Social Participation of People with Chronic Mental Health Needs: Building Horizontal and Vertical Forms of Social Capital

,


Introduction
Mental health services in Australia and across the globe are increasingly recovery-focused, moving away from traditional medical models that were primarily concerned with diagnosis and treating illness [1].Mental health recovery is defned as an individual and unique journey towards "a way of living a satisfying, hopeful, and contributing life even within the limitations caused by illness" [2].Social participation is positively related to the mental health recovery and wellbeing of people with mental health needs [3][4][5].Burns-Lynch et al. [4] found a positive association between higher levels of recovery, wellbeing, and the number of days of participation (intensity of participation) and the range of participation areas (breadth of participation).Tey concluded that more numbers of participation in areas that people considered important, and to the extent that they desired, were positively related to recovery and wellbeing.Nevertheless, people with mental health needs often experience fewer opportunities for social participation in settings of their choice [5][6][7][8].
Social participation requires making social connections with people beyond health and social care services [9].Chang et al. [10] proposed a conceptual model of social participation that builds on the defnition of participation as "involvement in life situations" of the World Health Organization's International Classifcation of Functioning, Disability, and Health framework [11].Chang's model includes participation in social life areas (e.g., participating in social groups and keeping in touch with family and friends), leisure life areas (e.g., going to the gym, to the movies, to a restaurant), and productive life areas (e.g., participating in learning activities and work for pay).
Most research on social participation has investigated the engagement of mental health consumers (see (i) in Additional Points) in diferent domains of community participation, such as with families and friends, employment, education, and sport (e.g.[7], age-related expectations for community participation [5], the efectiveness of interventions to promote social participation of adults with mental health problems in their wider community [8], the efectiveness of interventions aimed at expanding mental health consumers' social networks [9,12], or the validation of community participation measurement tools [3,10]).
Fewer studies examine the characteristics and implications of social participation that mental health organisations promote through their services.Tey have not examined the areas of social participation (social, leisure, or productive) that mental health services facilitate and the implications of this type of support to improve social participation for people with mental health needs.Addressing this research gap can help service providers to adopt interventions that increase the intensity and breadth of mental consumers' social participation and therefore their opportunities for recovery and wellbeing [4].
Tis paper applies key concepts from the network perspective on social capital (bonding, bridging, linking, horizontal, and vertical) to re-examine the fndings about the experiences of social participation of people with severe and persistent mental health needs (consumers) that were reported in the evaluation of an Australianintegrated intervention aimed at improving consumers' wellbeing [13].Tis social capital perspective ofers a lens to examine the breadth and intensity of participation experienced by the consumers taking part in the support program.Te article adds to the literature about how service providers can improve social participation and therefore consumers' opportunities for recovery and wellbeing.It does not test specifc quantitative or qualitative measures of social capital, which were not in the remit of the original evaluation.
Te article has four main sections.Section 1 explains the concept of social capital and its relevance to social participation.Section 2 reports on the methods of both the original study and the analyses in this paper, and Section 3 reports on the article's fndings.Section 4 discusses the study fndings in the context of the literature on personal recovery and suggests implications for interventions that seek to promote social participation of people with mental health needs.
1.1.Social Capital.Social capital is a complex and compound construct that is studied in multiple disciplines (including sociology, social epidemiology, and political sciences), with diferent perspectives [14][15][16][17].Within this diversity of approaches and defnitions, Burt [18] proposes that overall social capital is commonly used as a metaphor about advantage.Social capital theories propose that people who do better in society do so because they are somewhat better connected [18].From this perspective, social capital theories are complementary to human capital explanations of people's advantages, which focus primarily on people's personal skills and characteristics, rather than their social context.
In this research, we apply the network approach to social capital, which conceptualises it as the actual and potential resources embedded in people's social networks and social relations [19][20][21].Tis approach argues that people access their resources of social capital through the direct and indirect social ties that form their social networks and relations [20].Overall, social capital is a function of the number of social ties that make up people's social networks, the quality of the ties, and the amount and quality of the resources the ties can give access to [22].
Te network approach stresses the importance of both horizontal ties, that is, associations between people within the same social group, and vertical ties, that is, associations between people belonging to diferent social groups [23].Te literature further distinguishes between bonding, bridging, and linking social capital.Bonding social capital refers to horizontal connections within groups or communities characterised by similar sociodemographic background and strong close relationships.Examples include family members, close friends, and neighbours.Bridging social capital can refer to both horizontal and vertical connections that link people across groups, communities, or organisations that do not share common characteristics, for example, because they come from diferent cultural (e.g., ethnic and religious) or socioeconomic backgrounds.Linking social capital refers only to vertical connections but more specifcally connections between people who have diferent social positions or power in society [24].Examples are the relationships between a community-based organisation and government but also people's connections with welfare services, educational organisations, and doctor-patient relationships.Bonding, bridging, and linking social capitals all contribute to recovery and wellness and provide pathways to participation in society [25].
In this research, we use these concepts from the network approach to social capital as analytical tools to investigate the social participation promoted by service providers.We do this by applying these concepts to the fndings of a three-year evaluation of an Australian-integrated program aimed at improving the wellbeing of people with chronic mental health needs.Applying the network approach as an analytical tool is novel and has two main advantages compared to other ways to understand participation and social capital in mental health services.First, it avoids the lack of theoretical clarity that often characterises empirical social capital research [26,27].It also avoids the conceptual overlap that can occur when the concepts of social capital and social participation are operationalised using similar indicators [28].Second, applying the network perspective helps to assess the breadth and intensity of the consumers' social participation by investigating the characteristics and implications of the social networks promoted through the service providers' activities.

Methods
Te article applies the network approach to social capital as a framework to analyse the fndings of a longitudinal mixedmethod evaluation of an Australian program, CLS-HASI (Community Living Supports and Housing and Accommodation Support Initiative), in New South Wales (NSW).Te article's research question is as follows: what types of social capital were generated by the social participation activities of the service providers for people with chronic mental health?
Te aim of the evaluation that generated the data for this research was to assess the outcomes of the CLS-HASI program (outcome evaluation), identify what worked well in the program (process evaluation), and determine whether the program was cost efective (economic evaluation).Social participation was one of the main outcomes pursued by the CLS-HASI program and the object of a focus report in the evaluation study [13].In this section, we introduce the CLS-HASI program, describe the evaluation methods and its data sources, and fnally the specifc methods, samples, and data sources applied in this paper to reanalyse the evaluation's fndings on social participation using the concepts of the network approach to social capital.Te approach is to tailor support to the person's needs and goals.Te services and activities include support with daily living activities like shopping, meeting people in the community, learning new skills, maintaining accommodation and tenancies, and accessing other services such as clinical mental health and disability services.In many instances, these activities are organised by support workers, whose competency, skills, and training are often contributory factors in the achievement of social participation.
CLS-HASI is informed by the personal recovery approach, which is described as a person's own path toward achieving a fulflling and meaningful life within the constraints of their mental health condition (see Introduction).Tis approach aims to move beyond the limitations of traditional medical models primarily concerned with diagnosis and treating illness.Te personal recovery approach has been criticised for its individualistic focus, which can overlook the importance of sociostructural disadvantage [29]-such as poverty, homelessness, and discrimination on the basis of race-and interpersonal relationships in the recovery process [30].Jorgensen et al. [31] highlighted the tensions between the healthcare professionals' desire to meet and respect the mental health users' perspectives while living up to the strategic objectives that characterise hospital systems subject to societal norms and neoliberal development and focused on diagnosing and treating mental health consumers.Overall, these criticisms highlight the focus of the primary frameworks of personal recovery, such as the CHIME framework-connectedness, hope and optimism, identity, meaning and purpose, and empowerment [32,33]-on intrapsychic states such as hope, empowerment, and the personal journey toward a recovered identity [34], rather than structural factors that can limit the agency of mental health consumers.

Evaluation Methods.
A longitudinal mixed-method evaluation of the program ran from 2017 to 2021 [13].Ethics approval was from the ethics committees of UNSW Sydney, the NSW Aboriginal Health and Medical Research Council (AH&MRC), the South Western Sydney LHD, Corrective Services NSW, the Justice Health and Forensic Mental Health Network (JH&FMHN), Department of Communities and Justice (DCJ), and the NSW Population and Health Services Research Ethics Committee (P&HSREC).Peer researchers with lived experience of mental health needs and Aboriginality contributed to the evaluation design, data collection, and analysis.

Qualitative Methods.
Te longitudinal qualitative data were two rounds (in the second half of 2018 and 2019) of individual and group interviews with 111 participants, including 50 consumers, 2 family and carers, 31 service provider staf and managers in three sites (city, regional and rural), and 28 statewide organisational stakeholders in 5 focus groups.Te three sites were selected to cover variation in local context.Maximum variation sampling was used to include consumers with varying degrees of engagement, outcomes, and changes in their wellbeing and service providers and statewide organisational stakeholders with different roles in the program.Te fnal longitudinal qualitative sample size was sufcient to answer the evaluation questions, allowing the evaluation team to understand the complex interplay of contextual factors, and identify potential challenges and success stories within the program.
Participation was confdential and voluntary.Recruitment of consumers and families was at arm's length from the researchers through the local service providers.Interviews were semistructured and explored the person's experience of the program.Te questions were designed according to the program logic and advice from the lived experience researchers to measure change in outcomes and efectiveness of the program [13].Interviews were audiorecorded transcribed and entered in the software NVIVO, which was used to assist with the coding of the transcripts and their thematic analysis [35].Te fndings of the evaluation and a full explanation of its qualitative methodology were reported in a fnal report on the efcacy of the program in relation to its aims and objectives [13].

Quantitative Methods.
Te longitudinal secondary quantitative program data included consumer demographics, support, and outcomes in the program's Minimum Data Set (MDS) during the evaluation (2017-2019).Tese data included broad social participation types and standardised mental health outcome measures: Mental Health Outcomes and Assessment Tools MH-OAT [36]: Life Skills Profle 16 (LSP-16), Kessler 10 (K10), and Health of the Nation Outcome Scales (HoNOS).Two versions of MDS were analysed MDS Version 1 (MDSV1, November 2017 to April 2019) and MDS Version 2 (MDSV2, May to September 2019).Other quantitative-linked data for the outcome evaluation were from statewide data sources about consumers, such as health, housing, and corrective services.Te datasets were developed into a longitudinal time series framework to compare consumers' service use and outcomes before, during, and after support (2015-2019).Full analysis of these data is available in [13].
All consumers in the program were included in the quantitative data sample for MDS and linkage data: 5,533 consumers, including data about 4,619 consumers in MDSV1, 2,880 consumers in MDSV2, and some consumers in both datasets depending on their time in the program.
Consumers were relatively even by gender and with an average age of 42 years at program entry (Table 1).Around 44% had a primary diagnosis of schizophrenia with a similarly high level of coexisting conditions and risk factors.Levels of advanced education and work capability were low.Aboriginal and Torres Strait Islander consumers were overrepresented with over 15% of consumers compared to the current state Aboriginal population of around 3.5%.Te average length of time consumers stayed in the program was 10.7 months.Te characteristics of the qualitative sample were similar to the total quantitative sample.

Research Methods.
For this article, the authors reanalysed the interviews with consumers, family members and carers, and service provider staf and managers to examine the consumers' experiences of social participation and inclusion.Te authors did not use other data from the evaluation because they were not directly relevant to the focus of this article.
Taking a similar approach to that used in the original evaluation, the authors generated a preliminary list of nodes including the areas of social participation (social, leisure, and productive) and the concepts from the network perspective on social capital (bonding, bridging, linking, horizontal, and vertical).Te article's frst, second, third, and sixth authors divided the interviews between themselves and set up a plan to ensure a systematic and consistent approach to coding.Te plan entailed meetings and discussions to ensure a shared understanding of the coding framework, including doing a pilot coding of a small subset of the interviews to identify potential ambiguities or discrepancies in the coding process.Any disagreements or discrepancies were resolved by discussion, and notes and clarifcations were added to the nodes in NVIVO.
Te frst, second, third, and sixth authors started coding the interviews using the agreed analytical framework, adding new nodes to it to capture wider aspects of the activities the consumers engaged with.Examples of the new nodes were nodes about whether the consumers engaged in group activities or individual activities, whether the activities were in the community or in the premises of the service providers, the people who were involved (e.g., family, friends, and case workers), and whether they were associated with positive or negative experiences.
Regular coding consistency meetings were conducted throughout the analysis process.In these meetings, the authors engaged in ongoing refexive discussions, openly exploring their perspectives and potential infuences on the coding decisions.Tese discussions helped the research team to refect on their positionality throughout the analysis process.
Te qualitative data analysis was supplemented with quantitative data from the original evaluation to understand the types of support used and some relevant outcomes for the much larger sample of all consumers in the program.

Findings
Te fndings of the analysis are organised around the types of support that CMOs provided to promote participation in social, leisure, and productive life areas: social and leisure support (group, community, and family) and productive support (education, training, and work).For each type of support, we analyse the type of social capital that the social participation generated or not, as summarised in Figure 1.

Support Use and Wellbeing Outcomes Related to Social
Participation.Before presenting the specifc analysis about social capital, it is necessary to describe the types of support received and the wellbeing outcomes for all consumers in the program.Tese supports and outcomes are relevant because research shows that they might be associated with changes in social participation and social capital (see the introductory section on social capital).Each consumer received multiple support types according to their needs (Table 2).Te number of hours for each support type were tailored to the consumer and were adjusted up or down over time.
Te analysis for this article could not investigate the direct impact that support for participation in social, leisure, and productive life areas had on the consumers' wellbeing and recovery because the program took a holistic approach of multiple support types for each person.However, the study found that, overall, consumers liked the support they database showed that about one-third of consumers had a clinically meaningful improvement in mental health measured by the K10 scale (30%) and the HoNOS scale (33%) based on efect size [13].For most of the remaining consumers, scores on the two scales improved but not enough to be clinically meaningful.At the time of the frst round of interviews, it seemed that many group activities were taking place within the premises of the CMOs.Some consumers felt that their participation in the group activities was mandatory and that they could not voice their concerns regarding taking part in them.By the second round of feldwork, it appeared that CMOs were ofering more group activities that were meaningful to consumers and ofered more activities in the community, for example, walking groups.Some CMO workers reported that such changes in group activities occurred in response to informal feedback from the consumers.On average, CMOs provided 8 hours of group support per month per consumer (see (ii) in Additional Points).

Social
From a social capital perspective, overall, the group activities that the CMOs supported were an opportunity for consumers to foster horizontal ties, that is, connections and support networks with people in similar circumstances, particularly other consumers, and others associated to them, e.g., the CMO workers.Some consumers described friendships they made through the group activities.Tese  Tese reports from the consumers show that the horizontal ties that were promoted through the group activities did not always generate bonding social capital, that is, strong ongoing close relationships.

Community Engagement.
Te social capital impact of support for community engagement seemed to be limited on the consumers' horizontal and vertical ties.CMOs promoted the consumers' participation in the community by supporting their engagement with a wide range of activities, based on the consumers' personal goals and preferences.Most consumers reported the signifcant impact that this support had on their wellbeing: Well, it's changed my life, I just used to sit at home and smoke cigarettes day in, day out, being really depressed and mentally not well.CLS-HASI have just changed my life, they take me out nearly every day exercising, which was one of my main goals was to get back in to being healthy again.Tey gave me a membership with a gym and a pool to swim (Consumer).
Te main help I get these days is getting me from out of bed onto a trip around ten-pin bowling or fshing, we all do that kind of thing, you know?But that's where they help me most.(Consumer) In some cases, support workers supported the consumers with their daily errands until they felt more comfortable to attend independently.For consumers who lacked independent travel skills, such support included providing guidance and taking public transport together with the consumer until they were confdent in their capacity on their own.Consumers and support workers reported that providing support with transport was often a key factor to facilitate community participation.MDS data showed that the average number of support hours for travel to and from consumers varied considerably depending on the level of support they needed.It ranged from 35.9 hours per month for consumers with high-level support needs to 5.9 hours for consumers with medium-level support needs and to 1.8 hours for consumers with low-level support needs (Table 2).
Overall, consumers-regardless of their support needs-and support workers reported experiences of community engagement which seemed to be mostly casual and temporary in nature.Many consumers talked about how the support they received in relation to community engagement helped to grow trust in their support worker, which was often described as essential to their social engagement.Te support provided also helped people to enter and feel safer being in community settings.However, they did not report developing strong bonds with other people outside the CMOs through their community engagement, and any engagement they had with people other than CMO staf was often transactional rather than social, such as ordering cofee or paying for groceries.
I've made friends, I say hello and talk while I'm here, but not while out of there.(Consumer) For some service users, this might be the frst step to build confdence socialising in the community.I think it helps.It does, yes.Yes, it's good to be able to socialise and to be able to get out into the real world and try to do normal things.Otherwise, you lock yourself in a house in a room and you just go downhill.But by getting out into the real world with support it's a good thing.(Consumer) From a social capital perspective, the types of community engagement that consumers and support workers reported in the interviews had the potential to foster both horizontal ties and vertical ties, that is, connections with people that belonged to diferent social groups compared to the consumers.However, the often temporary and casual nature of the community engagement described by the consumers and support workers suggests that it did not commonly generate strong ongoing close relationships (bonding social capital) or strong connections with people from other groups, communities, or socioeconomic backgrounds (bridging social capital).

Maintaining or Restoring Family Connections.
Te impact of support to maintain or restore family connections seemed to be to strengthen their bonding social capital.Te program supported consumers to strengthen their family connections when that was a goal they agreed with.Family relations were a key focus of the program, so they engaged with families to be actively involved in the program when that was relevant to a consumer.6 Health & Social Care in the Community Te latest MDS data showed that about half of the consumers had connections with a family member or carer during their time in CLS-HASI3 (see (iii) in Additional Points).Most consumers who took part in interviews did not have strong connections with their families.
Where appropriate, CMOs supported consumer recovery through reconnecting with family: Because I have a big family, when I frst got started . . .they asked who was in our close circle and then they got to know my mum.She was pretty important to me.Tey got to know my Nan.Tey know my family that's close to me... (Consumer) Sometimes, however, there was family history or behaviour that made reconnection problematic for the consumer: I love my family, but the thing is they all drink alcohol and do drugs.(Consumer) Overall, most CMOs appeared to defne family as limited to immediate blood relatives.However, some CMOs embraced a broader concept of kinship or community ties, as often found in various non-Anglo-Saxon cultures.Some LHD and CMO staf members emphasised the signifcance of connecting with the extended families of Aboriginal consumers: We look at the family . . .So, we can bring them in too, if they need it. . .You need to cast the web a bit further, just not on the individual.(Staf) Te CMOs assisted consumers to reconnect with their families in two ways according to the staf.First, the services supported consumers in ways that they would otherwise seek from their families and carers.Tis support enabled consumers and their families to develop relationships outside of the cared for/carer dynamics: Te carers that I have spoken to who have got loved ones in the scheme really appreciate it.Tey feel that it takes a lot of the support burden of them . . .but also allows perhaps a bit of space for their loved one to develop without feeling too much oppressive support from a family member.(Stakeholder) Second consumers were able to re-establish and foster family relationships when staf supported them by organising meetings between consumers and their families in safe spaces.Safe spaces were, for example, a public place or a private room at the CMO location.
Te support for consumers to strengthen family connections was a small proportion of the total support that staf provided to consumers.MDS data showed that, on average, CMO workers provided 0.3 hours of support for family connections per consumer per month, ranging from 0.1 hours for consumers with low-level support needs to 0.4 for consumers with medium-level support needs and 2.5 hours for consumers with high-level support needs (Table 2).
From a social capital perspective, the support to improve family connections resulted in about half of the consumers in the program having an opportunity to strengthen their bonding social capital.However, the focus of the support on immediate family suggests that it was likely to reinforce a small number of relationships.Support for connections to wider kinship or community networks remained a potential source of bonding capital for most consumers.

Social Capital from Support for Productive Participation.
Te second type of support was to increase productive participation.Te social capital impact of support for education, training, and work seemed to be limited in generating "weak ties" that bridged between groups of people horizontally and vertically.Consumers were supported by the program to be involved in education, training, and paid or unpaid work.MDSV2 data showed that consumers received about one hour per month of support for educational or vocational activity or work, with no substantial diference across consumer subgroups.CMOs often focused on providing support for other social and daily living skills, according to the CMOs, which they felt were necessary to enable consumers to do formal learning and work in the future: We look at housing, fnances, education, their social interaction, any cultural things that they want to do.Tey are our main four things that we look at and then from there we work out what it is.It could be education, we have had some clients that want to go back to study.About 10% of consumers were employed, mostly part time (8% in MDSV2).More than 20% were looking for work, also mostly part-time work (17% in MDSV2).On average, consumers had a high score, that is, a negative score, on the question about capacity to work in the consumers' Life Skills Profle (LSP-16), with a high proportion of consumers who were "capable only of sheltered work" or "totally incapable of work."A consumer said, "I do struggle with daily tasks," explaining why work was not possible for now.
Te interviews with consumers showed that those who did work in the community enjoyed it: And I could be of site with (charitable organisation) ... it is not so much the money, I just want something to get up in the morning for (Consumer).
Te state and sector stakeholders who participated in the interviews saw volunteer work as a pathway to paid employment or as a goal for recovery.However, this was not evidenced in the data or in the consumers' interviews.
Overall, the support that consumers received to engage with education, training, and work was limited, with only a small proportion of consumers engaging with them.Te person-centred and goal-driven nature of the CLS-HASI, together with the high scores on the LSP-16 regarding the consumers' capacity to work, suggests that this might be because, for most consumers, participating in education, training, and work was a long-term goal.Tis assumption is supported by the fnding that LSP-16 scores decreased slightly following entry to CLS-HASI, although this change was not statistically signifcant (see (iv) in Additional Points).Te analysis still showed a clinically meaningful improvement in 26% of consumers.Te study also found that the support level reduced for many consumers during their time in CLS-HASI [13].
From a social capital point of view, participating in education, training, and work can help to generate weak ties, that is, bridges between closely knit groups of friends (intercliques), both horizontally and vertically.Te literature shows that these links can be an important source of information, including fnding work at the time of unemployment [37].Overall, only a small proportion of consumers had the opportunity to generate bridging and linking social capital by participating in education, training, and work.

Discussion
Tis study examined the types of social capital generated by the social participation support of the service providers for people with chronic mental health needs in an integrated program that aimed to support their recovery and wellbeing.Te research applied a network perspective of social capital (bonding, bridging, linking, horizontal, and vertical) to analyse the consumers' social participation in their social, leisure, and productive life areas.Te fndings showed that most consumers participated in social activities such as group activities in the service.Some consumers restarted connections with family members, where that was their goal.Most consumers participated in leisure activities based on their personal preferences, again mainly with other consumers in the service.Few people participated in education or training.
Tese fndings, seen through a social capital lens, reinforce the current literature that service providers tend to target the types of social activities within the consumer group that strengthen bonding and horizontal social capital [7,26].Tey were less focused on extending the support to activities that promote bridging, linking, and vertical social capital.Tese fndings are not surprising in a service context where support workers manage limited resources and community stigma against people with chronic mental health.Teir approach has implications for consumers' current and future social participation in social, leisure, and productive life areas, because the other forms of social capital have the potential to contribute broader recovery and wellbeing.However, these fndings may also show the limits of the personal recovery framework that, as discussed earlier, has been criticised when it focuses primarily on intrapsychic states and overlooks interpersonal relationships [30] and structural factors that can limit the agency of mental health consumers [31,34].
In the social and leisure areas of participation, the support the consumers received with ties to family and friends focused primarily on blood relationships and other mental health consumers.Ties with wider groups of people through community activities were mostly temporary and casual.From a social capital perspective, these fndings suggest that support for other relationships with kinship or friends from other groups or socioeconomic backgrounds remained a potential source of bonding and bridging social capital for most consumers.Tese potential wider ties would be particularly important for consumers whose former relationships with family and friends were not constructive for their recovery and wellbeing [12,38].
In the productive area of social participation, the study found that few consumers received support to engage with education, training, and voluntary or paid work.Te literature shows that participation in education and work can help to expand people's social networks and therefore social capital by generating opportunities for and from ties with diferent circles of people [18,37].Te consumers' low engagement with social participation in productive life areas does not preclude the possibility of their long-term goals in these directions.In fact, mental health and social capital literature suggests that encouraging people to envisage these productive goals can be a motivation to engage in social and leisure participation as incremental steps towards education and work [25].
Te fndings about social capital have implications for productive participation.Most of the consumers had weak connections with people from diferent backgrounds (bridging social capital) and with people in diferent social positions (linking social capital).Yet, the literature emphasises that for participation in productive life, people who have wider social contacts (friends of friends) have more job opportunities and a higher chance of success in the job market [18,37,39].From this network perspective of social capital, the fact that only a small number of consumers participated in education, training, and work which implies that support could be extended to both participate in productive life areas and promote activities that encourage bridging and linking social capital and the multidirectional benefts that they can generate [27,37,40].
Tis study has implications for how people are supported in the community to engage in social, leisure, and productive life areas.Giummarra, Randjelovic, and O'Brien [41] found that improved social and community participation requires purposeful strategies that identify meaningful participation preferences.Our research found that while people were being supported to access the community, there were unclear pathways as to how this access translated into meaningful or sustainable connections.Te need to better defne community engagement and to develop strategies to ensure that people are supported to build meaningful and sustainable connections has implications far beyond this study.
In the Australian context, many people receive their psychosocial support through specialist disability support services, the National Disability Insurance Scheme (NDIS).Te NDIS aims to support social participation of people with disability and to sustain participant improvements in employment and social/community activities [42].Te development of ways to identify good practice and clear indicators of when meaningful and sustainable community engagement is achieved has the potential to transform social capital for many people with chronic mental health.
Te fndings also have implications for how people are supported within training and employment settings.Tis study found that while some people believed they did not have the capacity to access employment and training, others were keen to engage or had earmarked it as something to work towards.People who were supported well to engage in paid or unpaid work or training had positive experiences.Limitations to access appeared to be a combination of the person's capacity, experience, or confdence and system blocks in employment and training settings.Projects such as the current National Workplace Initiative, which aims to create a nationally consistent approach to mentally healthy workplaces in Australia [43], are needed at local, state, and industry levels to ensure supports are in place for people to be able to engage in work.Tese kinds of interventions are particularly important to complement the sometimes limited focus of the personal recovery approach on structural factors and its capacity to generate an inclusive environment for the mental health consumers who have a desire to participate in education and/or employment.
Te limitations of the analysis are that it relied on qualitative and quantitative data that were not originally collected to examine the consumers' experiences of social capital but to answer the CLS-HASI's evaluation questions.Consequently, social capital could not be operationalised or measured as part of the original study.Te main limitations of the original evaluation were that it adopted a longitudinal mixed-method, before, and after research design because a control group in the statewide intervention was not possible.Also, the qualitative samples were limited to three sites and statewide stakeholders.Other variations between sites may have been missed.Further research could attempt to broaden the samples, sites, and opportunities for a comparison group.As reiterated throughout the article, the purpose of using a social capital lens was to explore how support changes participation.Te evidence that we use in the article is a good start to show the relevance of our approach and justifes further studies.

Conclusions
Tis research has used key concepts from the network perspective on social capital as a lens to critically assess the activities that consumers engaged with as part of the support they received in the program.Te fndings relied on the qualitative data because specifc measures linking support and activities to outcomes were not available.
Although the study found that, overall, the program met its goals, using a social capital framework helped to identify that service providers need to direct greater attention to facilitating a variety of social participation activities that can extend mental health consumers' horizontal and vertical social capital and so further contribute to their current and future recovery and wellbeing.

Data Availability
Te qualitative and quantitative data used to support the fndings of this study are restricted by all research ethics committees that gave approval to conduct the study.Tey include the following research ethics committees: UNSW Sydney, NSW Aboriginal Health and Medical Research Council (AH&MRC), South Western Sydney LHD, Corrective Services NSW, Justice Health and Forensic Mental Health Network (JH&FMHN), Department of Communities and Justice (DCJ), and the NSW Population and Health Services Research Ethics Committee (P&HSREC).Te restriction is in order to protect patient confdentiality and that of nonpatient interview participants.Data are accessible only to project researchers.Te linked data are available from the Centre for Health Record Linkage (CHeReL) for researchers who meet the criteria for access to confdential data.compared to what is reported in the MDSV1 (38%).Tis diference might be because in the MDSV2, the question on family was made mandatory and reworded to "family and carer involvement during the reporting period" from "family and carer involvement."(iv) Tere were low numbers of scores at program entry.

2. 1 .
Community Mental Health Intervention.CLS-HASI aims to improve health and wellbeing, including social participation.Te integrated intervention is funded by the NSW Government and delivered across the state by NSW Health through Local Health Districts (LHDs) and Community Managed Organisations (CMOs).Te program focuses on personal recovery through psychosocial support, accommodation services, clinical mental health support services, and housing support.CLS-HASI supported 5,533 consumers from 2015 to 2019 and is ongoing.

Figure 1 :
Figure 1: Types of program support for social participation and types of social capital generated.

I
'm a bit nervous of being in public by myself.But usually or normally they ring me . . .and I almost come straight out.But I just need that at the moment because I'll get more confdence, but they take me round to the bank and I've been shopping with them.(Consumer) Meeting people.Yeah, the staf help me . . . the shops are amazing up there [referring to local shopping mall] . . .Tey take me up and drop me on the streets and I talk to people.(Consumer)

(CMO) 3 . 3 . 1 .
Education and Training.Only a few consumers said they were enrolled in training or educational courses (3% in MDSV2).Similarly, few consumers in the feldwork sites participated or were about to start in formal learning activities such as TAFE: I was studying in TAFE College in [place], and now I'm studying in [College name].(Consumer) Actually, I'm starting up a TAFE course this month.(Consumer) Several consumers interrupted their education due to mental health illness.One participant expressed concerns about the possibility of academic stress exacerbating their mental health illness.Some consumers expressed an interest in pursuing formal and vocational education in the future or when they felt sufciently well.I wouldn't mind [doing] a computer course when I can down the track.(Consumer) I've been thinking about going to TAFE, but I don't know what to do.(Consumer) Health & Social Care in the Community I don't think I'll front up to university but maybe TAFE or something, but I just have to be physically functional on deck and feeling much better than what I am at the moment.(Consumer) 3.3.2.Work.Te program supported consumers who were interested to engage with work, either paid or unpaid.For example, consumers received support with paperwork and with referrals for volunteer and paid opportunities.

I
got a job when I was with [previous program provider], and that's a job [in a factory] . . .I've still got the job. . . .How long have I been working there for?About two years . . .Oh, I enjoy it . . .It's always good.(Consumer) (i) Mental health consumer is the current preferred selfadvocacy term in Australia.(ii) Data on group support were collected only after the introduction of the second version of Health & Social Care in the Community the Minimum Data Set.(iii) More consumers had family and carer involvement in the most recent MDSV2 (52%),
Capital from Support for Social and Leisure Participation.Te program promoted three forms of support to encourage consumers' social and leisure participation, which are analysed below: group activities, community engagement, and maintaining or restoring family relationships.
Group activities were organised around the consumers' social and recreational interests.Examples included cooking classes, craft groups, and cinema visits.Te aim of the group activities was to provide an opportunity to consumers to build social relationships because most of them did not have friends or relationships with their families:We have groups here that help [consumers] build relationships or friendships with other participants because most of our participants that we do have, they've got no family support or any friends or anything.Basically, some of them have only got us as support.(CMOworker) I run into [other people doing same group activity] but we've all got [poor] social skills-we don't say hello unless we're in the group.(Consumer)

Table 2 :
Social participation average support hours per consumer per month.
[13]ce: MDSV1 (n � 4,619) and MDSV2 (n � 2,880) monthly averages to January 2020 (total n � 5,533 including double counting of consumers who were in both MDS versions).Support-level group as defned by program managers: low � under 5 hours per week, medium � 5 hours per day to 5 hours per week, and high � more than 5 hours per day.Adapted from Purcal et al.[13].Note.Support level based on average support per consumer over the support period.