Social exclusion is a major risk for people living long-term in adversity and on welfare benefits. Over many decades, one of the explicit goals of government social policy in Sweden has been social inclusion of citizens at the bottom of the income distribution, and minimizing marginalization [1–3]. Social exclusion is therefore ideologically important, as recognized also by the WHO Commission of Social Determinants of Health  and the British Marmot review . Income inequalities have widened in recent times in Sweden  resulting in a growing proportion of the population at-risk of poverty.
In Sweden, means-tested social assistance serves as a last resort safety net. In 2010, 4.7% of the population received social assistance at some occasion during the year . Social assistance can be received as a main source of income or as a supplement to other incomes which fall below a specified level. More than one-third (37%) of all recipients aged 18 years and older had social assistance for at least 10 months in a year . Recent Swedish studies [2, 8] suggest difficulties in getting off social assistance. Young adults, single mothers and immigrants (especially newly arrived) are common groups among recipients [2, 8]. For example 24 per cent of Swedish lone mothers received social assistance on some occasion in 2010 .
During the severe economic recession in the 1990s, unemployment increased as did the number of social assistance recipients. Social policies became more restrictive, including cutbacks in public benefits and tighter eligibility criteria to means-tested benefits . The average duration of social assistance recipiency has increased from 4.3 months in 1990 to 6.6 months in 2011 . Hence, the situation of long-term recipients deserves special attention, as their situation may be seen as an indication of the state of welfare in the country.
Levels of social assistance in Sweden as in the other Nordic countries, have been higher than in other welfare state models . Kuivalainen and Nelson , however, show the recent development of the erosion the adequacy of benefits in Finland and Sweden. In 2007 Sweden provided less adequate benefits than, for example, the Netherlands and Germany . Kuivalainen and Nelson [9, 11] state that current benefit levels do not provide enough money for families to escape poverty in the Nordic countries; in terms of benefit generosity and poverty outcomes Nordic social assistance has moved closer to international patterns.
Studies have shown that both individual and structural factors are associated with receipt of social assistance . Recipients of social assistance in Sweden and Norway, for example, tend to be worse off than non-recipients in terms of educational achievements, financial resources, access to material goods, employment, self-reported health and quality of social relations [1, 12]. Exit from long-term social assistance to work is higher for men than for women, and higher for younger people [1, 13], but generally low among long-term recipients , suggesting that the welfare-to-work programs may be less effective for long-term recipients .
Qualitative studies among social assistance recipients in the Nordic welfare contexts have highlighted that shame, feelings of being stigmatised, anxiety, isolation and feelings of insecurity are common [15, 16]. Living on social assistance for a long time, in ‘chronic adversity’ , also means living with limited material resources, lack of autonomy, and feelings of ‘otherness’ , with consequences for health and well-being [15, 16]. Some of those already suffering deprivation may accumulate further disadvantages over time, increasing the risk of marginalization and exclusion from mainstream society [18–20].
The concepts of social exclusion and poverty in its widest sense overlap. By poverty some researchers refer to a lack of resources, especially income , or to relative aspects , whereas in social exclusion the multi-dimensional character is emphasized; components of personal welfare like employment, education, financial resources and health are seen as interrelated and mutually reinforcing [23–25]. Sen  discusses poverty in terms of poor living, capability deprivation, and being excluded from social relations which can lead to other deprivations and further limit a person’s life chances. Social exclusion takes place in a specific context; it is relative and dynamic in that not only the present, but also the past and the future are relevant factors in the process of social exclusion . Agency, another aspect of social exclusion , is a focus in this article and therefore discussed in more detail. Agency, or the capacity of people to act, has recently gained increasing attention in research related to social exclusion.
People who are excluded as active agents in their lives
The idea of agency is usually to see individuals as “autonomous, purposive and creative actors, capable of a degree of choice” (page 125) . Giddens  highlights that agency refers not to the intentions people have in doing things but to their capability of doing. According to Sen , functionings concern what a person manages to do and capabilities what a person can do, or can choose in her situation. The focus is on the positive, what kind of life people will be able to live, and how people are able to flourish . Sen  states that understanding the agency role is central to recognizing people as responsible persons: “not only are we well or ill, but also we act or refuse to act, and can choose to act one way rather than another” (page 190). The central question is the extent to which wider social, economic and political structures enable or constrain the agency of different groups and how their agency can have an impact on the wider structures in society .
Lister  discusses agency in relation to four mechanisms: ‘getting by’, ‘getting back at’, ‘getting organised’ and ‘getting out’. ‘Getting by’ refers to how people manage and adapt to their circumstances, deal with processes of stigmatization and exclusion and cope with personal trauma and tragedies of past and present [18, 31]. ‘Getting back at’ refers to the resistance people in poverty show, for example when dealing with welfare authorities and against negative labelling. ‘Getting organised’ is a collective act, involving the capacity of groups living in poverty for political activism and collective self-help . ‘Getting out’ refers to attempts to move out of poverty and exclusion, which according to Lister , can include both individual actions of the poor and the non-poor and economic and social processes and policies. The links between individual agency and the social environmental context has also been highlighted in the resilience literature.
Agency and resilience
Often, resilience is conceptualised as an individual trait or characteristic: people are either resilient or they are not. There is a small but growing body of literature, however, that conceptualises resilience as a process: “a dynamic process encompassing positive adaptation within the context of significant adversity” (page 543) . In this literature, resilience is not a condition of individuals alone but related to vulnerability and protective factors in individuals´ environments [32, 33]. Ungar and Liebenberg  highlight the importance of both external and internal resources for positive development, which in turn are influenced by culture and context. The focus is on conditions enabling individuals to develop normally despite adverse life conditions [32, 35].
The importance of creating settings where individuals have the opportunity for positive development has also been acknowledged [33, 36, 37]. In the health assets literature , several other concepts like sense of coherence  and self-efficacy  have been discussed, as related to the concept of resilience. Resilience is a relative, culturally and contextually dependent concept; for example, for resilience, important resources like self-esteem and efficacy are more or less valued in different settings [35, 36]. In the study reported here we maintain the definition employed in our previous studies, in which resilience is conceptualised as: “the process of achieving positive and unexpected outcomes in adverse conditions” (page 238) .
People who are poor or socially excluded are often described in research as passive recipients of welfare services [18, 42]. Little is known about the ways in which long-term recipients of social assistance are active agents in their lives, or their experiences of the obstacles and possibilities to cease welfare. From a public health perspective it is important to gain deeper knowledge about what in a long-lasting adverse situation makes the situation better or worse for those experiencing it.
The aim of this study was to explore and describe the ways long-term recipients of social assistance manage long lasting adversity and act as active, rather than passive, agents in their lives. We were also interested to examine social and health services, and their potential role as either supporting or undermining the well-being of social assistance recipients. We discuss our findings from a resilience perspective.