'What works here doesn't work there': The significance of local context for a sustainable and replicable asset-based community intervention aimed at promoting social interaction in later life.

Interventions that harness local assets to benefit a community are increasingly being promoted to improve health and well-being. In practice, we know little about how local contexts or reliance on local resources affect the sustainability and scalability of asset-based community developments. This qualitative case study documents the development and implementation of a novel asset-based community development project. Based in a large mainly rural county in North East England with relatively high levels of socioeconomic deprivation, the project aimed to prevent social isolation among older people, using a range of food-related activities. Twenty-one semi-structured interviews were conducted with service users, volunteers, project partners, project development workers and senior staff. Interviews explored the project's design and implementation process, outcomes for participants and the wider community, and project sustainability and scalability. Thematic analysis of the data identified four factors likely to be important for creating sustainable and replicable asset-based community projects. These factors are (a) recognising and harnessing assets among local people who may be otherwise marginalised due to age, geographical isolation and/or socioeconomic deprivation; (b) identifying assets that can be provided by local businesses; (c) genuine project co-production to develop activities that meet local needs and inspire enthusiasm among all stakeholders; and (d) ongoing organisational support to meet the challenges to sustainability that exist in socioeconomically deprived areas. We conclude that successful asset-based community projects require extensive community input and learning captured from existing programmes can facilitate the replicability of programmes in other community contexts.

What is known about this topic:

Asset‐based community approaches are increasingly being used to improve health and well‐being.

Asset‐based programmes harness resources within communities with the aim of tackling locally rooted problems.

Mapping local resources is the first stage in the development of an asset‐based programme.

What this paper adds:

Identifying and harnessing the unrealised resources of marginalised groups and local businesses builds community resilience and improves programme sustainability.

Genuine programme co‐creation, creating activities that meet actual community needs and engender enthusiasm, is necessary for creating sustainable programmes.

Co‐creation also requires organisational responsiveness and willingness to adapt to what works best ‘on the ground’.

INTRODUCTION

Asset‐based community development is gaining traction as a means of improving population health and well‐being (Foot, 2012). In contrast to traditional ‘deficit models’ where external agencies seek to ‘fix’ communities’ problems and deficiencies (Lam, Zamenopoulos, Kelemen, & Hoo Na, 2017), asset‐based interventions build on communities’ resources (Jakes, Hardison‐Moody, Bowen, & Blevins, 2015). Localism is at the heart of this approach (Bunt & Harris, 2010), the rationale being that by building on local knowledge one maximises the chances of addressing issues whose determinants are culturally, socially and economically rooted within a particular geographical area or community (e.g., neighbourhood levels of deprivation or variation in service provision) (Jakes et al., 2015). In practice, however, we know little about whether asset‐based community interventions deliver their theoretical benefits. It is also unclear how local context shapes these initiatives, and how reliance on local resources affects an intervention's viability and sustainability, particularly in socioeconomically deprived areas. In addition, because asset‐based initiatives are carefully tailored to the communities in which they are developed, the potential mechanisms for ‘rolling‐out’ successful initiatives are unknown (Bunt & Harris, 2010). This article documents the implementation of an asset‐based community project, which used food‐related activities to prevent social isolation among older adults. This work aims to strengthen the evidence base on the implementation of asset‐based community approaches, using qualitative methods to address the research question: what factors affect a project's sustainability and scalability?

Background

Age‐related reductions in social networks and valued relationships mean that social isolation and loneliness can be a problem for older adults (Courtin & Knapp, 2017). In the UK, 3.5 million people aged over 65 live alone and nearly a quarter of adults over state‐pension age go out socially less than once a month (Carr, Councell, Higgs, & Singh, 2014). A wide range of interventions have been developed in response to the problems of social isolation and loneliness among older people, the most common of which appear to be ‘social interaction interventions’ that seek to facilitate reciprocal interaction with peers to provide mutual benefits (Gardiner, Geldenhuys, & Gott, 2018).

Attempts have been made to review the effectiveness of interventions to tackle social isolation and/or loneliness (Cattan, White, Bond, & Learmouth, 2005; Cohen‐Mansfield & Perach, 2015; Dickens, Richards, Greaves, & Campbell, 2011; Findlay, 2003; Gardiner et al., 2018). These reviews have, tentatively, identified characteristics potentially shared by effective interventions. For example, most recently Gardiner et al. (2018) have suggested three common characteristics of interventions that demonstrated effectiveness: first, adaptation to local context; second, the adoption of a community development approach; and third, productive engagement. However, aspects of effectiveness identified to‐date tend to be general rather than specific and based on limited evidence (Bartlett, Warburton, Lui, Peach, & Carroll, 2012). Indeed, due to the difficulties in evaluating highly heterogeneous and complex interventions with diffuse outcomes, we know little about the effectiveness of interventions to tackle social isolation and/or loneliness and the factors that contribute to their success (Bartlett et al., 2012; Dickens et al., 2011; Gardiner et al., 2018). According to Cohen‐Mansfield and Perach (2015), one explanation for the limited evidence of effectiveness may arise from the methodological limitations inherent in attempts to quantitatively evaluate complex interventions targeting potentially vulnerable participants (e.g., randomisation presents both practical and ethical difficulties). The complexity of community‐based interventions makes them particularly challenging to evaluate, be it in relation to establishing outcomes or in matching outcomes to specific aspects of an intervention (Bartlett et al., 2012). Further, changes in quantitative outcomes cannot necessarily differentiate between aspects of an intervention or identify aspects that are highly valued by stakeholders. Cohen‐Mansfield and Perach (2015) suggest that ‘untangling’ the complexity of interventions will require both more research and alternative approaches. Qualitative approaches, such as that adopted in this study, have much to offer: qualitative research can explore interacting elements and the context in which interventions are conceived, developed and delivered (Gardiner et al., 2018). Moreover, qualitative research can describe the experiences of intervention stakeholders, thereby identifying a range of impacts (Moffatt, Steer, Penn, & Lawson, 2017).

The ‘Come Eat Together’ project

The ‘Come Eat Together’ (CET) project was designed by Age UK County Durham, a voluntary organisation working for older people. County Durham is a geographically large (2,230 km2) county in North East England with a population of just over 500,000 (Durham County Council, 2017). The county is predominantly rural, with remote, sparsely populated areas in the west and former coal‐mining communities in the centre, north and east. There are 12 major urban centres, the largest of which is Durham city (population nearly 48,000). The county is the 75th (out of 326, with 1 being most deprived) most deprived Local Authority in England (County Durham Council, 2015). County Durham also has an ageing population: the number of residents aged 65 and over increased by 13.5% between 2010 and 2015, while the number of working‐age residents fell by 1.2% (Durham County Council, 2017).

The CET project was developed in response to a consultation commissioned by NHS County Durham and Durham County Council in 2011 to find out how best to improve nutrition in later life. Older people identified barriers to accessing healthy food, including a lack of motivation to cook for oneself, difficulty carrying shopping and limited affordable or accessible transport. Many older men reported limited cooking skills, while financial hardship and driving cessation were also a concern, particularly in rural areas (Age UK County Durham, 2012). Social isolation and loneliness were also identified as problems by older adults responding to the consultation (Age UK County Durham, 2012). Loneliness among older people has been highlighted as a concern in County Durham, with 20% of adults over 65 reporting feeling lonely and 11% feeling intensely lonely (County Durham Council, 2014). Building on this consultation, Age UK County Durham designed and piloted the CET project in 2012, subsequently rolling it out in 2014 across most areas of the county.

The CET project's overarching aim is to promote social interaction by bringing older people together around food‐related activities (see Table 1). The implementation and monitoring of CET is overseen by a team of eight Age UK County Durham employees. Activities are targeted at people aged 55 and over who are at risk of being socially isolated (i.e., of having limited social contacts (Valtorta & Hanratty, 2013)), who might feel lonely (i.e., would like to have more people with whom they can socialise and/or develop relationships (Valtorta & Hanratty, 2013)) and/or who have difficulty accessing healthy food, but whose health‐related needs – be it mobility or cognitive functioning – do not preclude them from taking part in community activities. Participants pay between £1 and £10 per session depending on the activity and venue. Between 1st April 2014 and 31st July 2017, CET activities were accessed by 3,485 people and involved 367 volunteers and 67 ‘project partners’ (used here to designate lunch club venues, caterers and providers of community transport). Demographic data collected by Age UK County Durham indicate that the majority of service users are aged between 70 and 85 and two‐thirds are female.

Table 1

Come Eat Together project activities

Activity name	Description of activity
Lunch Clubs	Monthly lunch clubs take place in 'non‐traditional' settings, including community colleges, local government premises, a leisure centre, a sports ground, cafes, pubs, restaurants and hotels and local business canteens. Lunches are generally followed by an activity (e.g., a quiz or a talk).
Men's Breakfast Club	Monthly men‐only breakfast clubs, including a cooked breakfast followed by an activity.
Dining Circles	Involve older people going out together for a meal, cooking one together for themselves or having one delivered to a group.
Healthy Eating for Older People course	An accredited course that helps to raise awareness of how nutritional needs change with ageing. The courses are delivered in community venues and include cookery demonstrations and food tastings to introduce ideas for simple, nutritious meals.
Help with Shopping	A door‐to‐door accessible minibus service provides fortnightly transportation for shopping club trips to local supermarkets. Volunteers accompany services users on the bus and help with shopping if required.
Learn to Shop Online	Trained volunteer IT tutors deliver training to help service users gain the confidence and skills needed to shop online. Training is delivered in community venues and in individuals' homes.
Grow to Eat	Sessions delivered in the community providing practical advice on growing fruit and vegetables at home (even without a garden). Each session includes seeds and plants to take home. Delivered by volunteers using the ‘Grow to Eat’ toolkit.

DATA AND METHODS

We adopt a case‐study approach to conduct a detailed analysis of the specific context in which the CET programme took place (Yin, 2013). The research method took the form of individual semi‐structured interviews. Individual interviews, rather than focus groups, were conducted to allow participants to share their views privately, in their own words and free from group dynamics. Further, from a practical perspective, most participants preferred to be interviewed in their homes or, in the case of project partners, at their place of work. Ethical approval for the conduct of this study was secured from the University ethics committee (1,402/15,544/2017).

Participant sample and recruitment

We aimed to obtain a sample that covered four stakeholder groups central to the delivery of the CET project: service users, volunteers, Age UK County Durham staff and project partners. Acting as gatekeepers, Age UK County Durham staff identified potential participants from their database of contacts. Participants were purposively selected to reflect the diversity of service users, volunteers and project partners (see Table 2). Age UK County Durham provided participants (by post or in‐person) with an information pack containing an information sheet outlining the project and explaining the interview process, together with a form consenting to be contacted by a member of the project team to arrange an interview. Completed forms were returned to Age UK County Durham who passed the contact details of those who agreed to take part (22 of the 23 contacted) to the research team. The rationale for participants returning their consent forms to Age UK County Durham was that this would minimise uncertainty and inconvenience, particularly for older service users who have developed trusting relationships with the organisation. Participants could indicate their consent via a range of options: by phone, by posting their consent form to Age UK County Durham or by handing it to a project worker. The study information made clear that choosing not to participate would in no way affect participants’ relationship with Age UK County Durham. Twenty‐one semi‐structured interviews were conducted between August 2017 and January 2018 (Table 2). Only one person, a project partner, was unable to commit the time for an interview.

Table 2

Participant information

Participant code	Description	n
SU	Service User	8	Female n = 5, Male n = 3, Age range 70–86
V	Volunteer	5	Female n = 5
S	Staff	4	Senior management n = 2
			Project development worker n = 2
PP	Project Partner	4	Pub/restaurant n = 1
			Sport/leisure business n = 1
			Community leisure venue n = 1
			Education provider n = 1

Data collection and analysis

Two project partners were interviewed by telephone; all other interviews were conducted in person at a time and place convenient for participants. Interviews followed a topic guide, developed through discussion with Age UK County Durham, which covered participants’ involvement in CET; perceived strengths and weaknesses of the project and suggestions for improvement. Interviews lasted between 15 min and 1 hr 16 min (average: 34 min), were digitally recorded, transcribed verbatim, fully anonymised and managed using NVivo software (QSR International Pty Ltd, 2017). Data analysis was thematic (Green & Thorogood, 2014): the two interviewers initially coded the transcripts according to a scheme which followed the topic guide: information on how the project is run; positive aspects of the initiative; obstacles and potential barriers; and suggestions for the future. Findings from this second analytical phase were shared and discussed with the full research team. During this discussion, the dimension of local context – understood as the ‘circumstances or events that form the environment within which something exists or takes place’ (Poland et al., 2006: p56) – was identified as a thread throughout the three domains of implementation, sustainability and replicability, and across stakeholders’ accounts. Following this discussion, the two interviewers returned to the coded dataset and conducted a line‐by‐line analysis, with a focus on eliciting participants’ views on how local context had shaped CET and how it might shape its future. Table 2 lists participant codes.

FINDINGS

Table 2 reports characteristics of the final sample. Reflecting the population accessing CET, service‐user participants were aged between 70 and 85 years and the majority were female. The service‐user sample reflected the project's intended recipients, with six participants living in areas classed as among the 50 percent most deprived in County Durham (three lived in areas in the most deprived 20 percent). Interviewees’ account of the implementation and development of the project revealed two main themes pertaining to the project's sustainability and scalability: (a) the benefits of harnessing the often overlooked resources of local business and; (b) the importance of genuine programme co‐creation.

The overlooked potential of local businesses

Identifying and harnessing community assets is the first stage in developing an asset‐based intervention. Local businesses are largely unrealised local assets, providing both physical and social capital (Fisher, Geenen, Jurcevic, McClintock, & Davis, 2009). An important innovation in the CET project is the development of ‘non‐traditional’ lunch clubs located in existing businesses as an alternative to the ‘traditional’ model of community hall‐based lunch clubs. A key project aim is to promote older people's inclusion in their communities. The fact that project partners served the wider community, and not exclusively older people, was a priori identified as a vector of social inclusion.

The decision to organise non‐traditional lunch clubs was also motivated by the aims of sustainability and scalability, involving the pragmatic search for a service‐delivery model that would enable the relatively small voluntary‐sector team to run many different activities across a large county. Unlike traditional lunch clubs that rely on a fluctuating supply of volunteer cooks, CET project partners operate as food providers on a day‐to‐day basis. Involving local businesses also addressed the paucity of community hubs (e.g., community centres or church halls) in small, isolated communities. However, achieving sustainability meant that project partners needed to benefit from their involvement with CET. For this reason, identifying the ‘business case’ was the first stage in involving project partners:

‘It's a case of identifying the venue, working out with them what it is that they want to get from it … It's got to be a win‐win, so we work out what's going to be the win for them’. (S1)

Some project partners were participating in CET as part of a Corporate Social Responsibility (CSR) agenda (either formally or, in the case of smaller businesses, informally as part of a desire to help their community). It has been argued that asset‐based approaches to CSR, especially for small and medium‐sized businesses who rely heavily on their local environment, can result in mutual gains for both business and society (Fisher et al., 2009). Project partners reported a range of benefits from involvement, including positive publicity and increased trade from return visits by service users and their families. Mutual benefits from collaborative working were perceived to be a promising indicator of future sustainability. Should funding for the programme reduce or cease, the initiative could be continued with minimal organisational input, relying on committed project partners who had a vested interest in being involved. This view was shared by project partners, who were positive about remaining involved in the project long term.

However, relying on partnerships with local project partners could be challenging – particularly in the most socioeconomically deprived areas. Staff confidence in the sustainability of CET was strongest for activities underpinned by stable resources, such as a regular pool of attendees and engaged project partners. Changes in project partners’ circumstances, commonly in the form of a change of venue manager or owner, or where partners ceased trading, were identified as threats to sustainability. In areas with limited infrastructure (e.g., no local newspapers or community centres, or no suitable partners available as a local back‐up if a partner left the project), the continued running of activities was identified as being particularly challenging. It is acknowledged that community initiatives will need at least some level of ongoing support to flourish and develop (Puttick & Ludlow, 2013) and staff felt that some activities would always require ongoing organisational support.

Co‐creation of the CET programme

While much has been written on the process of mapping communities’ assets (Jakes et al., 2015), less attention is given to the processes involved in harnessing assets to actually ‘do something’ (O'Leary, Burkett, & Braithwaite, 2011). Once assets have been identified, co‐creation involves local people in the design and implementation of programmes. In the first stage of co‐creating the CET project, Age UK County Durham conduct extensive consultations with residents across the county. These identified variations in demands, resources and attitudes, leading to the decision to approach areas with distinct needs and expectations as separate entities. Adapting activities to fit the needs of potential participants was identified by staff as integral to the project's sustainability, with staff stressing that simply reproducing activity formats across different settings was unviable:

You’ve got to be adaptable and flexible to your way of working because what works here doesn’t work there. So you can’t go in with this prescriptive sheet where you think, ‘Well, I’ll follow this pattern’. It doesn’t work like that. (S4)

Staff credited extensive community consultations with allowing them to pre‐empt potential barriers to implementation and uptake. Because County Durham is a large, mainly rural county, access to activities via public transport was identified by all stakeholders as a key determinant of success. However, equally important were the sociocultural dimensions of community trust and openness. Perhaps due to previously negative or disappointing experiences with official agencies or ‘authority figures’, some communities, especially in areas of high socioeconomic deprivation, were challenging to engage, with a staff‐member noting that:

‘Some communities are easier to break into than others. There's a lot of mistrust, I think, still … So you've got to break down the barriers and build that rapport’ (S4)

The repair and strengthening of relationships between communities and organisations tasked with improving outcomes has been identified as central to building the community resilience needed to develop sustainable initiatives (Friedli & Carlin, 2009). Age UK County Durham is, itself, a community asset. The organisation is a well‐embedded trusted local face in the community with a wide network of community connections (Bunt & Harris, 2010). Staff identified in‐depth community knowledge as central to the project. Project development workers were drawn from County Durham and acted as ‘gappers’ (individuals employed by an organisation, but whose ‘practice and understanding and spirit is in the community’ (O'Leary et al., 2011: p23)), bridging the gap between Age UK County Durham as an organisation and the wider community.

Co‐production can harnesses people's natural enthusiasm, resulting in outcomes unlikely to have been achieved through central development (O'Leary et al., 2011). Similarly, community‐located initiatives can capitalise on a community's capacity to innovate and create imaginative solutions to local challenges (Bunt & Harris, 2010). The range of venues putting themselves forward as CET project partners demonstrates imaginative community engagement: museums, work‐place canteens, further education colleges, car showroom cafes, a sports stadium, a leisure centre and a ‘flight simulator’ attraction, in addition to the more obvious pubs, cafes and restaurants, were all acting as project partners. Engaging socially isolated older men has been identified as a particular challenge (Beach & Bamford, 2014) and, in an example of imaginative innovation, the (subsequently hugely popular) men‐only Breakfast Clubs were instigated by a project partner in response to a need he recognised through his employment.

The centrality of people: involving a community's less tangible assets

In the development of asset‐based interventions, the assets of potentially marginalised groups such as older people (particularly in areas of deprivation) can be overlooked (Benenson & Stagg, 2016; Ennis & West, 2010). A strength of the CET programme is in highlighting the visibility of older people – both as volunteers and service users – as a community asset and central to the programme's delivery and sustainability (e.g., as customers with spending power or as a source of human and social capital) rather than as passive recipients of help (or worse, as a burden). Staff identified the project's central aim as creating ‘opportunities [for older people] to share skills…feel valued [and] socially included’ (S3). The CET project was providing a ‘means to action’ for participants (Bunt & Harris, 2010), with a project partner describing their lunch club as ‘creating a lot of active citizenship’ (PP4), both for service users and volunteers.

Tangible community assets, such as built and financial capitals, are easiest to quantify and therefore most often recognised or valued (O'Leary et al., 2011). However, asset‐based approaches also build on less tangible, harder to quantify assets such as local networks and existing community relationships (Bunt & Harris, 2010; Lam et al., 2017) that are crucial to individual and community well‐being (O'Leary et al., 2011). The CET strategy for achieving sustainable reductions in social isolation centred on building social networks to create resilience in later life, especially among older people in isolated and/or socioeconomically deprived communities. Service users described shrinking social networks accompanying ageing, as one man described it: ‘you run out of friends when you get old, they either move away or they don't get out anymore’ (SU4). A number of service users described feeling very socially isolated, with one describing having ‘no friends whatsoever’ (SU5). Wider social networks were identified as a particular benefit of the CET programme:

I’ve met people that I wouldn’t otherwise have done. I don’t go out very much, because I don’t know anybody really … It certainly is getting people together. I’ve made the start of two friendships. These people I would never have met, because there’s nowhere to go. (SU2)

County Durham has a relatively settled community and activities located within communities were helping to rekindle old acquaintanceships: ‘you come across people you haven't seen for years … the local people you lose touch with, they crop‐up again.’ (SU4).

Volunteering, although not currently framed as such, is an important asset‐building strategy, including potentially marginalised members of a community both as recipients and participants (Benenson & Stagg, 2016). An important success factor identified in a review of learnings from asset‐based initiatives is the provision of a range of options for involvement that enable people to join in at their comfort level (Bunt & Harris, 2010). Participants identified a blurring between the roles of volunteer and service user, with older people at risk of social isolation yet unready to participate in activities as service users participating instead as volunteers:

I've got quite a few volunteers that are above 50, some are 60, 70, and they're eligible for the project but they don't feel as if they need the project so they help with the project instead. (S2)

I love [volunteering]…I would be lost if it wasn’t for going out to do the volunteer work. I would be stagnant at home…it gives me something to go out for. I get a lot out of it and as much as the [service users] do (V5)

Project volunteers were most often themselves older people living locally and acting as ‘peer volunteers’ whose community embeddedness and social networks had been harnessed as an asset.

Enabling service users to create activities to meet their particular requirements is likely to be an important aspect of sustainability. A staff member described the process by which activities were developed by men attending a breakfast club. By enabling real ownership of the activity, the men's ingenuity and unique skill sets could be harnessed for the benefit of the group:

Very quickly, the men themselves actually, developed the group and the men said, ‘Right, okay, let’s have a talk, but this is the sort of talk we want.’ And then they said, individual members of the group said, ‘Well, actually, I can give a really interesting talk about the Class 5 trains.’ Or, ‘I’ve worked in a power station, I can talk about working in a power station.’ So … it very quickly developed quite an interesting programme of talks that the men thought were interesting. (S3)

It has been suggested that social isolation interventions that promote productive, rather than passive, engagement are more effective (Gardiner et al., 2018). Both staff and service users identified the Healthy Eating course as particularly successful. Service users valued the interactive small‐group format that encouraged them to get involved and share their opinions and knowledge:

It wasn’t a great big class … everyone could chip in and give a little talk, it was really good … [the staff] treated you like a normal person … they engaged you in conversation. They asked your opinion. (SU1)

Co‐production requires genuine organisational willingness to learn and, often, a shift in organisational attitudes (O'Leary et al., 2011). A strength of Age UK County Durham staff was their ability to respond to what actually worked (the outcome) rather than imposing their vision (the process) upon programme activities. Staff described an initial contradiction between some of the outcomes from the community consultation to develop activities and what older people ‘on the ground’ actually wanted. An example was an online shopping activity, which had been identified as popular in the consultations. In practice, service users were unenthusiastic; it was the social aspects of food shopping that were highly valued by older people:

Internet shopping is no good for me. For me on my own, I want to go out to meet people. Everybody else said the same, who was [at the Healthy Eating Course]…the majority of them said they would never do Internet shopping. For us, going to the shops is a trip out. You might see somebody you know, which for me is marvellous. With living on your own, it’s dreadful (SU1)

Co‐production had been a learning curve for staff. Creating self‐sustaining activities required a new way of working, with staff developing the confidence to hand activities over to the groups. Co‐production of activities meant service users felt supported rather than directed (identified as an important aspect of encouraging community involvement (Bunt & Harris, 2010)). Staff and volunteers both reported that in many venues volunteers had taken ownership of the activity:

We practically took it over. I am now competent to organise things like quizzes. [Project Development Worker] is quite confident that we can run it ourselves now. (V5)

Staff emphasised how crucial the process of learning from experience had been for CET, and how anyone envisaging the delivery of a similar project would benefit from going through the same ‘learning curve’ experience to properly establish community needs.

DISCUSSION

Asset‐based projects are rooted in unique local contexts. Communities face different challenges stemming from geography and socioeconomic deprivation. The CET project appears to be successful in reaching potentially marginalised groups, including older people (especially older men) and communities in remote rural and/or areas of socioeconomic deprivation. The service user sample comprised people from areas of high socioeconomic deprivation and all reported feelings of loneliness and/or social isolation. This study suggests that the success of a project depends on how well it is tailored to the communities within which it is implemented. Factors highlighted as important for creating sustainable and replicable asset‐based community projects were (a) harnessing assets that can be provided by local businesses to create new social networks and build community resilience; (b) recognising and harnessing the less tangible assets of local people, many of whom were marginalised; and (c) implementing genuine co‐creation to develop activities that local people want and feel enthusiastic about. Feedback data collected by Age UK County Durham indicated that 97% of service users who completed a baseline and follow‐up survey agreed that they had met new people and friends through the CET project. This is supported by service users (and volunteers) participating in this study who all reported having wider social networks.

In times of constrained public spending, asset‐based approaches are attractive as potentially low‐resource interventions (MacLeod & Emejulu, 2014). Our study highlights that to adopt an asset‐based approach in areas where assets are limited, significant investment is needed from the organisation delivering the intervention. County Durham has areas of high socioeconomic deprivation and the challenges facing project partners in deprived areas posed a potential threat to the CET project's sustainability and required continued organisational involvement. Ongoing practical support, an in‐depth understanding of the different facets of the local context, and having time to develop trusting and reciprocal relationships with project partners were all important.

For organisations planning a sustainable asset‐based community project, perhaps the most important lesson arising from this study is the need for genuine co‐design and co‐production. Local solutions rely on specificity, local ownership and the ability of groups to tailor solutions to the particular local context (Bunt & Harris, 2010). Co‐produced interventions help build communities, with active participation creating new social networks (Boyle & Harris, 2009). It has been observed that ‘the contribution of citizens is rarely made visible and rarely included in the planning’ of interventions (Foot and Hopkins, 2010: p.15). Yet, ‘top‐down’ centrally led initiatives often fail when problems are closely linked to how people live their lives (Bunt & Harris, 2010). Participants in this study were very clear: what works in one area will not necessarily work in another. Development of programmes targeted at specific populations and communities needs to build‐in extensive community consultations to create activities that people want and feel enthusiastic about. This required in‐depth local knowledge acquired through consultations with individuals embedded within their communities. It is oft‐repeated that involving older people in the development of interventions is likely to be central to their effectiveness (Findlay, 2003). However, planning typically takes the form of surveys, focus groups and consultations with community representatives (Joseph Rowntree Foundation, 1999; NHS Confederation, Local Government Association, NHS Clinical Commissioners, & NHS Providers, 2018). Although helpful, these strategies are less intensive and suffer from the risk of reaching only usual responders rather than people who will necessarily use the intervention. As we find in this study, what is articulated in consultations is not always what is wanted on the ground. Services that genuinely meet the needs of users may challenge the assumptions of what professionals believe that people want and organisations must be willing to learn and adapt (NHS Confederation, Local Government Association, NHS Clinical Commissioners, & NHS Providers, 2018). The NHS community care ‘vanguards’ project identified the need to empower staff and encourage innovation, along with the ability to fail and learn lessons, as central to a project's success (NHS Confederation, LOCAL Government Association, NHS Clinical Commissioners, & NHS Providers, 2018). This study supports the importance of constant learning and genuine organisational responsiveness to what worked best ‘on the ground’. However, although extensive efforts were made to involve service users in the co‐creation of the CET project, nevertheless, it remains the case that the extent to which CET is co‐created will always be limited by the fact that those who actively participate will be a self‐selecting group and, not necessarily, the most marginalised.

Strengths and limitations

The strength of this work lies in the depth of interviewees’ accounts and in the diversity of the sample. Sampling allowed us to reach service users and volunteers living in some of the county's most socioeconomically deprived areas. While we accessed the views of the four stakeholder groups involved in CET, we acknowledge that our study design did not allow us to capture the views of all the people involved in CET. Participants were asked to provide their views on a service with which they were closely involved. This allowed us to gain insights, but may have resulted in under‐reporting of negative aspects.

CONCLUSION

In response to the problems of social isolation and loneliness, the UK government has recently published its first ‘loneliness strategy’ (Prime Minister's Office, 2018). This strategy aims to bring together a variety of stakeholders, including from business and the voluntary and community sectors, to tackle social isolation and loneliness by building more resilient communities and providing funding to increase the number of community spaces. A key part of the strategy is to strengthen the evidence base on what works. This current study suggests that an asset‐based community approach to tackling social isolation among older people is feasible, provided that the approach harnesses a communities’ unrealised resources and is genuinely co‐created. This approach is appropriate for areas where older people experience high levels of socioeconomic deprivation. We recommend that evaluators consider both the extent to which an intervention's effectiveness is dependent upon the local context in which it is implemented and its potential to be adapted to work in other settings.

ETHICAL APPROVAL

Ethical approval for the conduct of this study was secured from Newcastle University Faculty of Medical Science Ethics Committee (1,402/15,544/2017).