The findings detailed in the body of this report highlight the precarious socio-economic situation for many women in mid-age and beyond. Their poorer economic security relative to men has been well-documented in the literature as being founded in women's traditionally poorer attachment to the labour force in their younger years and the resulting disadvantage in regard to retirement income, particularly for those who are separated or divorced. The findings also point to close links between paid work and women's health, and the clear need not only for greater support for carers, who are at risk of poorer health themselves, but also information for potential caregivers about the implications of withdrawing from the labour force. They highlight the importance of education for women, including the need for women to be able to access ongoing further education and training in their mature years. Cultural gendered expectations may be indicated in the extent to which mid-age women are involved in childcare and the influence the birth of a grandchild has on women's retirement decisions. By implication, the findings also suggest that present generations of younger women need evidence-based information and advice about financial planning.
Experiences of paid work
ALSWH longitudinal data reported here confirm that mid-age women's attachment to the labour force is strengthening, with more participants taking up employment between 1996 and 2004 compared with those who gave up paid work. A clear association between employment and women's health is revealed in the data. Participants who had been continually in paid work between 1996 and 2004 reported better mental and physical health compared not only with women who had remained out of the labour force, but also with those who had been employed intermittently during this time. In 2004, approximately one quarter of all women in the mid-age cohort were carers, but caregivers were less likely to be consistently in the workforce.
There seem to be specific patterns of paid work that are related to optimum health. Working part-time, around 16 to 24 hours per week, was associated with better mental health at all four time points, while working very long hours (more than 49 hours per week) remains prejudicial to women's mental health. Mid-age women appear physically able to cope with longer hours of paid work even as they age. It is important to remember that the data reported here do not explain the direction of this association, i.e. whether good physical health makes it possible to work long hours, or whether working long hours contributes to good physical health. Better health is also associated with working the number of hours one prefers, irrespective of how many hours this is.
Planning for retirement
The data show that mid-age women are very uncertain about the timing of their 'retirement', bearing in mind that many women have difficulty with the meaning of this term. The problematic nature of the concept is underlined by comments made by ALSWH participants, which confirm that retirement has a very different meaning for many women and does not coincide with the traditional 'gateway' for retiring men. Almost half of the women who were not retired in 2004 said that they did not know when they expected to retire. There was also a mismatch between expectations for retirement age, and women's preferences for when they would retire.
Uncertainty about retirement age was particularly associated with some forms of non-standard work, less education, and lower status occupations. Consistent with the point made above, women who would like to work beyond age 60 were more likely to have a tertiary education. These findings highlight the importance of education for women, and the importance of further education and training to women throughout their lifecourse.
Women who had not retired were more likely to be separated or divorced than married or widowed, have more educational and occupational qualifications, and to have dependent children still at home. These findings indicate contrasting motivations for women to remain in employment: the need for an income for many women, including the many previously partnered women who find themselves on their own in their middle years; and the desire of women with more qualifications to continue work that they find satisfying. Such 'career' women are more likely than women with less education and fewer qualifications to have had their children at a later age, and these children may still be living at home or financially dependent when their mother is in her late forties and fifties. Area of residence was not significantly associated with retirement status.
A greater percentage of women who had not yet retired compared with those who were already retired indicated that they will be looking to some form of government support, i.e. an age pension, in their retirement. Women who were separated, divorced or widowed will be more likely to be reliant on the government than those who were currently partnered, or those who had never had a partner.
The two factors that were equally of most importance in women's decision to retire were their own health and their financial security. Being able to access superannuation was only slightly ahead of the desire for a lifestyle change (presumably including the 'sea change' or 'tree change') as a motivating factor for retiring. The need to provide care was also important for over 60% of women.
Clearly, family relationships affect retirement decisions. The birth of a grandchild was linked to retirement. Longitudinal ALSWH data show that the retirement of a spouse or partner, whether recent or more long-standing, is significantly associated with women's retirement, as is a partner's poor health. Although some women retired when their partner was made redundant, or when the partner died, others moved into work, possibly due to the need to bring in replacement income.
Patterns and predictors of women's labour market attachment
Finally, factors associated with changes in labour force attachment and retirement status over the four surveys between 1996 and 2004 were investigated. The models developed include factors that may be either a cause or an effect of a particular work pattern.
'Early retirees', compared with women who were 'always in paid work', were more likely to exhibit characteristics that included being a carer; having poor health; having concerns about money; and having a partner who retired in the previous year. They were less likely to have dependent children; feel satisfied with their work achievements; be more educated. While 'intermittent workers' (compared with women always in work) also had increased odds of being a carer, having concerns about money, and being in poorer health, they were also more likely to live in a rural or remote area.
The characteristics of later entrants into the labour market, or 'late starters', compared with women who were not in paid work at any time point in the longitudinal study, were investigated. Factors associated with increased odds of being a late starter included having more education; being divorced or in a de facto relationship or, in particular, being separated; being satisfied with work achievements in life so far; and experiencing regular time pressure. 'Late starters' were less likely to be carers, to have money concerns, or to be in poorer health.
As already noted, these factors may exist either prior to, or conversely be the result of, a particular work pattern. For example, the relationship between 'early retirement' and having financial concerns is likely to reflect income after giving up work, rather than before.
Future research directions
The results from this investigation give rise to many questions that are worthy of further examination, including a closer look at women's health and use of health services as they are associated with their experiences of work and retirement. As an example, we might ask whether the deleterious effect on this generation of mid-age women's mental health of continuing to work long hours in employment, translates into greater health service costs or retirement earlier than preferred.
There is also the potential for further investigation of associations between caring and retirement. For example, it is of note that working longer part-time hours is associated with poorer mental health than either shorter hours or full-time paid work. These hours often translate into virtual full-time work, and it may be that women with this pattern of paid work are also struggling to cope with substantial caring responsibilities. Such investigations may examine the extent to which caring responsibilities affect or change retirement planning, or conversely how experiences of paid work and retirement influence the strategies employed to cope with issues around caring.
Further information on women's experiences of paid work and retirement will be available as more data are collected in the Australian Longitudinal Study on Women's Health. Retirement questions are included in Survey 5. These data will offer the opportunity to look at changes in women's experiences of paid work and planning for retirement as they move further into their late fifties and early sixties.