Social exclusion of older people in deprived urban communities of England

Abstract

Addressing the causes and consequences of social exclusion represents a key theme in European social policy, reflecting growing awareness of the social costs which arise when individuals, families and communities become cut off from wider society. Conceptually, however, social exclusion remains underexplored in gerontology. The article suggests that exclusion represents a useful means of exploring the situation of older people in different environmental settings. Social exclusion in old age is conceptualised as a multi-dimensional phenomenon comprising of: exclusion from material resources; exclusion from social relations; exclusion from civic activities; exclusion from basic services; and neighbourhood exclusion. Drawing on a survey of 600 people aged 60 and over in deprived neighbourhoods of three English cities, the article develops indicators to represent each dimension of exclusion and seeks to assess the nature of social exclusion faced by older people in deprived neighbourhoods. Results reveal a considerable proportion of older people experiencing social exclusion in at least one form. The sample fell into three categories: 33% were not excluded on any of the five domains; 31% experienced exclusion on a single domain; 36% were vulnerable to the cumulative impact of multiple forms of exclusion. Multiple social exclusion was significantly correlated with respondents’ ethnic origin, educational status, housing tenure, perceived health status and quality of life. It is concluded that social exclusion represents a useful means of depicting disadvantage experienced by older people living in deprived urban neighbourhoods, and that it would be useful to extend the analysis to other types of residential setting.

Introduction

In the United Kingdom, the rising average incomes of older people in recent years mask a widening gap between those who are better off and a substantial minority who continue to live in poverty (Darton and Strelitz 2003). Evidence of growing income inequalities is matched by research highlighting the enduring nature of geographic inequalities (Hills 1995; Strelitz and Darton 2003). In their most acute form, such area variations characterise the distribution of good health and, ultimately, of life expectancy (Griffiths and Fitzpatrick 2001a; Shaw et al. 2001; Leyland 2004). In some of Britain’s most socially deprived urban communities, average male life expectancy at birth is up to 10 years lower than that in more affluent areas (Griffiths and Fitzpatrick 2001b). Although research evidence points to a growing geographic influence on individuals’ life chances, the impacts on older people of related inequalities have been underexplored in social gerontology. This article seeks to contribute to debates about the nature of environmental influences on ageing by exploring the characteristics of older people who live in urban communities marked by multiple forms of social disadvantage.

Academic and policy debates increasingly use the concept of social exclusion to address such area-based disadvantage (Scharf et al. 2002). While social exclusion remains a contested notion (Silver 1994; Levitas 1998), it succeeds in drawing attention to the social costs which can arise when individuals, families and communities become disengaged from wider society. As such, exclusion discourse reaches beyond the traditional concerns of poverty research to incorporate a range of additional domains. While there are clear overlaps between the concepts of poverty and social exclusion (Bhalla and Lapeyre 1997; Bauman 1998), the multi-dimensional nature of exclusion is emphasised in policy and research (Atkinson 1998; Atkinson and Davoudi 2000; Phillipson and Scharf 2004). For example, the British government defines social exclusion as ‘... a shorthand term for what can happen when people or areas suffer from a combination of linked problems such as unemployment, poor skills, low incomes, poor housing, high crime, bad health and family breakdown’ (Social Exclusion Unit 2001, p. 10). Berghman (1997) disaggregates the idea of exclusion, conceiving social exclusion in terms of the non-realisation of citizenship rights within four key societal institutions—the democratic and legal system, the labour market, the welfare system, and the family and community system. Underlying several approaches to conceptualising social exclusion is a concern with issues of place and space (Madanipour et al. 1998; Glennerster et al. 1999; Lupton and Power 2002; Murie and Musterd 2004). This reflects evidence of a growing overlap between socially excluded people and socially excluded places (Forrest 2004). The spatial dimension of exclusion is particularly important, given the way in which individuals’ self-identities are shaped by their residential context, with Marcuse (1996) suggesting that the neighbourhood defines ‘who a person is and where he or she belongs in society’.

While the concept of social exclusion has become a central feature of research and policy debates in recent years, such debates have focused most strongly on the needs of children, young adults and those of employment age (e.g. Department for Work and Pensions 2001; Hills et al. 2002). However, though underexplored, exclusion also represents a useful tool for social gerontologists seeking to research disadvantage (Scharf et al. 2001). There are at least three areas in which recent social exclusion discourse might reflect better the situation of older people. First, where social inclusion is identified as being related to labour market participation (Levitas 1998), the question emerges of the potential for older people who have withdrawn from their occupational roles to be integrated within society. Second, several studies draw attention to the dynamic nature of social exclusion (Byrne 1999), with individuals and families moving in and out of exclusion as their circumstances change (Leisering and Walker 1998; Burchardt 2000; Burgess and Propper 2002). While the boundaries of exclusion may be fairly fluid for younger people, older people’s situation is potentially less amenable to change. For example, older people who lack adequate material resources are likely to be unable to escape such disadvantage without additional financial support from the state. Third, the neighbourhood dimension of exclusion may represent a much more important aspect of exclusion for older people than for other age groups. While evidence points to the high degree of younger people’s mobility within urban areas (Neighbourhood Renewal Unit 2003), there is still a strong tendency amongst older people to age in place (Callahan 1992; Phillipson et al. 1999). Not only do older people tend to spend more time than do younger people in their immediate residential setting (Walker 2004), but many will have spent a substantial period of their lives in one and the same neighbourhood, deriving a strong sense of emotional investment both in their home and in the surrounding community (Young and Willmott 1957; Phillipson et al. 2000). Indeed, Rowles (1978, p. 200) suggests that the ‘selective intensification of feelings about spaces’ might represent ‘a universal strategy employed by older people to facilitate maintaining a sense of identity within a changing environment’. While it may be relatively straightforward for older people to maintain a sense of self in stable neighbourhoods which change organically, some residential settings may present problems. In our view, this is more likely to apply to urban than to rural settings. Ageing individuals living in rapidly changing urban communities may experience greater difficulty in maintaining their self-identity. This could apply to ‘zones of transition’ marked by a rapid turnover of people and buildings, and to unpopular urban neighbourhoods characterised by low housing demand and abandonment by all but the poorest or least mobile residents (Power 2000). Disadvantaged urban neighbourhoods, and the people who reside in them, may be also prone to ‘institutional isolation’ (Gans 1972) as services and agencies withdraw. For residents, especially those with restricted mobility, this might ultimately create problems in accessing even the most basic services, such as energy, food retailing, telephones and banking (Speak and Graham 2000).

In light of this discussion, social exclusion can potentially represent a flexible and multi-dimensional tool for examining the degree to which older people in different environmental settings simultaneously experience varying forms of disadvantage. However, there is a need, first, to disaggregate the idea of exclusion and, second, to generate suitable indicators of the dimensions of exclusion which are most relevant in later life. In this context, it is useful to draw on two previous studies which have sought to operationalise social exclusion in a way which acknowledges the concept’s multi-dimensional nature. Burchardt et al. (2002) identify four social exclusion domains which relate to the individual’s ability to participate in ‘normal’ social activities: consumption activity (the ability to consume up to a minimum level the goods and services considered normal for society); production activity (engagement in an economically or socially valued activity); political activity (engagement in some collective effort to improve or protect the immediate or wider social or physical environment); and social activity (engagement in significant social interaction with family or friends, and identifying with a cultural group or community). Similarly, Gordon et al. (2000) highlight four dimensions of social exclusion: impoverishment; non-participation in the labour market; lack of access to basic services; and exclusion from a range of social relations. This latter component is further subdivided into elements such as individuals’ non-participation in common social activities, social isolation, a perceived lack of support in times of need, lack of civic engagement and an inability to ‘get out and about’ (Gordon et al. 2000). Synthesising these approaches, the spheres of consumption/impoverishment, political/civic activity and social activity represent important dimensions of social inclusion and exclusion. Moreover, these represent some of the traditional concerns of social gerontology (e.g. Townsend 1957, 1981; Phillipson 1998; Walker 1998; Minkler and Estes 1999; Phillipson et al. 2000). Since access to basic services is important in later life, not least as a way of maintaining independence in the home (Means 1997; Tanner 2001), this aspect should also be incorporated in a study of social exclusion in old age. Absent from previous attempts to operationalise exclusion and, as argued above, especially relevant to debates about the social exclusion of older people, is the neighbourhood dimension. Against this background, it can be argued that there are at least five forms of social exclusion relevant to the circumstances of older people living in deprived urban communities:

• Exclusion from material resources, identifying the central role played by income and material security in determining individuals’ ability to participate in society;

• Exclusion from social relations, reflecting the importance attributed to the ability to engage in meaningful relationships with others;

• Exclusion from civic activities, recognising the need for individuals to be able to engage in wider aspects of civil society and in decision-making processes which may in turn influence their own lives;

• Exclusion from basic services, drawing upon the key role played by access to services in and beyond the home in terms of individuals’ ability to manage everyday life;

• Neighbourhood exclusion, reflecting the contribution made by the immediate residential setting to an individual’s sense of self and, potentially, their quality of life.

Following this discussion, a number of research questions arise. First, there is the issue of the degree to which older people in disadvantaged neighbourhoods experience social exclusion in its various forms. Second, are there types of social exclusion which occur more often than others? Third, to what extent do the different dimensions of social exclusion overlap? Fourth, what are the characteristics of socially excluded older people? In the absence of existing data sources, this paper uses data drawn from an empirical study undertaken in some of England’s most socially disadvantaged urban communities in order to address these questions.

Methods

Sample

The first stage of the empirical study involved selecting appropriate study locations. Liverpool, Manchester and the London Borough of Newham were identified as the three most deprived English local authorities in the 1998 Index of Local Deprivation (Department of the Environment, Transport and the Regions 1998). To account for within-area variation in relation to the intensity of deprivation, the three most deprived electoral wards in each local authority were selected. All nine wards were ranked amongst England’s 50 most deprived wards in 1998. The study areas vary in terms of their proximity to the respective city’s central core, socio-economic structure and population profile. However, they share key characteristics associated with intense urban deprivation, including above-average rates of unemployment, relatively poor housing conditions, a loss of amenities and services, and a high incidence of crime (Social Exclusion Unit 1998).

Face-to-face interviews were conducted using a standard questionnaire with 600 people aged 60 and over. Recruitment of participants occurred in two ways. A first group was randomly selected through local electoral registers, using a coding classification which assigns people to age bands according to the likelihood that their first name belongs to a particular birth cohort. In all, 501 respondents were recruited in this way (response rate of 42%). A second group was recruited from the largest minority ethnic group in each electoral ward, drawing on relevant community organisations and researchers’ local contacts. This strategy reflected the need to generate a sample which better reflected the characteristics of the older population of deprived urban communities in England. It was necessitated by the absence of a suitable sampling frame for minority ethnic older people (Sin 2004), and the fact that a purposive sampling might yield small numbers of respondents from a very diverse range of minorities. The non-purposive approach aimed to generate a sufficiently large sample from each minority group to facilitate statistical analysis. Using snowballing techniques, 99 older people from four different minority groups (Black Caribbean, Indian, Pakistani and Somali) were recruited, with interviews being undertaken in the language of respondents’ choice. For the purposes of this article, analysis focuses on 581 respondents for whom there is complete information regarding the different dimensions of social exclusion (described below).

Measures

Measures of exclusion

Exclusion from material resources encompasses the related concepts of multiple deprivation and poverty. In relation to deprivation, the research adapted Evandrou’s (2000) ‘index of multiple deprivation’. This comprises seven items considered especially important to older people (e.g. central heating, use of a telephone, access to a car), categorising people according to the degree of deprivation faced. Categories range from no deprivation (not disadvantaged on any of the seven characteristics) to high deprivation (disadvantaged on at least five characteristics). In relation to poverty, the research followed Gordon et al. (2000) by identifying as poor those respondents who said they were unable to afford what 50% or more of British people view as basic necessities. Drawing on the results of a national survey of the adult population (Gordon et al. 2000), respondents were presented with a list of 19 material items which included such basics as two meals a day, home contents insurance, and the ability to replace worn-out furniture. People lacking two or more items because they could not afford them were judged to be in material poverty. These indicators were chosen in preference to a simple income measure, not only because of their conceptual strength, but also because of the high degree of non-response associated with income questions in survey research (in this study, 26% of respondents failed to provide income data). To focus attention on respondents who were likely to be experiencing greatest difficulty in making ends meet, the composite indicator of exclusion from material resources counts as excluded those older people who are experiencing both material poverty and medium to high levels of deprivation.

Exclusion from social relations draws on indicators of social isolation, loneliness and non-participation in common social activities, and seeks to acknowledge the distinctive nature of these concepts (Victor et al. 2000). A social isolation measure reflected the availability and frequency of contacts with family, friends and neighbours. Individuals scored 1 for each of the following characteristics: has no relatives or children OR sees a child or other relative less than once a week; has no friends in neighbourhood OR has a chat or does something with a friend less than once a week; has a chat or does something with a neighbour less than once a week. Scores on the isolation measure range from 0 (no isolation) to 3 (high isolation), with a score of 2 or more being adopted as an indicator of social isolation. Loneliness was measured using the 11-item De Jong Gierveld loneliness scale (De Jong Gierveld and Kamphuis 1985). Adopting the cut-off points suggested by the scale’s authors (De Jong Gierveld and van Tilburg 1999), respondents scoring 9 or higher out of 11 were classed as (very) severely lonely. Non-participation in common social activities is an indicator derived similarly to the poverty measure outlined above. It encompasses seven common activities perceived as necessities by a majority of the adult population, including having friends or family around for a meal and celebrating special occasions (Gordon et al. 2000). An inability to participate in two or more common activities on the grounds of lack of income was adopted as an indicator of this sub-dimension of exclusion. The summary measure of exclusion from social relations recognises the conceptually distinct nature of each of its component indicators. Consequently, individuals were assessed as excluded on this domain if they were socially isolated, or were (very) severely lonely, or were unable to participate in two or more common social activities.

Exclusion from civic activities was assessed with reference to two indicators chosen to reflect different types of civic participation. First, respondents were asked whether they attended (1) religious meetings or (2) meetings of community groups. Response categories ranged from (1) yes, frequently to (3) no, never. Second, respondents were presented with a list of 11 civic activities, and asked whether they had undertaken any of the stated activities in the 3 years preceding interview. The list encompassed a broad range of activities, reflecting a wide variety of types of civic engagement. Included were activities such as voting in elections, writing a letter to a newspaper and taking part in fundraising initiatives. In order to provide a parsimonious measure, individuals were judged to be excluded from civic activities if they never participated in meetings of religious or community groups and did not take part in any type of civic activity.

Exclusion from basic services was measured in terms of individuals’ access to and use of a range of services both within and beyond the home. In the home, respondents were asked whether in the previous 5 years they had used less water, gas or electricity or had used the telephone less often in order to save money. Beyond the home, an indicator of service exclusion was derived from the non-usage in the year preceding interview of three services regarded as particularly important to older people (post office, chemist, and bus service). The two sub-dimensions of this exclusion domain are conceptually distinct. As a result, the adopted indicator of exclusion from basic services encompassed those respondents who had used less of three or four services in the home in order to save money, or who had not used two or more key services beyond the home.

Neighbourhood exclusion was assessed through two indicators: individuals’ perceptions of their neighbourhoods, and their feelings of security when moving around in the neighbourhood. Neighbourhood perceptions were addressed in two separate questions, with a summary indicator chosen to reflect the greatest degree of neighbourhood dissatisfaction. Respondents who (1) could identify only something which they disliked about their neighbourhood, or (2) were ‘very dissatisfied’ with their neighbourhood were deemed to have a negative perception of their neighbourhood. A measure of neighbourhood security was derived from responses to the question ‘How safe would you feel if you had to go out alone in this neighbourhood after dark?’ Potential response categories ranged from (1) very safe to (4) very unsafe. People who responded that they would feel ‘very unsafe’ under such circumstances were identified as being most at risk of neighbourhood exclusion. The composite indicator of neighbourhood exclusion reflects a parsimonious approach and seeks to identify those people experiencing the greatest disadvantage. It counts as excluded those individuals who expressed negative perceptions of the neighbourhood and who also reported that they would feel very unsafe in their neighbourhood after dark.

Individual variables

One purpose of this article is to assess the degree to which different groups within the population of older people in deprived urban neighbourhoods experience forms of social exclusion. In this respect, analysis encompasses a range of individual variables which were distinct from the exclusion indicators. In addition to age, gender, ethnicity and marital status, variables relating to education level, length of residence in the neighbourhood, housing tenure and conditions, experience of crime, health status and quality of life form part of the analysis.

Education level was determined according to the age at which respondents had left school, with those leaving school at 15 years of age or younger regarded as having a low educational status. Length of residence in the community sought to distinguish between those who were relatively recent movers to the neighbourhood (0–4 years in community), those who were relatively stable residents (5–19 years in community) and those who had lived in the neighbourhood for a longer period of time (20 or more years). Housing tenure distinguished between owner-occupiers, and those who rented from social and private landlords. A measure of housing conditions was derived from individuals’ responses to a set of questions asking whether they had any of nine housing problems (such as dampness, inadequate heating, lack of space). A similar approach was adopted to determine individuals’ recent experience of crime. This was based on questions asking whether respondents had been victims of any of seven types of crime (including burglary, street crime, assault) in the 2 years preceding interview. Health status was assessed on the basis of a self-rated measure of general health, ranging from (1) very good to (5) very poor, and a standard measure of limiting longstanding illness. A single-item quality of life measure, which has been shown to correlate closely for this sample with indicators of subjective well-being (Diener et al. 1985) and morale (Lawton 1975), asked respondents to rate their quality of life from (1) very good to (5) very poor (Smith et al. 2004).

Results

Participant characteristics

The sample ranged in age from 60 to 96 years (M=71.6 years; SD=8.0). Older residents of deprived urban areas differ from nationally representative population samples in a number of ways. While the gender distribution of the deprived areas sample broadly reflects the national pattern, it differs in relation to other socio-demographic factors (Table 1). Compared with national data, significantly fewer older people in deprived areas were married or living as a couple, and there were higher proportions who were widowed, divorced or separated or who had never married. The proportion of those who live alone is higher in the deprived areas sample, and their health—especially as reflected in the reporting of limiting chronic illness—is generally worse. Linked in part to the recruitment method, but also reflecting the geographic concentration of minority ethnic groups in disadvantaged neighbourhoods, the sample is also more diverse than is the general older population in terms of its ethnic composition. While 92% of people aged 65 and over in Great Britain as a whole describe themselves as white (Office for National Statistics 2004, p. 18), the equivalent proportion in the deprived areas study was 74%. In all, 13% of respondents described themselves as Black Caribbean, 6% as Somali, 3% as Pakistani, and 4% as of Indian origin. The proportion of respondents who owned their own homes was significantly lower than the national average, especially amongst those aged 70 and over. Conversely, older people in deprived areas were much more likely to rent their homes from social landlords than was the case for those in Britain as a whole. This highlights the concentration of survey respondents within lower socio-economic groups. Finally, it is also important to note that the study population displays a remarkable degree of residential stability. Length of residence ranged from less than 6 months to 96 years (M=39.2 years; SD=22.0). Of those aged 70 and over, just 14% had lived in their present community for less than 20 years. In the general population, the equivalent proportion is 28%. This ‘ageing in place’ is also evident within the 60–69 age-group, around three-quarters of whom are long-term residents of their neighbourhood.

Table 1.

Sample characteristics: older people in deprived areas and Great Britain (data sources: General Household Survey 2000–2001 and Deprived Areas Survey 2000–2001)

Sample characteristics		General Household Survey 2000–2001		Deprived Areas Survey 2000–2001
Age		65–74 years	75 and over	65–74 years	75 and over
Sex	Male	47	38	43	37
	Female	53	62	57	63
Marital status	Single	6	7	10	8
	Married/cohabiting	66	41	46	27
	Widowed	21	49	32	60
	Separated/divorced	7	4	12	6
Household composition	% living alone	29	50	41	59
Health	% reporting longstanding illness	57	64	66	68
	% reporting limiting longstanding illness	37	47	47	61
Base		(n=4,719)	(n=3,888)	(n=242)	(n=201)
Age		60–69 years	70 and over	60–69 years	70 and over
Years resident in neighbourhooda	0–4 years	9	8	6	3
	5–19 years	23	20	20	11
	20 or more years	68	72	74	87
Base		(n=1,124)	(n=1,404)	(n=256)	(n=311)
Housing tenureb	Owner-occupier	76	65	52	37
	Social renter	20	29	39	55
	Private renter	4	6	9	8
Base		(n=3,195)	(n=4,559)	(n=256)	(n=311)

^aCoulthard et al. (2002)

^bGeneral Household Survey 2002 data

Experience of different forms of social exclusion

Using the previously highlighted measurement approach, exclusion from social relations affected around two-fifths of respondents, nearly three out of ten were prone to exclusion from material resources, and almost one-quarter experienced exclusion from basic services (Table 2). Exclusion from civic activities and the neighbourhood affected 15 and 13% of the sample respectively.

Table 2.

Proportion of older people in deprived urban neighbourhoods excluded on different domains

Domain of social exclusion	Indicator of exclusion	% excluded on this indicator	% excluded on this domain
Exclusion from material resources	Material poverty (lacking two or more necessities)	41	29
	Multiple deprivation (deprived on three or more characteristics)	60
Exclusion from social relations	Social isolation (isolated on two or more characteristics)	20	41
	Loneliness (severely or very severely lonely)	16
	Unable to participate in two or more common activities	17
Exclusion from civic activities	Non-participation in civic activities	24	15
	Never attends meetings of religious or community organisations	48
Service exclusion	Has restricted use of at least three of four basic services in the home	13	23
	Has not used at least two of three key services beyond the home	10
Neighbourhood exclusion	Expresses very negative views about the neighbourhood	22	13
	Would feel ‘very unsafe’ when out alone after dark	45

Experience of the different dimensions of social exclusion was statistically linked to a number of individual level variables (Table 3). For this sample, it is noteworthy that exclusion tended not to vary significantly according to age or gender. The exceptions were that those in the oldest age group (75+) were more likely to be excluded from basic services and the neighbourhood, and that women were more likely than men to be excluded from civic activities. Marital status was related to exclusion from material resources and from civic activities. In each case, married people were least likely to be excluded, and widowed or divorced/separated people were most likely to be excluded. Vulnerability to each dimension of exclusion varied significantly according to individuals’ ethnic origin. While white respondents were most likely to be excluded from civic activities and the neighbourhood, Pakistani and Somali elders were especially prone to exclusion from material resources, social relations and basic services. Older people of Indian or Black Caribbean origin tended to be less vulnerable to the different forms of exclusion than were each of the other ethnic groups. Respondents’ age on completing full-time education was associated with exclusion from material resources, civic activities and the neighbourhood. Interestingly, length of residence in the neighbourhood was not statistically linked to any of the five dimensions of exclusion, suggesting that short-, medium- and long-term residents of deprived communities face similar risks of exclusion. Owner-occupation tended to be associated with a significantly lower risk of exclusion from material resources and civic activities. The presence of one or more housing problems was related to exclusion from material resources, social relations, basic services and the neighbourhood. With the exception of the civic activities measure, those people who had experienced one or more crime in the 2 years preceding interview were much more likely to be excluded on the different exclusion domains. While the presence of a limiting chronic health condition was related to exclusion from basic services and the neighbourhood, self-reported health was significantly related to all forms of exclusion. In each case, those reporting (very) poor health were much more vulnerable to exclusion than were those with (very) good health. A similar pattern applied to the quality of life measure. Older people who reported a (very) good quality of life were significantly less likely to be excluded on each domain than were those reporting a (very) poor quality of life.

Table 3.

Proportions of older people in deprived urban neighbourhoods affected by different forms of social exclusion^a

	% of respondents in each category excluded from:
	Material resources	Social relations	Civic activities	Basic services	Neighbourhood
Age	n.s.	n.s.	n.s.	***	*
60–74 years	29	41	14	19	11
75 years and above	32	39	15	31	18
Gender	n.s.	n.s.	*	n.s.	n.s.
Female	31	39	18	22	14
Male	27	43	11	25	13
Marital status	***	n.s.	**	n.s.	n.s.
Never married	27	46	14	17	19
Married	20	39	10	25	15
Widowed	37	39	20	21	13
Divorced/separated	39	46	23	27	5
Ethnic origin	***	***	*	***	**
White	27	37	18	19	17
Black Caribbean	25	37	7	21	3
Indian	24	29	5	14	5
Pakistani	33	63	11	33	4
Somali	69	77	11	69	11
Age on completing education	***	n.s.	***	n.s.	*
15 years or less	33	41	17	24	15
16 years or more	6	40	4	17	6
Years in neighbourhood	n.s.	n.s.	n.s.	n.s.	n.s.
0–4	41	56	26	26	7
5–19	29	48	14	29	10
20 or more	29	39	15	22	14
Tenure	***	n.s.	***	n.s.	n.s.
Home owner	12	36	8	21	12
Social renter	44	43	20	24	14
Private renter	33	51	22	31	18
Housing conditions	***	***	n.s.	*	*
No housing problems	20	32	16	19	11
One housing problem	28	44	16	23	12
Two or more problems	45	54	12	29	19
Experience of crime	*	***	n.s.	***	***
No crimes	26	34	16	17	10
One or more crimes	35	50	13	31	19
Self-reported health	***	***	*	**	***
Very good or good	22	32	11	18	10
Neither good nor poor	34	45	16	21	12
Very poor or poor	37	50	20	32	19
Health status	n.s.	n.s.	n.s.	**	*
No limiting condition	27	38	14	18	11
Limiting condition	32	44	16	28	16
Quality of life	***	***	***	***	*
Very good or good	23	30	10	17	11
Neither good nor poor	38	53	21	23	15
Very poor or poor	43	64	23	45	21
All (n=581)
% of sample excluded on this dimension	29	41	15	23	13

^aTest on differences based on Chi-square tests: n.s., not significant; *, p<0.05; **, p<0.01; ***, p<0.001

Multiple exclusion

Further analysis (1) points to the overlapping nature of the different dimensions of social exclusion, and (2) highlights the potential for individuals to experience simultaneously multiple forms of exclusion. In relation to overlaps, a number of patterns could be identified (Table 4). For example, exclusion from material resources was statistically related to each of the other forms of exclusion. Of respondents excluded on the material dimension (n=171), 59% were also excluded from social relations, 20% from civic activities, 33% from basic services and 19% from the neighbourhood. Exclusion from social relations was related to each of the other dimensions of exclusion, with the exception of the civic activities domain. Exclusion from civic activities was linked to exclusion from material resources and basic services. Neighbourhood exclusion was most closely associated with exclusion from social relations, and also showed a relationship with exclusion from material resources.

Table 4.

Relationship between different forms of social exclusion^a

% also excluded in relation to:	% of older people excluded in relation to:
	Material resources	Social relations	Civic activities	Basic services	Neighbourhood
Material resources	-	43***	39*	42***	41*
Social relations	59***	-	49	58***	58***
Civic activities	20*	18	-	21*	18
Basic services	33***	33***	32*	-	27
Neighbourhood	19*	19***	16	16	-
Base (number excluded on each domain)	(n=171)	(n=237)	(n=87)	(n=134)	(n=78)

^aTest on differences based on Chi-square tests: n.s., not significant; *, p<0.05; **, p<0.01; ***, p<0.001

Drawing these findings together, significant proportions of older people in deprived urban areas of England appear prone to the simultaneous experience of multiple forms of social exclusion. This is perhaps unsurprising, given the extreme deprivation levels of the respective neighbourhoods. While 33% of respondents were not excluded on any of the five domains of social exclusion, 31% were excluded on a single domain, and 36% experienced two or more forms of exclusion. Exploring the characteristics of those affected by such disadvantage (Table 5) indicated that the socio-demographic variables of age, gender, marital status and length of residence were not statistically linked to the experience of multiple exclusion. However, individuals’ ethnic origin and educational level were related to exclusion. Exclusion was most pronounced for Somali and, to a lesser extent, Pakistani older people. By contrast, Indian and Black Caribbean elders were less vulnerable. White older people were more likely to be excluded on a single domain, but less often affected by multiple exclusion. Relatively few respondents who had left school at the age of 16 or older were prone to multiple social exclusion. In relation to other variables, factors which appeared to protect individuals from exclusion were home ownership, having good housing conditions, not having been a recent victim of crime, reporting good or very good health, and the absence of chronic health conditions. Renting from a social landlord, having two or more housing problems, recent experience of crime, poor or very poor health and a limiting longstanding health problem were all closely associated with the experience of multiple exclusion. Multiple exclusion was also linked to individuals’ self-rated quality of life. While 42% of respondents who reported a (very) good quality of life were not excluded, this applied to just 11% of those identifying their quality of life as being (very) poor. By contrast, 65% of individuals reporting their quality of life to be (very) poor were excluded on two or more domains, compared with 23% of those with a (very) good quality of life.

Table 5.

Proportions of older people in deprived urban neighbourhoods experiencing different levels of social exclusion^a

	% of respondents in each category:
	Not excluded	Excluded on one domain	Excluded on two or more domains
Age (n.s.)
60–74 years	36	30	34
75 years and above	27	33	39
Gender (n.s.)
Female	33	29	37
Male	33	34	34
Marital status (n.s.)
Never married	31	35	35
Married	38	33	30
Widowed	30	30	40
Divorced/separated	27	27	46
Ethnic origin (***)
White	33	34	33
Black Caribbean	44	37	29
Indian	52	29	19
Pakistani	30	15	56
Somali	6	14	80
Age on completing education (***)
15 years or less	31	31	38
16 years or more	48	34	18
Years in neighbourhood (n.s.)
0–4	19	37	44
5–19	34	26	40
20 or more	33	32	35
Tenure (***)
Home owner	43	34	22
Social renter	34	29	46
Private renter	21	29	47
Housing conditions (***)
No housing problems	40	33	27
One housing problem	34	28	39
Two or more problems	21	30	49
Experience of crime (***)
No crimes	39	34	27
One or more crimes	25	27	48
Self-reported health (***)
Very good or good	43	34	24
Neither good nor poor	29	33	39
Very poor or poor	23	27	50
Health status (***)
No limiting condition	36	36	28
Limiting condition	30	26	44
Quality of life (***)
Very good or good	42	35	23
Neither good nor poor	24	28	48
Very poor or poor	11	24	65
All	33	31	36
	(n=192)	(n=181)	(n=208)

^aTest on differences based on Chi-square tests: n.s., not significant; *, p<0.05; **, p<0.01; ***, p<0.001

Discussion

The concept of social exclusion clearly represents a useful means of exploring forms of disadvantage in later life which reach beyond some of the traditional concerns of social gerontologists with phenomena such as poverty, deprivation and social isolation (Sheldon 1948; Townsend 1957; Tunstall 1966). While paralleling the approach of classic studies, the new focus on exclusion encourages gerontologists to extend their gaze towards an examination of the complex linkages between different forms of disadvantage experienced by older people. Alongside concerns with poverty and loneliness, exclusion debates encompass issues such as older people’s participation in civic society, and access to services and amenities. In particular, the research presented here suggests that exclusion can be helpful when addressing the specific impacts on older people of growing spatial inequalities within society as a whole.

The discussion should, however, also be framed within the context of several limitations associated with the research. The initial description of the deprived areas sample of older people illustrates some of the key ways in which this group differs from the older population of Great Britain in general. While this confirms the need to explore aspects of the experience of ageing in different environmental settings, it also emphasises the limits to which it is possible to generalise findings from this study to the older population as a whole. Moreover, while the achieved response rate of 42% in relation to the purposive sample represents a good outcome in geographic areas sometimes neglected by social researchers on the grounds of concern about high population turnover and interviewer safety, it is not possible to comment on the characteristics of the 58% of non-respondents. Limitations also arise from the non-purposive approach used to generate a sample of older people belonging to minority ethnic groups. People recruited in this way may not be representative of the respective minority populations. These points should be borne in mind when interpreting and discussing the research findings.

From the conceptual discussion and the range of data presented, it is nevertheless possible to make a number of observations about the nature of social exclusion and the way in which it affects older people in deprived areas of England. The characteristics of this population group point to the potential experience of disadvantage. Compared with national samples of older people, those in disadvantaged neighbourhoods are more likely to have a lower socio-economic status as reflected in lower rates of owner occupation and higher proportions of people belonging to ethnic minority groups. The deprived areas sample also displayed higher rates of chronic ill health. The fact that the study sample had a relatively high average length of residence in the community points to the tendency of those in disadvantaged neighbourhoods to age in place. Such communities therefore represent unique environmental settings in which to address the experience of ageing.

In terms of the different dimensions of social exclusion addressed in this research, older people in deprived areas appear to be especially vulnerable to disadvantage arising from the nature of their social relations and the lack of access to material security. Such disadvantage tended to occur regardless of individuals’ socio-demographic characteristics. For example, while age and gender correlate closely with poverty in national populations (Arber and Ginn 1991, 1999), this pattern did not occur in the study reported here. A potential explanation for the absence of traditional gendered and age-related explanations of disadvantage may arise from the overriding impact of the study population’s low socio-economic status. While clearly meriting further research, the combination of low working-life incomes and long-term residence in a low status neighbourhood potentially evens out some of the anticipated influences of gender and age in later life. Although individuals’ ethnic origin did correlate with the different forms of social exclusion, the study revealed important variations between minority groups. Some caution should also be expressed when interpreting these data, given the relatively low sample sizes for each group. Somali and Pakistani elders were most likely to be excluded on the material resources, social relations and basic services domains, but were rather less likely to be affected by exclusion from civic activities and the neighbourhood. In relation to the latter dimensions, this reflects such groups’ regular participation in religious activities and a positive perception that the local neighbourhood is gradually adapting to meet the needs of these particular communities. Indian and Black Caribbean older people were least likely to be excluded on most indicators of exclusion. By contrast, white respondents were most disadvantaged in terms of the civic activities and neighbourhood dimensions of exclusion. This finding challenges the idea that minority groups experience disadvantage in the same ways, lending weight to the need for a more differentiated view of the ageing of minority ethnic groups, as highlighted in a number of recent studies (Butt and Moriarty 2004; Nazroo et al. 2004).

The research presented here also highlights the importance of paying attention to environmental influences on ageing. Where comparable data are available, it is evident that older people in deprived urban areas appear to be more vulnerable to the experience of different forms of social exclusion than are those living in the UK as a whole. This can be illustrated, for example, with reference to the risks of poverty, loneliness and neighbourhood insecurity. Using the same methodology, Gordon et al. (2000) identified 21% of older people as being poor. In this study, the equivalent proportion of those unable to afford two or more material items or to participate in common social activities due to a lack of income was 45%. Loneliness studies in Britain have consistently reported proportions of between 7 and 10% of older people being very lonely (Victor et al. 2002). While the methodology adopted in this study was somewhat different, the proportion of (very) severely lonely older people was higher, at 16%. Where similar methods have been used to assess loneliness in other countries, the rates are also significantly lower than that reported for deprived areas of England (De Jong Gierveld and Van Tilburg 1999). In relation to neighbourhood insecurity, identical questions were posed in the nationally representative British Crime Survey 2000 and the study presented here. While the national survey reported that 43% of people aged 60 and over would feel a bit or very unsafe when leaving their homes after dark (Chivite-Matthews and Maggs 2002), the equivalent proportion in deprived areas was 66%. It will be necessary to conduct similar studies in other types of geographic location in order to identify the full extent of the spatial divide which marks old age in advanced industrial societies.

Importantly, the research presented here suggests that significant proportions of older people in deprived areas face multiple risks of social exclusion. Using a specially developed measure to reflect the multi-dimensional nature of social exclusion, two-thirds of respondents were classed as excluded in relation to at least one aspect of their lives. For over one-third, the experience of exclusion in one area was compounded by vulnerability to additional types of exclusion. The risk of being affected by multiple forms of social exclusion was greatest for elders belonging to some minority ethnic groups and for those with a relatively low level of education, and was closely associated with poor health and a poor quality of life. The research also points to the existence of connections between the different domains of exclusion. In particular, there was a strong relationship between exclusion from social relations and exclusion from material resources. This tends to confirm the findings of earlier studies which emphasise the ways in which poverty and deprivation can combine to restrict participation in a range of informal social relationships (Townsend 1979; Mack and Lansley 1985; Gordon et al. 2000). These results may also represent the outcome of disadvantages experienced earlier in the life course. Analysis of qualitative data arising from the same study emphasises the impacts of relationship breakdown and the loss of close family members and breadwinners in generating exclusion from material resource and social relations in later life (Scharf et al. 2005). In terms of the neighbourhood dimension of exclusion, this was most closely associated with exclusion from social relations. It is not possible to disentangle cause and effect using the data presented here, and clearly there is scope for further analysis of these data. However, the difficulties associated with living in deprived urban environments—as reflected in area dissatisfaction and feelings of neighbourhood insecurity—may combine to produce social isolation, loneliness and an inability to participate in common social activities. The fact that different types of exclusion co-exist suggests an urgent need for the co-ordination of policy responses to promote the social inclusion of older people (Phillipson and Scharf 2004).

Finally, this research highlights the need to bring urban issues to the forefront of gerontological research. Although deprived urban neighbourhoods are not typical of urban areas as a whole, the argument remains that the future of old age will, to a large degree, be determined by the extent to which living in cities is made a tolerable and enjoyable experience. While cities can be disabling and threatening environments at any age, the research presented here demonstrates that the associated risks increase with age. The key point is that at 75 or 85, people may feel an even greater sense of being trapped or disadvantaged by urban decay, and that this may limit their ability to maintain a sense of self-identity. The multi-dimensional concept of social exclusion represents a useful way of exploring such area-based disadvantage in old age. There is, however, a need to extend the analysis undertaken here in relation to deprived neighbourhoods of England to a variety of other residential settings in order to identify whether similar patterns exist elsewhere. It would also be worthwhile developing research which explores the degree to which older people in rural areas may be prone to similar forms of disadvantage.