Exploring the relationship between housing conditions and capabilities: a qualitative case study of private hostel residents

Abstract While housing can facilitate many of the freedoms associated with a ‘well-lived’ life, the Capabilities Approach (CA) is yet to have transformed housing research and evaluation. This paper explores the relationship between housing conditions and well-being, using Nussbaum’s version of the CA as the basis for analysis. It draws on data from a UK-based qualitative study of the experiences of individuals residing in privately-run hostels in the North of England. The analysis reveals much diversity in terms of the ways in which the residents perceived their housing conditions and the impacts of these on their exercise of key functions, despite all living in similar environmental conditions. This highlights the highly subjective and complex nature of the relationship between housing conditions and well-being. It is argued that a more robust understanding of the key factors that mediate the relationship being investigated is needed if the potential of the CA to advance housing research and evaluation is to be further realized.


Introduction
Dissatisfaction with traditional income-based measures of individual well-being and societal progress over recent decades has resulted in attention turning to alternative approaches, with the most prominent developments coming from thinkers working in the areas of subjective well-being and the capabilities approach (CA) (Binder, 2014;Evans, 2017). The CA (which is the focus of this paper) advocates that rather than focussing on levels of wealth and material resources (or even, desire satisfaction or preference fulfilment), assessments of well-being should focus on the opportunities that individuals have to lead the kinds of lives they have reason to value (Batterham, 2019;Nussbaum, 2003). The inclusion of measures focussed on opportunities (or 'capabilities') are now increasingly commonplace in national and international strategies and evaluations of well-being (Diener & Tov, 2012;Kimhur, 2020). While the strengths, but also complexities, of using capabilities as an evaluative space have already been debated in a range of policy and practice contexts (see Evans, 2017;Hartley et al., 2005;Hickel, 2020), this article addresses this debate in the context of housing. Specifically, it explores the impact of objectively poor housing conditions on experiences of well-being, and the implications of the findings for current thinking in the field and future housing research. This debate is highly pertinent in the context of growing levels of homelessness and diminishing access to decent and affordable accommodation across welfare states (Baptista & Marlier, 2019;Fitzpatrick et al., 2019).
The paper draws upon the findings of a UK-based study of the lived experiences of a group of individuals living in privately-run hostels, operating at the bottom end of the housing market. This is a property type noted for poor physical standards in the UK (Barratt et al., 2015;Davies & Rose, 2014;Gousy, 2016;Ward, 2015). Nussbaum's (2003) version of the CA and specifically, her list of ten 'functions' considered central to a 'well-lived' life, provides the main organizing framework for analysis. The paper begins with a critical review of the theoretical underpinnings and operationalisability of the CA, followed by a discussion of its relevance to and application within the field of housing. The paper then presents the findings of the empirical study and aims to advance understanding of the relationship between housing conditions and well-being using the language of capabilities and functionings. The analysis reveals much diversity in terms of the ways in which the residents perceived their housing conditions and the impacts of these on their exercise of key functions, despite all living in similar environmental conditions. This highlights the highly subjective and complex nature of the relationship between housing conditions and wellbeing. It is suggested that the diversity found is likely to reflect the mediating role played by a range of personal and social factors. As such, the paper advocates the utility of using capabilities as an evaluative space in housing research but argues that a more robust understanding of the nature and ways in which key factors mediate the relationship between housing conditions and well-being is needed if the potential of the CA to advance housing research and evaluation (and policy development) is to be further realized.

The capabilities approach
As the CA provides the main organizing framework for this paper, this first section will briefly outline the key features of the approach, its main strengths and suggested limitations. The CAoriginally developed by Sen in the 1980sis now an internationally acclaimed and widely accepted approach for conceptualizing, measuring and evaluating well-being at the individual and societal levels (Robeyns, 2006). Broadly speaking, the main premise of the CA is that assessments of well-being and societal arrangements should not focus on resources or people's mental states but the extent to which they have the opportunities needed to lead the kinds of lives they have reason to value (Alkire, 2005;Clapham et al., 2018;Robeyns, 2006). Central here are the notions of 'functionings' and 'capabilities'. 'Functionings' refers to the achievement of states of being and doing that a person has reason to value. 'Capabilities' refers to the substantive opportunities that one has to lead the kind of life they have reason to value (Evangelista, 2010;Jaseevank-Rysdahl, 2001). A further important though perhaps slightly less discussed aspect of capability scholarship is the concept of 'conversion factors'. The concept refers to factors that may either inhibit or enhance the ability of individuals to turn resources and opportunities into functionings, recognizing that everyone will have different abilities and needs. Conversion factors are typically understood to be personal (relating to someone's personal characteristics), social (relating to social norms or government policy, for example) or environmental (relating to the provision of public goods) in nature (Nambiar, 2013). As a normative framework, proponents of the CA advocate that the focus of public policy and services should be the provision of opportunities to enable individuals to lead the kinds of lives they have reason to value, but the concept of conversion factors has led some to argue that policy and services should not only focus on opportunities, but should also be concerned with the extent to which individuals exercise key functionings (Kimhur, 2020;. Considering its place in terms of the broader well-being literature, the CA has been argued to assume a useful 'middle ground' between purely objective and subjective conceptualisations of well-being, which have long dichotomized thinking (van Staveren, 2015). Similar to much theorizing on subjective well-being, the CA places 'the person'their freedoms, wishes, differences and agencyat the centre of analysis (Carpenter, 2009;Evans, 2017). In this respect, it can be argued to bypass concerns over paternalism and individual/cultural difference, which are long-standing criticisms of objectivist approaches to well-being (Binder, 2014;Clark, 2009;Sen, 2004;van Staveren, 2015). Although there is no consensus over how subjective wellbeing fits with the CA, many agree that it should be considered central (Binder, 2014;Clapham et al., 2018;Coates et al., 2015). The CA is nonetheless rooted in an objectivist approach to well-being, with the external conditions of people's lives considered of paramount importance (Binder, 2014;Clark, 2009). Research into the subjective well-being of those living in adverse objective conditions indicates that disadvantaged individuals often exhibit 'adaptive preferences', adjusting their expectations downwards to ensure a level of well-being (Biswas-Diener & Diener, 2006;Clapham et al., 2018). This suggests that some fundamental entitlements should be provided for, independent of the preferences of individuals (Nussbaum, 2003).
Perhaps the most enduring philosophical debates within the capabilities literature concern the specific capabilities which are central to well-being (or 'human flourishing') and who should decide these. Sen has long been reluctant to commit to a list of central capabilities, arguing that the capabilities that are important to individuals are likely to be highly contextual. As such, any lists developed need to be context-specific and arrived at through processes of deliberative democratic reasoning (Sen, 2004). Following much philosophical and empirical enquiry, however, Nussbaum (2000Nussbaum ( , 2003 produced a list which has proved to be highly influential, arguing that there are certain functionings and capabilities that are important to all and that a universal list is not only possible but is necessary as a basis for directing government action and enabling comparative judgements about individual and societal well-being to be made. Nussbaum's list of central functions is 'life', 'bodily health', 'bodily integrity', 'senses, imagination and thought ', 'emotions', 'practical reason', 'affiliation', 'other species', 'play' and 'control over one's environment'. There is a high degree of overlap between this list and those developed by other capability and broader well-being scholars (see Forgeard et al., 2011;Haller€ od & Seld en, 2013;Robeyns, 2003). Since the list was first produced, there has been ongoing debate about whether all of the functions are valid and of equal importance. For example, Sen has argued that 'survival' is the ultimate human function. Others argue that several of Nussbaum's functions are more fundamental than others. Empirical analysis resulted in  arguing that life, bodily health, bodily integrity, senses, imagination and thought, affiliations and control over one's environment should be considered fundamental, while emotions, practical reason, other species and play should be considered of secondary importance (see also Vallentyne, 2009). Extending their analysis further, they usefully developed the concepts of 'fertile functionings' and 'corrosive disadvantage'. The former refers to a situation whereby the attainment of one capability supports the attainment of others. The latter refers to the loss or lack of a particular capability subsequently undermining the exercise of others. Nussbaum (2003) has defended her list on the basis that all ten functions are qualitatively distinct and thus cannot be reduced without distortion but acknowledges that the idea of 'fertile functionings' and 'corrosive disadvantage' may provide grounds for prioritizing some capabilities over others in public policy terms (Nussbaum, 2011cited in Batterham, 2019. While the CA is not without criticism, these typically centre on its operational rather than philosophical merits. Linked to the above, the CA has been described as incomplete and under-specified (Alkire, 2005;Robeyns, 2006), for a number of decisions are needed before it can be applied. These include whether research and evaluation should focus on functionings or capabilities, the selection of and process for selecting functionings or capabilities, the weightings of these used and the impacts of choices made on the research and evaluation results (see Robeyns (2000) for a full discussion). While these are valid points, the under-specified formulation of the CA can equally be considered a strength, making the approach applicable to a wide range of contexts, and remaining credible as long as the parameters of its application are justified in each case (McCallum & Papadopoulos, 2020;Robeyns, 2006).

Housing and capabilities
The CA has proven highly influential across a range of fields including development education, poverty, education, welfare, public health, disability and gender studies (Anand et al., 2005;Burchardt, 2004;Carpenter, 2009;Evans, 2017;Robeyns, 2006;Schischka et al., 2008;Vehmas & Watson, 2014). It has also impacted on policy and practice in both the developed and developing worlds, providing the basis for the OECD Better Life Initiative, the UN Human Development Index, the UK's Equality Measurement Framework and the World Happiness Report (Alkire et al., 2009;Durand, 2015;Fukuda-Parr, 2003;Sachs et al., 2018), to give just a few examples. Within housing studies, the CA is yet to have transformed housing research and evaluation (or policy) (Kimhur, 2020;Lawson, 2020). This is somewhat surprising as the relevance of the CA to housing studies is clear. While not expressed using the language of capabilities and functions, a wealth of research across a range of disciplines has long evidenced the role of housing in facilitating many of the opportunities associated with a 'well-lived' life, including life satisfaction, physical and mental health, physical safety and security, opportunities for social relations and a sense of control over one's life (see Camfield et al., 2006;Coates et al., 2015;Evans, 2003;Krieger & Higgins, 2002;Mallett, 2004;Manturuk, 2012;Stevenson & Wolfers, 2013;van Praag et al., 2003). The relative contributions and ways in which specific housing attributes impact upon well-being have long been a major research topic in disciplines such as psychology, planning and geography, with over 100 different housing conditions identified as relevant (van Poll, 1997). While these studies have yielded conflicting results (regarding the relationship between tenure and well-being, for example), this has often been attributed to the mediating effects of 'the person' (Am erigo & Aragon es, 1990;Clark & Davies Withers, 1999;Moos, 1987;Roberts & Robins, 2004;Tomaszewski & Perales, 2014). These explanations largely reflect the capabilities literature on 'conversion factors' (Robeyns, 2003;Sen, 1999). Furthermore, for some time, there have been calls for a multi-faceted framework for analysis that places the concept of well-being at the heart of housing debate (Clapham, 2010;King, 2009). A number of models for assessing the adequacy of housing conditions and outcomes exist, but in most cases, these are located within discussions of housing quality, with housing quality framed as an end in itself, rather than a route to wellbeing. Where a concern with well-being is discussed, the concept is often reduced to a focus on physical health (van Kamp et al., 2003). In addition, housing research and evaluation has often favoured 'expert' assessments over the subjective assessments of users. This is despite much research highlighting a mismatch between objective and subjective evaluations of housing. Research into the slum clearance programmes of the 1960s is particularly relevant here (Harrison, 2004;Heywood et al., 2002;Murie, 1983). These studies indicate that while useful, physical property conditions and satisfaction can be misleading informational bases for housing evaluation (and policy) and an alternative informational basis is needed.
There has been a noticeable surge of interest in the CA amongst housing scholars in recent years however, with its utilization as both a conceptual lens and basis for empirical study becoming more commonplace (see Batterham, 2019;Evangelista, 2010;Gilroy, 2007;McCallum & Papadopoulos, 2020;Morris, 2012;Nicholls, 2010;Tanekenov et al., 2018;Watts et al., 2018). Critically, these studies have established the centrality of housing to a 'well-lived' life and as suggested by Batterham (2020), this matter should now be beyond question. There is also growing consensus that capabilities are a highly valuable informational base for the evaluation of housing outcomes, providing researchers with a framework which extends the traditional boundaries of research and more effectively captures the plurality of ends which users value (Clapham et al., 2018;Foye, 2021;Kimhur, 2020;Watts & Fitzpatrick, 2020). It is important to note, nonetheless, that despite its ethical merits, few studies have evaluated housing outcomes using 'capabilities' (rather than functionings), as these are largely abstract hypothetical states and thus difficult to measure (Foye, 2020). Linked to this and building upon the seminal work of King (2003), the application of the CA as a conceptual lens seems to be fostering more progressive (and explicitly normative) debate about what housing ought to enable us to do and be, what housing and homelessness policy should aim to achieve and how these aims might translate into practice (Batterham, 2019;Kimhur, 2020;Nicholls, 2010;Watts & Blenkinsopp, 2021;Watts & Fitzpatrick, 2020). But at present, there is limited agreement about how a normative housing research (and policy) agenda should proceed, and this is perhaps the most pressing issue for housing researchers with an interest in the CA. Specifically, opinion is divided over the need for a list of housing-relevant functionings. Kimhur (2020) recently made a case for the potential merits of this and even suggested what this list might include. Taylor (2020), on the other hand, has suggested that it is unnecessary to develop a separate framework for the application of the CA in the context of housing. Instead, it is suggested that the CA is used as a broader normative evaluative approach to all policy issues that impact on the ability of individuals to act as effective agents (of which housing is just one). If this latter approach is employed, it seems important to develop greater understanding of which aspects of housing are relevant to central capabilities in particular contexts (Batterham, 2020), which is one of the intentions of this paper. The concluding section will offer some reflections on the links between aspects of housing and central capabilities, the utility of this approach and the priorities for future research in this area.

Methodology
Before moving on to the findings and analysis, it is first appropriate to discuss the study which the paper draws upon. Specifically, the paper draws upon data collected from a sample of individuals living in privately-run hostels, operating at the bottom end of the housing market, in the North of England. Low-cost Houses in Multiple Occupation (HMOs) are an expanding sub-sector of the UK housing market (DCLG, 2016), but are a relatively neglected area of housing research. In this respect, the paper presents novel insights into this aspect of the housing market, as well as having much conceptual value. The limited evidence base suggests, however, that the subsector typically offers challenging physical and psycho-social environmental conditions (Barratt et al., 2015;Davies & Rose, 2014;Gousy, 2016;Ward, 2015). The study is thus a useful case for exploring the relationship between housing conditions and well-being. The data was collected through in-depth semi-structured (and partly 'life history') interviews with 13 residents, residing at different hostels. Qualitative methods are best placed to map the contours of people's biographies that are fundamental to understanding experiences of well-being and housing as personal and social constructs (Cieslik, 2019;Clapham, 2003). The qualitative approach further adds to the uniqueness of the paper, with much research into housing and well-being traditionally adopting a quantitative approach, involving the analysis of large datasets (Clapham et al., 2018). As a largely unknown and concealed population group, 'gatekeepers' from local authority housing and regulatory teams, homelessness charities, crisis support services, addictions services, criminal justice agencies and welfare, employment and general advice agencies were integral to the identification and recruitment of participants. Of the 13 interviewed, 11 were male and two were female. They ranged from the ages of 25 to 55. All were living in the hostels alone (without dependent children), were White British and all but two originated from the North of England. Eight were hostel residents at the point of interview and five had since moved to other forms of accommodation. All regarded themselves as 'otherwise homeless', having no other housing options available to them at the point of entry. The length of time which the residents had spent in the properties ranged from two months to several years.
The interviews sought to elicit information about the residents' biographies, their reasons for entry into the hostels, the nature of the physical, psychological and social conditions, and the impacts of these on their well-being. The schedules were designed in such a way that they began with simple, 'factual' and less intimate questions, only moving onto more personal and challenging questions once a level of rapport had been established. Furthermore, the specific questions asked and the ordering of these during the interviews varied in response to the flow of the interview dialogue and the nature and experiences of the participants (Bryman, 2012). The key sections covered and broad ordering of the questions were: basic demographic information; information about the physical, psychological and social property conditions within their hostel; their experiences of housing and homelessness; their 'private' lives (family and friendship networks, physical and mental health, substance misuse and significant life events); their 'public' lives (education, employment, engagement with the criminal justice system and contact with support services); the impact of living in the properties on various well-being domains; and, their lives since moving on from the hostels (where relevant). It is important to acknowledge that at the point of undertaking the interviews, the CA had not been finalized as the main analytical frame for the study. Nonetheless, the questions asked enabled its application, with the exception of a discussion of 'other species'. The interviews took place in environments which the participants considered safe and comfortable-typically, the premises of the gatekeeper organizations. Most lasted between 1.5 and 3 hours, depending upon the availability and openness of the participants.
With consent granted in all cases, the interviews were recorded using a digital recorder, and subsequently transcribed. The data was principally analyzed thematically, using the CA as the basis for a coding framework. The process followed Braun & Clarke's (2006) recommended six-phase approach to analysis. A key disadvantage of thematic analysis, however, is the difficulty of retaining a sense of continuity and contradiction through individual accounts, with the contradictions and consistencies across individual accounts often being highly revealing (Braun & Clarke, 2006). Accordingly, a process of 'narrative analysis'an approach to the elicitation and analysis of data that is sensitive to temporal sequences in people's lives (Bryman, 2012)took place also.
To be clear about the application of the CA within this paper, firstly, the focus is on 'functionings'that is, the extent to which the residents exercised various states of being and doing while living in the properties. As noted earlier, this is commonplace within applied capability studies as outcomes information is generally more feasible to observe and assess. Functions are, however, widely accepted to be useful proxy indicators for capabilities on the basis that most people will seek to achieve their capabilities as far as possible. They often serve, too, as capabilities for other functionings (Durand, 2015;Foye, 2020;Robeyns, 2006;Schischka et al., 2008;Wolff & de-Shalit, 2013). Secondly, Nussbaum's (2003) list of functions was used as an organizing framework for analysis. It was not feasible to develop a list of important functions with the participants due to the time-constraints of the study, but I was open to the inclusion of other functions should any have been a key feature of discussions. Thirdly, none of the functions discussed are weighted more heavily than others or aggregate calculations produced. Due to in the belief that all of Nussbaum's function have value and the presence of one does not necessarily compensate for the absence of another.

The residents' experiences of Nussbaum's central functions
This next section considers the residents' experiences of Nussbaum's central functions within the hostels (excluding 'other species').

Life
This function refers to the avoidance of premature death and considering one's life worth living. As per the broader housing and well-being literature (Ineichen, 2003;Krieger & Higgins, 2002), the analysis confirmed a clear relationship between the residents' housing conditions and their awareness of and active negotiation with their own mortality, but the specific nature of this relationship was complex, with diverse viewpoints expressed by different residents. At the most basic level, the provision of shelter, basic amenities, and safety and security measures impacted positively on some of the residents' sense of living a life free from the risk of premature death and increased their sense of having a life worth living. Several reported being 'happy' living in the properties and not wishing to move on. Commenting on the value of their accommodation to the life function, one resident stated: However, it was clear that for many, the sheer provision of shelter and basic amenities and facilities was insufficient to reassure them about a life of normal life expectancy. The unsanitary nature of some of the amenities within the properties generated a reluctance to use them and levels of disrepair further rendered some unusable. Coupled with broken or poor quality safety features, frequent thefts of food and significant levels of violence, some of the residents reported that living in the properties either did not support or actively undermined their likelihood of having a life of long duration. Furthermore, most reported increased mental health problems while living in the properties. One reported being consistently depressed and another described their housing situation as 'hitting rock bottom', preferring to sleep rough than remain in their hostel long term. Furthermore, when former residents were asked about their futures if they had remained within the properties, two thought they would have died due to the use of drugs as a coping mechanism for the unsanitary and unsafe property conditions. Here, one commented, 'To be totally honest, I'd probably have ended up dead. Found in the gutter or something like that. When I see how bad I got, it was just unreal'. This highlights the importance of the quality and not just the provision of housing to well-being (Ayala & Navarro, 2007).

Bodily health
This refers to good health through the fulfilment of basic needs, such as nourishment and shelter. There was much overlap between the data on 'life' and 'bodily health' (and 'bodily integrity', which is discussed next). Through the provision of shelter, amenities and facilities, the properties provided the residents with a certain level of warmth, access to safe drinking water, food storage and cooking facilities, and hygiene facilities. Nonetheless, the extent to which the residents experienced good bodily health as a result of living in the properties varied significantly, even among those living in the same hostel.
Several of the residents interviewed acknowledged the role that living in the properties played in terms of bodily health and for some, the ability to meet their basic health needs was fundamental to their well-being. When asked about the best thing about their hostel, one referred to the cooking and food storage facilities, while another discussed the sleeping facilities, commenting: Being able to sleep in a bed, in the warmth … the first night I was there, when I got a good night's sleep, was the best thing … being in from the cold, having a roof over your head.
However, for roughly half, living in the properties resulted in new or exacerbated existing physical health conditions. One reported chest problems as a result of the lack of cleanliness within their accommodation. Another developed a skin condition as a result of an infestation of bed bugs. Thefts of food from the kitchens and the unsanitary cooking conditions made it difficult for some to maintain a healthy diet and their physical health deteriorated. Others reported difficulties sleeping because of high levels of noise and adverse effects on their health because of this (see Krieger & Higgins, 2002, for similar findings). Here, one resident said: Where my room is situated, it's got like stairs next to it. And I can hear them going up and down the stairs. They don't walk up … they either stamp up or run up, and you end up with creaking in the floorboards, you know, I hardly get any sleep.
A further key theme was the impact of living in the properties on the residents' engagement in substance misuse. The majority who had histories of addiction reported either recommencing or increasing their use of dangerous substances while living in the properties, due to the influence of peers, the widespread availability of drugs and alcohol within the properties and/or using substances as a means of coping with the property conditions. This reflects much of the broader literature on the challenges of living in supported accommodation and low-cost shared accommodation (Barratt et al., 2015;Davies & Rose, 2014;Gousy, 2016). Not all residents, however, made a link between increased engagement in substance misuse and a decline in their bodily health.

Bodily integrity
This primarily refers to freedom from violent assault. The residents' views on the impacts of their housing circumstances on their bodily integrity were highly polarized. Five reported a positive sense of bodily integrity while living in the properties and attributed this to having a roof over their heads, a lockable bedroom door and/or the presence of staff and other residents within their accommodation. One explained: You're safe enough, d'you know what I mean. It's just like a community, you've got your community round you. It's only if you were starting something like, you'd have to get like dealt with.
However, all remaining residents reported a heightened sense of insecurity or absence of bodily integrity due to absent or poor quality security measures and the behaviours of other residents. Incidents of violence were widely reported, as well as drinking, drug abuse, high levels of noise and a more general sense of chaos within the properties. Here, one resident said: You don't feel comfortable and safe, like. You always have people knocking at your door, asking you for things. Drugs, or baccy, or drink, or money, whatever.
Much research has evidenced the risks of violence, abuse and even death which individuals are subject to within home environments (Blunt & Varley, 2004;Goldsack, 1999;Mallett, 2004). Violence or the threat of violence was a trigger for several residents leaving the properties.

Senses, imagination and thought
This refers to freedom of expression, pleasurable experiences and the avoidance of non-beneficial pain. Highly varied experiences of freedom of expression through speech were reported and were dependent upon the nature of the residents' relationships with landlords, managers and other residents. While almost half described relaxed, comfortable and friendly relationships with others within the properties, the remainder discussed feeling highly intimidated and trying to avoid encounters with others as far as possible. Some took extraordinary measures to do so, including vacating the properties as frequently as possible and only returning for brief periods to sleep, where necessary. A further key means of self-expression is 'home-making' (Blunt & Varley, 2004). Just one regarded their accommodation as 'home' and only three talked explicitly about decorating and furnishing their bedrooms with possessions. In some cases, residents did not have any possessions, or the financial resources required to engage in such a process. However, in other cases, residents explicitly reported choosing not to do this due to concerns over the security of their possessions in light of poor security measures and the behaviour of other residents within the hostels. Others were fearful of engaging in home-making in case this undermined their efforts to move on from the properties. As such, some residents chose not to fully pursue this function in order to maximize their practical reason (discussed shortly) (Somerville & Bengtsson, 2002).
There was limited discussion about the relationship between living in the properties and the avoidance of pain, but it is reasonable to assume that living in the properties facilitated this by offering protection from some external threats and some opportunities for the satisfaction of basic needs. The residents were more forthcoming, however, about the relationship between the properties and pleasure, with the amenities and facilities within the properties (such as the televisions in the communal areas), having a private space and opportunities to socialize with other residents being cited as key benefits of living in the properties by some. Others valued the space that the properties offered for engagement in substance misuse. It was clear, however, that there was a tension for some between this function and bodily integrity and bodily health. Some enjoyed the sense of escapism derived from engagement in substance misuse but conceded that it was likely to be damaging to them, physically, cognitively and emotionally. This complex intersectionality has been similarly found in research with rough sleepers (see Nicholls, 2010).

Emotions
This refers to having attachments to things and people. Roughly half of the residents had estranged relationships with family and friends. As such, no relationship between living in the properties and relations with others was found in these cases. But for the remainder, the reputation of the properties, poor d ecor and unsanitary conditions left them feeling embarrassed about their housing circumstances and as a result, several had withdrawn from relationships in order to avoid personal disclosures. Here, one resident said: Even though my room was clean and tidy, and like, me ma bought a hoover and that, I still wouldn't let anyone in, just the building itself. I didn't tell no-one I was living there. Embarrassment and fear of judgement also stopped the formation of new relationships (see Lowry, 1990;Phillips et al., 2005, for similar findings). The organizational aspects of the propertiesnotably, the rules around visitorscombined with noisy communal areas also made it difficult to sustain existing relationships with family and children.
Perhaps counter-intuitively, however, some reported that living in the properties had positively affected their relationships with others, whereby the poor aesthetics of the buildings, poor quality amenities, organizational restrictions and the behaviours of other residents had resulted in them spending more time with their family and friends outside of the properties. Furthermore, through meeting new people, one resident had started a relationship with another resident in their hostel.

Practical reason
This refers to having a conception of a 'good life' and actively working towards this. Most of the residents had hopes for the future and were making positive strides towards the realization of these while living in the properties. For some, what they considered to be challenging environmental conditions within the hostels had made their conception of a 'good life' clearer and were motivating factors for change. For some, access to facilities and amenities and the ontological security which they derived from the properties enabled them to start taking steps towards a 'better' life (see Sixsmith, 1986). For one, meeting and socializing with new people had increased their confidence to look for employment, which they saw as a key route towards a better future. As such, practical reason was clearly linked to the residents' functioning's in the areas of life, control and affiliations.
There were nonetheless several residents who were unable to form a conception of a 'good life' while living in the properties. For these residents, the property conditions were so challenging, they felt unable to think about their futures. Here, one resident explained: I was just really depressed where I was living. I was really down … I couldn't see any future, do you know, it was a really depressing place and intimidating and like I say, at night time and stuff … you couldn't really sleep properly because you're always worrying about your door going in and the people that were there.
Rather than addressing the sources of stress, several of these residents responded by focussing on short-term plans such as funding and maintaining a drug habit. A resident recalled: I felt every day was just the same … from morning til night, I knew everywhere I was going on that day, I knew exactly what I was doing, so … you haven't got a life, you cannot plan something, the rest of your life, staying in there. I had a monkey on my back … heroin, crack, crack cocaine, and that was to do with all, you know, the hostel and that. I would go and do anything to make money, so when I go back on a nighttime, I could have some drugs … it was just a vicious circle altogether.
Of concern is that these short-term responses were only serving to further undermine their exercise of practical reason. For others, the ability to move on was hindered by a lack of information and support services within the properties, and securing and sustaining employment was considered unrealisti, due to an absence of routine, negative peer influences, prolific drug use and/or difficulties sleeping while living in the properties. The only resident who had worked while living in one of the hostels lost this following relapse into addiction due to the stressful nature of the environmental conditions (see Davies & Rose, 2014, for similar findings).

Affiliations and play
This refers to living well with others and the enjoyment of laughter, play and recreational activities. The residents' experiences of affiliations within the properties varied greatly. Some reported 'getting on well' with other residents. Two even regarded the opportunity to socialize and the camaraderie had with others to be the best aspects of living in the properties. As a result of the affiliations developed, one had started to think of their accommodation as 'home' (Easthope, 2004). Key recreational activities within the properties included talking, watching television, listening to music and drinking alcohol with others. In some cases, an informal system of sharing and paying back limited resources such as money, food, tobacco and alcohol formed the bedrock of friendships and created a sense of community. In other cases, physical proximity and similar past experiences fostered a sense of group cohesion. Others found commonality in terms of concerns within the hostels, and they united to challenge unfair eviction practices, maintenance issues and thefts. Overall, the affiliations developed clearly provided some with a range of benefits including survival, enhanced personal safety, happiness and emotional resilience. (see also Garc ıa et al., 2005;Nicholls, 2010). Dunn (2000) explored the 'buffering' and 'direct' effects of social support, with peer support potentially reducing the perception that a situation is stressful and peers being able to provide practical assistance in stressful situations. It was often the case, however, that where a sense of community was present, this centred on risk-taking behaviours. Accordingly, for some, the presence of affiliations was likely to be undermining other key functions such as bodily health, bodily integrity, and practical reason. For others, affiliations and play were not key features of their daily lives within the hostels. Some felt highly intimidated by the manner and unpredictable behaviour of other residents and violence within their accommodation. Others simply did not identify with other residents.

Control over one's environment
This refers to participating effectively in choices that are central to one's life. By virtue of living in insecure and shared accommodation, the residents lacked a degree of control over their lives, with challenges posed not least in relation to levels of privacy and opportunities to seek quiet, refuge and sanctuary (Barratt et al., 2015). In addition to this, the residents discussed a lack of control in terms of their immediate environments due to externally imposed rules, being reliant on landlords and managers for provisions and repairs, and having limited say over who they lived with. However, in most cases, it was the impact of living in the properties on other functionssuch as bodily integrity and emotionswhich resulted in some feeling that they did not have control over their lives.
Nonetheless, a minority of residents suggested that living in the hostels had heightened their sense of personal governance, due to being able to decide when to wash, eat and sleep. What's more, for one, the lack of financial responsibility that came with living in the properties was highly positive and gave them the 'mental space' needed to deal with other matters (see Ineichen, 2003, for similar findings). Overall, however, this function was not discussed in detail by the residents.

Discussion
Three substantive discussion areas emerged from the analysis and are summarized here.

The relationship between housing conditions and central functions
Firstly, a clear association was found between housing and nine of Nussbaum's (2003) central functions. The analysis thus further substantiates the centrality of housing as both an enabling and destabilizing force in the experience of a 'well-lived' life and supports calls for housing to have greater status in discussions of basic justice and well-being (see also King, 2003King, , 2009Nicholls, 2010;Taylor, 2020). At a more practical level, the analysis advances previous capability-informed housing research by identifying relationships between specific housing conditions and different central functions (Batterham, 2020). While the importance of access to housing and minimum quality thresholds for particular physical housing conditions to well-being is well established, the analysis clearly illustrates the centrality of psycho-social conditions to well-being also. In particular, the characteristics and behaviours of fellow residents were fundamental to experiences of bodily integrity, affiliations, and play. The study thus supports calls for greater consideration to be given to psycho-social housing conditions in the evaluation of housing outcomes and policy discussions, with their role currently under-emphasised especially in so far as they relate to well-being (Eyles & Williams, 2008).
The analysis also signals that some functions are likely to be influenced by a much greater number of housing conditions than others. For example, in this study, the functions of life and bodily health were related to a significantly greater number of housing conditions than bodily integrity and emotions. As such, some functions could be seen to have more or less complex relationships with housing conditions than others and are therefore potentially more straightforward or challenging to support from a housing policy and management perspective. Complicating matters further, the qualitative data indicated different strengths of association between the various housing conditions and functions discussed. The impacts of some conditions were discussed at length and so appeared central to the residents' experience of different functions. Others were less frequently or saliently discussed suggesting that they play more peripheral roles in experiences of well-being. Furthermore, reflecting the findings of much research into the relationship between housing and well-being (Diaz-Serrano, 2009;Francescato et al., 2002), there were highly conflicting results in terms of the ways in which particular conditions affected functions at the individual level. For some residents, particular conditions positively affected their experience of functions. For others, the impacts were negative. For others still, no relationship was apparent. This highlights the highly subjective nature of housing experiences and critically, the mediating effects of 'the person' on the relationship in question (discussed shortly).

Relationships between functions
Secondly, the analysis highlighted the complexity of relationships between functions. As identified through previous research (Durand, 2015;Nicholls, 2010;Robeyns, 2006;Schischka et al., 2008), clear interdependent, as well as competing, relationships between the functions could be seen. It was often the case that the conditions within the properties enabled the exercise of some functions, while simultaneously undermining the exercise of others. For example, some enjoyed the opportunities afforded by the properties for affiliations and play, but the space to engage in substance misuse simultaneously undermined the bodily health. The relationship between the various functions was not necessarily bi-directional. Furthermore, some functions seemed to be highly important in their own right, while others seemed to be functions or indicators of others. For example, the exercise of 'life' seemed to reflect the extent to which the residents enjoyed a range of other functions (notably, bodily health and integrity), rather than being a function which the residents actively pursued as an end in itself. Conversely, bodily health and integrity, practical reason and affiliations seemed to be actively pursued by most of the residents. In light of the significant conceptual overlap and conflict between the functions discussed, the analysis suggests that greater consideration should be given to the relevance and value of Nussbaum's list (both broadly and in relation to housing), and the idea of weightings. Additionally, while the functions most widely discussed by the residents largely reflect the list of key functions developed by Wolff & de-Shalit (2013), the analysis suggests that practical reason may be more fundamental to a 'well-lived' life than previously thought.
There was also some evidence of the clustering of functions as discussed by Wolff & de-Shalit (2013), with some residents being acutely focussed on the achievement of both bodily integrity and practical reason, and others preferring to focus on life, bodily health, affiliations and play, for example. The empirical evidence is too limited to extend the literature in this specific respect, but it does give further credence to the notion of clusterings and suggests the need for more research into this with a much larger sample.

The diversity of the residents functionings
Thirdly, one of the most significant points to emerge from the analysis was the extent to which each of the residents was leading 'well-lived' lives in the context of their particular housing circumstances. Despite living in similar housing circumstances, some of the residents reported actively exercising all the functions being explored while living in the properties, while others reported exercising only few, if any. Thus, contrary to past assumptions (Hatuka & Bar, 2017), the analysis indicates that no simple relationship exists between the objective quality of housing conditions and well-being. Furthermore, the varying levels of value and priority given by the residents to different functions suggest that assumptions cannot be made about the extent to which an individual can be considered to be leading a 'well-lived' life by taking into account the number of functions being exercised alone.
The analysis further suggested that the residents' experiences of key functions were mediated by a number of personal and social factors. Indeed, the extent to which the residents reported exercising particular functions was often accompanied by insightful explanations. For example, several explained that their housing conditions compared favourably to past episodes of rough sleeping or prison. This suggests that some of the residents had actively adjusted their housing expectations downwards and highlights the conflict between objectivist and subjectivist interpretations when individual expectations have been suppressed (Biswas-Diener & Diener, 2006;Sen, 1999). For others, the exercise of certain functions was only achievable through a reliance on the external support of friends and family. This highlights the importance of 'personal' conversion factors. Linked to this, the analysis also raised questions about the extent to which some of the residents (particularly those with multiple and complex needs) had the capacity to assess their functionings (Hills & Argyle, 2001;Ryan & Deci, 2001). There were several cases where the residents provided vivid descriptions of objectively inadequate housing conditions but went on to report their enjoyment of living in the properties. This was sometimes in the context of discussing enduring mental health problems and substance misuse issues. This further highlights the importance of taking into account the characteristics, past experiences and resources at the disposal of individuals and their ability to convert resources into functionings in housing research. Without this information, there is considerable scope for incorrect assumptions about the adequacy of particular housing situations and the likely outcomes of policy and practice developments (Batterham, 2020;Kimhur, 2020).

Conclusion
While it should now be beyond dispute that housing is an important site for and source of well-being and capabilities are a highly useful informational space for the evaluation of housing outcomes, both of these points were highly apparent in the analysis presented in this paper. The paper clearly evidences that residence within the hostels had profound impacts on the extent to which the residents were leading 'welllived' lives and furthermore, this was comprised of a plurality of ends which extended beyond issues of resources and utility. While important points to make, the paper extends the literature on housing and capabilities in several respects. Firstly, as stated earlier, there is ongoing debate about how a normative housing research agenda using the CA should proceed. The approach employed in this paper was to apply Nussbaum's list of central functions to the empirical data and further explore the links between aspects of housing and central capabilities. In doing so, it has revealed but also provided insights to the extreme complexity of the relationship between housing conditions and well-being. This includes the range and types of housing conditions which are relevant to different functions and the centrality of particular housing conditions to particular functions. As this study included interviews with just 13 residents, much more extensive empirical investigation is needed, but the study should nonetheless be seen to provide an important basis for further investigation.
Secondly, the paper suggests the need for a re-appraisal of the relevance of Nussbaum's (2003) list of central functions in the context of housing. This is particularly in light of evidence of the 'clustering' of functions, the interdependent but also competing nature of the relationship between some functions, and the different weightings and value ascribed by the residents to different functions. As Foye (2020) recently purported, the question of 'who draws up the list of capabilities and how?' is highly pertinent for housing scholars. The paper adds weight to the suggestion by Watts & Blenkinsopp (2021) that lists of capabilities (or functionings) should be empirically informed, with efforts made to involve those in a diversity of living situations in the construction of such lists. It is likely that those with different past experiences, needs, wants and housing constraints (amongst other factors) are likely to have different views on valuable functionings and the weightings of these.
Thirdly, and perhaps most significantly, the paper revealed much diversity in terms of the extent to which the residents were living 'well-lived' lives within the hostels, despite the properties all offering similar objectively poor housing conditions. It is highly likely that this reflects the different abilities of the residents to convert the opportunities afforded by the properties into functions, with these being mediated by the presence of absence of a number of conversion factors. A failure to understand these could easily lead to misunderstandings about the adequacy of housing conditions and the effects (or effectiveness) of policy. Thus, echoing the recent writings of Kimhur (2020), Batterham (2020) and Watts & Blenkinsopp (2021) not least, the paper suggests that a key priority for future housing research is to uncover the causal mechanisms through which housing influences well-being, not just the conditions that impact on this. But while a focus on conversion factors is likely to yield useful insights, the analysis suggests the need to also draw on broader concepts from across social science, such as that of 'person-environment fit' and 'standards of comparison' (Alkire, 2005;Robeyns, 2006;Schischka et al., 2008). Greater understanding in this area (and the two areas discussed above) could yield highly nuanced and practicable insights in relation to which individuals are most likely to respond positively to (or at least be able to cope with) particular housing situations and why, and which individuals are most likely to over-or under-rate the adequacy of their housing conditions, by how much, and in what ways, thus enabling the development of more targeted, person-centred and effective housing policies and interventions.