Social relations and experiences of social isolation among socially marginalized people

Study context 

The Danish welfare state belongs to the group of universalistic welfare states (Esping-Andersen, 1990; Esping-Andersen, 1999). The universalistic system promotes redistribution of goods and status because all citizens are endowed with similar rights, irrespective of class or market position (Esping-Andersen, 1990). This means that all Danish citizens, regardless of their labour market position, have equal access to extensive welfare benefits and services, e.g. health care, education and benefits in case of unemployment (Bambra, 2005). The universal benefits and services of the Danish welfare state coexist with social marginalization, understood as an accumulation of poor living conditions and lack of opportunities to participate in several key areas of life (Larsen, 2009; National Board of Social Services, 2008). Our study population comprises people who are socially marginalized and who use welfare services such as shelters and drop-in centres. These provide accommodation, meals, and support to homeless people, substance users, the mentally ill, the poor and others. In the following, we use ‘socially marginalized people’ as a joint term to refer to users of such welfare services.    

Social relations

Social relations can be defined according to their structure and function (Due et al., 1999). The structure has to do with the formal and informal relations that individuals are involved in and the type, size, frequency, diversity and reciprocity of the relations. The function of social relations covers interpersonal interactions and is defined as social support (e.g. emotional, instrumental or informational support), relational strain (conflicts and excessive demands produced by relations) and social anchorage (feelings of belonging to formal and informal groups) (Due et al., 1999).

Social isolation

The concept of social isolation is not unambiguously defined in the literature and definitions vary according to different disciplines, e.g. psychology, sociology and nursing research (Locher et al., 2005; Stewart et al., 2009; Warren, 1993). In this study, we use the definition put forward by Cornwell and Waite (2009), identifying two forms of social isolation: social disconnectedness, defined as lack of social relations and low levels of participation in social activities, and perceived isolation, defined as a subjective experience of feelings of loneliness and perceived lack of social support (Cornwell & Waite, 2009).

Social relations and social isolation combined in a typology

From studies of general (not socially marginalized) populations, it is evident that the relationship between social relations and social isolation is not clear-cut. People lacking social relations do not necessarily perceive themselves as isolated and lonely, and people who perceive themselves as isolated are not necessarily without networks (Cornwell & Waite, 2009; Gierveld, van Tilburg, & Dykstra, 2006; Meeuwesen, Hortulanus, & Machielse, 2001). How social relations and social isolation are associated among the socially marginalized and whether the complex relationship between them also holds true among the socially marginalized has, to our knowledge, not been explored previously. The combination of the two concepts, social relations and social isolation, provides the possibility of creating a typology taking into account the above mentioned complexity. A similar typology of social contacts has been used to analyse social relations in the general population (Meeuwesen et al., 2001). Figure 1 shows our definitions and usage of the two key concepts, social relations and social isolation. The arrows indicate that we view the concepts as interrelated.

OPEN IN VIEWER

Figure 1.

Adjusted conceptual framework of the concepts ‘social relations’ (Due et al., 1999) and ‘social isolation’ (Cornwell & Waite, 2009).

Goal

In this study, we aim to explore the relationship between social relations and social isolation among socially marginalized users of shelters and drop-in centres in Denmark using a typology combining the two concepts. Hereby, we recognize the complexity of social relations and social isolation, as the typology makes it possible to take into account the structure and function of social relations as well as the perceived social isolation and explore these collectively.  

The interview setting

Interviewees were recruited from 11 different shelters and drop-in centres in three Danish cities and all interviews were conducted at the shelters/centres. Shelters are run by public services under the Act on Social Services. This places municipalities under an obligation to provide temporary accommodation in shelters to individuals who do not have a home or are unable to stay in their own home. Drop-in centres are run by public services or voluntary organizations, e.g. the Church Army. Some drop-in centres are specifically directed at people using substances or with social psychiatric problems; other centres are directed at socially marginalized people in a wider sense. Shelters provide accommodation whereas drop-in centres do not. Besides this, shelters and drop-in centres provide similar services: meal services, counselling about economic and social matters, addiction treatment possibilities and assistance in maintaining contact with the municipality or other authorities. Additionally, they offer social support and activities, practical help and care. The shelters/centres employ people with various professional backgrounds, e.g. social educators, social workers, nurses and voluntary workers. In Denmark, shelters and drop-in centres provide services not only to homeless people but to a broader spectrum of socially marginalized people such as, among others, substance users, the mentally ill and the poor. Thus, shelters/centres are used by people covering a continuum of social marginalization, from the severely marginalized or socially excluded who need various services (shelter, food, care, nursing or a break to regain some energy) to the less socially marginalized who may primarily seek a cup of coffee and some company. 

Data collection procedures

Interviews were conducted in two phases: 25 interviews from July to September 2008 and 21 interviews with new interviewees in December 2010 and January 2011. The first round of interviews aimed to obtain a broad understanding of the health and wellbeing of shelter/centre users. Interviewees were interviewed about their general health and wellbeing, social relations, experiences of social isolation and living conditions. The second round of interviews were focused directly on social relations and experiences of isolation using the questions from the first round of interviews. This study is based on narratives from both rounds of interviews.  

Interviewees were reached similarly in the two rounds. The selection of interviewees occurred with the help of a contact person from the shelter/centre. The contact persons had a pivotal role in the data collection process because of their familiarity and well-established trust with the users. The contact person was aware which of the users were not feeling well or were too substance influenced to participate in an interview on the day in question.

Face-to-face interviews were conducted by the first author and carried out as partly semi-structured, partly open interviews based on a thematic interview guide. Interviewees were asked about their social relations, who constituted their network, and feelings of isolation. The interviews did not aim to exhaustively elicit every possible social relation in the life of the interviewee but rather to capture the persons and relations that were important (either in a positive or a negative way) to the interviewee.

Establishing trust between the interviewer and the interviewee was extremely important because of the interviewees’ vulnerable situation. In many cases, the interviews were stories of neglect, failure and negative experiences and, thus, difficult to tell. The interviewer tried to establish and secure the interviewees’ feelings of trust by conducting the interviews in their own environment and ‘home’, where they were surrounded by familiar faces, and by informing them thoroughly about the research project in an intelligible way. Each interview ended with a debriefing. Some of the interviewees were deeply affected by telling their stories, but no one said they had felt uncomfortable when asked at the debriefing. Conversely, some expressed that it had been a positive and beneficial experience to get the opportunity to ‘unburden’ themselves. One interviewee refrained from answering questions about his family relations because it was too difficult emotionally for him to talk about.       

The length of the interviews varied from 16 minutes to one hour and 30 minutes, primarily due to the interviewees’ tiredness, concentration difficulties or drug or alcohol influence.

Participants

In all, seven women and 39 men were interviewed, ranging in age from 22 to 64. Six interviewees were born outside Denmark (in Greenland, Iceland, Germany, Iran or Somalia). All interviewees lived on social security benefits. At the time of interview, 32 were homeless, whereof 29 lived at shelters and three lived on the streets. In total, 14 had their own apartment but visited shelters/centres daily. In all, 11 had a romantic partner, whereof two were married; others were single, divorced, or widowed. In total, 31 of the interviewees had children. The interviewees were not asked directly whether they used substances or had any mental illnesses. However, during the interview, the interviewees brought such issues up and 35 told stories of substance use (e.g. alcohol, cocaine, heroin and/or marijuana). A few were former users of alcohol and/or drugs and thus ‘clean’ at the time of interview. In all, seven interviewees had been diagnosed with a mental illness, such as schizophrenia or manic depressive psychosis. Two of the women had previously worked in prostitution.

Data preparation and analysis

In immediate continuation of the interviews, the interviewer wrote down her reflections on the interview setting, the atmosphere and the relationship between the interviewer and the interviewee. All interviews were recorded digitally (with the interviewees’ permission) and transcribed verbatim by the interviewer. Names and place names were anonymized. Nvivo 9 was used to code and analyse the data. To understand and derive meaning from data, the interviewer coded the interviews in multiple rounds (Miles & Huberman, 1994). Codes and analytical themes were discussed thoroughly with the co-authors and revised accordingly. First, a within-case analysis for each interview was performed, coding data with predefined, descriptive codes, e.g. ‘contact with parents’. Subsequently, cross-case analyses were conducted and during this, the idea of creating a typology combining the concepts of social relations and social isolation emerged. We created five groups and this typology provided a meaningful structure and frame for the analysis.

Of the 46 interviews, 31 contained comprehensive and sufficient information to place them in one of the five groups. Eight interviews did not contain enough information on social relations and social isolation to form part of the five groups; however, they were used more peripherally and helped inform the initial analyses. Seven interviews were of poor quality and not usable, either because the interviewee could not be defined as part of the target group or because the interviewee did not deliver a coherent story, did not answer the relevant questions, made digressions, evaded issues or refrained from answering the questions.

The socially related and content

Within this group were nine men and one woman, ranging in age from 28 to 62. While two had become homeless recently and did not use substances, the lives of the remaining were characterized by severe social marginalization, e.g. long-term homelessness, long-term use of alcohol or hard drugs, prostitution or mental illness.

Typically, the socially related and content had satisfying social relations. They had relatively broad networks and social support came from several sources: parents, siblings, children, romantic partners, the family of partners, friends and professionals. Generally, feelings of satisfaction with their social lives characterized the group. Accordingly, the interviewees did not perceive themselves as isolated and did not express feelings of loneliness. On the contrary, they expressed happiness and satisfaction.

The following excerpt exemplifies the satisfactory relations of this group. Here, a 62-year old male found support among his neighbours at the shelter where he had lived for a great period of years:

Interviewee: Those I’ve known for a long time … I can turn to. I’ve got some where I can say: “Hey man, I’ve got a problem”.

The socially related and content spoke of close ties to professionals. In these cases, professionals played a pivotal role in the interviewees’ lives, as they supported them both emotionally and instrumentally. Thus, interviewees talked to professionals about issues bothering them or affecting them emotionally. Additionally, the professionals helped interviewees with administrating finances, getting apartments, getting new clothes, handing out and dosing medicine, and getting in contact with doctors, hospitals, the social services and other public services and authorities. Generally, interviewees said that the professionals helped increase their wellbeing and quality of life. The interviewees who described having such relations with professionals also described having satisfying relations with family members or friends. Thus, the social relations with professionals appeared to be additions to their existing web of relations – in some cases important additions, as some described that in terms of emotional support, the professionals made up their key source.  

The pivotal role of the professionals was evident in the interviewees’ narratives and the positive wording used to describe them:

They take care of me … They are the people I know, and they know me … If there are things I don’t know how to handle, well, then my contact person knows and I’ll say “what do we do here?” That comforts me and puts me at ease. (52-year-old male)

Interestingly, there did not seem to be an association between being socially related and content and the degree of social marginalization. As mentioned, the lives of most interviewees in this group were characterized by severe social marginalization. Nonetheless, they described themselves as happily married, being in romantic relationships and being closely related to some parts of their family or the family of their partner. Thus, they appeared to have access to, and to engage in, meaningful and supportive social relations while, simultaneously, living a life on the edge.

For example, one 46-year-old substance-using male lived at a shelter and sometimes lived as a vagabond. His long-term marginalized life seemed to have affected his health negatively and significantly. Despite of this, he unequivocally answered in the affirmative to the question of whether he felt satisfied with his life and added that he ‘almost always was in a good mood’. He appeared untroubled. His network consisted of his wife, family, ‘drinking buddies’ and other shelter residents.

A 44-year-old male whom we interviewed had been diagnosed with paranoid schizophrenia and had used drugs and alcohol since early childhood. He stayed in a rented apartment with his wife. He described that he felt very satisfied with his present social network and especially emphasized the significance of his wife. He said that it was his wife who had recently made him take disulfiram and stop drinking alcohol. As the excerpt shows, the interviewee found that his wife played a pivotal role in his wellbeing and life satisfaction:

I drank 25 beers a day, right. And Liz [wife] thought that was too much. So, I took disulfiram… I’ve done it for four months. Simply to make her happy, right. Otherwise, she would have left me … and I would die if that happened. For sure. I can’t live without her. I can’t. She supports me. And I support her. We complete each other, right, and… we have a really great relationship. (44-year-old male)

The satisfied loners

All five interviewees in this group were men, and all but one were in their 50s or 60s and had lived in social marginalization for many years. Characteristic of this group was the absence of social relations combined with the perspective of social isolation as a positive, rewarding experience.

Typically, these interviewees viewed social isolation positively, as rewarding and a much needed source of peace and quiet. They did not feel lonely when asked directly; however, they seemed fully aware that their social relations were very limited in both structure and function. They either did not have any contact with their families, had very sporadic contact with family members, or their family members had died. Friendships appeared absent in their lives and access to emotional support seemed non-existent or very limited. Their social relations mainly consisted of sporadic contact with other shelter/centre users and professionals who primarily provided practical support.

Social relations seemed synonymous with conflicts, quarrels and excessive demands. Thus, being alone became synonymous with peace and quiet. They were loners, kept mostly to themselves and thrived this way. A 53-year-old male interviewee said that he preferred being alone because of the relational strain he experienced from his existing social relations. His network primarily consisted of his ‘drinking buddies’ who constantly wanted his money or beers or who needed help with translation from Greenlandic to Danish in their contact with staff members or the social services. Also, he described his sporadic, practical relations with professionals in negative terms. He said that he felt like the shelter staff were ‘out to get him’ and that they did not support him. Thus, he preferred his own company. This positive attitude towards being alone is illustrated by the following excerpts: 

It’s the best thing … walking in the woods and being completely alone. But when I get home [the shelter], then everybody’s there [other shelter residents]. I’m sick and tired of them. (53-year-old male)

Two of the elderly interviewees had lived on the streets for a number of years. In this period, they said, they were more or less ‘completely alone’. They slept outdoors and went to shelters for free food and showers, without having much contact with the staff or other shelter users. One said that he did not receive any social security benefits because he was ‘outside the system’. He survived on deposits from collecting bottles. Despite their markedly isolated situations, the two interviewees both described this period in very positive terms. One, who stayed in a tent in a park for a number of years, said:

It was a damned great time. Damn, I felt good staying out there. I really did. And I had a bloody nice tent, you know … Then I pilfered a chair downtown and there was a huge stone. So, I had a table and a chair and everything … I stayed there for so long that one of the fallow deer, well, it became bloody tame and came begging for food when I sat outside eating. (62-year old male)

The satisfied loners described having sporadic contact with professionals at shelters and drop-in centers, primarily getting practical support. However, in some cases, these more practical relations between professionals and the satisfied loners did not appear insignificant at all. The 62-year-old male mentioned above told a story of increased wellbeing due to minor but significant actions from the professionals:

There was this woman who ran the drop-in center in X Street. She was one to be trusted. She was fantastic, or is. She’s still there. I met her when I stayed out there [in the park]. Then I told her that I didn’t get any morning coffee. Now, she thought that was a damned shame (laughs). So, she bought me a thermos and informed the staff that if I was there by 8 pm … they closed at 8 … and then she said there had to be enough coffee left for me to fill my thermos (laughs). She has helped a lot of people. (62-year-old male)

The socially related but lonely

Within this group were three men and one woman, ranging in age from 34 to 57, all of whom had a history of living in severe social marginalization with long-term homelessness, long-term substance use and poor physical and mental health. Despite involvement in social networks, perceptions of isolation still prevailed in this group. Typically, the social relations of the interviewees in this group were with romantic partners, professionals and other shelter/centre users. As was the case with the socially related and content, the socially related but lonely described relations with the professionals in very positive terms. Professionals were described as significant others; they played a pivotal role and made up key sources of emotional and practical support. Interviewees typically had very sporadic contact with family members, including children and grandchildren, and this had been the case for several years. Essential to understanding the characteristics of this group seemed to be that family members were not part of their current relations; however, interviewees did have sources of support in their lives, e.g. professionals. But apparently the quality of these relations could not overcome perceptions of being socially isolated.

A 57-year-old homeless woman living at a shelter said that her current network consisted only of professionals who helped her with emotional, practical and financial issues. She described a very close and personal relationship with the shelter staff and said that they took care of her and nursed her, and that she felt good when she was around them. They talked and drank chocolate together, took trips together, and they helped her buy new clothes and look for apartments. When asked whether she had anybody else in her social network, she responded negatively. 

Interviewer: Would you say you have any friends?

One 34-year-old homeless street dweller answered, when asked if he ever felt lonely,

Yeah, even though you’re with a lot of people, you do feel alone because it’s hard to share all the stuff that’s in your head … those I hang out with, I don’t want to share stuff with them because I can’t be sure that I’ll hang out with them next week, right. Maybe they hang out with somebody else next week. So, you spend a lot of time on your own. (34-year-old male)

The stories of the interviewees in this group entailed descriptions of the volatile and transient nature of their relations with other socially marginalized people, in romantic relationships or as people to hang out with. Typically, the other shelter/centre users were described as being socially marginalized themselves and as having challenges of their own, such as homelessness, substance use and/or mental or physical illnesses. Their lives were too chaotic and they seemed too low on resources to emotionally help and support the interviewees. Thus, the interviewees’ stories indicated that while social relations were present structurally, they were more or less not present functionally, and the quality of these social relations was described as very poor. Apparently, spending time with other shelter/centre users or having relatively close relations with the professionals could not overcome their feelings of isolation.

The socially isolated

Nine interviewees, ranging in age from 37–55, could be characterized as socially isolated. Of these, two were women. The degree and severity of social marginalization varied greatly within this group. Some had a long history of living with homelessness, substance use and a poor physical and mental health; others had only recently become homeless, did not use substances and appeared to be in good shape.

Common in the stories told by this group were widespread and comprehensive feelings of social isolation. Some interviewees did not have any contact with their families, including children, and had not had contact for a long time. Others had very sporadic contact with family members or had lost family members. Additionally, interviewees only had a few acquaintances or did not have any at all, and most said that their network consisted of themselves only. Some described getting practical support from professionals but apart from this, stories of access to significant social support were absent.

The following excerpts illustrate the all-encompassing isolation experienced by interviewees in this group. When asked about significant people in his life, and whether he had someone whom he trusted, a 55-year-old male said:

I used to have that, but … you know, I wouldn’t have thought it, but when you become homeless all of a sudden … well, then everything disappears. Your entire network. It totally breaks apart. So, all those people you thought you could care about … they’re gone. I met someone yesterday … yeah, he avoided me. One of my old friends … So, my network consists of myself. (55-year-old male)  

On the question of whether she ever felt alone, a female interviewee answered,

Yeah, you bet I do. You bet I do. I do. I could lie in here [in my room] and OD without anyone finding out. Those who live here just sit in their rooms all day. Day in, day out, watching TV. They don’t open the door, if you knock ... It’s sick. They aren’t normal. (46-year-old female)

Contrary to the satisfied loners, who structurally and functionally experienced a similar lack of social relations, the socially isolated said they felt sad, hurt and ashamed by the lost relations to their families. A homeless male explained that he no longer had any contact with his brothers because they had (in his words) ‘disowned’ him. He felt he had been ‘excluded from the family’ and said this made him feel sad because he missed them.

Similar to the satisfied loners, the socially isolated occasionally received practical support from professionals. However, the socially isolated expressed negative views towards the professionals. They said that they felt watched, pinned down and too dependent. A homeless and mentally ill male stated that sporadic, practical contact with professionals made up his only social contact. He expressed his skeptical views about the relationships between professionals and shelter/centre users, directing attention to the power held by the professionals. To him, they represented the municipality and, thus, could have other agendas than improving the wellbeing of the shelter user in question.

The in-betweens

Three interviewees, aged between 22 and 34, belonged to the group of in-betweens.  One was a woman. One interviewee could be characterized as more severely marginalized than the remaining, living with homelessness, severe substance use and poor health. All seemed to come from disrupted and relatively disadvantaged families lacking resources and energy. Interviewees described having family members who had ‘drunk themselves to death’ or died of an overdose. Generally, the interviewees in this group were difficult to categorize both in relation to their degree of social relatedness, their access to social support and in relation to their perceived isolation. In many ways they were socially related, and they did not feel isolated, but nonetheless they appeared troubled and unsatisfied with their social lives.

Similar to the socially related and content, they had broad networks and retrieved social support from several sources: family, friends, professionals and other shelter/centre users. However, even though the interviewees spent time with their family and friends, these relations, typically, did not seem to be used for emotional support. Instead, the professionals often appeared to be sources of emotional support and were used for talking about problematic issues. The professionals were also described as sources of increased wellbeing and quality of life. A 34-year-old homeless woman, who had previously lived on the streets but at the time of the interview was staying at a shelter, said that she had ‘become good friends’ with the staff at the different shelters where she had stayed. She said,

I remember once, I had taken an overdose … and somehow, I had shit my pants … then I walked all the way from the church to the drop-in center … I remember, they washed my clothes. That made me really happy. Otherwise, I would’ve had to go around in shitty clothes. They washed it right away … that was really nice. (34-year-old woman)    

Simultaneously, the in-betweens said they did not perceive themselves as isolated. However, despite descriptions of having social relations both structurally and functionally, and despite not perceiving themselves as isolated, the in-betweens, unlike the socially related and content, could not be described as untroubled. Narratives of being used to being alone seemed to prevail, and they expressed prosaic views on how they had been alone most of their lives and they were used to taking care of things on their own. They seemed to perceive social isolation as a familiar inevitability. The following excerpts illustrate this:

I’ve always been more or less alone, you know. (30-year-old male)

Discussion of main findings

We found great variation in the social relations and experiences of social isolation among the interviewees, indicating an ambiguous and complex relationship between the two, which is in accordance with findings of other studies of general (non-marginalized) populations (Cornwell & Waite, 2009; Gierveld et al., 2006; Meeuwesen et al., 2001).

The group of in-betweens underlines the complex relationship between social relations and social isolation. This group was characterized by being socially related to some extent and not feeling isolated, but nonetheless appearing troubled and unsatisfied with their social lives. Interestingly, interviewees in this group were relatively young (22–34 years), and  age probably played an important role in trying to understand their in-between position. They all seemed to have hopes and dreams for the future related to living an ‘ordinary’ life. More was at stake for them when their social relations failed them compared to the satisfied loners, who were primarily elderly, long-term socially marginalized people who did not perceive themselves as isolated, despite having few or no social relations, and who seemed to have come to terms with their social identity as loners.

Overall, it was cause for optimism to find that some interviewees had satisfying social relations, did not experience social isolation and – even in some cases of living in severe social marginalization – seemed to have meaningful and supportive social relations. Our typology indicates that family relations, when present, played a pivotal role in  interviewees’ social wellbeing and life satisfaction. The socially related and content in particular had access to support and resources to cope with the challenges of life through their involvement in family relations. Other studies confirm the importance and supporting role of family relations among socially marginalized people (Hawkins & Abrams, 2007; Padgett et al., 2008). When family relations were not present or not functional, they often seemed to affect interviewees’ social wellbeing negatively. Thus, there seemed to be an association between involvement in family relations and the perceived social isolation experienced by the interviewees. In particular, the socially related but lonely and the socially isolated were characterized by having very sporadic or no contact with family members, and both groups perceived themselves as isolated. Somehow, interviewees seemed to attach a sort of mythical meaning to the family as something that was associated with the idea of a possible ‘ordinary’ life – having ‘ordinary’ relations. When family relations were few and contact very sporadic it brought on strong feelings of isolation and despair and seemed to illuminate their marginalized situation even more.        

The professionals helped interviewees with administrating finances, getting apartments and maintaining contact with public authorities and, thus, tried to help interviewees uphold their position in the social system and avoid further marginalization. Other studies support the finding of significant social relations between service providers and socially marginalized people (Dominguez & Watkins, 2003). Briggs (1997) defines the resources stored in social relations as social capital and distinguishes between the use of these resources to get by or to get ahead: that is, for social support or for improving one’s life circumstances. Access to resources to get ahead is particularly important for socially marginalized people (Briggs, 1997). In our study, involvement in family relations, as described above, helped some interviewees get by, and through relations with the professionals, interviewees had access to resources to get ahead and help improve their life circumstances. However, our study also shows that in some cases professionals played a pivotal role, as they supported interviewees both emotionally and instrumentally and helped increase interviewees’ wellbeing and quality of life. Thus, relations with the professionals were not solely used by interviewees to get ahead (to improve life circumstances) but also to get by (for social support). Our results indicate that the professionals had a position in which they were able to influence the lives of the interviewees to a great extent and influence the decisions they made, both positively and negatively. Thus, a few interviewees (among the socially isolated and the satisfied loners) indicated the possible disadvantages and drawbacks of the sometimes close contact between the users and the professionals. As our analyses show, the networks of the interviewees sometimes consisted only or mainly of professionals; in these cases, the influential roles of the professionals may have been even stronger. This analytical point of our findings may be generalizable to other shelter/centre users and may be particularly relevant in a welfare state such as the Danish one, with a well-developed social security system entailing social relations between the users of and the professionals in the system. These relations may be more or less close or more or less enforced, but they are inevitable and necessary in order to meet the needs of the socially marginalized.

Our typology also shows that a social network composed mainly of professionals, as was the case with the group of the socially related but lonely, could not overcome perceptions of social isolation. This may be explained by the importance of making a distinction between naturally occurring and deliberately constructed social relations when it comes to socially marginalized or excluded groups (Crow, 2004). Relations between professionals and interviewees were constructed, conditional relations and although important because of their ability to secure social support and improve life circumstances, they could not reduce feelings of isolation.

Interviewees experienced social isolation very differently. Interestingly, some interviewees to a great extent experienced positive dimensions of social isolation. Other studies have found that, compared to the general population, homeless people experience the positive, growth-enhancing and enriching aspects of loneliness to a lesser extent (Rokach, 2005). Our findings are interesting because in health-related research, social isolation and having few social relations is often associated with negative health and wellbeing effects, while involvement in social relations is considered beneficial to one’s health and wellbeing (Berkman & Glass, 2000; Cattell, 2001; Due et al., 1999). However, as our study suggests, when living in socially marginalized conditions, involvement in social relations may sometimes be considered burdensome and conflict-ridden, and being alone may add a positive dimension by bringing peace, quiet and quality of life. Initiatives aimed at improving the social lives of socially marginalized people may benefit from taking into consideration their individual needs, such as the need for privacy and ‘loner time’. Not respecting the need for privacy of some socially marginalized people may contribute to further stigmatization and marginalization.

Furthermore, we found that the association between the five groups and the degree of social marginalization experienced by the interviewees was complex. For example, the socially isolated did not automatically consist of the socially most marginalized interviewees. Other studies support this finding and conclude that material, social and psychological needs may interact in a horizontal manner to create overall wellbeing, and that a success in one may help counterbalance a failure in another (Biswas-Diener & Diener, 2006; Padgett et al., 2008). Our findings in this regard highlight the importance of looking at the social and structural conditions in which social relations are embedded and shaped (Berkman & Glass, 2000). Additionally, our findings stress the importance of continuously working on improving the social wellbeing of socially marginalized people, regardless of their degree of social marginalization and regardless of whether improving their social wellbeing means reestablishing contact with family members or giving priority to ‘loner time’. Our typology of social relations and social isolation shows that social wellbeing is important and indeed meaningful despite a life of severe social marginalization. This may be important knowledge in relation to health promotion initiatives and social initiatives targeted at socially marginalized people.

Study limitations

The qualitative interviews made possible an in-depth analysis providing a greater understanding of the characteristics of social relations and perceptions of social isolation among 46 individuals living in social marginalization in Denmark. We acknowledge the non-representativity of our data; however, our findings — regarding the complex association between social relations and social isolation, the pivotal role of the professionals and the importance of improving social wellbeing despite social marginalization – may be analytically generalizable to other users of shelters and drop-in centres, particularly in welfare states with well-developed social security systems.

Our study population included only seven women. Generally, most users (75%) of shelters in Denmark are men (National Social Appeals Board, 2010); a Danish survey of shelter/centre users had a similar gender-wise skewness (Pedersen et al., 2011).

Interviewing people who are homeless, mentally ill or under the influence of drugs or alcohol may at times be challenging, and the ‘rules’ and setting of the interview may be different from other interview situations. Often the interviewees led chaotic lives, making it difficult to arrange and complete the interviews, and following an interview guide was sometimes complicated by interviewees’ drug or alcohol influence. Thus, as indicated previously, not all of the 46 interviews were relevant or of good quality. In addition, our results may have been affected by interviewees’ mental illness and substance use. For example, interviews were conducted with shelter/centre users in different phases of intoxication and with users of different drug types. This may have influenced responses to questions regarding their social wellbeing and may have produced more positive or negative attitudes depending on the phase of intoxication. We recognize this potential bias, but we also consider it an ‘unwanted side effect’ of studying socially marginalized people, who are very difficult to recruit and whose subjective experiences of social wellbeing will inherently be affected by the often chaotic and unstable context in which they take place. 

The role of the interviewing researcher also needs to be addressed here: it is possible that the interviewees may have regarded the interviewer as a professional or a ‘system representative’, and this may have influenced our results. The interviewees may have thought they could benefit or get help from participating in the interview, or, on the contrary, may have withheld or altered information about themselves and their situation if they perceived the interviewer as a representative of the social system.

Furthermore, we face the risk of having recruited the best-off of the socially marginalized shelter/centre users because only the best-off wished to participate, and only the best-off were able to convert their life experiences into coherent stories. Thus, we reached only a few street dwellers, and our sample did not include those who do not use the services of shelters and drop-in centres. However, our interviewees generally represent a broad spectrum of social marginalization, from the socially vulnerable to the socially excluded, reflecting the variety of people using shelters and drop-in centres in Denmark.