Increased personal and public interest in people's retirement decisions and experiences in their retirement years is linked to the notion of retirement from paid work and the means of support in old age. In this sense, the years of retirement are a metaphor for later life and how older people will manage their income, health care, housing and social needs. Individuals and governments are particularly concerned about the balance between the capacity to provide for one's needs in old age ('self provisioning'), and relying on government and community assistance to achieve a satisfying and productive life in the 'retirement' years (Rowland 1991; Clare and Tupule 1994; Cornman and Kingston 1996; Disney 1996; OECD 1996). The experience of retirement and later life years is increasingly seen as connected to an individual's family as well as personal circumstances (Szinovacz, Ekerdt and Vinick 1992; Moen 1996).

A life-course transition

According to O'Rand, Henretta and Krecker (1992:83), retirement is a life-course transition 'that is meaningful only when it is socially and temporarily situated in families', as well as in persons' lives'. Moen (1996:132) refers to the 'dynamic interplay between personal and family events, retirement and wellbeing' and locates retirement within a life-course framework of 'an individual's prior occupational trajectory and ongoing situational and structural exigencies and opportunities'. The experience of retirement as described by these authors will be affected by economic factors such as accrued income or pension eligibility, health status and availability of health care, as well as career and family histories and the roles and relationships among immediate and extended family members.

Szinovacz and Ekerdt (1995:383) believe that family circumstances define and structure later life. 'The experience of retired life varies depending on the sort of kin convoy that accompanies individuals over the retirement process - whether people are married, have children and grandchildren, have surviving parents, or rely on ties to siblings'.

The availability of any family support (emotional, financial and practical) in the post-retirement years is often pivotal in governments' calculations of public provision of health care, housing , community services and income support (Victorian Parliament 1997; McCallum and Geiselhart 1996; Thomas and Suzman 1995; Clare and Tupule 1994). Concerns are often centred on the changing family demographics of those now aged around 50 that may alter the potential networks of family support in their later years of retirement (Matras 1990). These demographic changes include: later age at marriage, lower fertility rates, higher rates of divorce, second marriages and single parenthood, and increased longevity.

Age and concept of retirement

In Australia, as in other industrialised countries, the age of retirement or exit from paid employment has been decreasing, particularly for men. According to the ABS (1997a) in April 1997, the labour force participation rate was 72 per cent for men aged 55-59, and 43 per cent for women. At ages 60-64, the participation dropped to 45 per cent for men and 19 per cent for women. By the age of 65 and over, only 10 per cent of men and 3 per cent of women were in the labour force.

The trend toward earlier retirement has been attributed to rising income levels; changes in pension, social security and superannuation benefits or eligibility; and an increase in employer early retirement schemes related to labour market restructuring (Rosenman and Warburton 1996; Quadango and Hardy 1996).

The concept of retirement, however, is not clearly defined. Retirement may be considered a process and a state (Atchley 1982). According to Cornman and Kingston (1996) retirement 'embraces a wide range of arrangements, activities and uses'. Some people leave their long-term paid employment careers, voluntarily or involuntarily, and never again enter the paid labour force. For others, there may be a transitional phase where they may work reduced or part-time hours in their same job, or retire from long-term jobs and re-enter the labour force for shorter intermittent intervals, sometimes taking on different types of work. Some people say they will never retire, usually meaning that they defy the 'end of meaningful activity' association with that term, and avidly pursue education, volunteering or other avocations (Matras 1990; Monette 1996).

Many women, particularly in older cohorts, may consider themselves retired from the time they left the labour force after the birth of children. Other women may relate being retired to the time of their husband's official retirement (Monette 1996; Arber and Ginn 1995).

Studies have found generally high levels of life satisfaction among those who have retired (Queensland Department of Families, Youth and Community Care 1996; Shanahan 1994). Quality of life or wellbeing in retirement has been shown to be influenced by attitudes toward work, retirement and timing of retirement (Honig 1996; Reitzes, Mutran and Fernandez 1996; Maule, Cliff and Taylor 1996), health of self (and, if married, one's spouse), financial status, level of family and wider social integration and personality traits (Moen 1996; Harlow and Cantor 1996; Reis and Gold 1993). The consequences of ageing and health deterioration over time have been shown to affect satisfaction as the years of retirement extend into older age (Sharpley, Gordon and Jacobs 1996; Encel and Studencki 1996).

The overall aim of the Later Life Family Study was to examine the role of family and family relationships in the lives of Australian men and women aged 50-70 years. The retirement decisions and future 'lifestyle' expectations of this age group that are associated with family relationships is important because of their public policy impact on the provision of a range of community resources and income supports (Clare and Tulpule 1994; OECD 1996). From a family perspective, the age and timing of retirement has financial and practical consequences for levels of intergenerational exchange. These may encompass expectations about family caring: caring for grandchildren or as the carer of a frail spouse or parent(Szinovacz and Ekerdt 1995).

To gain an understanding of the salience of family relationships in retirement, respondents in the study were asked a number of questions. These included the amount of time spent with family, the importance of family, the influence of adult children or elderly parents on later life decision making, the levels of practical and emotional support between the generations, and satisfaction with family life.

Specific aspects of retirement canvassed included: age and timing of retirement, reasons for retirement, sources of income, planning for retirement, housing intentions, ease and difficulty of retirement. Feelings about the future and general life satisfaction were also measured. Comparisons were made between those still in the labour force and those who were retired on several dimensions: timing of retirement, sources of income, planning and aspects of family and life satisfaction.

Study sample

The information presented in this paper is drawn from the Australian Institute of Family Studies' Later Life Families Study, a random national telephone survey conducted in late 1996. The study sample comprised 721 respondents aged 50-70 years, 43 per cent of whom were men and 57 per cent women. Just 34 per cent of men (104) and 47 per cent of women (195) classified themselves as retired.

In the study, another 14 per cent of women (57) and 4 per cent of men (13) classified themselves as not working, not looking for work and not retired. The meaning of 'retired' appears somewhat ambiguous for women who have not been in paid work for many years. This is a more common situation among today's older women, perhaps, than younger cohorts.

Over a third (39 per cent) of women who said they were not in paid work or retired indicated they had never been in the labour force since having children; however, the others had been employed in combinations of full- and part-time work since having children. Around 19 per cent of women who described themselves as retired also indicated they had never worked since having children. Some of these women may consider themselves to be retired if their husband has retired, and some may even consider themselves retired from intense child-rearing responsibilities.

Of those who were retired, 54 per cent of men were aged 65 and over, and 77 per cent of women were aged 60 and over. However, of all retired men, 78 per cent were aged 60 and over.

Almost half (47 per cent) of both retired men and women in the study had been retired between two and five years, while over a quarter (27 per cent) including more women than men, had been retired for six to ten years. Equal proportions of men and women (14 per cent) had been retired between 11 and 21 years. Approximately 18 per cent of men, but only 7 per cent of women, had been retired less than a year.

The proportion of study respondents still in the labour force was fairly comparable with national statistics (ABS 1997). For example, 52 per cent of men and 23 per cent of women in the study aged 60-64 years were in the labour force compared to 45 per cent of men and 19 per cent per cent of women nationally. Somewhat higher proportions of men and women over age 65 in the study were still in the labour force (15 per cent and 11 per cent respectively) compared to national figures (10 per cent and 3 per cent respectively).

The sample of retired people used in this analysis is based on the 104 men and 195 women who classified themselves as retired. As the meaning and experience of retirement is considered to be different for men and women (Moen 1996; Arber and Ginn 1995), the analysis was done separately for each.

This paper will begin by describing some of the major factors deemed to influence ease in retirement and life satisfaction for study respondents. The role of family will then be explored in more detail.

The age at which any individual retires may reflect choices and constraints. Until recently, many employment careers would have come to an end at the compulsory retirement age; in other cases, retrenchment may have been the impetus. Depending on financial status and security, retirement may have to be deferred or earlier retirement seen as a viable option. Timing of withdrawal from the labour force may be influenced by one's own health and, if married, a partner's health and caring responsibilities (Moen 1996).

Age at retirement

Prior to the rescinding of compulsory retirement ages in most States in the early 1990s, the traditional ages of compulsory retirement and eligibility for the old age pension in Australia were age 65 for men and age 60 for women (Encel 1996). In the Later Life Family Study, as Figure 1 shows, only 20 per cent of men had retired at age 65 or older, and 35 per cent of men had retired between ages 60 and 64. Thus, just over half of the men in the study retired near or at the traditional compulsory age of retirement. Still considered to be an early retirement age, 36 per cent of men retired between ages 51 and 59. Less than 10 per cent of men indicated they retired under age 50.

The ABS (1995) survey of retirement of those aged 45 or older reported similar figures: 26 per cent of men had retired at age 65 or over, 32 per cent at ages 60-64, 30 per cent at ages 50-59 and 13 per cent aged 49 and younger. Overall, 45 per cent of men had retired prior to age 60. Encel and Studencki's (1996) three-State Australian study reported 34 per cent of men had retired before age 60.

The pattern for women in the study was quite different. Only 4 per cent of women retired at age 65 or older, although 21 per cent had retired between ages 60 and 64, around the compulsory retirement age. Nearly 40 per cent had retired between ages 51 and 59. Nearly a third of women, however, indicated they had retired before age 50. Of these, 18 per cent had retired prior to age 41, reflecting for this generation a tendency to leave the labour force permanently with the onset of marriage and family responsibilities. Only 2 per cent of men had retired at this early age.

Comparisons with the ABS (1995) retirement survey for women is limited by the definition of retirement as withdrawal from full-time work only, whereas many women will have retired from part-time, not full time, work and therefore would not be counted in this survey as retired. Approximately three-quarters of women considered retired by the ABS had done so by age 60.

Figure 1. Age at retirement by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Timing of retirement

The timing of retirement, particularly whether it has been voluntary or involuntary, has been associated with an individual's satisfaction and adjustment to retirement (Encel and Studencki 1996; Maule, Cliff and Taylor 1996). As Moen (1996:135) states, 'Culturally grounded norms and frames shape individual expectations and beliefs about the "right" time for retirement'.

When asked whether they would have preferred to retire earlier or later than they did, Figure 2 shows that just half of men and almost two-thirds (63 per cent) of women said they had been happy to retire when they did. A small proportion of men (11 per cent) and women (6 per cent) would have preferred an earlier retirement. On the other hand, nearly 40 per cent of men and 31 per cent of women would have liked to defer retirement.

For some, this wish for continued labour force participation may have been related to a lack of choice in timing. For example, those who gave their own or a partner's health as a reason for retirement were most likely to say they would have preferred to retire later than they did. Community mores and employment restrictions on married women working may have contributed to the early retirement of the older women. Nevertheless, the desire for continued participation in paid work by many of these men and women reinforces current policy initiatives to encourage the retention of older workers in the labour force as a means to improve financial viability in later life (Victorian Parliament 1997).

For those in couples, similar proportions of men and women (59 and 52 per cent respectively) indicated their partners were happy to retire at the time they did. Men (21 per cent) were significantly more likely than women (8 per cent) to say their partner would have liked to retire earlier. Conversely, women were twice as likely as men (40 per cent and 21 per cent respectively) to say their partners would have preferred a later retirement.

These differences by gender tend to conform to the belief that men are more attached to a work identity and more likely to be the primary income earner than women, many of whom, as in this sample, retired early or withdrew from the labour force to assume family caring roles. As younger cohorts of married women remain in the labour force for longer periods of their lives, decisions on timing preferences for couples may become more complicated (Arber and Ginn 1996; Rosenman and Warburton 1995).

Figure 2. Retirement preference by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Anticipated retirement age

Men and women in the Later Life Family Study who were still in employment, as shown in Figure 3, appear to be planning to remain in paid work until later ages than those who had already retired. Over two-thirds (67 per cent) of men and almost half (47 per cent) of women in the labour force indicated they would retire after age 65. Only 12 per cent of men and women were thinking of retiring before age 60.

These shifts are likely to be related to changes in the ages of compulsory retirement for men and women, the increased labour force participation of women, and concerns about income adequacy in later life as life spans lengthen (McCallum and Geiselhart 1996; Quadango and Hardy 1996).

From a family perspective, the age and timing of retirement, and the reasons for retirement, can affect relationships between husbands and wives, and between parents and their adult children and grandchildren. Time for couple and family activities, levels of family caregiving responsibilities and availability of financial support may alter during the years of retirement (Atchley 1992).

The later anticipated age of retirement of respondents currently in the labour force suggest that these family dynamics will be subject to change as men and women remain for longer years in employment. In turn, the age and timing of retirement and altered family circumstances have consequences for government policies related to income support, health care provision, and other community resources for those in later life (Szinnovacz, Ekerdt and Vinick 1992).

Figure 3. Age intend to retire by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Retired or not retired?

Retirement for some people may not mean total withdrawal from all paid employment, but only retirement from a specific work career. In the Later Life Family Study, 9 per cent of men and 10 per cent of women over 50 who declared they were retired or not in paid work did some occasional paid work. The types of jobs done were varied, from running a cattery for boarding cats, cleaning houses, fixing toys and teaching sewing to working as a census collector, counselling, tutoring and editing.

The Canadian General Social Survey (Monette 1996) reported that 13 per cent of people aged 50 and over who had ever retired went back to work afterward. Men who had retired under age 59 were the majority. Wanting to occupy their time and wanting to work were the main reasons given for reentering the labour force followed by financial reasons. A higher proportion of those with higher education and professional or managerial occupations had returned to work compared with those with less education.

In their study of retired persons, Encel and Studencki (1996) found 15 per cent of respondents had been in paid employment for periods since their exit from the labour force. Most worked part-time in a variety of jobs, often different from their previous occupations. High educational level and professional skills were the strongest predictors of post-retirement employment.

The reasons given for retiring at any time, as noted earlier, will reflect voluntary and involuntary circumstances and be influenced by individual, family and institutional factors. The varied reasons given for retirement by men and women in the Later Life Family Study are shown in Figure 4.

Family

Retirement, as noted earlier, can provide time for family activities and involvement, and these may even be an incentive for retiring. Having more time for the grandparent role is a reason women, particularly, may retire earlier (Millward 1997). In some situations this time may be needed to care for family members. Although recent research has highlighted some of the benefits of combining employment with elder caregiving for women (Sharlach 1994; Phillips and Bernard 1995), family caregiving needs of a spouse or elderly parent have also been associated with early retirement for women (ABS 1994; Murphy et al. 1997).

When it came to family reasons, 12 per cent of women indicated they retired to care for family members, and 14 per cent to spend more time with family. Less than 2 per cent of men offered these reasons. The majority of these women had retired before age 50 and family life may have always been their major activity. Similarly, in the ABS (1995) survey, 20 per cent of women and 3 per cent of men gave family reasons as the main reason for early retirement.

Figure 4. Reasons for retirement by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Health

Health issues were a dominant reason given for retirement. A third of men and a quarter of woman gave their own or a partner's poor health as the reason for retirement. Health reasons were more likely to be given by men aged less than 60 years than by older men, while for women, retirement for health reasons did not vary significantly by age. Only 3 per cent of men and women gave the health of other family members as a reason. As a reason for leaving employment, the health of a marital partner or, in the case of adult children, the health of their parents, may have a profound affect on family relationships and responsibilities (Moen 1996).

Poor health has been consistently mentioned in the literature as a reason for retirement, particularly before recent trends toward early retirement resulting from corporate and pension/social security incentives (Howe and Manning 1987; ABS, 1995; Monette, 1996). An ABS (1995) survey of retirement reported that 48 per cent of men and 25 per cent of women gave their ill-health or injury as the reason for retiring before the standard retirement age. A Victorian survey of retirement adjustment (Sharpley, Gordon and Jacobs 1996) found that 15 per cent of men and 11 per cent of women gave illness as a reason for retirement. Howe and Manning (1987) also observe that some older workers may be encouraged to seek a health-related retirement in the face of redundancy if benefits are likely to be more advantageous. For some older workers, a health reason for retirement may be seen as more acceptable to self-esteem.

Redundancy/employment

Earlier retirement may be increasingly involuntary (Rosenman and Warburton 1996) in a climate of high unemployment and business and government 'downsizing'. Around 14 per cent of men and 5 per cent of women gave being made redundant or dismissed as their reason for retirement. Similarly, around 14 per cent of men but only 1 per cent of women indicated that retirement came about by accepting an early retirement package. In the ABS (1995) retirement survey, 7 per cent of men and 3 per cent of women who had been in full-time work were retrenched or took a package.

A small percentage of men and women (4 per cent) indicated they retired because they could not find any work. Selling their business led to retirement for 4 per cent of men and women. For a few men and women (7 per cent) having enough income prompted retirement, while 8 per cent of men and women said they did not like work.

Given the fairly high proportion of men and the majority of women who had retired prior to the pension ages of 65 or 60, it is not surprising that only 16 per cent of men and 6 per cent of women gave reaching pension age as the reason for retirement.

Other reasons

Few respondents specifically mentioned quality of life reasons for retirement, such as wanting to pursue different activities. However, the 12 per cent of women and men who said they 'wanted to do other things' may have assumed family involvement among these things. In the ABS (1995) survey, 18 per cent of men and 25 per cent of women gave 'more leisure time' as a reason for retirement.

Partner considerations

For those who are married, their partner's work status may be another family consideration in retirement decisions. In dual earner couples, the retirement of one partner may affect the timing decisions of the other. Women who did not enter the labour force until after children were in school or older may prefer to remain in paid work even after their husbands retire. Some couples will prefer joint retirement regardless of work history, while for others, decisions will be influenced by anticipated family income and potential retirement benefits. Thus, retirement timing in couples may be sequential or synchronised (Henritta, O'Rand and Chan, 1993; Monette, 1996).

Of the men and women in the study who were retired and married, 13 per cent of women said they retired when their partner was about to retire, but only one man indicated this was his reason.

Married respondents were also asked why their partners had retired. Family considerations were again more relevant for wives than husbands. While 20 per cent of husbands indicated that their wives had retired because they wanted to spend more time with family members, only 6 per cent of women suggested their husbands had retired to increase family togetherness. No one indicated caring for family members had been a husband's motive. Around 11 per cent of men indicated their wife had retired because the husband was retiring, compared to 3 per cent of women whose husbands had retired because of their wife leaving the labour force. Of the men who indicated their wives were retired, 39 per cent said their wives had retired prior to age 50, compared to only 8 per cent of women whose husbands had retired at this early age. Many of these wives would have 'retired' from paid work when they had children and remained at home to care for them.

Implications for family caring responsibilities were evident given that over a third (38 per cent) of women said their husbands had retired because of health reasons, and a fifth (21 per cent) of men said the same for their wives. Their own or their partner's health was included in this response.

According to their wives, equal proportions of husbands (13 per cent) had retired when they reached pension age, had been dismissed or made redundant, or had wanted to do other things. Similarly, 9 per cent of men said their wives had retired at pension age, or because they wanted to do other things.

The circumstances surrounding leaving paid employment may have a long-term effect on family life in the retirement years. Loss of anticipated income or poor health, for example, may change plans for travel, housing or levels of economic support to and from adult children and grandchildren (Maule, Cliff and Taylor 1996; Reitzes, Mutran and Fernandez 1996).

Studies on life satisfaction in retirement find that the majority of people are satisfied with their lives (Atchley 1982; Monette 1996; Reitzes, Mutran and Fernandez 1996). However, health problems and inadequate finances have been strongly associated with stress in retirement (Bosse et al. 1991; Szinovacz and Washo 1992; Sharpley, Gordon and Jacobs 1996; Atchley 1982; Commonwealth Department of Social Security 1992; Monette, 1996). Reis and Gold (1993) focus on an individual's personality attributes and the concept of emotional hardiness as potential factors contributing to life satisfaction in retirement. From a life course perspective, satisfaction in the retirement years cannot be separated from past and current, personal, familial and social trajectories and circumstances (Moen 1997; O'Rand, Henretta and Krecker 1992).

Source of income and financial situation

Financial problems are consistently mentioned in the literature as one of the major predictors of dissatisfaction in post-retirement life. Therefore, retired respondents in the Later Life Family Study were asked several questions to ascertain their financial situation: their main income source, lump sums, and satisfaction with their income and standard of living.

As shown in Figure 5, just under half (49 per cent) of retired men and almost two-thirds of women (62 per cent) indicated their main source of income was a pension or benefit. Superannuation, other investments and annuities were the main source of income for 44 per cent of men and 29 per cent of women. Just 6 per cent of men and women claimed they had no personal income but were dependent on a spouse or others, and 2 per cent still received an income from a salary or business.

Of those respondents who received a pension, the most common form was the Age Pension received by 41 per cent of men and 54 per cent of women. A Disability Support Pension was the next most common pension for men (14 per cent) and a Veteran's Pension (10 per cent)for women. Only two people received a Carer Allowance as their main income source.

These findings on income in retirement are similar to the ABS (1995) survey. It reported that the main source of income at retirement for half (50 per cent) of men and 35 per cent of women over 45 who had retired from full-time work had been a government pension or benefit. The figure for women is lower than in the study as it does not include women who had been working part time. Over 40 per cent of respondents in Encel and Studencki's (1996) retirement survey were pensioners.

Figure 5. Main source of income in retirement: retirees by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Government and private superannuation schemes have also enabled beneficiaries to take their benefits in the form of a lump sum instead of as a pension. Individuals could then spend or invest their lump sum as desired, which, in some cases, could have enabled them to obtain a part Age Pension. Insurance and worker's compensation payments can also be in the form of a lump sum.

Nearly two-thirds of men (64 per cent) or their partners and over half (54 per cent) of women or their partners who were retired or no longer in paid work indicated they had received lump sum benefits. When asked their main use of this money, 59 per cent of men and 39 per cent of women had invested in shares and other annuities, 20 per cent of men and 10 per cent of women in real estate, 11 per cent of men and women had paid off their mortgage. Women (20 per cent) were significantly more likely than men (6 per cent) to use their lump sum for general living expenses. Much smaller proportions of men and women had used their lump sum payments for clearing other debts, making home improvements, and paying for a holiday or car. Only one person had provided gifts for family members.

Satisfaction with income

In response to a question about their standard of living, 95 per cent of women and 89 per cent of men said they were satisfied with their standard of living (see Figure 8). However, when respondents were asked to describe their financial situation at the present time, 59 per cent of retired men and 48 per cent of retired women said they were 'managing well' (for example, able to put money aside for holidays, house improvements and so on) or 'doing very well' (Figure 6). A third of men and 42 per cent of women indicated they were 'coping' ( for example, with some money left but used for irregular bills) and 9 per cent of men and 10 per cent of women were finding it quite or extremely difficult financially.

The ability of those whose main source of income was the pension to cope financially was compared with those whose income came from superannuation or investments since the Age Pension and other pension rates have been considered modest compared to average weekly earnings.

Figure 6. Financial adjustment in retirement by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

For men, although the differences were not significant, those who relied mainly on pensions or benefits were less likely to say they were managing well (49 per cent) than those on superannuation or investments (65 per cent). They were also more likely to say they were having difficulty financially (12 per cent and 7 per cent respectively). A third of those on either source of income said they were coping. Women who relied on a pension or benefit were significantly less likely to say they were managing well (34 per cent) compared with women who relied on superannuation and investments (68 per cent). However, around half (51 per cent) of women on pensions or benefits indicated they were coping financially.

Encel and Studenki (1996) also reported that only a small proportion of retired respondents (10 per cent), including pensioners, found it difficult to make ends meet, although 45 per cent said they were 'just managing'. However, a much higher proportion of those on superannuation (37 per cent) compared to pensioners (6 per cent) had 'no money problems'. According to the Department of Social Security's (1992) retirement survey, equal proportions (43 per cent) of retired pensioners, and those of pension age but not on a pension, reported their current financial situation was 'worse off' than before retirement. Around 20 per cent in both categories reported they were now 'better off'. Half of pre-pension age retirees also said they were 'worse off' since retirement.

The generally positive financial outlooks for those relying on superannuation and investments have been queried in studies done by investment management firms (Migro 1996) and social policy analysts (Cass 1996; Rosenman and Warbuton 1996; McCallum and Geiselhart 1996). They suggest that for many people, superannuation payments may not provide an adequate income retirement. For example, the increased unemployment and early redundancy trends for men in their 50s, the patterns of part-time, intermittent and lower paid work histories of women, and service sector jobs (many of which are not highly paid), will result in lower superannuation payments. Rosenman and Warburton (1996) make the point that 'the significant non-cash benefits that accrue to those with pension status' may mean some age pensioners are better off financially in retirement than those on the same level of income from superannuation and self-funding sources.

A dominant public policy concern as the population ages is how the income needs of older people will be financed. Governments are reviewing pension and social security benefits and eligibility. They are investigating ways of expanding employer and self-employed superannuation coverage, and have abolished compulsory retirement ages to encourage longer labour force participation and less dependency on government benefits in old age (Encel 1996; Rosenman and Warburton 1996).

Health status

The majority of study respondents appeared to be in good health, not surprising given their age range of 50-70 years. When asked about their health status, almost three-quarters (72 per cent) of retired men and women said their health was good or excellent. Married women who were retired were less likely to say their partners' health was good (63 per cent) than the retired married men, 80 per cent of whom had a partner in good health.

The majority of all men (70 per cent) and women (76 per cent) in the Later Life Family Study who stated they were retired said their retirement was easy for them as shown in Figure 7.

Somewhat surprisingly, given the association between poor health and lower satisfaction in retirement in the literature, the majority of men appeared to find retirement easy regardless of their health status. Nearly three-quarters of men who claimed their health was 'good' and two-thirds who said their health was 'poor' indicated that adjustment to retirement was easy for them. For women, health status was significant; 82 per cent of women who felt their health was 'good' found retirement easy, whereas only 60 per cent who were in 'poor' health felt this way. Perhaps leaving a stressful working environment resulted in improved health for some of these men and women.

Those who maintained they were managing well or coping financially generally also found retirement easier than those who said their financial situation was difficult. However, this was only significant for women. The small number of men who said their financial situation was difficult prevented further analysis.

Figure 7. Ease of adjustment to retirement by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Feelings about retirement were associated with source of income. Those men and women whose main source of income was government pensions or benefits were significantly less likely to say their retirement was easy than those whose source of income was from superannuation or investments of various kinds (67 per cent and 82 per cent respectively).

Ease with retirement appeared to be associated with a person's satisfaction with their life in general. Overall, as Figure 8 illustrates, almost all of those who were retired (89 per cent of men and 96 per cent of women) claimed to be very satisfied or satisfied with their life as a whole. In terms of general feelings of confidence and self-esteem, again retired respondents seemed generally robust. Some 94 per cent of men and women said they were satisfied with the sense of purpose and meaning in their life, 93 per cent with their personal and emotional life, 89 per cent with the extent to which they were the kind of person they would like to be, 90 per cent with what they were accomplishing in life, and 96 per cent with the way they handled problems.

Nevertheless, there were differences in aspects of life satisfaction between those who felt retirement was easy, and those who said it was difficult or had mixed feelings. For example, being less satisfied about the purpose and meaning in their life, their personal/emotional life and, for men, what they were accomplishing, meant feeling less sanguine about retirement.

Retired men who were not currently married were significantly less satisfied with their personal emotional life and sense of purpose than married men and all women. Szinovacz and Ekerdt (1995) have observed that the post-retirement experiences of unmarried men are less positive than for married or widowed women who are retired. They suggest that this was due to the apparent greater dependence of men on their spouses for emotional and social support.

However, marital status did not appear to significantly affect whether men and women felt their retirement was easy or difficult. Approximately 70 per cent of men, whether married or not, said their retirement was easy. Similarly for women, 73 per cent of married women and 81 per cent of those widowed, divorced or never married indicated retirement was easy.

Baltes and Carstensen (1996) have attributed the high levels of satisfaction reported in later life to older people's ability to selectively redefine or moderate their goals and expectations to what is possible for them to attain. As part of this process, individuals appear to find ways to adapt to, accommodate and compensate for the changes that may occur with ageing. They also optimise opportunities in other domains including different ways of being involved with and enjoying family and community.

Figure 8. Aspects of life satisfaction: retirees by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

The future

As one ages, concerns about health and physical capacities, and the loss of a partner or friends may intrude on the initial pleasures and satisfactions of the earlier retirement stage of later life (Pearlin and Skaff 1996). When respondents were asked how they felt about the future, two-thirds of retired men (65 per cent) and 71 per cent of women said they felt mostly positive about the future. While only a small proportion of men (11 per cent) and women (6 per cent) felt mostly negative about their future, around one in four admitted to mixed feelings.

A positive outlook in retirement would have been reinforced by respondents' feelings about maintaining their independence and perceptions of usefulness (Baltes and Carstensen 1996). Almost all retired women (98 per cent) and men (90 per cent) agreed they had plenty to keep them occupied when they were old. Around three-quarters (73 per cent) of men and women (76 per cent) who felt this way also believed their retirement had been easy for them. 

Similarly, three-quarters of men and women who felt they would be able to do most things for themselves (between 88 and 90 agreed with this statement) had found retirement easy. In a less positive vein, around one in three men and women were concerned about being alone. Only the men, however, were more likely to have found retirement more difficult than those who were not concerned about being alone. This is a comment, perhaps, on women's tendency to be more connected over their life course to networks of family and friends than men (Szinovacz 1992).

Some studies (Encel and Studenki 1996; Monette 1996) report that levels of satisfaction declined after the first years of retirement due to expectations of these years not being met, or to diminished health and social networks. In the Later Life Families Study, men who had been retired six years or more were somewhat more likely to say that retirement was easy for them (79 per cent) than those in the first five years of retirement (61 per cent). Regardless of years retired, around three-quarters of women said retirement was easy for them. Length of retirement made no difference to satisfaction with their life as a whole for men and women, and over 90 per cent were satisfied. Given the age range of respondents, and the majority who were in good health, these results are not surprising.

Similar findings on retirement satisfaction have been reported in the literature. Encel and Studencki (1996) found general satisfaction with retirement. Sixty per cent of their sample agreed that retirement was an agreeable situation, 13 per cent found it difficult and 15 per cent said it was both agreeable and difficult. Positive aspects of retirement mentioned included more time with family and opportunities for travel and leisure. Finances presented the greatest source of difficulty, although only for 12 per cent of respondents. Sharpley, Gordon and Jacobs (1996) also found that finance was a source of stress (15 per cent) and their health (9 per cent) for men and women who were retired.

A survey of older people's retirement income decisions (Commonwealth Department of Social Security 1992) reported that over 70 per cent of retired older people were satisfied with the way things have gone since their retirement. The main reasons for dissatisfaction with life in retirement related to their financial position, particularly for pensioners (42 per cent) but also for those relying on other income (28 per cent), and health problems (20 per cent). Nearly half (47 per cent) of Canadian retirees said they enjoyed life more in the year after retirement, while almost 40 per cent felt their lives were the same (Monette 1996).

Contact with friends, family and community provide social participation in later life that has been shown to promote physical and psychological health (Teshuva, Stanislavsky and Kendig 1994; Victorian Parliament 1997).

Retirement seemed to provide respondents with personal time and time for friends. As shown in Figure 9, the majority of men and women said they had enough time for themselves (80 per cent) and friends (84 per cent). Contact with friends was considered very important to 51 per cent of men and 57 per cent of women, and 'somewhat important' to the others.

For many retired men and women, volunteer work in the community or participation in community clubs and organisations replaces their former involvement in paid work (ABS 1997b). Community involvement was considered 'very important' to 29 per cent of men and 41 per cent of women who were retired; however, it was 'somewhat important' to 47 per cent of men and women. Although the nature of their activities was not identified, the majority of these retired men and women thought some level of community involvement was important in their lives.

An ABS (1997b) survey of voluntary work revealed that 20 per cent of men and women aged 55-64 did some kind of unpaid voluntary work, as did 17 per cent of those aged 65 and more.

As Atchley (1992:145) observes, 'The effect of the retirement of one or both spouses on the couple is an important potential determinant of individual perceptions of retirement and retirement adjustment'. Decisions and reasons surrounding the timing of retirement, health concerns, financial situation and family relationships are commonly mentioned factors affecting couples in retirement. The retirement of one spouse is bound to affect the other spouse and usually other family members: adult children, elderly parents and grandchildren. Patterns of time use, activities, obligations, plans for the future, and the dynamics of the marital and family relationship may change with the shift in employment status of either partner (Wolcott 1997a).

Shared decisions

Given the potential impact on a couple's life, whether or not the decision to retire is made jointly or individually may be an important consideration. Respondents in couples were asked whether they and their partner had made decisions and planned for retirement together. Over three-quarters (77 per cent) of men and two-thirds (67 per cent) of women indicated they had made these decisions together.

Couples' adjustment to retirement

Respondents in the study who said they were retired and in a couple were asked how they felt about their retirement and how their partner felt about it. They were also asked, when applicable, how they felt about their partner's retirement.

Figure 9. Satisfaction with amount of time: retirees by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Figure 7 shows that around two-thirds of men and women (68 per cent) indicated that their own retirement has been very easy or easy for their partner, 6 per cent had mixed feelings, and a quarter (26 per cent) admitted it has been difficult for their partner. For those who had a partner who was retired, 79 per cent of men and 73 per cent of women appeared to find the situation easy to adjust to, while 21 per cent of men and 27 per cent of women had some difficulty adjusting to the new couple dynamic.

Their own or a partner's health status had some impact on adjustment in retirement, but less than might be anticipated given the emphasis in the literature on the negative impact in retirement of poor health (Bosse et al. 1991; Szinovacz and Washo 1992; Sharpley, Gordon and Jacobs 1996). Around half (57 per cent) of women) and 40 per cent of men who said they had retired because of their own or a partner's poor health said their retirement was difficult for their partner. Where their partner had retired for these health reasons, 40 per cent of women and half of men found the partner's retirement difficult for themselves. When a partner's health was poor, both men and women said their own adjustment to retirement was more difficult.

Poor health in retirement may disrupt the plans partners had for this stage of their life. If caring responsibilities are demanding, it may place added strains on the couple relationship and alter relationships between extended family members (Long and Mancini 1990). Studies of older couples, however, point to both enhanced closeness and increased tension in spouses' adaptation to illness (Brubaker 1991).

Marital satisfaction

The retired men and women in the study appeared to be happily married, and 95 per cent of men and women described their marriage as very happy or happy. A number of researchers (Vinick and Ekerdt 1991; Askham, 1992) suggest that retirement has little effect on the marriage, and that the quality of married life prior to retirement was the best predictor of post-retirement marital satisfaction.

Division of household work

Although some studies suggest that after retirement the gender division of household chores may become less rigid, the majority of studies suggest continuity in the way couples carry out tasks over the life course (Cliff 1993; Dorfman 1992). The survey's retired couples followed a fairly traditional division of labour, and men and women said the female partner took the main responsibility for doing household chores in 59 per cent of cases. Nevertheless, over a third (35 per cent) also said that both shared household tasks. When it came to who took main responsibility for being the economic provider, over half (55 per cent) of men and women said 'we both do', and over a third (38 per cent) the male partner. Management of finances was shared in 40 per cent of couples.

More than two-thirds (69 per cent) of partnered men and women agreed that they each did their fair share around the house, and 88 per cent felt appreciated for what is done. Such perceptions of equity appear to facilitate satisfaction in retirement (Dorfman 1992). 

Retirement of one or both partners may provide the opportunity for couples to renegotiate the division of household labour and there is an assumption that men, particularly, may increase their share in household work upon retirement (Dorfman 1992).

Comparing the men and women in couples who were retired with those who were still in paid work, few differences emerged regarding the division of household tasks. Employed men were more likely to say their partner had main responsibility for chores (63 per cent) and did more than their fair share around the house (39 per cent) than retired men (53 per cent and 18 per cent respectively). There were no differences on these dimensions between retired and employed women.

According to Szinovacz, Ekerdt and Vinick (1992:1) the retirement experience of individuals 'affects and is affected by their social networks and relationships, foremost among them their ties to family members'. The role of family in retirement has already been discussed in relation to the joint decision to retire, family reasons as the impetus for retirement, and how those in a couple have adjusted to their own or their partner's retirement. Study respondents were also asked several questions about how family influenced decision making and the importance of family contact in their lives.

Time with family

The time available for personal, social and family activities expands when the hours previously taken up with paid employment cease. For women in later life, the hours previously devoted to intensive child rearing will also leave time for other pursuits. The centrality of family involvement and contact for those who are retired, however, will vary according to prior degrees and patterns of closeness, personal values and inclination, geographic proximity and actual and perceived needs of self and family members (Bengston and Schrader 1992).

When asked whether retirement made any difference to the amount of time they spent with their family, over a third (37 per cent) of men and 47 per cent of women said they now spend more time with family. Around 50 per cent said retirement had made no difference. As shown in Figure 9, around two thirds (67 per cent) of men and women felt the amount of time spent with family was 'enough'. However, despite being retired, over a quarter (29 per cent) of men and a third (34 per cent) of women indicated that the amount of time they had with family was 'not enough'. In some cases, the reason may be that their family does not live geographically close enough to see frequently.

A study of Canadian retirees (Monette 1996) reported that increased time with family and more time for leisure were the reasons retirees believed they got more out of life than before they retired.

Influence of family in decision making

Respondents were asked specifically whether the needs of their adult children influenced or would influence decisions about later life and growing older, and in what ways this might occur. Around a quarter (27 per cent) of men and a third (32 per cent) of women said adult children influenced their decisions. Their comments revealed that because of their children they would remain geographically close to be there for family companionship, and to provide care and support when needed.

There were, however, no significant differences between later life respondents who were retired and those still in the labour force (36 per cent of men and 43 per cent of women) regarding whether the needs of adult children would affect their later life decisions.

The needs of elderly parents were also an influence on later life decisions for 18 per cent of men and 39 per cent of women who had a living parent or in-law. Again, comments revealed the desire to remain in close proximity to provide support and care if necessary. Parents also provided a guide, either to replicate or modify, for how this generation would like to prepare for their old age, particularly in being better prepared financially for their retirement years and taking advantage of being healthy and fit while they could.

Given the family concerns of those already retired, it is not surprising that the needs of elderly parents were also an influence on later life decisions for those aged 50-70 years still in the labour force (29 per cent of men and 38 per cent of women).

Importance of family connections

While the majority of respondents did not claim that their adult children or elderly parents specifically influenced their decisions about retirement, nevertheless family contact was important to most retired respondents. Figure 10 shows contact with family members was said to be 'very important' for 85 per cent of retired women, significantly more so than for men (65 per cent). However, 28 per cent of men indicated family contact was 'somewhat important' compared to 16 per cent of women.

In a more specific way regarding the place of family in their lives, 85 per cent of women and 64 per cent of men who were grandparents indicated that being a grandparent was very important to them. The remaining grandparents, men and women, said it was 'somewhat important'. The role of grandparent was also at least somewhat important to 95 per cent of men and 98 per cent of women who were in the labour force.

For a quarter of retired women and 14 per cent of men, time was taken in being a main carer, mainly of their spouse or older family member.

Although the proportion of men and women who specifically mentioned that family influenced their retirement and later life decisions and activities was not large, the levels of intergenerational support, emotional, practical and financial, between all study respondents and their adult children and older parents was substantial (Wolcott 1997b). While family connectedness was important to most respondents, at this stage of their retirement, self-sufficiency and independence in their living arrangements appeared to be a priority. No one stated that they had moved in with their children when they had retired.

Looking to their future, approximately 14 per cent of retired men and women indicated they would move in with their children when asked what they would do if they could not to live on their own later in life. These responses confirm other research (Finch and Mason 1993; Aldous 1997; Kendig 1986) that suggests a desire on the part of parents to retain autonomy in later life while maintaining close and supportive family relationships.

Figure 10. Importance of social contact: retirees by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

As stated initially, a life course approach to retirement takes into consideration a range of individual and family transitions and trajectories, and aspects of the social/political environment that shape life circumstances during the later years (Moen 1996). A number of dimensions that appear to contribute to ease and satisfaction in retirement have already been described. A logistic regression analysis was undertaken to explore some of the variables mentioned in the retirement literature that appear to predict an easier retirement.

In accord with other studies (Encel and Studencki 1996; Sharpley, Gordon and Jacobs 1996; Bosse et al. 1991; Commonwealth Department of Social Security 1992), the best predictors to finding retirement easy were: coping financially, retiring at the time preferred, and being satisfied with life as a whole. The model successfully predicted the outcome for 80 per cent of cases.

Contrary to some of the results of these and other studies that examine families and retirement (Szinovacz and Ekerdt 1995), neither health status of self and partner, being a main carer, or the importance and amount of time with family and friends were reliable predictors of ease in retirement. Since the majority of respondents appeared to be satisfied with their amount of family involvement, were in reasonably good health and satisfied with life overall, this result may not be surprising for this sample.

The Victorian Parliament's Inquiry into Planning for Positive Ageing (1996:18) comments on the paradox,'that so much effort is devoted to preparing for some phases of life and so little for the life stage which has dramatically increased in length, the period of old age'. Planning for retirement, if it is pursued, appears to generally focus on financial issues (Sharpley, Gordon and Jacobs 1996; Commonwealth Department of Social Security, 1992; Young and Schuller 1991).

Retired respondents in the Later Life Family Study were asked whether they had prepared for their retirement. More than one in three (37 per cent) men but only one in five (22 per cent) women said they had actively prepared for their retirement. The lower preparatory rate of women may be related to older married women's definition of retirement and reliance on a partner's income (Onyx and Watkins 1996).

Participation in retirement information/planning sessions provided at their workplace, followed by attendance at planning sessions by other organisations or consultants, were the most common pre retirement preparations for both men and women as shown in Figure 11. Reading books and articles, watching TV programs, talking informally to friends or colleagues, and making investments were other ways people indicated they had planned for retirement.

Changing attitudes toward retirement planning is suggested by the responses of later life men and women who were still in the labour force. Seventy per cent of men and 63 per cent of women said they were preparing for their retirement. However, confirming other studies on retirement, Figure 12 shows that financial investment was the main form this preparation took. Of those men and women who said they were preparing for retirement, 40 per cent mentioned investing. Even for those who said they were doing 'other' things, the majority of comments were financially related, mainly about 'saving money'.

Figure 11. Preparation for retirement: retirees by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Figure 12. Preparation for retirement: workers by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

Sharpley, Gordon and Jacobs (1996) reported that only 19 per cent of men and 13 per cent of their retired sample had received any pre-retirement education, and where they had it was financial information only. The importance of taking a 'life planning' approach to later life or the post-retirement phase that incorporates aspects of personal health, long-term care arrangements, social connectedness, intellectual and spiritual growth as well as financial planning has been recommended (Victorian Parliament 1996).

Anticipated retirement income

According to the ABS (1995) retirement survey, there has been a shift in thinking toward greater individual responsibility to provide for the retirement years and less expectation of reliance on the Age Pension. Respondents in the study currently in employment were asked what their main income source would be once they were retired (Figure 13). Almost one in three women (33 per cent) and one in five men (19 per cent) indicated that the age (or other government) pension would be their main retirement income. However, over a third of men and half of women indicated it would be one source of income in retirement. Conversely, superannuation would provide the main income for approximately one in three men (34 per cent) and one in five women (22 per cent); however, 10 per cent of women would be relying on their husband's superannuation. Nearly two-thirds of men and over half of women would also have superannuation as an additional income source. Over a third of men (37 per cent) and a quarter (27 per cent) of women were counting on interest from investments, real estate or savings for their main income.

Figure 13. Expected main source of income in retirement: workers by sex

Source: Australian Family Life Course Study, Australian Institute of Family Studies, 1996.

The ABS (1995) survey reported that of those aged 45 and over in full-time work in 1994, 32 per cent of men and 19 per cent of women expected their main income at retirement to come from superannuation. Twenty-eight per cent of men and 29 per cent of women would be relying on the age or other government pension and benefits (for example, invalid, sole parent, widow). By 1996 (ABS 1996) the proportion of women who expected superannuation to be their main retirement income source had risen to 23 per cent, reflecting their increased labour force participation rates.

The retired men and women in the Later Life Family Study were between the ages of 50 to 70 years, a period sometimes described as the 'third age' (Young and Schuller 1991). According to Laslett (1989) this time should be an 'era of personal fulfilment', a time before the frailties of the 'fourth age' begin to appear in earnest. Overall, the men and women in this study indicated they were in reasonably good health.

The majority of these retired men and women found adjusting to their own retirement easy. For those in a couple, the majority did not find their partner's retirement difficult. Being able to manage on their income contributed to their sense of wellbeing in retirement, as did the perception that their life as a whole was satisfactory. Those who found retirement difficult or had mixed feelings about this stage of their life were also generally less satisfied with their life as a whole.

Family contact and involvement were clearly important in post-retirement life for most women and a majority of men. As parents and grandparents, they provided support for their adult children and received support from them. Friends and community activities were also important aspects of these retirement years.

For these retired men and women, the majority who would be characterised as in the younger range of later life years; planning for retirement did not appear to have been a major concern. For those aged 60 and over, especially the men, perhaps a more secure employment history accompanied by the assurance of the Aged Pension made planning less vital to their vision of a comfortable retirement for themselves and their family.

The majority of the retired men and women in the study were also married, satisfied with their marital relationship and had children. This provided a family situation that included high levels of intergenerational support. Today's environment of high divorce rates, lower marriage and fertility levels, and an increase in insecure employment may mean that coming generations of men and women may not have these somewhat benign life course trajectories as they enter later life.

Overall, 39 per cent of men and 31 per cent of women in the Later Life Family Study indicated they would have liked to retire later than they did. (Just over half of the men and a quarter of the women had retired at age 60 or over). Later life respondents, men and women who were still in the labour force, however, intended to remain in paid work until a later age. Over two-thirds of men and almost half of women in the labour force indicated they would retire over age 65. These trends will have an impact on men and women in mid and later life, and on the lives of their adult children.

For families, expectations about the timing of retirement and post-retirement lifestyle will be influenced by the changing economic, employment and social security environment. Expectations and levels of intergenerational care and practical support may be altered by new patterns of participation in paid and non-paid work across the life course as each generation enters later life.

The extension and costs of education, high housing costs and more fragmented work experiences and relationship trajectories of adult children may result in adjustments to later life parents' retirement plans. Early retirement may not be a feasible option if parents continue to contribute to their adult children's or grandchildren's financial support. Rising medical and nursing home costs may mean a delayed retirement for later life adults so they can contribute financially to their elderly parents' care. Changes in pension entitlements and the availability of aged care resources may also require men and women to delay retirement to ensure adequate provision for their own old age.

If men and, particularly, women do remain in paid work up to or past traditional retirement ages (thus altering the present trend toward early retirement), their capacity for family involvement and caring, either grandchild care or care for frail elderly and ill family members, will be constrained. Such a trend will affect the provision of community resources such as home and community care and child care.

Most of these scenarios assume a healthy later life and the availability of employment that will enable men and women to continue in paid work and accumulate resources for their older years. If present trends of underemployment and insecure employment persist, the ability of younger cohorts to ensure an adequate income in the retirement years may be jeopardised.

It would seem that the concept of phased retirement - the reduction in hours or days worked as one nears the traditional ages of retirement - may be an alternative to explore for people who would like to continue in paid employment, at least in their early later life years. The organisation of phased retirement would, additionally, entail adjustments to taxation, superannuation and pension systems to ensure adequate income in these pre- and post-retirement years (Patterson and Wolffs 1997; Rosenman and Warburton 1996; Reday-Mulvey 1997). Such a pattern of paid work would, however, enable later life workers to combine aspects of their employment careers and income earning with desired family involvement and caring responsibilities.

Planning for later life in all its aspects will become more essential as the life course transitions between employment and retirement become less rigid. Only a small proportion of this group of retired people had planned for their retirement and this was mainly financial. Although greater proportions of those who were still in employment were planning for retirement, this too was mainly related to future income. Women's continued participation in paid work after having children may also mean decision making around retirement and post-retirement becomes more complex for couples.

For the generation about to enter their 50s, their later age of marriage, lower birth rates, higher rates of divorce and increased longevity may also modify the potential constellations of family support in their later years of retirement (Matras 1990; Millward 1997).

The prospect of increased longevity and potential extended periods of later life in and not in paid work, and more complex family structures and patterns of combining work and family life over the life course, mean that wider community access to 'whole of life' education and planning resources for later life is necessary. Retirement planning should include discussions about expectations of family support and intergenerational exchange in later life.

Additional research is necessary to understand how early, delayed or phased retirement from paid employment may affect intergenerational supports and transfers among family members in later life. Changes in family composition and structure over time are important dimensions to consider. Concomitantly, research needs to explore how the circumstances surrounding retirement reasons and timing affect the linkages between intergenerational supports in later life and public provision of income and community care resources.