but not outside with others. As a community, ethnic diversity means that what is appropriate for one group might not be appropriate for another and might not result in any take-up. Because participation involves funding and therefore redistribution of resources (economic capital), its meaning will inevitably be contested, both at the level of rhetoric and in social practice.

The question of activities being ‘culturally sensitive’ to be worth investing one's energy in was not the only issue, however. The quote below demonstrates this point. MB (46, woman, on job activation scheme, Lebanese) has no education and has been in Denmark for 15 years. Over the years, the local job centre sent her on several council-run job activation schemes in which she, as an unemployed person, is forced to attend remedial job-seeking courses in exchange for her unemployment benefits, but her investment had not resulted in a regular job. Lack of contact with the labour market and the resulting marginalisation has since persisted. MB discusses the pressure from the Community Renewal project to become involved in activities:

For MB, social capital made up of linking ties anchored in formal organisations, such as social services, ideally impacts her capacities and opportunities. In her case, however, this linking tie is negative (Wacquant 1998: 29) as it does not affect her life chances, such as finding a job, but keeps her in a marginal and dependent position. Unpaid community participation is considered exhausting, and her statement reinforces concerns that the government's strategy may place too much responsibility on people from disadvantaged communities. MB's quote demonstrates that, for her, the broader context of vertical structures imposes limits on the way knowledge, resources, and power can be deployed across her networks. The linking ties established between participating residents and the local service providers and government representatives had not developed into linking networks that could have facilitated exchange of information, building of capacity, and promotion of sociability and social leverage. Although all respondents mentioned casual acquaintanceship with high-status individuals, none had been able to achieve higher status through these linking ties.

As stated in the Introduction, the Danish Government (2010) is concerned with the ‘emergence of parallel societies with deviant norms of behaviour and values’ which apparently ‘constitute(s) a democratic threat to social cohesion’. This indicates that even universalistic welfare states like those of Scandinavia which have best resisted the rise of what Wacquant calls advanced marginality, with its ‘territorial stigmatization linked to the emergence of zones reserved for the urban outcasts’ (Wacquant 2007: 68), face increasing processes of spatial concentration of dispossessed households. Given the current situation, Denmark is seeking ways to generate social capital, since it is considered to be a diminishing resource, along with economic development tools and a strategy for poverty alleviation (Daly and Silver 2008). In their conception of social capital and social support, both social cohesion and social capital are used interchangeably. Moreover, whereas improved health is an explicit goal, neither health nor health outcomes are defined. The health-related influence of participation therefore remains unclear.

Other studies (e.g. Blokland and Savage 2008; DeFilippis 2001; Ferlander 2007) suggest that SNA can be used to inform our understanding of both the inclusive and the exclusive aspects of social capital. To my knowledge, the present study is one of the first to empirically further our understanding of the kind of social capital created in neighbourhood renewal initiatives and ‘ghetto strategy’ by considering who participates in which activities as well as how and why residents participate and the effects on density and quality of social connections (social capital) of living in a low-status neighbourhood with low income (low economic capital) and low educational level (low cultural capital). The respondents – all possessing a high degree of centrality – discussed how opportunities to participate, a sense of agency and strategy, access to bonding and bridging capital and sharing the value of participation are involved in decisions to participate. A lack of vertical and bridging links connecting participants to social networks outside the community not only hampers the flow of material resources, but is likely to impede the information and ideas that also circulate through social networks. Bridging capital can (ideally) enable people and groups to control deviancy and reinforce positive health norms (Ferlander 2007). The bridging capital identified in this study appears to be a vehicle through which participants learn to relate to each other across ethnicities. Trust is seen as a core element in the social capital literature. However, as shown, it is not a stable characteristic in a ‘ghetto’ area such as Auburn but a relational process associated with inequalities of class, race, and gender within which the residents are situated. Social capital involves contestation of space and, as shown, relationships created within particular communities can isolate or possibly even harm groups, such as the non-participants in Auburn.

This study reflects the classic argument of Bottomore (1954) that levels of participation in social and civic community life in urban settings are significantly influenced by individual socioeconomic status. Although not all respondents in this study weakened their social ties, their lack of opportunity to create linking social capital in vulnerable areas has been documented in previous research (Warr 2005). Seventeen groups with a total of 133 members were identified, thus only a small fraction of the 1800 residents in Auburn, and perhaps only the relatively resourceful, were included. The areas on the Danish ghetto list are, by definition, made up of diverse voices and interests across age, gender, and ethnicity which may not be represented by those who participate in community initiatives (Carpenter 2007; Gilchrist 2007). Moreover, in her research on the impact of area-based intervention on social mix in deprived neighbourhoods, Christensen (2015) finds that difficulties in increasing social mix are due to strong residential selection in moving patterns. In continuation hereof, this study shows that patterns of interaction are driven by preferences for similarity. McPherson et al. (2001) note that networks based on race/ethnicity create the strongest divides in a person's environment and have many implications for information transfer, attitude formation, and social interaction and experiences. The assumption underlying many Danish community-building projects is that the more community members are supported to take control by being involved in the design and implementation of activities, the more likely their health is to improve. Other research has shown that active involvement in a range of community activities did increase perceived physical health benefits (e.g. Callard and Friedli 2005; Ziersch and Baum 2004). In Auburn, very few of the 1800 residents participated in community-building activities, and the ones who did were mainly women who had lived in Auburn for more than 10 years. These were described as ‘the usual suspects’ by the renewal plan staff. Some of them experienced disapproval and criticism from other residents, which posed a risk to their well-being, which, in this case, accounted for participants’ deliberate self-exclusion from participation. Another negative aspect of the social capital created is that non-members of the networks are, by definition, precluded from the resources provided by the project funders. A focus on promoting self-reliant capacity building and even community control can therefore lead to exaggerated expectations of what can be achieved simply by strengthening social networks (Carpenter 2007).

Social difficulties and barriers suggested by the people I interviewed reveal the everyday ways in which power relations operate through affecting access to resources. The costs involved are experienced differently by the different groups and are not always given due consideration by initiators of participatory initiatives. Certainly, these issues and their impact on social capital are not addressed in Putnam's work, since it excludes the conflicts or opposing interests that underlie the ‘universal’ values of society, including solidarity, trust, and togetherness. For Bourdieu, an important aspect of the social reality of competition for resources is the exclusion of members of other groups from access (Bourdieu 1984), which makes his definition of social capital more appropriate when addressing ‘parallel societies’, since the kind of social capital produced must retain a connection to economic capital. However, while in this study the networks identified are networks based on race/ethnicity that indeed create boundaries, in fact, it appears that these groups find useful forms of support and benefit from these social networks and therefore an understanding of each other's problems.

With the rhetoric of ‘Ghetto Strategy’, the government officially sanctions the stigmatisation of the neighbourhoods that are labelled as such. This undermines the potential for social solidarity, which might be an impetus for collective action to challenge and resist the demonisation of a neighbourhood (Wacquant 2008). Social and health policies are therefore only effective if they are informed by an understanding of the processes by which people make decisions about their own social and civic participation and how these decisions are constrained by their social and economic circumstances (Baum et al. 2000). Local actions at the community level must be accompanied by broader-scale public policy interventions, because, in struggles for power, place itself is a sign of distinction (Stephens 2008).

In this paper, I have discussed community participation in a deprived neighbourhood in Denmark subject to neighbourhood renewal. Through the lens of social capital and social network theory, I have explored which residents participate and why they participate in initiatives that aim to promote social cohesion and health and the resources they gain from their participation. I found that there are some unintended negative consequences of this type of state-led renewal programme in the form of self-exclusion and exclusionary dynamics in the neighbourhood. Based on the theoretical concepts of Bourdieu, I found that these dynamics centre on power struggles, which lead the least powerful to opt out. Thus, the Danish ‘Ghetto Strategy’ which aims to increase local community participation and volunteering could have the unintended consequence of increasing social and health inequalities rather than reducing them. It therefore seems vitally important that policies advocating participation as a means of addressing social inequality consider factors which promote or hinder the likelihood of participation by socially excluded groups. In addition, a prerequisite is that residents play an active and direct part in the initiative and have the power to determine the direction and actions taken. This was not the case in Auburn, where only activities politically assessed as ‘new’ and ‘innovative’ were funded. Since community participation has become a statutory obligation for many Danish communities, these findings are important in the debate on how to best address social problems in the Scandinavian universal welfare states.

Many thanks to two community developers, Vikki Thygesen and Vibeke Kvist, for their work on the project. I am also indebted to community residents who willingly gave their time to be interviewed as part of this study. Special thanks to John Scott for sharing his insight into sociological theory, stratification, and social network analysis and for his excellent substantive comments on an earlier version of this paper.

No potential conflict of interest was reported by the author.

Anette Lykke Hindhede is associate professor at the Department of Learning and Philosophy, Aalborg University. Her earlier work deals primarily with sociology of identity processes, risk, work and the body, studies of social stratification, and encounters between patients and the welfare state.