Scholarly article on topic 'Pathways into living alone in mid-life: Diversity and policy implications'

Pathways into living alone in mid-life: Diversity and policy implications Academic research paper on "Sociology"

CC BY
0
0
Share paper
Academic journal
Advances in Life Course Research
OECD Field of science
Keywords
{"Living alone" / Mid-life / Pathways / "Policy implications" / "Baby-boom cohort" / "Partnership trajectory"}

Abstract of research paper on Sociology, author of scientific article — Dieter Demey, Ann Berrington, Maria Evandrou, Jane Falkingham

Abstract This paper adopts a life course approach to investigate the pathways into living alone in mid-life in Britain and how these vary by gender and socio-economic status. The rise in the proportion of people living alone over the past three decades has been well documented. However, much of the focus of the existing literature has been on either people living solo in young adulthood or in later life. Mid-life has received surprising little scholarly attention, despite the fact that living arrangements in mid-life are changing rapidly, and that household composition and socio-economic circumstances in the period immediately prior to retirement are strongly associated with living arrangements and associated sources of support in later life. This paper therefore aims to fill this gap. We begin with a review of previous research on living alone and present a conceptual framework of the pathways into living alone in mid-life. Data from the United Kingdom Household Longitudinal Survey (UKHLS) are used to analyse the partnership and parenthood histories and socio-economic characteristics of those currently living alone in mid-life. The findings indicate that the dissolution of a marriage with children is the dominant pathway into mid-life solo-living, but that there is also a substantial group of never partnered men living alone. These never partnered men are split between those with low and high socio-economic status. Distinguishing between different groups of individuals living alone in mid-life is important for policy as these groups of men and women will have different social and financial resources as they enter later life. Mid-life men living alone who have not had children, have no educational qualifications, are not economically active and who live in rented housing are likely to be most at risk of needing a social and economic ‘safety net’ in old age.

Academic research paper on topic "Pathways into living alone in mid-life: Diversity and policy implications"

Contents lists available at SciVerse ScienceDirect

Advances in Life Course Research

journal homepage: www.elsevier.com/locate/alcr

Advances in Life Course Research

Pathways into living alone in mid-life: Diversity and policy implications

Dieter Demeya'*, Ann Berringtona, Maria Evandrou b, Jane Falkinghama

a ESRC Centre for Population Change, University of Southampton, Highfield, SO17 1BJ Southampton, United Kingdom b ESRC Centre for Population Change, Centre for Research on Ageing, Social Sciences, University of Southampton, Highfield, SO17 1BJ Southampton, United Kingdom

CrossMark

ARTICLE INFO

ABSTRACT

Article history:

Received 19 September 2012 Received in revised form 15 January 2013 Accepted 10 February 2013

Keywords:

Living alone

Mid-life

Pathways

Policy implications

Baby-boom cohort

Partnership trajectory

This paper adopts a life course approach to investigate the pathways into living alone in mid-life in Britain and how these vary by gender and socio-economic status. The rise in the proportion of people living alone over the past three decades has been well documented. However, much of the focus of the existing literature has been on either people living solo in young adulthood or in later life. Mid-life has received surprising little scholarly attention, despite the fact that living arrangements in mid-life are changing rapidly, and that household composition and socio-economic circumstances in the period immediately prior to retirement are strongly associated with living arrangements and associated sources of support in later life. This paper therefore aims to fill this gap. We begin with a review of previous research on living alone and present a conceptual framework of the pathways into living alone in mid-life. Data from the United Kingdom Household Longitudinal Survey (UKHLS) are used to analyse the partnership and parenthood histories and socio-economic characteristics of those currently living alone in mid-life. The findings indicate that the dissolution of a marriage with children is the dominant pathway into mid-life solo-living, but that there is also a substantial group of never partnered men living alone. These never partnered men are split between those with low and high socioeconomic status. Distinguishing between different groups of individuals living alone in mid-life is important for policy as these groups of men and women will have different social and financial resources as they enter later life. Mid-life men living alone who have not had children, have no educational qualifications, are not economically active and who live in rented housing are likely to be most at risk of needing a social and economic 'safety net' in old age.

© 2013 Elsevier Ltd. All rights reserved.

1. Introduction

One of the most salient changes in family life across the industrialised world since the Second World War has been the steady rise in one-person households (Fokkema &

* Corresponding author. Tel.: +44 02380597988. E-mail addresses: D.Demey@soton.ac.uk (D. Demey), A.Berrington@soton.ac.uk (A. Berrington), Maria.Evandrou@soton.ac.uk (M. Evandrou), J.C.Falkingham@soton.ac.uk (J. Falkingham).

1040-2608/$ - see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.alcr.2013.02.001

Liefbroer, 2008; Goldscheider & Waite, 1993; Hall, Ogden, & Hill, 1997; Jamieson, Wasoff, & Simpson, 2009; Prioux, 2002; Wall, 1989). Living alone in Europe is particularly common among women in late-middle and old age following the death of a spouse (Prioux, 2002; Wall, 1989). However, previous studies have found that more men than women live on their own in early and middle adult life (Prioux, 2002; Wall, 1989). For instance, in Northern and Western Europe in 2008, at ages 30-49 around one fifth of men were living alone compared to one tenth of women, whereas at ages 50-69 slightly more

women than men were living alone (lacovou & Skew, 2011). Since the 1980s, there has been a rise in living alone across Europe in the young and middle age groups, especially among middle-aged men (Demey, Berrington, Evandrou, & Falkingham, 2011; Fokkema & Liefbroer, 2008; Prioux, 2002). At the same time, the proportion of women living alone in later life has decreased as a result of improvements in male life expectancy (Macunovich, Easterlin, Schaeffer, & Crimmins, 1995; Prioux, 2002; Tomassini, Glaser, Douglas, Broese van Groenou, & Grundy, 2004). As a consequence, while in the past a considerably larger number of women than men lived alone, men have closed the gap in recent years (Prioux, 2002).

The rise in living alone in mid-life over time in part reflects recent changes in demographic behaviours and in the pathways into solo-living. Demographic changes commonly associated with the so-called Second Demographic Transition (Lesthaeghe, 1995) - such as the delay of family formation, the decrease in marriage rates and the diffusion of cohabitation, rising divorce rates and the rising incidence of childlessness - have led to a diversification of life course trajectories over time, with more people living without a partner or co-resident children. The magnitude of this shift is further underlined by the size of the cohorts currently in mid-life in Britain, reflecting those men and women born during the baby-booms of the late 1940s and early 1960s. ln 1985 there were 20 million persons aged 35-64 in the United Kingdom; this rose by nearly a quarter to 24.7 million in 2010 (Office for National Statistics, 2011). The familial and economic resources of these mid-life men and women will be important determinants of future later life outcomes, such as living arrangements and care needs (Gaymu et al., 2006; Martikainen, Nihtila, & Moustgaard, 2008; Mutchler & Burr, 1991; Pendry, Barrett, & Victor, 1999; Tohme, Yount, Yassine, Shideed, & Sibai, 2011). It remains the case that the majority of social care in later life is provided by co-residential spouses or children (Pickard, Wittenberg, Comas-Herrera, King, & Malley, 2007). Marital disruption has been shown to result in an increased loss of support (Glaser, Tomassini, Racioppi, & Stuchbury, 2006) and receipt of formal social care services in later life have been shown to be disproportionately concentrated on those older people living alone (Evandrou & Falkingham, 2004). Thus understanding the demographic and socioeconomic composition of the currently middle-aged population is therefore important in its own right and is also a key element for policy makers both for ensuring appropriate services for this age group today and in planning the future provision of elderly care and housing as these groups enter old age.

Despite the rise in the prevalence of solo-living in mid-life, there has been little scholarly attention regarding the different pathways into living alone in this phase of the life course, and how these are in turn related to gender and socio-economic status, or on the policy implications of such a trend with regard to social and economic outcomes later in life. Furthermore, previous studies have mainly focussed on the legal marital status of those living alone, which is increasingly recognised as being unsuitable for assessing current partnership status as well as partnership history given the increases in cohabitation and

re-partnering as well as Living-Apart-Together (LAT) (Haskey & Lewis, 2006). This study aims to fill these gaps by investigating the partnership and parenthood trajectories of men and women currently living alone in mid-life in the UK and how these trajectories differ by socioeconomic status.

This study contributes to the literature on living alone in mid-life, adding value to previous research in a number of ways including: (i) by examining actual partnership status rather than legal marital status and taking cohabitation into account; (ii) by investigating the presence of non-residential children; (iii) by adopting a gender perspective and considering both men and women; and (iv) by stressing the policy implications of an increasingly heterogeneous population living alone in mid-life.

We address the following three sets of research questions:

1. What are the partnership and parenthood trajectories into living alone among those men and women currently in mid-life (aged 35-64)?

a. What proportion has never partnered, ever partnered and ever re-partnered?

b. What proportion has ever had children?

2. How do these vary across mid-life i.e. between individuals in early (35-44), mid (45-54) and late (55-64) mid-life?

3. How do the socio-economic characteristics of those living alone in mid-life compare with those living with a partner? And how do they vary according to the partnership trajectory into living alone?

To answer these research questions, we analyse data from a new, very large national survey carried out in the UK in 2009 and 2010 which provides retrospective information on partnership and parenthood trajectories with detailed current information about living arrangements, children living outside the household and socio-economic attributes. ln Section 2 we define what we mean by mid-life before reviewing the previous literature on living alone. ln Section 3 we discuss the different pathways into living alone in mid-life and their interplay with socioeconomic status and gender. ln Section 4, we describe the data sources, sample and measures whilst the main findings are presented in Section 5. ln Section 6 we conclude by summarising the main findings, drawing out the policy implications of familial and economic resources in mid-life for support and care needs in later life, and discussing the limitations of the study and opportunities for further research.

2. Previous research on living alone in mid-life

Mid-life or middle age is a phase in the life course which in the literature has commonly been situated between the end of the childbearing years and the onset of old age. Mid-life has been associated with several life course events, transitions and social roles particularly within family, employment and occupational trajectories, such as the growing up of children, the empty-nest period, or women's

return to work following childrearing for young children. The structuring of age can be formal, at the level of social structures and institutions, or informal, at the level of individuals, and may differ by gender, cohort, socioeconomic position, culture and over time (Settersten & Mayer, 1997). For instance, the official retirement age varies between countries and is lower for women than for men in some countries; similarly the average age at becoming a parent differs between cohorts and between educational and occupational categories. Conceptions of the timing of mid-life have been found to vary, among others, by gender, education and income in the United States (Toothman & Barrett, 2011). As a consequence, the boundaries of mid-life are difficult to establish and have been varyingly defined in empirical research depending upon the research questions or data availability. ln this study we use a broad age range to encapsulate different stages (early, mid and late) of mid-life. We purposefully include younger mid-life men and women who may be living alone as a consequence of either postponing or relinquishing partnership formation. Living alone in early mid-life is uncommon in the UK as compared to other Western and Nordic countries (lacovou & Skew, 2011) largely due to the relatively early age at entry into first partnership (Stone, Berrington, & Falkingham, 2011). Living alone in early mid-life is selective of both very highly educated people and those who are socio-economi-cally disadvantaged. We therefore use 35 as the younger age cut-off. Since the State Pension Age (SPA) in the UK for men is currently age 65, with the SPA for women currently in the process of being harmonised to this age we use this as our upper age limit. Retirement is traditionally seen as a stage of the life course associated with old age. Therefore, we focus on those aged 35-64, distinguishing those women and men in early mid-life (35-44), mid-life (4554) and late mid-life (55-64).

Much of our current understanding concerning the determinants and consequences of living alone is based on evidence from the later part of the life course. More elderly women than men live alone as a consequence of gender differences in the average age at marriage and life survivorship (Gaymu et al., 2006; lacovou & Skew, 2011; Prioux, 2002; Tohme et al., 2011; Wall, 1989), with more women than men making the transition into living alone following the institutionalisation or death of a partner. Research on living alone in later life in several countries shows that, among the non-institutionalised older population, living alone is associated with higher income, good health, being ever married and having children (Gaymu et al., 2006; Martikainen et al., 2008; Mutchler & Burr, 1991; Pendry et al., 1999; Tohme et al., 2011). This indicates that the capacity to live alone in old age is influenced by, among others, the ability to purchase professional services and the availability of adult children as these are one of the primary sources of informal support (Pickard et al., 2007). ln an extensive literature review on living arrangements and health in old age, Hays (2002) lists a number of studies which show that those living alone in later life have a higher use of home-based health-care and other services. For the UK, Glaser et al. (2006) found that there is a positive effect of the death of a spouse on using

domiciliary care services among the ever married population aged over 70, controlling for number of living children and socio-economic characteristics. A Swedish study found that never and ever married elderly adults living alone without children are more likely to use home-help services than the ever married with children, and are less likely to receive informal support (Larsson & Silverstein, 2004). Thus pathways into later life solo-living are to an important extent structured by the accumulation of (dis)advantage during the life course, as well as by family formation trajectories.

There are a limited number of studies in the UK and the US which have focussed on mid-life living arrangements in general and on living alone in particular. These have investigated the effect of rising income on the propensity to live alone (Pampel, 1983); transitions into and out of living alone and how these differ by gender, ethnicity, age groups, income and time period (Chandler, Williams, Maconachie, Collett, & Dodgeon, 2004; Richards, White, & Tsui, 1987); the influence of partnership status and transitions on employment patterns among middle-aged women (Austen & Ong, 2010; Moen, 1991); and the demographic and socio-economic characteristics of those living alone and how these have changed over time (Hall & Ogden, 2003; Hall et al., 1997; Jamieson et al., 2009). Evidence for the UK indicates that among those aged 20-59 living alone, more men than women have never married or are divorced, whereas more women than men are widowed (Hall et al., 1997). The larger proportion of men compared to women living alone in early mid-life has been explained in the literature by the fact that most dependent children remain with the mother after partnership breakdown, so men transition into living alone whereas women become single parents (Fokkema & Liefbroer, 2008; lacovou & Skew, 2011; Prioux, 2002). Since there has been a rise in partnership dissolution rates, this would also explain the sharp increase in the proportion of men living alone in early mid-life (Fokkema & Liefbroer, 2008). The gender gap in living alone narrows by age and by late middle age slightly more women than men are living alone due to gender differences in life expectancy (lacovou & Skew, 2011; Prioux, 2002).

Whereas higher economic resources, good health and the availability of kin are characteristic of those living alone in later life, there is some evidence from the UK that living alone in mid-life is associated with lower socioeconomic status in terms of higher unemployment rates and renting in the private and social sector as well as poor health (Hall et al., 1997). For the United States, Lin and Brown (2012) found that the socio-economic composition of the unmarried - who are not necessarily living alone -aged 45-63 varied by marital status and gender, with widowed women and never married men being most economically disadvantaged in terms of educational level, employment, income and health insurance. Furthermore, transitions into living alone in the US have been found to be most common among the young and older people while transitions from living alone to other living arrangements are more common in early middle age than in late middle age (Richards et al., 1987). This suggests that those who enter solitary living are likely to remain in this living

arrangement throughout mid-life. For instance, Chandler et al. (2004) found for England and Wales that 63 per cent of men and 74 per cent of women aged 35-44 living in a one-person household in 1981 were still living alone in 1991. These figures rose to 72 per cent of men and 79 per cent of women for those aged 45-54 in 1981. The study also showed that the most common household origin of those living alone in mid-life in 1991 compared to ten years earlier is a one-person household.

3. Pathways into living alone in mid-life

Understanding the different pathways into living alone in mid-life is not only important for understanding the composition of the population living alone in this age group, but also for projecting how it might change in the future as certain pathways become more or less dominant. Given recent demographic changes we might expect more single people with differential trajectories in this phase of the life course in the future. It is also important as partnership and parenthood status of those living alone in mid-life are likely to be important predictors of these states in later life.

As Fig. 1 illustrates, we define two main classes of trajectories into living alone in mid-life, namely never having experienced a co-residential union and ever having experienced a co-residential union which has dissolved. These pathways can then be further differentiated by parenthood status and, among the ever partnered, by dissolution type.

3.1. Never partnered

A first possible pathway into living alone in mid-life is to have never experienced a co-residential partnership. Kiernan's (1999) analysis of data from the Fertility and Family Surveys (FFS) shows large variation in the proportions never partnering within Europe as well as between men and women: among women aged 30-34, the proportion never partnered ranges from less than ten per cent in Northern and Western Europe to 17 per cent in Italy. Never partnering by age 30-34 is more common among men with estimates ranging from ten per cent or less in the Northern countries to 35 per cent in Italy. Those who have never partnered may consist of those who are delaying union formation, are unsuccessful in finding a partner, as well as those who have a preference for solo-living, although a very small minority regard remaining single as a desirable option (see for instance Thornton & Young-DeMarco, 2001).

Previous research shows that the experiences of delaying or relinquishing union formation differ between socio-economic groups. For instance, Ermisch (2008) shows for the UK that higher educated women are more likely to delay marriage than low educated women. Similar evidence has been found in other developed countries for both men and women (Heard, 2011). The latter study also found that marriage rates are higher among the higher educated by early mid-life, while men in the lower income groups are considerably less likely to have ever been married than men in the higher income groups. There is

Fig. 1. Pathways into living alone.

also strong evidence that men's socio-economic status influences both cohabitation and marriage. For instance, Kalmijn (2011), analysing data from the European Community Household Panel (ECHP), shows that the probability of entering cohabitation or marriage is the highest among men who are employed, have built up work experience, are in the higher income groups, have good health and are higher educated. Stone et al. (2011) have found that unemployed men in the UK are increasingly delaying family formation into their late thirties and forties and are likely to remain living alone. We hypothesise that the two most important pathways into living alone in mid-life among those who have never experienced a co-residential partnership are: first, the delay of partnership formation into early mid-life among those with relatively high socio-economic status (evident among those living alone in early mid-life (aged 35-44)) and second, persistent singlehood among those with relatively low socioeconomic status (evident among those in late mid-life (aged 55-64).1

Most children are born to parents who are living together and for this reason only a small minority of the never partnered will have non-residential children. However, the proportion of children born to parents who are not living together is higher in Britain than in most other European countries and has been estimated at 15 per cent among children born in 2000. This proportion has increased over time, and not all fathers start living together with the mother following the birth (Kiernan, 2006). Women who have a child outside of a co-residential union are substantially younger on average and have a lower educational level than women living with a partner (Kiernan, 2006). These children may leave the maternal home when the mothers are still relatively young, so a third pathway into living alone in mid-life among those who have never partnered, particularly for women, could be following the departure of children after the entry into single motherhood at a relatively young age (''empty-nest single parent''). At the same time, some never partnered men will be non-residential fathers (''non-residential parent with dependent children living elsewhere'').

3.2. Ever partnered

Since most people have ever experienced a co-residential union, we hypothesise that the most common pathway into living alone in mid-life is through partnership dissolution. This may be directly, for instance when moving out following a divorce and forming a single person household or after a partner dies, or indirectly, for instance after the children leave the parental home when having lived as a single parent for some period following a partnership dissolution. Trajectories into living alone among the ever partnered can be differentiated by the dissolution type and the presence of children.

1 We recognise that is not possible without partnership intentions data

to identify those who intentionally remain single from those who are

unable to find a partner despite intending to do so.

An important pathway into living alone is following separation or divorce. Previous research across Europe has not found a consistent relation between socio-economic status and dissolution risks (for a review see Lyngstad & Jalovaara, 2010). However, evidence from the UK suggests that dissolution from marriage (but not from cohabitation) is more common among those from poorer socio-economic backgrounds (see for instance Berrington & Diamond, 1999; Steele, Kallis, Goldstein, &Joshi, 2005). They will also be more likely to be parents at the time of dissolution since there is a positive relation between educational level and the incidence of childlessness (Kneale & Joshi, 2008). Among those who have children at the time of dissolution, men will be more likely than women to make the transition into solitary living since dependent children usually stay with the mother after separation (Fokkema & Liefbroer, 2008; lacovou & Skew, 2011; Prioux, 2002). ln 2011, women accounted for 92 per cent of lone parents with dependent children (Office for National Statistics, 2012a). Divorced and separated mothers may subsequently start living alone once their children have left home, resulting in a narrowing gender gap in the proportions living alone towards late mid-life.

Ever partnered men and women may also make the transition to living alone following the death of a partner. This is more common among women due to their greater longevity, although the gender gap in life expectancy is closing as a consequence of the faster pace of improvement in male life expectancy (see for instance Gjonca, Tomassini, Toson, & Smallwood, 2005). Since there are very few respondents, especially at the younger ages, which have experienced widowhood, we are not able to identify these as a separate group from those who have experienced divorce or separation.

4. Data and methods

4.1. Data

The analysis uses data from the United Kingdom Household Longitudinal Survey (UKHLS) also known as Understanding Society (University of Essex, lnstitute for Social and Economic Research, & National Centre for Social Research, 2011). This is a new longitudinal panel survey of more than 40,000 private households in the UK (McFall, 2011). The data are collected in face-to-face interviews with all household members aged 16 and over. We use data from the full first wave collected between January 2009 and January 2011. The unique feature of UKHLS is its large sample size, which enables us to study relatively small groups living alone in mid-life, as well as a wealth of information on retrospective partnership histories and current partnership status and other demographic and socio-economic characteristics. This allows us to investigate the partnership history and parenthood status of middle-aged adults living alone, and to compare the socio-economic characteristics of those currently partnered with those living alone in mid-life. We select men and women aged 3564 who completed a full interview in wave one. The analysis includes everyone with non-missing values on the variables included in the analysis. We exclude proxy respondents

because they did not complete the retrospective partnership history and because information on overtime work and pensions is lacking. The mid-life living alone sample consists of 1725 males and 1624 females, and the currently partnered sample consists of 8078 males and 9426 females. The household response rate for eligible households is 57.6 per cent and the individual full interview response rate among co-operating households is 81.8 per cent. Household-level nonresponse is slightly higher in areas with relatively high proportions of single person households and individual-level nonresponse is noticeably higher among singles (Lynn, Burton, Kaminska, Knies, & Nandi, 2012). The data are weighted with the individual-level full interview only weight, which adjusts for unequal selection probabilities, sampling error, household level nonresponse and within-household nonresponse. The individual weights post-stratify the sample to population estimates and sex, age and geographical region.

4.2. Measurement of variables in UKHLS

4.2.1. Living arrangements

We consider nine different living arrangements: living alone; living with a partner and (a) dependent child(ren)2; living with a partner and (an) independent child(ren); independent child living with both parents; living with a partner and without children; living without a partner and with (a) dependent child(ren); living without a partner and with (an) independent child(ren) only; independent child living with one parent; and 'other' living arrangements.3

4.2.2. Living alone

Our assessment of whether a person is living in a single person household is based on the number of people in the household reported in the household grid. This grid includes members absent from the household at the time of the interview such as children living in halls of residence and those who are normally part of the household but are temporarily living in institutional accommodation.

4.2.3. Living with a partner

A person is considered to be in a co-residential union if they are living together with a spouse, civil partner, or with a cohabiting partner (including those who spontaneously mentioned that they are in a same-sex couple).

4.2.4. Partnership trajectory

Adults were asked details of their past co-residential partnerships: the partnership type (cohabitation or marriage), the start and end dates, and type of partnership dissolution (cohabitation ceased, separation, divorce or death). We identify whether someone has never partnered, ever cohabited but never married, and ever married, and for the latter we make an additional distinction between those who never and ever cohabited. These refer to free-standing

2 Aged 16 or over.

3 Other living arrangements include for instance, lone parents living with their parents; those who are living with a partner and their parents; or those who are living with a partner, children and their parents.

episodes of cohabitation, i.e. cohabitations not followed by marriage. There is no data on LAT-partnerships in UKHLS.

4.2.5. Parenthood trajectory

Adults were also asked to indicate whether they have any living relatives outside the household, allowing us to distinguish whether a person has a non-residential child. For the latter, since we are also interested in non-resident parents, we make a further distinction between those who have at least one non-residential child aged under 16 and those who have at least one non-residential child aged 16 or over. Family ties refer to biological, adopted or foster-relationships and exclude step- and in-law relationships.

4.2.6. Socio-economic status (SES)

In the UK context education, housing tenure and economic activity are widely used indicators of SES (see for example Grundy & Holt, 2000; Hall & Ogden, 2003; Sefton, Evandrou, & Falkingham, 2011).4 We use these attributes together with an indicator of whether the respondent belongs to an occupational pension to examine the socio-economic status of those living with a partner and those living alone. Highest educational qualification5 is coded as: no qualifications, some qualifications, higher education; current economic activity has three categories: employed full-time (more than 30 h per week, including normal and overtime hours), employed part-time (30 h per week or less, including normal and overtime hours), not employed (mainly unemployed, retired, or long-term sick or disabled); housing tenure distinguishes between: owner-occupier (owned outright or with mortgage), social housing (local authority or housing association), private renting and other; occupational pension: yes (member of employer's pension scheme or receiving a pension from a previous employer, from a spouse's previous employer, or a private pension or annuity), no (self-employed, not eligible for employer's pension scheme, not a member of employer's pension scheme, not receiving a pension from a previous employer, from a spouse's previous employer, or a private pension or annuity), unknown (missing).6

4.3. A comment on age-period-cohort effects

The main analytical sample used here are men and women currently aged 35-64 in 2009-10. These individuals were born between the mid-1940s and mid-1970s.

4 Sefton et al.'s (2011) study of older British women's personal incomes using the British Household Panel Survey (BHPS) shows for instance that women who worked full-time have higher personal incomes than women who worked part-time or were predominantly inactive. These differences in personal income are strongly related to differences in occupational pension income. A long full-time career also matters more for highly qualified women because more of them are in receipt of a private pension.

5 Some qualifications include GCSE-level and equivalent qualifications, A-level and equivalent qualifications, higher non-degree qualifications and other qualifications, while higher education includes those with degrees.

6 The questionnaire only contained questions about membership of an employer's pension scheme of the current employer. As a consequence, occupational pension membership for those who were active on the labour market but are unemployed is unknown.

Table 1

Living arrangements in mid-life, by ten-year age groups (35-64) and gender (column percentages).

35-44 45-54 55-64

Males Females Males Females Males Females

Living alone 11 6 15 10 15 19

Other 7 5 6 6 6 7

Living with a partner and (a) dependent child(ren) 56 54 32 21 5 1

Living with a partner and (an) independent child(ren) only 4 5 21 24 19 15

lndependent child living with both parents 2 0 1 0 0 0

Living with a partner and without children 17 10 20 22 51 53

Living without a partner and with (a) dependent child(ren) 1 15 1 6 0 0

Living without a partner and with (an) independent child(ren) only 1 3 2 9 2 5

lndependent child living with one parent 2 1 2 1 1 1

Unweighted N 3940 5189 3468 4584 3088 3760

Source: UKHLS (2009-10).

Notes: weighted percentages, estimates may not add up to 100 per cent due to rounding.

Observed differences between those in early (35-44) mid (45-54) and late (55-64) mid-life in the proportions who have never partnered, ever cohabited, ever married or ever re-partnered may be driven by age, period or cohort effects. For instance, the proportion that has never experienced a co-residential partnership may decrease with age. This may simply be because those in the older age groups have had more time to find a partner (age effect), or it may indicate a greater acceptance of persistent single-hood in the younger age groups compared to the older age groups (cohort effect), or the recent economic recession may have disproportionately influenced the income situation of those living alone, which in turn may have delayed moving in with a partner (period effect). lt is not possible to disentangle these age-period-cohort effects in a cross-sectional analysis such as this, and we will therefore mainly focus on differences within age groups rather than between age groups. We will return to this issue in the discussion.

5. Results

5.1. Living arrangements in mid-life

Our estimates of the proportions living alone by age in the UK (Table 1) are comparable to those reported in other surveys (lacovou & Skew, 2011; Office for National Statistics, 2012b). After living with a partner, living alone is currently the second most common living arrangement in mid-life in the UK, with the prevalence of solo-living being lower than that of the Nordic countries but higher than the prevalence in southern Europe. The only exception to this can be found in the group of women in early mid-life (aged 35-44) where the second most common living arrangement is as a lone mother with at least one dependent child. More men than women are living alone in early mid-life and mid-life, while more women than men are living alone in late mid-life (p < 0.01). This is also the case in most other European countries (lacovou & Skew, 2011). Other groups who are not currently living alone but who may thought to be at risk of living alone at later ages, such as a lone parent with an independent child or independent adult (i.e. middle aged) children living with at least one parent, are very

small. The remainder of the paper therefore focusses on those living alone, comparing their socio-economic characteristics with those mid-lifers living with a partner.

5.2. Partnership and parenthood trajectories of those living alone in mid-life

Table 2 shows the partnership history and parenthood status of middle-aged men and women living alone in 2009-10 by ten-year age groups. Within a particular age group, differences between the proportions for men and for women that are statistically significant (p < 0.05) are shown in bold. The top panel of the table shows that, at ages 35-44, one third of those living alone have never been in a co-residential partnership, and, among those who have ever partnered, the majority have ever cohabited but have never been married. Among the ever married a significant minority have cohabited with someone at some point across their life course.7 These findings illustrate that the partnership histories of those living alone in early mid-life are diverse, and also that this diversity would not be fully captured by focussing on current legal marital status alone. For instance, three quarters of those living solo aged 35-44 are never married, but most have ever experienced a co-residential partnership at some stage. Our analyses also show that the partnership histories of solo-living men and women in this particular age group (35-44) are very similar.

ln the 45-54 age group, more men than women living alone have never partnered (25 versus 19 per cent) or have ever cohabited but have never married (28 versus 16 per cent), while substantially more women than men have ever been married (65 versus 47 per cent). Among the ever married, two-thirds of men and four-fifths of women have never cohabited. One of the reasons for the latter gender differences is that men are more likely than women to repartner, and these are usually cohabiting unions.

7 It is important to note that for those who have ever been married and ever experienced a cohabitational episode, this cohabitation can have occurred before or after the marriage. In other words, they have cohabited, dissolved the cohabitation, and then married; they have re-partnered after marital dissolution; or a combination of both.

Table 2

Partnership history and parenthood status of those living alone, by ten-year age groups (35-64) and gender (column percentages).

35-44 45-54 55-64

Males Females Males Females Males Females

Partnership history

Never partnered 32 32 25 19 24 12

Ever cohabited & never married 42 40 28 16 13 5

Ever married 26 27 47 65 63 84

Ever married & never cohabited 16 16 33 52 51 69

Ever married & ever cohabited 10 11 14 13 12 14

Partnership history and parenthood status

Never partnered, no children 30 30 23 18 24 11

Never partnered, child(ren) 2 2 1 2 0 0

Ever partnered & never married, no children 28 34 21 11 10 3

Ever partnered & never married, child(ren) (at least one under 16) 11 2 4 0 0 0

Ever partnered & never married, child(ren) (none under 16) 3 4 4 5 2 2

Ever married, no children 9 17 9 19 13 12

Ever married, child(ren) (at least one under 16) 14 4 12 2 3 1

Ever married, child(ren) (none under 16) 3 7 25 44 47 70

Unweighted N 535 331 619 504 571 789

Source: UKHLS (2009-10).

Notes: weighted percentages. Within each age group, statistically significant differences (p < 0.05) between men and women are highlighted in bold.

In late middle age (age 55-64), where living alone is more common among women than among men (see Table 1), gender differences in partnership histories are most marked. Twice as many men as women have never partnered (24 versus 12 per cent respectively) or have ever cohabited but have never been married (13 versus 5 per cent respectively), while more than eight out of ten women have ever been married, compared to six out of ten men. Of those who had ever married, the majority of both men and women (around four-fifths) had never experienced a freestanding cohabitation (i.e. not followed by marriage).

The bottom panel of Table 2 shows the combined partnership and parenthood trajectories of those currently living alone by age and sex. Being a parent is defined as having at least one non-residential child and we make a distinction between those with at least one non-residential child aged under 16 and those with one or more non-residential children all aged over 16 (except for those who have never been in a co-residential union as very few have non-residential children). Those with independent nonresident children may have previously been a single parent or may have moved out from the household in which their child(ren) lived, for instance following partnership dissolution.

Focussing on the presence of non-residential children first, we see that, at ages 35-44, one third of men living alone have at least one non-residential child, of which most have at least one non-residential child aged under 16. In contrast, less than one fifth of women living alone in this age group have a non-residential child and very few have a non-residential child aged under 16. These findings indicate that men and women living alone in early mid-life are predominantly childless: this is especially the case for women, reflecting the fact that those women who have had children are more likely to be still living with them. Those women who have had children and are now living alone probably entered motherhood at a relatively young age. At age 45-54, one half of men and women living alone have non-residential children. Of those who have at least

one non-residential child, one third of men have at least one aged under 16 but very few women have. This suggests that these women make the transition into living alone once their children leave the maternal home. In the 55-64 age group, the proportion without non-residential children is almost twice as high for men than for women.

Second, there are substantial gender differences in parenthood status by partnership history which are indicative of different pathways into living alone in mid-life between men and women. In general, it is very uncommon to never have been in a co-residential union and to have a non-residential child. Furthermore, at ages 35-44, more men than women living alone who have ever been in a co-residential union have non-residential children, mainly young children. In particular, substantially more ever married women than men do not have children. At ages 45-54, more ever partnered but never married men than women do not have children, whereas more ever married women than men have no children. In the same age group, a substantially larger proportion of ever married men than women have at least one non-residential child aged under 16, while more ever married women than men have non-residential children aged over 16. The latter difference is even more marked in late mid-life.

5.3. Socio-economic status

In this section we compare the socio-economic status of those mid-lifers living with a partner and those living alone, and examine heterogeneity among those living alone according to whether they have ever partnered. Tables 3 and 4 show the percentage distribution by educational level, current economic activity, housing tenure and occupational pension status of middle-aged men and women living with a partner or alone at the time of the survey. Figures in bold indicate that the difference between those living with a partner or alone is statistically significant (p < 0.05).

Table 3

Socio-economic status of males living with a partner and males living alone, by ten-year age groups (35-64) (column percentages).

35-44 45-54 55-64

Partner Alone Partner Alone Partner Alone

Educational level

Higher education 30 27 26 23 23 19

Some qualifications 59 58 58 58 48 46

No qualifications 10 15 16 19 29 35

Current economic activity

Employed full-time 86 67 81 63 55 37

Employed part-time 5 4 5 6 10 10

Not employed 10 29 14 31 35 53

Housing tenure

Owner-occupier 77 50 83 54 86 53

Social housing 10 24 10 25 9 32

Rented 14 26 7 21 5 15

Occupational pension

Yes 44 34 46 35 52 38

No 46 37 41 36 32 28

Unknown 10 29 13 30 17 35

Unweighted N 3108 535 2608 619 2362 571

Source: UKHLS (2009-10).

Notes: weighted percentages. Within each age group, statistically significant differences (p < 0.05) between those living with a partner and those who are living alone are highlighted in bold.

Table 4

Socio-economic status of females living with a partner and females living alone, by ten-year age groups (35-64) (column percentages).

35-44 45-54 55-64

Partner Alone Partner Alone Partner Alone

Educational level

Higher education 30 43 22 24 15 15

Some qualifications 63 46 64 55 51 51

No qualifications 8 11 14 22 33 34

Current economic activity

Employed full-time 42 69 48 57 24 30

Employed part-time 33 9 31 11 26 14

Not employed 24 22 21 32 50 56

Housing tenure

Owner-occupier 79 54 84 55 87 64

Social housing 11 23 10 31 9 27

Rented 10 23 6 15 4 8

Occupational pension

Yes 39 46 45 40 46 53

No 36 30 35 28 41 34

Unknown 25 23 20 32 13 13

Unweighted N 3673 331 3127 504 2626 789

Source: UKHLS (2009-10).

Notes: weighted percentages. Within each age group, statistically significant differences (p < 0.05) between those living with a partner and those who are living alone are highlighted in bold.

Compared to those living with a partner, middle-aged men and women living alone are generally more likely to have no qualifications, to be not employed, to be in social housing or privately rented housing, and among men are less likely to be a member of an employer's pension scheme or receiving an occupational pension. Looking at the group of solo-living men and women in late mid-life, aged 55-64, i.e. those who are closest to entering later life and thus who are most at risk of needing economic and social support in the relatively near future, we can see that one third have no qualifications, over one half are not employed, and almost a third live in social housing. Just over one third of men living alone in late mid-life are

currently contributing to or receiving an occupational pension, compared to one half of women. This may reflect the fact that some widows receive a pension from their deceased spouse's previous employer.

Differences in socio-economic status between those living with a partner and living alone are relatively consistent across the age groups for men, but this is not the case for women. More solo-living women in early mid-life (aged 35-44) are higher educated, are working fulltime and are a member of their employer's pension scheme than partnered women. ln contrast, a considerably larger proportion of partnered women in this age group are working part-time than solo-living women (33 versus 9

Table 5

Socio-economic status of never and ever partnered males living alone, by ten-year age groups (35-64) (column percentages).

Educational level

Higher education 34

Some qualifications 49

No qualifications 17 Current economic activity

Employed full-time 68

Employed part-time 4

Not employed 27 Housing tenure

Owner-occupier 53

Social housing 27

Rented 21 Occupational pension

Yes 41

Unknown 27

Unweighted N 172

67 4 29

49 23 28

34 29 36

23 60 18

52 25 23

35 38 28

15 45 41

27 11 62

60 35 5

21 46 33

51 31 18

Source: UKHLS (2009-10).

Notes: weighted percentages. Within each age group, statistically significant differences (p < 0.05) between those who have never and ever partnered are highlighted in bold.

Table 6

Socio-economic status of never and ever partnered females living alone, by ten-year age groups (35-64) (column percentages).

35-44 45-54 55-64

Never Ever Never Ever Never Ever

Educational level Higher education Some qualifications No qualifications

Current economic activity Employed full-time Employed part-time Not employed Housing tenure Owner-occupier Social housing Rented Occupational pension Yes No

Unknown Unweighted N Source: UKHLS (2009-10).

Notes: weighted percentages. Within each age group, statistically significant differences (p < 0.05) between those who have never and ever partnered are highlighted in bold.

44 42 44

46 47 38

11 11 18

72 67 57

8 9 11

20 24 32

60 51 67

21 24 24 19 25 9

52 44 52

28 32 17

21 24 32

111 220 102

19 26 14

58 57 51

23 17 36

57 30 30

11 12 15

33 59 56

52 82 62

32 17 29

16 2 9

37 61 52

31 23 36

32 16 13

402 88 701

per cent respectively). These differences level off throughout mid-life, which suggests that they are driven by the presence (or absence) of children, which influence women's current and future employment patterns, and of a partner. In late mid-life, a considerably larger proportion of partnered women are working part-time, so that, overall, more partnered women than women living alone are employed in this age group.

Tables 5 and 6 compare the socio-economic characteristics of never and ever partnered middle-aged men and women living alone. Figures in bold indicate that the difference between the never and ever partnered is statistically significant (p < 0.05). We can differentiate

between two groups of men living alone in mid-life: the younger mid-life 'eligible bachelors' and the older mid-life 'loners'. On the one hand, among men aged 35-44, a higher proportion of never partnered compared to ever partnered men are higher educated (34 per cent versus 23 per cent). Fewer never partnered than ever partnered men aged 4564 are working full-time. For instance, less than three out of ten never partnered men aged 55-64 are working fulltime, compared to four out of ten ever partnered men.

Differences in socio-economic status between never and ever partnered middle-aged women are more marked than among men. In particular, a considerably higher proportion of never partnered solo-living women aged

Table 7

Mix of familial and economic resources among males and females aged 55-64 living alone (column percentages).

Males Females

Never partnered No children Not owner-occupier 10 2

Never partnered No children Owner-occupier 15 9

Never partnered Children Not owner-occupier 0 0

Never partnered Children Owner-occupier 0 0

Ever partnered No children Not owner-occupier 13 4

Ever partnered No children Owner-occupier 11 11

Ever partnered Children Not owner-occupier 24 29

Ever partnered Children Owner-occupier 28 44

Unweighted N 571 789

Source: UKHLS (2009-10). Notes: weighted percentages.

45-64 are higher educated, are owner-occupiers, and have an occupational pension than ever partnered solo-living women in this age group. This is in sharp contrast with the socio-economic characteristics of middle-aged men living alone: the findings thus suggest that never partnered men living alone in late mid-life are considerably more economically disadvantaged than women.

5.4. Mix of familial and economic resources in later mid-life

ln the final part of the analysis we examine the mix of familial and economic resources among men and women living alone in late mid-life (ages 55-64) in order to foreshadow their likely circumstances in old age. We focus on the 'presence' (i.e. existence) of children and whether a person is an owner-occupier as these are important indicators of future economic and social resources and the associated 'ability' to meet individuals' care needs in later life, and take into account partnership history as this is related to parenthood and housing trajectories. Previous research has demonstrated that informal care in later life is primarily provided by spouses or partners or adult children (Pickard et al., 2007). Table 7 shows the proportions of current 55-64 year old solo-living men and women who have (n)ever partnered, have (no) children, are (not) owner-occupier and combinations of these factors. Ten per cent of men aged 55-64 living alone have never experienced a co-residential partnership, have no children and are not an owner-occupier. A further 13 per cent are 'disadvantaged' in two domains, i.e. although they were ever partnered, they do not have children and are not owner-occupiers. Thus, over a fifth of the current cohort of men who are living alone in the ten years prior to State Pension Age, risk entering later life without significant family and economic resources. The picture among solo-living women in this age group is different as very few of these women have never partnered, have no children and are not owner-occupiers (just 2 per cent). The group that may be more 'at risk' of being poorly resourced as they enter later life are those women who have partnered and have had children, but whose partnership dissolved and who have no housing equity; nearly one in three (29 per cent) of solo-living women in late mid-life (aged 55-64) fall into this category.

6. Discussion, policy implications and future research

Living alone is currently the second most common form of living arrangement in mid-life in the UK. ln this study, we have argued that it is important to consider the heterogeneity of those living alone in mid-life. People experience different partnership and parenthood trajectories into living alone in mid-life, and these trajectories interact in complex ways with educational and employment careers, gender and social context. Using new data from UKHLS, we have provided important insights into the partnership history, parenthood status and socio-economic characteristics of middle-aged men and women living alone in the UK. The analysis goes beyond previous research in that we have been able to consider past partnership trajectories that include both legal marriages and cohabitations as well as children living outside the immediate household and provides a number of important results. First, partnership dissolution is the main partnership trajectory into living alone in mid-life, although a non-negligible proportion of men have never experienced a co-residential partnership. This is in turn reflected in the second main finding, namely that in late mid-life substantially more solo-living women than men have non-residential children. Third, those living alone in mid-life have relatively lower socio-economic status than those living with a partner; this is especially the case for never partnered men in late mid-life. Taking these findings together there appear to be two distinct groups who are lacking both familial and socio-economic resources: men living alone in late mid-life who do not have, and have never had, a partner or children and are not owner-occupiers and older mothers who have experienced partnership dissolution and who are not owner-occupiers.

The analysis of the retrospective partnership histories and of parenthood status clearly indicates that pathways into living alone in mid-life are diverse, and differ between age groups and by gender. Here we summarise the main findings by relating the results of the analysis to the conceptual diagram in Fig. 1, structuring the discussion by the three age groups (early mid-life, mid-life, and late midlife). ln early mid-life, many solo-living men and women have never lived together with a partner and these are either delaying or foregoing partnership formation. Still, most of those living alone in early mid-life have had a

partner, and among these, the majority do not have children. However, there is also a substantial proportion of 35- to 44-year-old men living on their own who are fathers, and in most cases, fathers of dependent children, which is not the case among women. Thus, for men, three common pathways into living alone in early mid-life could be identified: never partnering; ever partnering, no children and partnership dissolution; and ever partnering, children and partnership dissolution coupled with moving out of the household where the children are present. In contrast, for women only the first two pathways are common trajectories into living alone in early mid-life. Among those in their late forties and early fifties, three distinct pathways into living alone are observed. The first relates to childless (mainly men) who may have cohabited but are unlikely to have married. The second relates to mainly childless women who have experienced marital dissolution. The third, and most prevalent trajectory, is to have experienced the dissolution of a married couple family with children. In late mid-life, two main trajectories into living alone are observed. The first, more common among older men living alone, is to have never married (the majority of whom have never partnered). The second, and most common pathway, especially for older women relates to the dissolution of a married couple family with children. Thus men living alone in late mid-life are far more likely to be childless than women.

Since there are cross-national differences in the diffusion of cohabitation, dissolution rates and the timing of family formation, some of the pathways into living alone in mid-life found for the UK may be more or less common in other countries. For instance, in countries with lower levels of cohabitation (e.g. Poland, Spain and Italy), we might expect two pathways into living alone in early mid-life to be dominant, namely either never partnering or marital dissolution. Differences in the timing of family formation and dissolution will also affect which pathways are common in early mid-life, mid-life and late mid-life. Furthermore, as there are more lone parents, in particular mothers, in Britain than in other European countries (Iacovou & Skew, 2011), we might expect that it is less common in other countries to enter solo-living when children leave the parental home.

These findings have important implications for policy. Reports of fair or poor health and some disability in early old age are higher among the unmarried (Grundy & Holt, 2000). This, taken together with trends in partnership trajectories and kin availability, suggest that the demand for public care for those living alone in later life is likely to rise in the future, in particular to meet the care needs of solo-living men. The long term care system in the UK is reliant on unpaid or informal care provided by families and friends (Hancock et al., 2012). Our findings indicate that more than one fifth of men living alone in late mid-life will not be able to rely on children for informal support and might not have sufficient financial resources to purchase home-based health-care, as suggested by their housing tenure status which is strongly related to wealth. Furthermore, previous research has shown that those who are not home owners face a higher risk of admission to a care home (McCann, Grundy, & O'Reilly, 2012).

Among those with low economic resources but who do (potentially) have children, demand for public care might be higher among solo-living men than women. There is some evidence which suggests that divorced parents receive less support from their adult children than married parents, and in particular divorced fathers compared to divorced mothers (see for instance Kalmijn, 2007). However, contrasting findings have been reported for the UK by Glaser, Stuchbury, Tomassini, and Askham (2008). Furthermore, even among those who do have children, the availability or willingness of adult children to provide care may decrease in the future as a consequence of the decline in the average family size, the decrease in multigenerational co-residence and the increase in middle-aged women's participation to the labour market (Pickard et al., 2007). This will affect both men and women with children who are living alone.

Previous research has highlighted the importance of life course family and labour market experiences for incomes in later life (Dewilde, 2012), with women who have had time out of the labour market to care for children facing both a wage penalty on re-entry to the labour market (Sigle-Rushton & Waldfogel, 2007) and a pension penalty upon retirement (Evandrou & Glaser 2003; Sefton et al., 2011). Women who have partnered and had children but who have then experienced partnership breakdown may be furthered disadvantaged through the loss of a partner's pension, although this may be ameliorated through institutional structures including both the legal system and public transfers (Uunk, 2004). Thus women living solo in mid-life with no housing wealth, and who have had interrupted labour market histories as a result of having children, may be at risk of entering later life with low individual pension entitlements, and face the risk of a low resourced old age.

It is important to bear in mind that middle-aged men and women who are living alone may be in a relationship with a person not living in the household. Previous research has shown that, in Britain, a substantial proportion of those not in a co-residential relationship are LAT, ranging from about one third among 35- to 39-year olds to about one fifth among 50- to 59-year olds (Haskey, 2005). These non-residential partners could be an important source of support for those living alone in old age, but unfortunately there was no question on LAT-relationships in UKHLS.

It is also important to note that this study has examined the characteristics of those living alone in mid-life at a particular point in time using cross-sectional data and retrospective data to characterise their current position. As a consequence, individuals with short durations in solo-living may have been underrepresented, in favour of those groups with longer durations in living alone. Our understanding of the population living alone in mid-life could therefore be improved by using longitudinal data to study differences in the duration and incidence of living alone. Finally, as more waves of UKHLS become available, research examining the extent to which people living alone in mid-life are likely to go on to live alone in old age will provide new insights into the characteristics associated with persistent solo-living, its consequences and implications for policy. Mid-life has been a hitherto 'Cinderella'

phase of the life course; it is hoped that this study will stimulate further research on this important phase of the life course.

Acknowledgements

This study was carried out at the ESRC Centre for Population Change (CPC), a joint initiative between the Universities of Southampton, St. Andrews, Dundee, Edinburgh, Stirling and Strathclyde, in partnership with the Office for National Statistics (ONS) and the General Register Office Scotland (GROS) (now the National Records of Scotland, NRS). The Centre is funded by the Economic and Social Research Council (ESRC), grant no. RES-625-28-0001. The findings, interpretations, and conclusions expressed in this study are entirely those of the authors and should not be attributed in any manner to ONS or GROS/NRS. The United Kingdom Household Longitudinal Survey (UKHLS) is conducted by the lnstitute for Social and Economic Research (lSER) at the University of Essex. UKHLS data were accessed via the UK Data Archive. Neither the original data creators, depositors or funders bear responsibility for the further analysis or interpretation of the data presented in this study. The authors would like to thank two anonymous reviewers for their constructive comments on an earlier version of this paper.

References

Austen, S., & Ong, R. (2010). The employment transitions of mid-life women:

Health and care effects. Ageing and Society, 30, 207-227. Berrington, A., & Diamond, l. (1999). Marital dissolution among the 1958 British birth cohort: The role of cohabitation. Population Studies, 53, 19-38.

Chandler, J., Williams, M., Maconachie, M., Collett, T., & Dodgeon, B. (2004). Living alone: lts place in household formation and change. Sociological Research, 9.

Demey, D., Berrington, A., Evandrou, M., & Falkingham, J. (2011). The changing demography of mid-life, from the 1980 to the 2000. Population Trends, 145,16-34. Dewilde, C. (2012). Lifecourse determinants and incomes in retirement: Belgium and the United Kingdom compared. Ageing and Society, 32, 587-615.

Ermisch, J. (2008). The new dynamics of family formation and the explosion of childbearing outside marriage. ln J. Scott, S. Dex, & H. Joshi (Eds.), Women and employment. Changing lives and new challenges (pp. 133155). Cheltenham: Edward Elgar. Evandrou, M., & Falkingham, J. (2004) Demographic change in Europe: lmplications for future family support of older people. ln P. Kreager & E. Schroder-Butterfill (Eds.), Ageing without children (pp. 175-197). Oxford: Berghahn Books. Evandrou, M., & Glaser, K. (2003). Combining work and family life: The

pension penalty of caring. Ageing and Society, 23(5), 583-602. Fokkema, T., & Liefbroer, A. C. (2008). Trends in living arrangements in Europe:

Convergence or divergence? Demographic Research, 19,1351-1418. Gaymu, J., Delbes, C., Springer, S., Binet, A., Deesesquelles, A., Kalogirou, S., et al. (2006). Determinants of the living arrangements of older people in Europe. European Journal of Population, 22, 241-262. Gjonca, A., Tomassini, C., Toson, B., & Smallwood, S. (2005). Sex differences in mortality, a comparison of the United Kingdom and other developed countries. Health Statistics Quarterly, 26, 6-16. Glaser, K., Stuchbury, R., Tomassini, C., & Askham, J. (2008). The long-term consequences of partnership dissolution for support in later life in the United Kingdom. Ageing and Society, 28, 329-351. Glaser, K., Tomassini, C., Racioppi, F., & Stuchbury, R. (2006). Marital disruptions and loss of support in later life: A longitudinal study of the United Kingdom. European Journal ofAgeing, 3, 207-216. Goldscheider, F. K., & Waite, L. J. (1993). New Families, no families? The transformation of the American home Berkeley and Los Angeles: University of California Press.

Grundy, E., & Holt, G. (2000). Adult life experiences and health in early old age in Great Britain. Social Science and Medicine, 51,1061-1074.

Hall, R., & Ogden, P. E. (2003). The rise of living alone in lnner London: Trends among the population of working age. Environment and Planning A, 35, 871-888.

Hall, R., Ogden, P. E., & Hill, C. (1997). The pattern and structure of one-person households in England and Wales and France. International Journal of Population Geography, 3,161-181.

Hancock R., Malley, J., Wittenberg, R., Morciano, M., Pickard, L., King, D., et al. (2012, September). The role of care home fees in the public costs and distributional effects of potential reforms to care home funding for older people in England. Health Economics, Policy and Law, 1-27, FirstView Article.

Haskey, J., & Lewis, J. (2006). Living-apart-together in Britain: Context and meaning. International Journal of Law in Context, 2, 37-48.

Hays, J. C. (2002). Living arrangements and health status in later life: A review of recent literature. Public Health Nursing, 19,136-151.

Haskey, J. (2005). Living arrangements in contemporary Britain: Having a partner who usually lives elsewhere and Living Apart Together (LAT). Population Trends, 122, 35-45.

Heard, G. (2011). Socioeconomic marriage differentials in Australia and New Zealand. Population and Development Review, 37,125-160.

lacovou, M., & Skew, A. J. (2011). Household composition across the new Europe: Where do the new Member States fit in? Demographic Research, 25, 465-490.

Jamieson, L., Wasoff, F., & Simpson, R. (2009). Solo-living, demographic and family Change: The need to know more about men. Sociological Re-search14.

Kalmijn, M. (2007). Gender differences in the effects of divorce, widowhood and remarriage on intergenerational support: Does marriage protect fathers? Social Forces, 85,1079-1104.

Kalmijn, M. (2011). The influence of men's income and employment on marriage and cohabitation: Testing Oppenheimer's theory in Europe. European Journal of Population, 27, 269-293.

Kiernan, K. (1999). Cohabitation in Western Europe. Population Trends, 96, 25-32.

Kiernan, K. (2006). Non-residential fatherhood and child involvement: Evidence from the Millennium Cohort Study. Journal of Social Policy, 35, 651-669.

Kneale, D., & Joshi, H. (2008). Postponement and childlessness - Evidence from two British Cohorts. Demographic Research, 19,1935-1968.

Larsson, K., & Silverstein, M. (2004). The effects of marital and parental status on informal support and service utilization: A study of older Swedes living alone. Journal of Aging Studies, 18, 231-244.

Lesthaeghe, R. (1995). The second demographic transition in Western countries: An interpretation. ln K. O. Mason A.-M. Jensen (Eds.), Gender and family change in industrialized countries (pp. 17-62). Oxford: Clarendon Press.

Lin, l.-F., & Brown, S. L. (2012). Unmarried boomers confront old age: A national portrait. The Gerontologist, 52,153-165.

Lyngstad, T. H., &Jalovaara, M. (2010). A review of the antecedents of union dissolution. Demographic Research, 23, 257-292.

Lynn, P., Burton, J., Kaminska, O., Knies, G., & Nandi, A. (2012). An initial look at non-response and attrition in understanding society. Understanding Society Working Paper Series, 2. http://research.understandingsocie-ty.org.uk/publications/working-paper/2012-02.pdf Accessed 04.09.12.

Macunovich, D. J., Easterlin, R. A., Schaeffer, C. M., & Crimmins, E. M. (1995). Echoes of the baby boom and bust: Recent and prospective changes in living alone among elderly widows in the United States. Demography, 32, 17-28.

Martikainen, P., Nihtila, E., & Moustgaard, H. (2008). The effects of socioeconomic status and health on transitions in living Arrangements and mortality: A longitudinal analysis of elderly Finnish men and women from 1997 to 2002. Journal of Gerontology: Social Sciences, 63B, S99-S109.

McCann, M., Grundy, E., & O'Reilly, D. (2012). Why is housing tenure associated with a lower risk of admission to a nursing or residential home? Wealth, health and the incentive to keep 'my home'. Journal of Epidemiology and Community Health, 66,166-169.

McFall, S. L. (2011). Understanding society - UK household longitudinal study: Wave 1, 2009-2010 user manual. http://data.understandingso-riety.org.uk/files/data/documentation/wave1/User_manual_Under-standing_Society_Wave_1.pdf Accessed 04.09.12.

Moen, P. (1991). Transitions in mid-Life: Women's work and family roles in the 1970. Journal of Marriage and Family, 53,135-150.

Mutchler, J. E., & Burr, J. A. (1991). A longitudinal analysis of household and nonhousehold living arrangements in later life. Demography, 28, 375-390.

Office for National Statistics. (2011). Mid-1971 to mid-2010 population estimates: Quinary age groups for constituent countries in the United

Kingdom; Estimated Resident Population. http://www.ons.gov.uk/ons/ rel/pop-estimate/population-estimates-for-uk-england-and-wales-scotland-and-northern-ireland/population-estimates-timeseries-1971-to-current-year/rft—table-2-quinary-age-groups-constituent-countries. zip Accessed 04.09.12.

Office for National Statistics. (2012a). Families and households, 2001 to 2011. Statistical Bulletin 19 January 2012. http://www.ons.gov.uk/ons/ dcp171778_251357.pdf Accessed 04.09.12.

Office for National Statistics. (2012b). General lifestyle survey. Household reference tables. Table 3.4. http://www.ons.gov.uk/ons/publications/re-reference-tables.html?edition=tcm%3A77-226919 Accessed 20.12.12.

Pampel, F. C. (1983). Changes in the propensity to live alone: Evidence from consecutive cross-sectional surveys, 1960-1976. Demography, 20, 433-447.

Pendry, E., Barrett, G., & Victor, C. (1999). Changes in household composition among the over sixties: A longitudinal analysis of the Health and Lifestyles Surveys. Health and Social Care in the Community, 7,109-119.

Pickard, L., Wittenberg, R., Comas-Herrera, A., King, D., & Malley, J. (2007). Care by spouses, care by children: Projections of informal care for older people in England to 2031. Social Policy and Society, 6, 353-366.

Prioux, F. (2002). Recent demographic developments in France. Population (English Edition 2002), 57, 689-728.

Richards, T., White, M.J., &Tsui, A. O. (1987). Changing living arrangements: A hazard model of transitions among household types. Demography, 24, 77-97.

Sefton, T., Evandrou, M., &Falkingham,J. (2011). Family Ties: Women's work and family histories and their association with incomes in later life in the UK. Journal of Social Policy, 40, 41-69.

Settersten, R. A., & Mayer, K. U. (1997). The measurement of age, age structuring and the life course. Annual Review of Sociology, 23,233-261.

Sigle-Rushton, W., & Waldfogel, J. (2007). The incomes of families with children: A cross-national comparison. Journal ofEuropean Social Policy, 17, 299-318.

Steele, F., Kallis, C., Goldstein, H., & Joshi, H. (2005). The relationship between childbearing and transitions from marriage and cohabitation in Britain. Demography, 42, 647-673.

Stone, J., Berrington, A., & Falkingham, J. (2011). The changing determinants of UK young adults' living arrangements. Demographic Research, 25, 629-666.

Thornton, A., & Young-DeMarco, L. (2001). Four decades of trends in attitudes toward family issues in the United States: The 1960 through the 1990. Journal of Marriage and the Family, 63,1009-1037.

Tohme, R. A., Yount, K. M., Yassine, S., Shideed, O., & Sibai, A. M. (2011). Socioeconomic resources and living arrangements of older adults in Lebanon: Who chooses to live alone? Ageing and Society, 31,1-17.

Tomassini, C., Glaser, K., Douglas, A. W., Broese van Groenou, M. I., & Grundy, E. (2004). Living arrangements among older people: An overview of trends in Europe and the USA. Population Trends, 115, 24-34.

Toothman, E. L., & Barrett, A. E. (2011). Mapping midlife: An examination of social factors shaping conceptions of the timing of middle age. Advances in Life Course Research, 16, 99-111.

University of Essex, Institute for Social and Economic Research and National Centre for Social Research. (2011). Understanding society: Wave 1,20092010 [computer file] (2nd ed.) SN: 6614. Colchester, Essex: UK Data Archive [distributor].

Uunk, W. (2004). The economic consequences of divorce for women in the European Union: The impact of welfare state arrangements. European Journal of Population, 20, 251-285.

Wall, R. (1989). Leaving home and living alone: An historical perspective. Population Studies, 43, 369-389.