Conceptualizing the health and well-being impacts of social enterprise: a UK-based study

Summary Social enterprises–businesses that work for social benefit rather than for the maximization of financial returns to shareholders or owners–could potentially prove to be an innovative and sustainable way of tackling ‘upstream’ social determinants of health. However, empirical work focusing upon how, and to what extent, social enterprise-led activity may impact upon health and well-being is still relatively scarce. This study examines how social enterprises portray their impact, and how such impacts may be considered in health and well-being terms. Through analysing evaluative reports of the work of social enterprises in Scotland (n = 17) utilizing a ‘process coding’ method, we investigate both the self-reported impacts of the work of social enterprises and the mechanisms by which these are said to be derived. Revisiting previous conceptualizations in the extant literature, this work allows us to present an ‘empirically-informed’ conceptual model of the health and well-being impacts of social enterprise-led activity, and thus presents a significant advance on previous hypothetical, theoretically-based conceptualizations. It is considered that these findings further improve our overall knowledge of ways in which social enterprise and other parts of the third sector could be considered as potentially valuable ‘non-obvious’ public health actors.


INTRODUCTION
It has long been recognized that community-based organizations can tackle aspects of social vulnerability (Galea et al., 2005) which we increasingly understand to be critical to public health. One particular form of organization that has attracted considerable policy attention, particularly in recent times, is the 'social enterprise'-a business with social objectives, where surpluses are reinvested in social purposes, rather than for distribution to shareholders or investors (Galera and Borzaga, 2009). There are a number of prominent examples of social enterprises in the UK, including The Big Issue, Divine Chocolate and Fifteen-the restaurant founded by celebrity chef Jamie Oliver which provides opportunities for young people at risk of social exclusion. However, most social enterprises are small in scale, owned and operated by, and for the benefit of, local communities (Ridley-Duff and Bull, 2015).
When social enterprise has been discussed in relation to public health, it is most often in relation to its role or potential as a mechanism for delivering health and social care services, either as an alternative, or complement, to mainstream provision (Hazenberg and Hall, 2014) or as a mechanism for enhancing community involvement and service design, particularly in rural contexts (Muñoz, 2011;Farmer et al., 2012). However, there is still a significant gap in knowledge of how, and to what extent, such activity can impact upon the social determinants of health, particularly in relation to healthenhancing mechanisms and causal pathways.
Drawing upon externally audited reports of the activities and effects of the work of social enterprises, this paper aims to address this perceived gap by investigating the impacts of social enterprise-led activity and the mechanisms through which such impacts are said to be derived. Reflecting upon extant theoretical conceptualizations, our findings are then used to inform the development of a new 'empirically informed' conceptual model, and we conclude by reflecting on possible future research directions for researching the social enterprise/ public health nexus.

BACKGROUND
The idea that organizations led by social entrepreneurs may have a vital role to play in the development of health promotion through community-based action on the social determinants of health was mooted as far back as the end of the 1990s in the context of the Healthy Cities initiative (Catford, 1998;De Leeuw, 1999). Since then, the body of literature relating to the health and well-being impact of social enterprise-led activity has slowly developed, from grey literature written mainly by practitioners (McDermid et al., 2008;Boswell et al., 2009;Westwater, 2009) to theoretical or conceptual papers written by academics (Roy et al., 2013(Roy et al., , 2015b(Roy et al., , 2017Farmer et al., 2012;Muñoz et al., 2015), and we have only now reached a point where we are starting to see systematic reviews emerging. The recent systematic review by Mason et al. (2015) sets social enterprise within a wider context-as a form of 'social innovation'-and attempts to assess the ability of such innovations to address health equity, finding inconsistent evaluative evidence of impact. Some of the benefits they are able to identify, however, include 'the mobilization of latent or unrealized value through new combinations of (social, cultural and material) resources; growing bridging social capital and purposeful approaches to linking individual knowledge and experience to institutional change' (Mason et al., 2015: ii116). Roy et al. 2014, meanwhile, set out their hypothetical case for social enterprises-all social enterprises-to potentially be considered as a complex form of public health 'intervention' since they work to address aspects of social vulnerability at the local level, irrespective of whether they explicitly intend to have a health impact.
An important gap in evidence remains, however, in relation to empirical work which seeks to test this hypothesis. This paper aims to contribute towards filling that gap, to explore how social enterprise practitionersimplicitly or explicitly-conceptualize the impact of their activities upon the health and well-being of the individuals and communities they seek to support. From this analysis, we construct and present an 'empirically informed' conceptual model in order to provide a platform for future enquiry.
One method of gaining an insight into the various ways in which practitioners explain their impact is through an assessment of reports developed specifically for such purposes, namely 'social impact measurement' reports. Social impact measurement has had a chequered history as social enterprises have come under increasing pressure to evidence the 'social value' (Di Domenico et al., 2010) that they purport to create, particularly as a means of gaining legitimacy from stakeholders and funders (Arvidson and Lyon, 2014). The two most common methods of social impact measurement in the UK, among a great many that have proliferated in recent years, are Social Accounting and Audit (SAA) and Social Return on Investment (SROI) (Gibbon and Dey, 2011). Both methods share a number of similarities in that they: both undertake processes to account for activities of the organization; seek to incorporate the voices of a broad range of stakeholders; and consider the ways in which each activity impacts upon social, environmental and economic factors (Gibbon and Dey, 2011;Hall and Arvidson, 2014). In many cases their reports are then 'audited' (SAA) or 'assured' (SROI) by an external observer to independently assess the validity of the claims being made. Thus, these reports provide a valuable insight into the intentions, activities and perceived outcomes of social enterprise practitioners, data which we could then analyse in terms of their potential health and well-being impacts.

METHODOLOGY AND METHODS
SAA reports were sourced from the Social Audit Network website 1 (108 reports) while SROI reports were sourced from the Members' Area of the SROI Network website 2 (59 reports), which has since been renamed Social Value UK. Only 'assured'/'audited' reports were included in the analysis to lend a degree of quality, accuracy and external validity. Given that conceptions of social enterprise are 'politically, culturally, 1 www.socialauditnetwork.org.uk 2 www.socialvalueuk.org historically and geographically variable' (Teasdale, 2012, p. 100), we also decided to limit our sample to organizations based in a single polity with which we have significant knowledge, namely Scotland. Widespread acceptance of the definition of social enterprise by government and the sector means that Scotland has relatively coherent boundaries around the social enterprise concept. This definition requires organizations to aspire towards financial independence through trading, and contain an 'asset lock', meaning 'profits are reinvested in the business or in the beneficiary community and not distributed to owners/shareholders/investors' (Senscot, 2010).
Scotland has recently been described by Scottish Government politicians as having 'the most supportive environment in the world' for social enterprise (Roy et al., 2015a) and a recent report (Social Value Lab, 2015) estimated that there are 5199 social enterprises currently operating in Scotland, employing 112 409 individuals, turning over £3.6 Billion per annum, and holding £8.77 Billion in assets.
87 SAA reports and 46 SROI reports were disregarded as the organizations were based outside Scotland, while 7 SAA and 4 SROI were excluded for lacking a significant trading element. The remaining reports were examined in greater depth to establish whether their activities and institutional form met the accepted definition of social enterprise as detailed above, with 5 SAA and 1 SROI not containing the requisite 'asset-lock'. In the end, 17 reports (9 SAA, 8 SROI) were analysed. Data found in these reports, in addition to publically available information from the Office of the Scottish Charity Regulator (OSCR) and the Financial Conduct Authority (FCA), were used to inform details on the sample as shown in Table 1.
In an attempt to discover what it is that social enterprises claim to do, and what impacts they consider to be caused by such activities, data concerning the distinct processes and outcomes mentioned in each organization's report were identified through utilizing a form of coding called Process Coding (Corbin and Strauss, 2007, pp. 96-97;Saldan~a, 2013, pp. 77-81). This method identifies semantic clues within qualitative data to identify the ways in which an activity (or 'process') can lead to a final outcome. This can take the form of identifying gerunds (working, learning, interacting, etc.) and other distinct statements of cause and effect which could be identified within the reports. Processes, in this case, were identified as any activity of the social enterprise related to a particular health and well-being outcome, irrespective of whether or not the practitioners explicitly intended to have an impact along such lines.
The various distinct processes were grouped into broad themes ('themeing the data'-see Saldan~a, 2013) and coherent sub-themes in order to make sense of the data. Outcomes, on the other hand, were considered to be the result of one or more of the processes of the social enterprise on individuals and/or their communities. We identified a number of specific processes linked to each outcome. Some of the impacts related to the target group of beneficiaries, while some were employees within the organization itself or members of the community. We also discuss the societal effects at the level of 'systems'. To reflect the different recipients of these effects, impacts have been grouped in accordance with their perceived 'level of outcome', albeit certain outcomes were seen to be operating at different levels simultaneously. Table 1 outlines the name, location, constitutional form and a brief description of each of the organizations, as well as the type of report analysed, what period of time it covered and when it was published. The sample includes a broad spectrum of social enterprises, covering 10 years and nine local authority areas across Scotland, and are split almost evenly between SAAs (n ¼ 9) and SROIs (n ¼ 8).

FINDINGS
This section will briefly set out the context in which the various social enterprises operate and discuss the various social missions of the sample social enterprises and the range of vulnerabilities they seek to address. Following this, we consider the pathways through which the various social enterprises were seen to impact upon people's lives through a public health lens, first considering 'intermediate outcomes' and then longer-term health outcomes.

Context
One of the most common challenges cited was that of unemployment (1,2,6,7,9,10,11,14) with recognition that that areas that experience high levels of unemployment do not only suffer economic consequences (7, 11, 12), but are considerably more likely to experience other negative impacts including poor physical and mental health (7,9,10,15). Financial exclusion (2, 5), income deprivation (7, 10) and fuel poverty (6) were contributing factors to the social disadvantage (2) associated with a lack of financial means (6, 10). This had a knock-on effect on communities as areas developed social problems (7,9,11) and the demand to live there decreased (11), often due to community disrepair resulting  from inadequate physical amenities such as housing (7,11,15,16), community space or surrounding environment (3,8), and a lack of community spirit or responsibility for the upkeep of the community (7). Other elements associated with poverty in a community were the levels of drug and alcohol abuse (6,7,9,15) and high crime rates (3). There was also a lack of provision of services to support various 'target groups', including homeless (6, 7, 15), disabled and other vulnerable individuals (6, 15, 16), lone parents (6), ex-offenders (6) and asylum seekers (6). Social isolation, especially in the growing elderly population, meant that people struggled to get out and about and build a social network (2,7,9,14,15,16,17), negatively impacting upon confidence (17), independence (16) and, ultimately, health (15,16,17). Finally, there were often limited skills and capabilities held within communities, both concerning formal academic (6, 7), and other soft skills (17).
In order to mitigate these conditions, some organizations directly targeted these vulnerabilities, while others sought a more indirect route, whereby their work would impact upon other elements in the community.

Outcomes
As detailed in the previous section, the perceived effects of the work of social enterprises were grouped into themes. These themes include: Enhanced social connectedness; Employment, employability and meaningful work; Economic impact; Enhanced confidence and selfesteem; Improved sense of meaning and control; Positive spaces and environments; Access to services; and Improved health and wellbeing. Each theme is backed up by data, such as that explained in Table 2, and we discuss each theme in turn.

Enhanced social connectedness
'Wider Social Networks' was a core outcome that spoke very clearly and loudly through the in depth qualitative interviews. It is the experience of working closely together with others in the Clubhouse environment that produces, time and again, increased number of friendships and social activity.'

9-Scotia Clubhouse
The work of social enterprises was claimed to have resulted in the strengthening (10, 12, 16) and broadening (9, 12) of social networks in communities, increasing the number (6,9,10,12,16) and quality of individual relationships (6,9,10,11,12,13,15,16), and bringing diverse members of communities closer together (10, 12). This was also seen to reduce pressure on other forms of formal or informal support for people (12,14,15,16). The inclusion of individuals in the social groupings, activities and training related to formal employment was the main way through which this was achieved (6,9,12,14). In turn, this led to the development of other life skills including teamwork (9, 10, 14), communication (10,13,14,15) and coping skills (9, 13, 17), as well as the feeling of being involved in, and contributing to, the community (2, 10, 12, 14, 17) which were generally seen to derive from employment environments. Community spaces and activities provided a vehicle through which individuals could engage and contribute to their community (8, 10, 11, 12) while it was seen that improvements to housing could improve the quality of relationships within the family, and with others (11, 15, 16).  Employment, employability and meaningful work 'Real Jobs is supporting disabled people into sustained work which is important for the aims of reaching people furthest from the labour market and tackling inequality in employment'

Economic impact
'The majority of staff live in the local area and as a result of the employment that they are in, have salaries to spend with local business.'

8-Banff and Macduff Community Trust
The increase in individual income was attributed largely to gaining employment (3,6,7,8,11,13,17). However, the majority of financial impacts related to increased savings, supported by the provision of financial services (2, 5, 6, 7), as well as reducing bills and debt (2,5,6,8,13) and other overheads related to inadequate housing (7,11,16). The ability to manage money more effectively was seen to have improved (7) and it was also perceived that having pay automatically deducted and invested in an employer-based savings plan reduced employee absenteeism (5). The economic impact on communities centred around building a vibrant local economy (1, 7): generating (8, 10, 17), retaining (7, 8, 13) and spending (1, 2, 8, 11) money locally. This was achieved through the development of local trading entities (7, 8), improving the superficial look of the community (8) and building the skills and capabilities of local people through specialized business development services (1).
The work of social enterprises was seen to save on public expenditure. This was most widely claimed in the health sector (9,10,12,13,14,15,16) and unemployment services (6,7,8,13). The most common cause for this was the provision of employment opportunities to individuals (6,8,13,14). The provision of adequate housing (11,15,16) and outdoor spaces (12) was strongly correlated with savings to the health service (11,12,16), but was also seen to reduce the demand for the police (11, 13) and social service (13) involvement, as well as reducing council expenditure on housing services (11,15,16).
Enhanced confidence and self-esteem 'Seems more confident and initiates conversation now if she meets people from Milltown when she is out and about'

Improved sense of meaning and control
'Through membership, acceptance and shared ownership of tasks, individuals with severe & long term mental health issues, find meaning, stability, new roles and purposeful work.'

Positive spaces and environments
'Easthall does not have a focal point that its residents could identify with, take advantage of and take pride in. The Glenburn Centre is now complete and provides a place for people to formally and informally meet and generate a positive impact on the area.'

3-Easthall Residents Association
The perception of what made a space 'positive' was made up of the often overlapping factors of: improved superficial appearance (3,8,10,11,12); a place where different individuals and community groups could meet (3,8,10); and the improved ecological environment in the area (3,7,8,10,12). These spaces were also seen to provide safe places with reduced crime (3, 12), caused by a number of community-based initiatives including employment (8, 13), housing (11), social contact (10) and community facilities (3,12). In turn, the ability to create a welcoming environment free from crime led to a sense of ownership and the ability to support social networks (8), secure resources (8, 10) and amplify the voice of the community (8).
The processes generally perceived to lead to such environments were relatively few, consisting mainly of creating green spaces (3,10,11,12), developing local businesses and community initiatives (3,8,10,11,12) and maintaining environmentally friendly institutional policies (4,7,8,12). The attitude of the community and social enterprises towards environmental practices was seen to have changed due to taking steps to reduce waste and emissions and investing in gardening activities (4,5,6,7,10,11,12,17).

Improved access to services
'All were very happy with the level of service provided and the range of services available online' 5-Scotwest Credit Union Awareness of, and access to, services was generally facilitated through the social enterprise providing new services such as community-based educational initiatives (1,2,6,7,11,13,14,17), financial services (2, 5), employment support (6), recycling facilities (3) and the sale of fresh produce (10,12). This sometimes took the form of a local social enterprise collaborating with other organizations (4,8,10) to make services more accessible to people (10,15,17). The availability of the service was then communicated to local people through official communications (4,8,10) or engagement with the local community (3,4,8,10,11). Investing in and providing sufficient housing for all, particularly of good quality (7, 11) was seen as being responsible for a decrease in homelessness (7,11,15).

Improved health and wellbeing
As well as the intermediate effects detailed above, reports indicated that the work of social enterprises was having a direct impact on the health of individuals and communities.
'As a result of the skilled and fast response of concierge staff there has been 11 potentially life threatening incidents intervened in with successful outcomes during the period' 7-West Whitlawburn Housing Co-operative The provision of a pleasant (11), safe, relaxing (12) environment, especially one outdoors (3, 10, 12), was perceived to benefit both physical (3, 14) and mental health (6,9,11,12), as was being involved in a working environment (6,9,10,14). Employment, and especially outdoor work, was seen to encourage healthy eating (4, 9, 10, 12), physical activity (3, 9, 10, 12), personal hygiene (16) and a reduction in the use of medication (6, 9, 12), drugs (6, 13) and alcohol (13). Increased social interaction (9, 10) and support (9, 17) was seen to have an effect on mental health, while good quality housing (7, 11) and health education (9, 10) were seen to benefit physical health. Quality of life and general wellbeing was seen to be enhanced through the provision of financial (2) and housing services (15, 16), job satisfaction (8), the improved appearance of the community (11) and spending time outdoors (10).

DISCUSSION AND CONCLUSIONS
The intention of this study was to build upon existing conceptualizations and present a new 'empirically informed' model upon which to base future work. Drawing on the outcomes identified in the examined reports, as well as the claimed pathways through which they were developed, the model prepared is presented at Figure 1.
In previous conceptualizations (particularly Roy et al., 2014) the different types of intervention were conceived of consisting of (a) those social enterprises that involve the target beneficiaries within the work of the social enterprise, and thus could be said to directly intervene; and (b) those that trade in order to have an indirect impact upon a target set of beneficiaries. We do not explicitly distinguish between these indirect and direct impacts as we consider that a single social enterprise can have multiple impacts upon different individuals, both directly and indirectly, and at the same time. We have also attempted to convey the complexity of what is happening at the local level, recognizing that many different processes impact upon people at different levels, including at the level of the individual, the community, and broader societal systems.
There are obvious limitations of using SAAs and SROIs to construct a conceptual model. For example, there may be justifiable concerns regarding the quality of the evidence in the reports, and also who is doing the reporting, given that: (a) the outcomes were selfreported, albeit that the ones we chose were externally assessed, which injects a degree of validity; and (b) we know that many organizations use the evaluation process for promotional purposes (Arvidson and Lyon, 2014) and so may be tempted to over-claim their impact. Therefore, although not rigorously testing the extent of each impact (they are not intended to be the 'truth' by any means), using these claims to conceptualize and model the potential effects of social enterprises builds directly upon previous conceptual research, and can serve as a platform upon which to build future research into the health benefits of involvement with social enterprises. Potential next steps in this field of research will be to compliment these conceptual models with empirical evidence, and to develop the practical uses of these findings with regard to the evaluation of outcomes. If the activities of social enterprises can be categorized in the manner (if not precisely the groupings) listed above, the relative impacts on each of the intermediate effects, and upon health and wellbeing, may be considered in terms of a complex health intervention. Process evaluation (Moore et al., 2015) may be employed to examine the relative effect of each of the potential pathways in an effort to discover which are more or less effective in regard to health improvement. This may also shed light on how social enterprise activities differ from other private, public or third sector interventions and determine whether there is anything unique about the form within the health field. As is typical of research on the social determinants of health (particularly viz. Dahlgren and Whitehead, 1991), including in relation to contemporary discussions regarding health 'assets' (Morgan and Ziglio, 2007;Morgan et al., 2010), we employ the terms 'upstream' (or distal) and 'downstream' (or proximal) to explain the potential impact of the identified 'intermediate outcomes' in health and well-being terms. It could reasonably be argued that we do not, at least not explicitly within our model, take sufficient account of socio-political factors that we know are crucial to health (Navarro, 2008;Beckfield and Krieger, 2009;Mooney, 2012;Schrecker and Bambra, 2015). Indeed, Haugh (2012) argues that social enterprise practitioners perceive the importance or value of their work in the way that they organize their activities to solve particular societal problems. Inevitably such problems are socio-political in nature: social enterprise is 'inherently political' (Curtis, 2015) and can even involve positing 'an alternative economic culture that differs sharply from the market philosophy, centred instead around the provision of socially useful services, meeting need, ethical trade, and social/community empowerment and democratization' (Amin et al., 2003, p. 116). The impact of social enterprise at the level of the political economy (as far 'upstream' as one gets, it could be argued) also deserves further examination in the future.
Through analysing the work of social enterprises through a public health lens, we see that irrespective of whether a social enterprise considers their impacts to be explicitly 'health-focused', their work has clear implications for health and well-being. This study therefore advances our understanding of the role of actors that are not formally part of health systems and yet obviously have a role to play in creating the conditions for a healthy society. The intention of future research will be to explore the idea of 'non-obvious' public health actors in greater breadth and depth. If there is potential for social enterprises to benefit public health in developed economies, working in partnership to augment or enhance the work of well-resourced, efficient public health services, then this could even start to make an economic case for subsidy of social enterprises and for their recognition as potentially valuable public health actors. Such an eventuality may, of course, create future dilemmas. What may happen to social enterprises if they formally become part of health systems? What unintended consequences might this have for social enterprises, and for the independence of the third sector? However, if it can eventually be shown that investing in the work of actors located in the third sector, rather than in more conventional public or private health services, will yield 'better' results (howsoever determined) in the long term, then this may prove to have significant consequences for public health policy and practice.