most striking trends in living arrangements over the recent past has been the growth .... In this article we use a set of household 'building blocks' based on the .... Dead. N/A. 27 (627). N/A. 27 (627). Not contacted. N/A. 13 (305). N/A. N/A. Total.
Population Trends 105
Autumn 2001
The dynamics of living arrangements in later life: evidence from the British Household Panel Survey Maria Evandrou, Jane Falkingham, Katherine Rake and Anne Scott ESRC Research Group ‘Simulating Social Policy for an Ageing Society’ (SAGE), London School of Economics and Kings College London
Living arrangements are a key dimension of quality of life and well-being in old age. Availability of family care, as well as social and economic support, is in part a function of whom you live with. In order to be able to forecast future changes in household composition, particularly in relation to planning and targeting particular community care services, information is needed on the probability of a person experiencing a change in their living arrangements, and the life course events that may act as triggers for such changes. This article uses data from the British Household Panel Survey (1991–99) to analyse the dynamics of living arrangements in later life.
I NTRODUCTION Living arrangements are a key dimension of quality of life and wellbeing in old age. Availability of family care, as well as social and economic support, is in part a function of who you live with. One of the most striking trends in living arrangements over the recent past has been the growth of single person households, which has been most marked amongst older people. In 2000 there were an estimated 3.6 million persons over state pension age living alone, compared to 1.1 million in 1961. 1 This has policy implications for both health and social care. Older people living alone are more likely to enter an institution than those living with other people; 2,3 and amongst those living in private households, older people living alone are more likely to be in receipt of social and community care services, such as home help and meals-on-wheels, than those living in other household types.4 In order to be able to forecast future changes in household composition, particularly in relation to planning and targeting community care services, information is needed on the probability of a person experiencing a change in their living arrangements, and the life course events that may act as triggers for such changes. To date, research examining living arrangements in later life has largely been cross-sectional.5 Such analysis does not capture the extent to which people may change their living arrangements to adapt to new circumstances over the life course, such as widowhood, changed financial circumstances or deterioration in health. The majority of research investigating the dynamics of changes in the household structure of older people has been from the United States.6,7,8 The Sample of Anonymised Records (SARs) from the 1991 Census has recently been used to investigate changes in living arrangements in later life, and the extent to which they are related to the migration of the older person or the movement of younger people into the homes of older people.9,10 Several studies of household change amongst elderly
37
National Statistics Statistics
Population Trends 105
Figure 1
Autumn 2001
Number of people above state pension age living alone, 1961–2000
HOUSEHOLD, FAMILIES AND LIVING ARRANGEMENTS ACROSS THE LIFE C OURSE Although most people have an intuitive sense of what is meant by the term ‘household’, attempts to formulate a general definition which encompasses non-traditional lifestyles have been less than satisfactory, either in a theoretical or practical form.15 In the 2001 Census of Population in Britain, a household is defined as one person living alone or a group of people living together who share common housekeeping (at least one meal together each day) or a living room.
Millions 4.0
3.5
3.0
In official statistics and survey data, a distinction is usually made between the household and the family. A family is defined as a group of persons closely related to each other by blood or marriage and living in the same household, such as a married couple with or without nevermarried children, or a lone parent with never-married children. With this definition, a family will always form a household or part of one, but a household does not need to include a family.
2.5
2.0
1.5
1.0
0.5
0.0 1961
1971
1981
1991
2000
Source: ONS (2001) Social Trends.
people have been carried out using the Longitudinal Study.11,12,13,14 These provide useful insights into changes in living arrangements between two points in time ten years apart, although they cannot inform us about changes that may have taken place during the intervening years. Analysis of longitudinal data from the Health and Lifestyles Surveys (HALS) examined changes in the household composition of older people between HALS1 (in 1984/85) and HALS2 (7 years later).3 The study concludes that over a seven year period there is ‘great diversity of experience and a considerable amount of change in what is often considered a fairly stable and homogeneous section of the population’. This paper investigates the dynamics of living arrangements in households containing persons aged 60 and over, reporting initial findings from research using data from the British Household Panel Survey (BHPS) (1991–99). It documents how living arrangements vary with age, gender and marital status, and the extent to which life course events act as triggers for changes in living arrangements. The paper builds upon the existing body of research by examining both the likelihood of experiencing a change in living arrangements across an eight year period and the year on year probability of experiencing a transition. This provides a more comprehensive picture of the mobility in living arrangements in later life than has hitherto been possible with other datasets. It is useful to first consider the dynamics of living arrangements theoretically, conceptualising who lives with whom and what factors may intervene to change this. The article therefore begins with a discussion of the definition of households and how the living arrangements of any one individual may change across the life course, before turning to examine changes in the living arrangements of older people over time.
National Statistics
38
Analysis of household composition and living arrangements is more complicated than that of families for two major reasons. Firstly, the number of household types possible may be unlimited; whereas the number of family types in developed countries such as Britain is usually confined to a dichotomy along two axes; couple (married or cohabiting)/single, and with child(ren)/no child. Secondly, for dynamic (or longitudinal) analysis, the family will usually have a well-defined starting point and finishing point. In contrast, the household will often not have a well-defined starting or finishing point, and questions such as “is this household at time A the same as the one which existed at time B?” will not have a unique answer. In order to analyse and disaggregate changes in household composition over time it is useful to identify a range of units or ‘building blocks’ that may combine together to form a household. In 1985 Ermisch and Overton put forward a four-fold classification of Minimal Household Units16 and the concept of such building blocks has been widely used since then. In this article we use a set of household ‘building blocks’ based on the most important interpersonal bonds in Western societies: sexual relationships and those between parent and dependent child. The living arrangements of older people are diverse. Dale and colleagues developed a comprehensive typology of household composition with 23 different classifications5. This paper, in order to investigate changes in household composition over time, uses a simpler five-fold typology, distinguishing household members who are aged 60 and over: 1. Single person aged 60 and over (‘single’ includes those never married, divorced, separated and widowed); 2. Couple only (married or co-habiting), with one or both aged 60 and over; 3. Single persons aged 60 and over living with at least one other adult under 60 (these may be with or without dependent children); 4. Couple 60 and over living with at least one other adult under 60 (these may also be with or without dependent children); and 5. Multiple adults aged 60 and over, i.e. two or more adults aged 60 and over (could include combinations of couples and single persons). The concept of the life course provides a useful theoretical framework for analysing the dynamics of living arrangements, as the type of household in which people live necessarily changes as people move through the life course.17 Figure 2 shows the main pathways between the different household types and their relationship to different life course events. Changes in living arrangements can occur as a result of a transition between any of these household types, or a move into an institution or death. The magnitude of these different transitions is investigated below.
Population Trends 105
Figure 2
The majority of complex households involved one or both parents living with one or more adult children, although some older people also live with other relatives and non-relatives. In 1991, 86 per cent of original sample members aged 60 and over living in Type 3 households (singles with under-60s) were living with an adult child, compared to 91 per cent of those in Type 4 households (couples with under 60s). Of those in multiple 60 plus households (Type 5), 13 per cent were living with an adult child. Adults co-residing with an elderly parent range in age from 16 to 73. Thus the type of intergenerational complex household ranged from fathers in their 60’s with children in their late teens, to elderly mothers living with their adult children who themselves were aged over 60.
Key changes in living arrangements in later life
Multiple 60+
Bereavement
> Single 60+
Bereavement
Couple 60+
>
>
Younger leave
Institution
By Wave 8 in 1998, 27 per cent of the sample (627 individuals) had died, after a mean of 4.0 years in the study. Contact was lost with a further 13 per cent (305 individuals), after a mean of 3.8 years. The remaining 60 per cent were followed for 7 years to Wave 8. The overall mean length of follow-up was 5.8 years. Of those contacted at Wave 8 (1,392 individuals), 41 per cent were now living alone and 48 per cent were residing with a spouse only.
Younger return or older move in with younger
>
>
Younger return or older move in with younger
Single 60+ & under 60
Bereavement
