Introduction
After a preliminary description of the retirement housing sector, this paper addresses the findings of previous literature on the nature and causes of loneliness amongst older people, followed by a review of studies on measures to address older people's needs for social interaction and emotional support. We then present the fieldwork which examines what organised social activities take place in retirement housing, and how these activities measure up to the desiderata suggested by the literature. The fieldwork also considers the relative roles played by staff and by residents' own initiatives in organising social activities in retirement housing schemes.
Retirement housing (RH), sometimes known as sheltered housing, is the term used in the United Kingdom (UK) to describe specialised housing developments restricted to those over 55 (or sometimes 50 or 60, depending on landlord and area). The focus of interest in this study is how organised social activities in this type of housing can help to improve informal social support and avert loneliness, to which its residents may be particularly vulnerable. According to an analysis of 2008 data from the English Longitudinal Study of Ageing (ELSA), people in RH are especially likely to suffer from poor health, widowhood, low access to cars or report a shortage of money, all these being factors associated with loneliness amongst retired over 60s as a whole in the ELSA sample (Gray Reference Gray, Forsman and Nyqvist2015). Amongst 383 retired respondents over 60 living in RH (defined as housing with a ‘warden’) more than a quarter reported feeling lonely ‘much of the past week’, compared to only 12 per cent of 5,699 over 60s in ordinary housing.
Around 7 per cent of the English population live in housing where occupancy is restricted to older people (Pannell, Aldridge and Kenway Reference Pannell, Aldridge and Kenway2012). About 85 per cent of residents are over 65, with a quarter over 85 (Blood and Pannell Reference Blood and Pannell2012a ). Some schemes, offering on-site care services, are described as ‘extra care’ or ‘very sheltered’, but the term excludes residential care homes. About one-fifth of the RH sector is for leasehold purchase, the rest is social housing for rent, with entry criteria on the basis of housing need, including needs based on disability (Lock and Whittington Reference Lock and Whittington2006). These rental schemes are mainly provided by ‘registered social landlords' who own and manage social housing. Although there is no income condition for eligibility, home-owners would generally be directed to leasehold schemes, which are mainly offered by private-sector property developers but also by some registered social landlords, including the one surveyed. The RH sector also includes large retirement villages with 100–250 dwellings, often mixing tenures and support levels. Thus, estates can be characterised in two dimensions as ‘extra-care’ or ‘ordinary’ RH, and as rented, leasehold or mixed.
Retirement estates are intended by their providers to be supportive of neighbour interactions and collective social activities. Many have common rooms, and the estate community provides potential opportunities for making friends and for generating support between neighbours. Social rented estates specifically for older people historically had resident staff providing social and practical support to residents (Blood and Pannell Reference Blood and Pannell2012a ; King, Pannell and Copeman Reference King, Pannell and Copeman2009). The ELSA analysis by Gray (Reference Gray, Forsman and Nyqvist2015) defined RH as housing with a ‘warden’, the word used in the ELSA questionnaire, although they are now usually called ‘estate manager’. In recent years, estate managers have moved off-site and are now frequently part-time, sometimes replaced altogether by peripatetic support workers. The cuts in staff hours are said to have reduced their role in supporting social activities and generating community spirit (Gray Reference Gray2014; King, Pannell and Copeman Reference King, Pannell and Copeman2009).
Staff in RH often facilitate or organise gatherings, classes or outings to enhance residents' social lives, sometimes developing local partnerships to secure resources or volunteers (Blood and Pannell Reference Blood and Pannell2012a ). Residents may also organise activities for themselves with or without support from staff or outside volunteers. Staff often publicise activities in community centres, education centres, social clubs, etc., outside the housing scheme, to encourage more mobile residents to engage with the wider community. However, when advanced age reduces residents' mobility and energy, opportunities very close to home become increasingly important (Callaghan Reference Callaghan2008; Croucher, Hicks and Jackson Reference Croucher, Hicks and Jackson2006).
Extra-care schemes devote more resources to organising social activities than in ‘ordinary’ RH, with a much higher complement of staff hours per resident. Several evaluations of extra-care housing (Callaghan Reference Callaghan2008; Callaghan et al. Reference Callaghan2008; Croucher, Hicks and Jackson Reference Croucher, Hicks and Jackson2006) analyse the role of organised group activities in developing a sense of community. Kneale (Reference Kneale2013) identifies several potential mechanisms through which extra-care housing can lower social isolation; by its ethos, design, organised activities, sense of community, and through all of these by achieving improved health and mobility.
Several case studies of retirement villages, involving interviews and discussions with residents, show how they generate friendships and mutual aid between residents (Bernard et al. Reference Bernard, Liddle, Bartlam, Scharf and Sim2012; Croucher, Pleace and Bevan Reference Croucher, Pleace and Bevan2003). Retirement villages are typically over 100 dwellings, some rented and some leasehold, with facilities such as a café or restaurant, a gym or exercise room, craft workshops, a computer room and a library (Bernard et al. Reference Bernard, Bartlam, Sim and Biggs2007). Their size and capital receipts from leasehold sales permit greater investment in communal facilities than is usually affordable in social rented housing. For historical reasons, some have also benefited from charitable funding, and/or more public funding than most RH schemes. Retirement villages, unlike most smaller RH schemes, generally aim to mix residents who need ‘extra care’ with people who are still fit and active. Examples include two schemes managed by the Joseph Rowntree Housing Trust, where Croucher and Bevan (Reference Croucher and Bevan2010) and Croucher, Pleace and Bevan (Reference Croucher, Pleace and Bevan2003) surveyed residents; and Denham Village, established in the 1950s by the Licensed Victuallers Trust, where Bernard et al. (Reference Bernard, Liddle, Bartlam, Scharf and Sim2012) conducted a longitudinal study of residents using interviews and diaries. All three studies suggest that many of those who enter retirement villages are looking for neighbourly conviviality, and that this expectation is frequently realised. However, Croucher and Bevan (Reference Croucher and Bevan2010), in their study of a retirement village in Hartlepool, found some tension in the community between active, relatively healthy owner-occupiers and ‘extra-care’ residents who were mainly social tenants referred by the local authority. In a case study of 90 residents in a fourth village, Evans (Reference Evans2009) also reported little social interaction between the ‘fit’ and the ‘frail’, due mainly to the distance between the extra-care accommodation block and the leasehold dwellings bought by the ‘fit’.
Although most people who enter social rented RH do so simply because their previous home has become unsuitable or unavailable (Blood and Pannell Reference Blood and Pannell2012a ), several surveys report that residents also value its sociability; the opportunities for meeting neighbours through coffee mornings, lunches or other social events (Callaghan Reference Callaghan2008; Percival Reference Percival2001; Taylor and Neill Reference Taylor and Neill2009) and the help given by staff to organise such events and support tenants' social committees (King, Pannell and Copeman Reference King, Pannell and Copeman2009). However, living in age-segregated housing cuts residents off from neighbours of different ages, making it important to maintain links to the wider community and to other generations (Blood and Pannell Reference Blood and Pannell2012a ). Older people may often benefit from making and retaining younger friends, avoiding the loneliness that comes from being very old and having outlived several of one's peer group (Gray Reference Gray2006, Reference Gray2009; Luanaigh and Lawlor Reference Luanaigh and Lawlor2008). Younger friends may remain able to give practical support for longer than those neighbours in RH who are of a similar age.
Organisation of social activities in RH of all kinds tends to be based on custom and practice, residents' and staff initiatives, and on the internal policies of individual housing providers. Some providers offer good practice guides (e.g. Hanover 2010, 2011, 2012, 2013), as do support organisations for housing professionals such as Emerging Role of Sheltered Housing (EROSH; with good practice materials and case studies on a subscription-based website at http://www.erosh.co.uk/) and the Housing Learning and Improvement Network (an online resource at http://www.housinglin.org.uk/). These guides are generally based on small-scale case studies, often without systematic evaluation. This suggests that the choice of activities could benefit from a stronger evidence base about what works best to help residents make friends or overcome feelings of isolation or loneliness. However, little research on this issue has been done on those forms of RH which are social rented but not within the ‘extra-care’ category.
The forms and causes of loneliness amongst older people and their relationship to the retirement housing environment
Extensive literature suggests that loneliness is a significant hazard for older people (Beach and Bamford Reference Beach and Bamford2014; Bolton Reference Bolton2012; De Jong Gierveld, van Tilburg and Dykstra Reference De Jong Gierveld, van Tilburg, Dykstra, Vangelisti and Perlman2006; Victor et al. Reference Victor, Scambler, Bowling and Bond2005; Wenger et al. Reference Wenger, Davies, Shahtahmasebi and Scott1996). In a second interview with people over 65 drawn from the Great Britain Omnibus Survey sample in 2002, analysed by Victor et al. (Reference Victor, Scambler, Bowling and Bond2005), 7 per cent of the sample of 999 participants were ‘often’ or ‘always’ lonely within the 38 per cent (385 participants out of 999) who were lonely at least sometimes. An analysis of data from the 2008 wave of the ELSA found that 14.8 per cent of retired people over 60 (N = 6,913) reported feeling lonely ‘much of the past week’ (Gray Reference Gray, Forsman and Nyqvist2015).
We consider firstly how loneliness is conceptualised and measured, the factors pre-disposing people to loneliness and their relevance to RH. We then consider what can be learnt from studies of interventions or services which are intended to alleviate loneliness and isolation amongst older people, both in RH and other settings.
An important distinction needs to be made between loneliness (a subjective feeling) and isolation (the objective features of an individual's relationships or lack thereof). Table 1 summarises the methods and findings of six large-sample studies of loneliness and/or social isolation amongst older people in the UK, all with samples ranging from 534 to over 10,000. These studies are Beach and Bamford (Reference Beach and Bamford2014), Dahlberg and McKee (Reference Dahlberg and McKee2014), Demakakos, Nazroo and Nunn (Reference Demakakos, Nazroo and Nunn2006), Gray (Reference Gray, Forsman and Nyqvist2015), Victor et al. (Reference Victor, Scambler, Bowling and Bond2005) and Wenger et al. (Reference Wenger, Davies, Shahtahmasebi and Scott1996).
Table 1. Summary of six UK studies on loneliness amongst older people
Notes: ELSA: English Longitudinal Study of Ageing. EL: emotional loneliness. SL: social loneliness.
Three of the six studies presented in Table 1 use simple questions about how often people feel lonely, such as those used in Wave 4 of ELSA or extension to the Omnibus Survey by Victor et al. This type of measure is often described as ‘self-assessed loneliness’. Because some people are reluctant to admit to loneliness, especially men (Pinquart and Sörensen Reference Pinquart and Sörensen2001), many researchers prefer attitudinal scales with several elements. One of the most frequently used scales is that developed by De Jong Gierveld, van Tilburg and Dykstra (Reference De Jong Gierveld, van Tilburg, Dykstra, Vangelisti and Perlman2006). This uses 11 statements, five to test ‘social loneliness’ or the availability of people to talk to, lean on, call on, trust or feel close to; and six to test ‘emotional’ loneliness (in summary, missing a close friend, feeling emptiness, missing the pleasure of company or having people around, too limited a circle or feeling rejected). In ELSA Waves 2 and 6, the test statements used were broadly based on the De Jong Gierveld scale. Whilst it is frequently used as a combined scale to measure both aspects of loneliness as a single variable, Dahlberg and McKee (Reference Dahlberg and McKee2014) used the different elements of this scale to measure emotional and social loneliness separately. Wenger et al. (Reference Wenger, Davies, Shahtahmasebi and Scott1996) used a self-assessment measure of loneliness and also an eight-item scale which is somewhat different from De Jong Gierveld's. For both measures, the results of Wenger et al. were fairly similar; predictors in their multivariate models included health, household composition (reflecting mainly the effect of living alone) and also the type of support network the individual had – an important variable to which we return later.
One possible weakness of the De Jong Gierveld scale is that it does not refer specifically to feeling the lack of a partner, a kind of absence which is different and potentially more severe in quality than ‘missing a close friend’. Measurement of loneliness is generally hampered by differences between individuals in the meanings they may attach to the concept of loneliness or to particular test statements (Victor et al. Reference Victor, Scambler, Bowling and Bond2005). The quantification of the degree of loneliness into a ‘how often’ measure or a score in a multi-item scale also fails to differentiate between the different aspects of the problem in ways that might help practitioners suggest ways to address it. Loneliness may arise from someone's social network being too small, or because their network, even if large, does not satisfy the individual's needs – e.g. because of geographical distance or lack of emotional rapport; or where friends are very old or infirm they may be unable to socialise much or provide instrumental support. Measures of isolation can pick up some objective characteristics of networks, but two individuals might respond to an identical network or identical socialising opportunities in different ways, depending on subjective factors and their social skills (Stevens Reference Stevens2001).
The six studies summarised in Table 1 examine – in five cases using multivariate analysis – the effects of several factors including widowhood, living alone, poor health, age, socio-economic status and, in some cases, contact with relatives or type of social network. The first three factors are widely reported as associated with loneliness in international as well as British studies (De Jong Gierveld and van Tilburg Reference De Jong Gierveld and van Tilburg2010; Pinquart and Sörensen Reference Pinquart and Sörensen2001). All the studies in Table 1 identify poor health, having no partner or living alone as predictors of loneliness. They show more varied and complex conclusions about the effect of socio-economic status and of the number and type of social contacts. We briefly examine previous work on each of these factors, from these and other studies, to highlight correlates of loneliness which may be particularly relevant to identifying support needs amongst British RH residents.
Gender and partnership status
Partner loss is one of the main causes of loneliness for either gender (De Jong Gierveld and van Tilburg Reference De Jong Gierveld and van Tilburg2010; Victor et al. Reference Victor, Scambler, Bond, Bowling and Marston2006; Wenger and Burholt Reference Wenger and Burholt2004; Wenger et al. Reference Wenger, Davies, Shahtahmasebi and Scott1996). Although women are more at risk of widowhood, Stevens (Reference Stevens1995) shows that women adapt better to partner loss than men, through comparing a Dutch sample of 50 women aged 60–75 with a parallel sample of 31 men of similar age. In later work, Stevens showed how widows can be helped to adapt (Stevens Reference Stevens2001; Stevens, Martin and Westerhof Reference Stevens, Martina and Westerhof2006).
Wenger et al. (Reference Wenger, Davies, Shahtahmasebi and Scott1996) found both higher loneliness and greater isolation amongst those living alone or without a partner (whether widowed or never married). This was shown through multivariate modelling of the first wave (1979 data) of their 20-year longitudinal study of 534 people aged 65 and over in rural Wales. The number of years widowed was a significant factor leading to higher isolation, but not to greater loneliness. By 1999, 47 survivors of the original sample were still available for interview (Wenger and Burholt Reference Wenger and Burholt2004); this second study found that several who had been widowed did say they had become more lonely over time.
Older men living alone are more likely to be socially isolated than older women living alone (Bartlett et al. Reference Bartlett, Warburton, Lui, Peach and Carroll2013; Beach and Bamford Reference Beach and Bamford2014; Scharf and De Jong Gierveld Reference Scharf and De Jong Gierveld2008). Beach and Bamford (Reference Beach and Bamford2014) note that one in six men over 85, according to ELSA data for 2012, care for their partner; care-givers of either gender are found to be at high risk of loneliness in a study of 8,787 people over 65 across 12 European countries (Sundström et al. Reference Sundström, Fransson, Malmberg and Davey2009).
Age
Advancing age increases the risk of widowhood and of poor health, both important causes of loneliness. Wenger et al. (Reference Wenger, Davies, Shahtahmasebi and Scott1996) found in their 1979 data that age itself was not associated with self-assessed loneliness when controlling for health, widowhood or living alone. But some attrition of friendship networks over time was found amongst those who survived to be between 85 and 102 in the 1999 phase of their study (Wenger and Burholt Reference Wenger and Burholt2004). This could be due to participants’ declining health and mobility, leading to less socialising as they aged; or, as Pinquart and Sörensen (Reference Pinquart and Sörensen2001) point out, the last survivors of any age-cohort tend to outlive their friends. Victor et al. (Reference Victor, Scambler, Bowling and Bond2005) found that those aged 85 or more reported less incidence of loneliness than 65–84-year-olds. Whilst warning against response bias amongst the very old, they suggest a possible ‘survivor effect’ – those who avoid loneliness survive longest or are less likely to enter residential care; or an ‘adaptive’ effect – the oldest are those who have coped best with partner loss and/or health issues.
Socio-economic status and neighbourhood quality
Higher financial resources may permit more socialising, and are found to be associated with lower isolation in the meta-analysis of studies of isolation and loneliness by Pinquart and Sörensen (Reference Pinquart and Sörensen2001). The fact that organised social activities within RH schemes are generally free, or offered for a small contribution to cover costs of food, etc., is helpful to residents on low incomes. However, socio-economic status may also affect social networks outside the housing scheme which have developed over many years. Socio-economic factors found to be associated with less loneliness in multivariate models are education above basic level (Victor et al. Reference Victor, Scambler, Bowling and Bond2005) and higher income (Dahlberg and McKee Reference Dahlberg and McKee2014). However, several variables appear to protect against loneliness when they are examined singly without controlling for other factors: wealth (Demakakos, Nazroo and Nunn Reference Demakakos, Nazroo and Nunn2006), car ownership, home ownership and higher social class (Victor et al. Reference Victor, Scambler, Bowling and Bond2005). Predictors of less isolation are higher income (Beach and Bamford Reference Beach and Bamford2014; Dahlberg and McKee Reference Dahlberg and McKee2014) and a middle-class former occupation (Wenger et al. Reference Wenger, Davies, Shahtahmasebi and Scott1996).
Loneliness amongst the over 60s tends to be especially common in deprived urban neighbourhoods in both the UK and the Netherlands (Scharf and De Jong Gierveld Reference Scharf and De Jong Gierveld2008). Gray (Reference Gray2009), using the British Household Panel Survey for 1991 and 2003, found that good neighbourhood quality (measured by whether older people regarded their neighbourhood as a positive social environment) was amongst the most important correlates of older people feeling relatively rich in emotional and practical support. Perceptions of neighbourhood may focus on security and safety, or the level of trust between neighbours, or the social resources which neighbours offer in terms of support to older people, e.g. that neighbours in deprived areas are less likely to be car owners and more likely to be preoccupied with poor health or other problems. This may have implications for RH providers; a socio-economic and age mix with a reasonable proportion of car drivers and people with resources to lend, and of healthy older people who are mobile and have energy to chat, may foster good neighbour relations.
Health
Loneliness is associated in numerous studies with poor health (e.g. Cacioppo and Patrick Reference Cacioppo and Patrick2008; Iecovitch, Jacobs and Stessman Reference Iecovich, Jacobs and Stessman2011). Poor health has been identified both as a cause of loneliness (Victor et al. Reference Victor, Scambler, Bowling and Bond2005) and as its result. Holt-Lunstad, Smith and Layton (Reference Holt-Lunstad, Smith and Layton2010), in a meta-analysis of 148 papers mainly reporting studies of hospital patients, find many studies attesting an effect of loneliness itself on mortality independently of various health indicators. However, they find it hard to determine whether loneliness contributes to ill-health or illness restricts social relationships. In longitudinal studies which compare the survival rates of lonely and non-lonely, the effect of loneliness on mortality is inevitably tangled with health variables, and according to Tiikkainen, Heikkinen and Leskinen (Reference Tiikkainen, Heikkinen and Leskinen2008), the body of research on the relationship between poor health and inadequate social networks shows no clear conclusion as to causal direction.
What is clear is that ill-health and loneliness can become a vicious circle, whichever is the initial trigger factor. For example, Verstraten et al. (Reference Verstraten, Brinkmann, Stevens and Schouten2005) found that clients of a rehabilitation centre for people who had lost their sight were particularly vulnerable to loneliness, but those who did not report loneliness were able to access more social support and adapt better. This finding may imply that lonely people more generally have less capacity to access social support and cope with health disasters. James et al. (Reference James, Wilson, Barnes and Bennett2011) identified loneliness or social isolation as a risk factor for cognitive decline in the elderly, leading to further psychological and physical vulnerabilities. The overall implication for housing providers of these studies on the connection between health and loneliness seems to be that there is a ‘business case’ for arresting the ‘downward spiral’ in which health problems impede sociability, leading to depression, isolation and further decline both in health and in social support for the sick or immobile individual. Such a downward spiral creates not only unhappiness but a potential cost burden for the health service and for social care services.
Contact with children and social networks
There has been increasing concern in the UK about growing childlessness and smaller family networks as factors which raise the risk of isolation and lack of informal care for older people, making friends and neighbours increasingly important as sources of social support (Pahl and Pevalin Reference Pahl and Pevalin2005). Demakakos, Nazroo and Nunn (Reference Demakakos, Nazroo and Nunn2006) identified frequent contact with children (at least once a week, face-to-face or by phone) as an important safeguard against loneliness (see Table 1). Dahlberg and McKee (Reference Dahlberg and McKee2014) found less ‘social loneliness’ amongst those who saw their children or other close relatives at least weekly, and less emotional loneliness amongst those who received informal care for at least four hours weekly from a friend or relative. Wenger et al. (Reference Wenger, Davies, Shahtahmasebi and Scott1996) argued that contact with friends is more important in combating loneliness than contact with children, who are more valued for instrumental support. Wenger's team included social network type as a test variable in their models of loneliness and isolation. People whose network types included relatively many friends outside their family, and who showed engagement in community and voluntary groups, were found less lonely and less isolated than people whose social contacts were more limited to kin and neighbours. Because the ‘network type’ variable is a composite of different types of social contact, it is hard to draw any inference about the specific role of organised group activities. Concerning this, we turn to the work of Litwin and Shiovitz-Ezra (Reference Litwin and Shiovitz-Ezra2006, Reference Litwin and Shiovitz-Ezra2010) on formal and informal contacts, which is of key importance to our own research question, insofar as organised social activities within RH are an example of ‘formal activity’.
Litwin and Shiovitz-Ezra (Reference Litwin and Shiovitz-Ezra2006) took up an important distinction made in older literature between three types of social activity: informal (contact with family, friends and neighbours), formal (contact through classes, clubs, etc.) and solitary. They critically examined the finding of Lemon, Bengston and Peterson (Reference Lemon, Bengston and Peterson1972) that life satisfaction is enhanced by informal activity with friends, but not with other kinds of informal, formal or solitary activities. Litwin and Shiovitz-Ezra (Reference Litwin and Shiovitz-Ezra2006) report a multiple-stage factor analysis using a sample of 1,334 Jewish Israelis aged over 60 who had lived in Israel at least seven years, drawn from an Israeli government survey. They measured social relationship quality by four indicators, each self-assessed on four dimensions: satisfaction with contacts with respondents’ children, friends and neighbours, and self-assessed loneliness. They found that formal activity did not influence the quality of social relationships, whilst informal activity did. Their Israeli study also showed that frequent formal activity was correlated with frequent informal activity, although one cannot tell the direction of causation. If what housing practitioners can take from this is that formal activities do not in themselves improve overall social relationship quality, the question remains whether such activities can lead to more informal activity, or at least opportunities to engage in it, for individuals with relatively few informal contacts.
The same authors' later research on a multi-ethnic sample of 1,462 people aged 57–85 across the United States of America (USA) (Litwin and Shiovitz-Ezra Reference Litwin and Shiovitz-Ezra2010) confirms this connection between formal and informal activity in a more culturally diverse context than the Israel-based study. This has important implications for housing providers about the relative effect of group attendance (that is, ‘formal’ activity) on loneliness. Adapting the typology of social networks of Wenger et al. (Reference Wenger, Davies, Shahtahmasebi and Scott1996), Litwin and Shiovitz-Ezra categorised older people's social networks as family, friends, neighbour-focused, congregant (centred on religious groups), diverse (containing all of these) and restricted (relatively few contacts in any category). The dependent variable was ‘wellbeing’, which was measured on three dimensions: loneliness, happiness and anxiety. Although only diverse networks were significantly associated with lower wellbeing after controlling for background variables, the highest number of friends was found amongst those with ‘friend-based’ networks, who also attended groups more frequently than any people with any other network type, despite having fewer close relatives than the ‘family’ or ‘diverse’ types. These findings suggest that group attendance generates friendships, rather than simply being the outcome of other types of social contact.
Formal group activity may be regarded as a springboard which may or may not lead to supportive friendships. Opportunities to meet others are a necessary condition of developing helpful and satisfying social networks, but not a sufficient condition. They may not provide attachment figures or close confidantes. In particular, Van Baarsen (Reference Van Baarsen2002) found that neither emotional nor practical support from personal networks alleviated emotional loneliness following partner loss. However, bereaved individuals with relatively high self-esteem suffered less emotional loneliness after their partner's death, and those with more network support showed less social loneliness. This suggests that interventions which secure a good ‘background’ level of support from neighbours and friends, and which improve self-esteem, may help to prepare some people better for partner loss.
RH providers can offer organised group activities which can help people expand their social networks or deepen existing contacts with neighbours. But improving the quantity of residents' social contacts (in terms of the number of people they meet or how often), or even in some respects their quality, will not always lead to close friendships or practical support. This must be recognised as a limitation of interventions to bring neighbours together or help them access organised activities in the wider community beyond their housing estate. However, although organised social activities do not necessarily reduce loneliness, their importance should not be down-played as a way of expanding contacts from which friendship and support may sometimes, if not always, develop. Practical support from neighbours, such as help in times of sickness, appears to be more forthcoming on retirement estates which have an active ‘organised group’ social life (Gray Reference Gray, Forsman and Nyqvist2015). This finding came from a postal survey of residents accompanied by focus groups on eight retirement estates, run by the same housing association that hosted the survey of estate managers reported in this paper. Some related findings, presented here for the first time, address the question of whether organised social activities help residents make friends. The different estates were ranked by the frequency of organised common-room activities and outings, and this ranking was compared with their ranking on three indicators of friendliness and aid among neighbours. Residents were asked (a) whether they had more or fewer friends than during the two years before moving here; (b) whether they agreed with the statement that ‘it is quite easy to make friends on this estate’; and (c) who would help with food shopping or laundry if they were ill (neighbours, relatives, friends from outside the estate, care workers or ‘don't know’). Estates were given a friendliness score according to their combined ranking on these three questions. The ‘friendliness’ scores were in almost the same order as the ‘activity’ scores. (Two other potential ‘friendliness’ indicators were excluded from the ranking: a question on whether respondents ‘would like more companionship or contact with people’ yielded 37 ‘agree’ or ‘strongly agree’ responses out of 101, but there was little difference between estates. Also excluded was a question about whether the ‘friend whom you most rely on in difficult times' was a neighbour, since only 10 per cent of respondents said yes.)
Social activities and other interventions to address loneliness and isolation amongst older people, in retirement housing and elsewhere
A review of 32 interventions to alleviate loneliness amongst older people in various settings found that 86 per cent of participatory, activity-based programmes showed some positive outcomes; these programmes were more effective than home visiting or internet training (Dickens et al. Reference Dickens, Richards, Greaves and Campbell2011). The question is how different types of activity may facilitate the high-quality, continuing informal contacts which Litwin and Shiovitz-Ezra (Reference Litwin and Shiovitz-Ezra2006) argue are crucial to developing wellbeing.
In this section, we examine accounts of the range of social activities in RH, and their significance for addressing loneliness and isolation. We ask what an optimal ‘menu’ of social activities for RH might include, and what can best address different aspects of loneliness (social and emotional), especially for the most vulnerable groups of residents. These groups have been identified as including: those with sensory impairment or mobility problems (Callaghan et al. Reference Callaghan, Netten, Darton, Baumker and Holder2008); carers and un-partnered men (Beach and Bamford Reference Beach and Bamford2014); childless people (Gray Reference Gray, Forsman and Nyqvist2015); the very old who have outlived their friends (Wenger and Burholt Reference Wenger and Burholt2004); and those with low self-esteem or poor friendship skills (Stevens Reference Stevens2001).
Various desiderata emerge from the literature mentioned here about what the ideal ‘menu’ of group activities for older people should provide.
Provision of ‘something for everyone’
For example, some older people want active exercise like keep-fit sessions, line dancing or a walking group, whilst others need chair exercises (evaluated, for example, by Hobby Reference Hobby2006). Men may want different activities from women. Cue games (Hanover 2013), walking football (Beach and Bamford Reference Beach and Bamford2014) and ‘sheds’ for hobbies and DIY (Milligan et al. Reference Milligan, Payne, Bingley and Cockshott2013) have been suggested as activities which men enjoy and which help to counter-balance the predominantly female social milieu of many RH schemes.
A range of different forms of mental stimulation and physical exercise to suit various tastes
Remaining mentally and physically active is an important defence against memory loss. Lasting friendships are fostered by development of affinity through shared interests. Book clubs, art and craft activities, computer use, educational classes, talks, discussions, problem-solving games and various forms of physical exercise may achieve this better than ‘tea and talk’ or the ubiquitous bingo games. Activities found to be successful in alleviating loneliness include music (Hays and Minichiello Reference Hays and Minichiello2005) as well as art, group discussions and ‘therapeutic writing’ (Windle, Francis and Coomber Reference Windle, Francis and Coomber2011). Some UK case studies report successful gardening projects, providing both exercise and sociability (Hanover 2010, 2013).
Activities which will foster emotional support and informal contacts that develop into real friendship, addressing emotional loneliness, rather than just being with others for a couple of hours
Housing providers can help to create an environment conducive to generating informal contacts between neighbours, developing peer support and mutual aid; examples are provided in Blood and Pannell (Reference Blood and Pannell2012b ), Callaghan, Netten and Darton (Reference Callaghan, Netten and Darton2009) and Darton et al. (Reference Darton, Bäumker, Callaghan and Netten2011), and in the ‘good neighbour’ schemes reported in Hanover (2011). Contact with active, slightly younger people outside the housing scheme may help residents to find new confidantes and practical support. In extreme cases, isolated people may need one to one befriending or counselling to generate the confidence and motivation to reach out to new social circles. Some people who have suffered particular breaks in their social networks due to poor health, bereavement or moving district may welcome group counselling on how to rebuild personal friendships, in the style of the schemes described by Cattan et al. (Reference Cattan, White, Bond and Learmouth2005) and Stevens (Reference Stevens2001).
Breaking the vicious circle which leads through loneliness to poorer health, and thus to reduced capacity to engage with others and make new friends
Health promotion events can be social activities in themselves. The NAPA Life and Soul project (National Association of Providers of Activity for Older People (NAPA) 2012) trained volunteers amongst residents of sheltered housing to persuade other residents to eat well, take exercise and stay mentally active by engaging in positive leisure activities. They were also trained to set up group activities to help others achieve these goals. Housing scheme staff had an important role in supporting these volunteers and reinforcing their ‘healthy living’ messages. The project evaluation covering 1,700 residents and volunteers demonstrated an improved sense of wellbeing and new friendships, despite only limited success in inducing more exercise and healthier eating.
Supporting the most frail and especially those with mobility problems to take part in community life
Retirement villages, with a sizeable community of mixed degrees of frailty and dependency, may create a culture in which neighbours help each other with mobility issues, domestic tasks after discharge from hospital and other everyday favours (Biggs et al. Reference Biggs, Bernard, Kingston and Nettleton2000). Users of manual wheelchairs have particular difficulty accessing community events. Sometimes staff can push wheelchairs, but they may have insufficient time or strength, and are unlikely to be available during evenings and weekends. Volunteer ‘pushers’ can be arranged from within the estate community or from outside, e.g. from student groups.
Helping residents maintain links with the wider community, in particular links with other age groups and with healthier people
Integration of older people with the activities of younger generations within community centres has been recommended in a government strategy document on social inclusion of the over 50s (Social Exclusion Unit 2006). This can offer more choice of activities than age-segregated clubs and greater opportunities for developing intergenerational relationships and friendships based on participants’ real interests. Partnerships with schools, which are popular with RH residents, have been used to offer music from school choirs, conversation with children over school dinners (Hanover 2009), shared reading activities or helping with gardening (Hanover 2010). Other projects have incorporated conversation about older people's reminiscences into history lessons (Hanover 2009) or into a youth theatre production (Hanover 2012). Teenage volunteers can provide valuable stimulus to information technology learning and group leisure activities, as well as resources for individual befriending and organising food at events (Hanover 2012). Blood and Pannell (Reference Blood and Pannell2012b ) report a system of ‘buddy clubs’ bringing older people together with university student volunteers. Contact with people only slightly younger than RH residents is also important as a potential source of friendship and practical support; in an earlier study of day centres and of a befriending scheme in London, Gray (Reference Gray2006) found that when frail older people find support from non-relatives, they are generally friends or neighbours over 50. One day centre had an ‘early retired’ club open to all over 50, from which it drew several volunteers to help with activities for older clients. Blood and Pannell (Reference Blood and Pannell2012b ) review several good practice reports about turning a sheltered housing scheme into a ‘community hub’ open to all local residents over 50. However, our fieldwork revealed that where communal facilities in housing schemes are opened to non-residents, adequate security measures and charging systems are needed. Residents sometimes feel that security is compromised or that outsiders are ‘free-riding’ on something that they pay for through their management charge.
Encouragement of internet use as an important gateway to the wider social environment
In an analysis of the UK government's Understanding Society survey, risk factors for loneliness such as living alone, poor health status or mobility problems are also associated with lower internet use (Green and Rossall Reference Green and Rossall2013). They review several studies showing that internet use by older people can alleviate the psychological effects of isolation, as well as actually reducing isolation and loneliness. But whereas 80 per cent of those aged 55 use the internet, only 30 per cent do so at 75. Internet use is especially important for those with poor mobility, and to communicate with far-away family members, although only 22 per cent of elderly internet users communicate by Skype. These findings underline the importance of computers in common rooms, and lessons on using their internet functions which were only offered on two estates in our fieldwork. Cheap internet access through a communal broadband connection in housing schemes is also helpful, individual broadband connections being too expensive for some residents.
Offering specialised help to those who are too frail or immobile to leave their homes, or who have become withdrawn due to bereavement or crisis leading to rupture of social networks
Befriending and counselling schemes are an option particularly for the over 85s who, according to the fieldwork reported here, are least likely to participate in organised social activities within their housing scheme. Andrews et al. (Reference Andrews, Gavin, Begley and Brodie2003) reported generally positive results from a very small-scale study of a befriending scheme run by Age UK in Buckinghamshire. However, Cattan et al. (Reference Cattan, White, Bond and Learmouth2005), comparing several evaluations of befriending schemes, had mixed conclusions on their outcomes and benefits. Their review of several dozen programmes across Europe and the USA concluded that one-to-one visiting is less effective than group activities, variously involving group discussions about health topics, friendships and how to improve them, or peer-led group support after bereavement. One of the most successful was a Dutch ‘friendship programme’, offering 12 weekly ‘lessons’ in relationship skills to groups of 8–12 older women. This programme, reported in more detail by Stevens (Reference Stevens2001), was sufficiently promising to be replicated by 200 agencies with the use of standardised manuals. It was evaluated as successful even with some individuals who had suffered partner loss (Stevens, Martina and Westerhof Reference Stevens, Martina and Westerhof2006). Because it specifically addresses the issue of how to turn acquaintances into friends, it may help to address emotional loneliness as well as how to expand or change networks. Crucially, this type of programme may help to improve self-esteem, which Van Baarsen (Reference Van Baarsen2002) identifies as a factor helping people to overcome emotional loneliness.
Lester et al. (Reference Lester, Mead, Graham, Gask and Reilly2012) found that befriending is valuable only where there is good matching of befriender to client. They comment that befriending is most valued by those in residential care, where people may feel ‘lonely despite being surrounded by other people all day’ – in other words, they are experiencing emotional loneliness rather than merely a lack of contacts. This may also affect frail older people in extra-care housing. Gray's focus group work suggested that in mixed-dependency-level schemes, visiting of sick or frail neighbours often develops naturally (Gray Reference Gray, Forsman and Nyqvist2015).
Telephone befriending could, in principle, be organised as part of a remote or peripatetic support service within the array of programmes offered by housing providers. Two studies suggest that telephone helplines and befriending services are mainly valuable for periods of transition such as bereavement, and as a springboard to help individuals re-enter social networks. These are the national telephone helpline Silver Line (Callan Reference Callan2013) offering weekly befriending plus information and referrals, and an earlier pilot scheme in eight localities evaluated by Cattan, Kime and Bagnall (Reference Cattan, Kime and Bagnall2011) which helped older people to gain confidence and become more socially active in their communities. Some users of the Silver Line would have preferred face-to-face contact, however, that would clearly have raised costs.
Fieldwork methods
The fieldwork reported here is a survey of estate managers of retirement housing employed by one of the largest English providers of social rented housing for older people, whose properties also include some owner-occupied (leasehold) dwellings for the over 55s. The survey investigated how the frequency and nature of social activities is affected by factors such as the availability of common rooms, estate size and tenure mix, the amount and type of support given by managers and the role played by residents’ own initiatives, including formal residents’ groups. Earlier focus groups and a postal survey of tenants of the same housing association were reported in Gray (Reference Gray, Forsman and Nyqvist2015); the managers’ survey, conducted whilst that paper was in press, represents the final stage of this three-part project.
E-mails were sent in October 2013 to 450 managers yielding 326 replies about 445 estates. They managed a randomly chosen sample of social rented estates for people over 55 across England, owned by a major housing association. ‘Estate’ is defined here as a housing management unit, usually a single block of retirement flats or a group of linked blocks. The questionnaire provided for separate responses concerning up to five estates. The sampling frame included only managers of social rented estates, but because some had charge of additional estates, these included 102 leasehold estates and 62 of mixed tenure. Forty-six estates were ‘extra care’. Almost 30 per cent of respondents managed more than one estate, with one-third working part-time. Of the 445 estates covered, 56 per cent had 20–39 dwellings, 22 per cent less than 20 dwellings and 22 per cent at least 40 dwellings. Just over a third of estates shared their manager with one other estate, and 15 per cent with two or more others.
Empirical findings on the range and frequency of social activities in retirement housing
We now examine the findings about organised social activities from the survey of estate managers. We enquire how the activities on offer measured up to the desiderata outlined above, and what were the drivers of the range and frequency of activities.
The estate managers’ survey reported some organised social activities on 80 per cent of estates. But only 43 per cent had four or more different types of regular activity, and only 34 per cent had something happening every week. The three main factors affecting the frequency of activity were:
-
1. Availability of a common room. Only 56 per cent of estates had one but 97 per cent of common rooms had regular activities taking place. Common rooms were present in 85 per cent of the largest estates (those with 40 or more dwellings) and in 90 per cent of extra-care estates. But only 13 per cent of estates with less than 20 homes had their own common room.
-
2. The influence of a residents’ association or social committee. Residents’ groups were present in 28 per cent of all estates, and 39 per cent of those with a common room, but only 13.5 per cent of estates without one.
-
3. Managers’ efforts to organise events. Their support was crucial to widening the range of activities beyond the coffee mornings and bingo sessions that residents typically organised by themselves. Managers helped with activities on seven out of ten estates. They often encouraged the more frail or less mobile residents to attend activities, sometimes pushing wheelchairs for those who needed help. Managers frequently publicised local community centres, clubs and classes that residents could join. However, those with difficulty in walking or using public transport may find getting to outside activities difficult, unless they have their own cars or a lift. ELSA data from 2008 showed that only 54.8 per cent of RH residents have access to a vehicle when they need it (Gray Reference Gray, Forsman and Nyqvist2015).
In some cases, managers initiated and ran an activity which they thought would benefit residents, such as an exercise class or a party. In other cases, managers responded to residents’ expressed interest or supported activities initiated by a residents’ committee. Sometimes managers invited outside speakers, arranged instructors or volunteers, organised party food, obtained equipment, helped to raise funds for various costs or just publicised resident-organised events. Their suggestions and encouragement to residents about what to organise also increased the frequency and range of events. Extra-care estates were the most likely to have social activities supported by managers in some way at least once a month (78.3 per cent compared to 42.9 per cent of estates that were not ‘extra care’). Manager-supported events were also more frequent on estates with a common room, as shown in Table 2.
Table 2. Frequency of social activities by whether estates have a common room or a residents’ group
Note : Seven of the 445 estates covered by survey responses are excluded from this table because of insufficient information.
Source : Survey of retirement housing estate managers (N = 326 managers, covering 438 estates for which all the information in this table is available).
Table 2 shows the interactions between the three variables driving the frequency of organised social activities. Having a common room raised the likelihood of weekly activity by 60.7 percentage points, and having a residents’ group by 29.6 percentage points. But residents’ committees were more likely to form if there was a common room available, so amongst estates that have a common room, the committee's presence only raises the proportion of estates having weekly activities by 8.4 percentage points.
The presence of a residents’ group made at least as much difference to the frequency of manager-supported activities as it did to the frequency of activities organised by residents themselves.
The presence of residents’ groups had a marked effect in encouraging the manager to help organise activities in estates without a common room, where several managers had made efforts to accommodate residents’ gatherings in empty flats, guest rooms, foyers, gardens in summer or even in their own offices.
In a few cases, funding cuts or increased workload had curtailed activities or left the manager with little time to help residents organise social activities. This highlights both potential benefits of volunteers and partnerships with other organisations and of developing residents’ own capacity to run things for themselves. Several managers had initiated regular activities and found that later the residents themselves could take over. A few had been particularly proactive in developing residents’ capacity to run events, by raising funds for construction of a common room, taking minutes for the residents’ social committee or chairing its meetings. Supporting committees in these ways may help avert internal disputes (mentioned as an issue by a few managers) and overcome some residents’ resistance to committee work as a ‘chore’.
Only 28 per cent of estates had a formal residents’ group. They were more common on larger estates; almost 40 per cent of estates with at least 40 dwellings had one, but only 12.5 per cent of those with under 20 homes. Former residents’ groups had collapsed on one estate in four, most commonly through ‘lack of interest’ (33 per cent) or internal disputes (25 per cent). However, Gray (Reference Gray, Forsman and Nyqvist2015) suggested that another reason is often the declining health of organisers.
Managers’ help has more influence on the range and quality of activities than on their frequency. Table 3 shows the reported forms of social activities; respondents who managed two or more estates were asked about this for their first or ‘main’ estate. Activities involving partnerships with other organisations or shared outings with other estates crucially depended on managers’ efforts; they were rarely organised by residents’ groups alone. Managers arranged talks or volunteer help from the Royal Voluntary Service, Age UK, adult education colleges, churches, and health support groups like the Alzheimer's Society or the Campaign to Tackle Acquired Deafness. Some arranged for local traders to bring residents farm food tastings, demonstrations of mobility aids or fashion shows. Partnerships with schools were quite common, e.g. harvest festival events or inviting children to sing at Christmas or Easter.
Table 3. Type of activity by who organised it
Despite the ingenuity of some estate managers in engaging partner organisations to provide talks, classes, product demonstrations and entertainments, the activity menu in most estates was rather narrow. Typically the common room was used for coffee mornings or afternoon tea sessions, plus Christmas and birthday parties and, at most, just one other regular event. Book clubs, reminiscence sessions, computer classes and other adult education classes were rare. Art or craft sessions were mentioned by only 17.1 per cent of respondents, and exercise or dance sessions by 13.8 per cent. An unmet demand for exercise sessions was suggested by the postal survey of tenants in the earlier stage of this research project (Gray Reference Gray, Forsman and Nyqvist2015). The narrowness of the activities menu shown in the managers’ survey responses echoes King, Pannell and Copeman's (Reference King, Pannell and Copeman2009) national focus group study of 64 tenants across 25 sheltered housing schemes, in which residents wanted a more varied range of activities, and saw the support of on-site staff as crucial to providing them.
Conclusions
The fieldwork demonstrated the importance of mixed age and mixed capacity/dependency in RH, to secure an adequate number of people able and willing to lead residents’ groups and generate social support for the most vulnerable. This echoes Kneale (Reference Kneale2013), who argues that a sense of community and the capacity for informal aid to neighbours is fostered by a balance of highly dependent and less-dependent residents. Earlier focus groups and a postal survey of tenants of the housing association which hosted the estate managers’ survey (Gray Reference Gray, Forsman and Nyqvist2015) identified 66–75-year-olds as the age group providing most informal help to neighbours. The managers’ survey suggested that this age group is also the most active in attending and organising estate activities. As a housing scheme ages, so do its initial residents; the ratio of frail to active may rise, making it harder for the community to sustain adequate informal help between neighbours and an active social life. Introduction of volunteers or participants from the wider community then becomes increasingly important to sustain an attractive activity programme and new friendship opportunities. Estates which mix leasehold with rented units may attract a higher proportion of occupants who are car owners and in relatively good health than those which are all social rented. Leaseholders, being generally former home-owners, have on average higher lifetime income than renters, which is correlated with better health.
The estate managers’ survey shows staff playing an important role in generating and sustaining social activities, especially on estates without a common room. Many managers offer help to residents with social activities even amongst the one-third of respondents whose job description does not include ‘social support’ of residents. In 28 per cent of estates, managers organised all social events taking place such as coffee mornings, parties and outings; in 40 per cent they helped residents raise funds for events, outings or classes; and in 46 per cent they helped resident-organised events with logistic arrangements like wheelchair pushing, shopping for party food, or finding speakers, instructors or equipment. Over three-quarters helped to publicise events through estate newsletters or posters. All these forms of help were associated with a greater frequency of organised activities. Estate managers thus perform an important collective service to the retirement estate community which a peripatetic support worker contracted to visit individuals could not do. This finding is of concern since some housing associations are replacing on-site support staff with peripatetic or call centre-based staff whose relationship is with the individual client rather than the community. Some potentially valuable forms of intervention – such as counselling, ‘friendship skills’ programmes or ‘lessons’ in internet use – crucially require specialist skills which will often need to come from outside the RH community, so that they cannot rely on residents’ self-organisation.
The menu of activities frequently lacks stimulating options, consisting mainly of coffee mornings, Christmas and birthday parties. Where it is broadened to include outside speakers, films or entertainers, exercise or other classes, or visits from schools, this usually needs the manager's help even where there is an active residents’ group. Outside contacts are the aspects of activity organisation which residents are least likely to undertake for themselves. This reflects a perceived need for a professional, rather than a residents’ representative, to approach head teachers or managers of local organisations. A broadening of the menu might be expected to attract more residents into activities, with greater prospects of them making supportive friendships on the basis of shared interests, and a stronger basis for sustaining a residents’ association.
Managers are more likely to support social activities where there is a residents’ group they can work with and a ‘quantum’ of activities that they can publicise, encourage or fund-raise for. Some managers have sought to establish or support residents’ groups to run their own activities; a few report successes in handing over organisational tasks to them. However, only a minority of estates have a formal residents’ group, and these frequently collapse through internal disputes or the leadership's declining health as they age. A crucial question is how best to mobilise residents to form and sustain residents’ committees, to organise activities and to take a greater role in developing partnerships with external organisations.
What is especially important to the health services and social care system, and to housing providers, is that loneliness and isolation increase the older person's chances of needing formal care. Insufficient social stimulation means greater risk of memory loss and depression, which together with insufficient exercise contributes to a decline in both physical and mental health. As well as being a personal tragedy, this provides a ‘business case’ for housing providers to consider how their services will contribute to the quality of older people's social lives, in order to reduce the ever-growing costs of medical treatment and social care.
Acknowledgements
No external funding was involved in this study. The corresponding author's university provided computing facilities as part of normal staff support. The costs of staff time and postage involved in administering the postal survey of tenants and the managers’ survey were provided by the collaborating housing association, as were the staff time and facilities for data input and for the co-author's contribution to writing the paper. The focus groups and postal survey were submitted to the research ethics scrutiny process in accordance with the normal procedure in the corresponding author's university. The survey of housing managers was submitted to an internal process by their employing organisation's research team to consider issues of confidentiality, workload and inclusivity. Both authors have been substantially involved in writing this paper. No conflicts of interest are apparent. The second author is no longer employed by the collaborating housing association, having successfully sought a career move elsewhere.
