Will older workers change their retirement plans in line with government thinking? A review of recent literature on retirement intentions.
This paper reviews recent retirement and retirement intention literature, with a view to assessing the acceptability of growing calls for later retirement and the conditions that may lead to a change in present plans. The review finds broad consensus with regard to the key factors that enter the retirement decision, significant among which are that financial considerations are not always prioritised, and that high rates of involuntary retirement may hold the key to understanding the recent trend to early retirement. Across a broad range of studies, many external factors serve to disrupt retirement intentions, which exceed actual retirement by 1-3 years and desired retirement age by double that margin. The findings indicate an elasticity of around 6 years wherein actual retirement could now move up or down depending on how well revisions to retirement--and labour market policy--accord with the needs and interests of older workers. They also illustrate a related need for more information on the retirement intentions and circumstances of women, whose increasing labour force participation at older ages appears to account for the recent small increase in Australia's average retirement age.
An inexorable structural ageing of the populations of the world's industrialised nations is now occurring. Even conservative projections for these countries indicate a doubling of the proportion that is elderly, vis-a-vis decline in both the proportion and number that is young, over the next few decades. This changing ratio of old to young has significant implications for the labour markets of these nations, especially in terms of the respective numbers at labour market entry and retirement age. Most domestic labour markets are projected to contract in size, first as a proportion of their respective populations, and then numerically.
Recognising their long-term inability to increase the numbers of young labour market entrants, even via massive migration programs (World Economic Forum 2004, p. xx), the governments of most industrialised nations are now anxious to maintain older workers in the workforce. This official encouragement of labour market participation for older workers represents a significant shift in policy direction over recent years. At its centre the new approach involves revision of a broad range of pension and other non-employment income support schemes under which the changes can be approximately divided into two camps: economic 'carrots', through which pension wealth is increased as a result of pension scheme changes and longer participation in the labour force, and economic 'sticks', through which access and entitlement to various pension and superannuation schemes are delayed and restricted (Blondal and Scarpetta 1998; Henkens and van Dalen 2002; Taylor 2001, 2004). Notably, recent literature indicates greater policy activity around the latter, although in Australia it is fair to say that the most recent policy changes favour the 'carrot' approach. Workplace reform is also high on the agenda in most countries.
Little research has as yet been undertaken, however, to assess the extent to which the new direction may (or may not) match the contemporary retirement expectations and orientation toward work of older workers. As Henken and van Dalen (2002) note, early retirement has become almost standard practice for workers in the industrialised world. In the 1970s and 1980s, the early retirement of older workers was one way to alleviate unemployment among the young and, during the 1980s and 1990s, early retirement became a more attractive option than unemployment for a significant number of older workers also affected by the ongoing structural adjustments of the labour market. In other cases, the existence of unemployment and disability pension benefits where the work-test is waived for older workers has encouraged--and continues to encourage--early retirement (Blondal and Scarpetta 1998; Perry 2001; Duval 2003). Older workers' attitudes towards an 'appropriate' retirement age, therefore, are likely to have altered along with these trends, and their changed retirement plans may also now have been in place for several years, making them more difficult to change (Henkens 1999; Phillipson 2004).
Additionally, while recent policy initiatives have tended to focus on financial incentives and disincentives, a broad array of evidence suggests that retirement decisions are significantly motivated by non-monetary factors (Henkens 1999; Taylor 2001, 2004; Henkens and van Dalen 2002; AARP 2005). These include a complex mix of involuntary and voluntary forces, only some of which can be interpreted as 'intentions' or 'decisions'.
In Australia retirement intentions surveys have tended to be limited in the information they have sought or too recent for a comprehensive body of information to have been established. (2) Those that exist have typically canvassed the desire to work in various capacities after reaching pension age (and thus be semi-retired), but not amenability to change current plans. Most have been limited to specific geographic areas or industries, such as the Australian Public Service (e.g., Ministry of [Western Australia] Premier and Cabinet 2001; Commissioner for [ACT] Public Administration 2002; Management Advisory Committee Report 2003). Nor has the effectiveness of recent policy reforms to alter retirement intentions or behaviours as yet been tested in Australia, again a reflection of their recency. It could also be said that there has yet been little engagement with the concept of retirement itself and the extent to which it can now be analytically operationalised (cf. Borland 2005).
This paper reviews a selection of recent retirement intention literature with a view to assessing the acceptability of later retirement and the conditions, including recent policy changes, that may lead to a change in any present plans and/or encourage longer participation in the labour force. It does not attempt an inclusive coverage of the very large background literature as much of it pertains to the still-recent period when early retirement was being encouraged; much also traverses the same ground. Overall it finds broad evidence that the much promulgated 'early retirement crisis' may have had less to do with retirement intentions or decisions than with involuntary retirement and neo-liberal workplaces and policies.
A useful review of a broad range of factors affecting retirement behaviour across 12 OECD countries over the period 1998-2000 (Taylor, Tillsley, Beausoleil, Wilson and Walker 2000) identifies three major themes. The first is the complexity of the inter-relationship between the economic, social, structural, organisational and attitudinal variables associated with the retirement decision; the second, that all countries covered by the review were in the process of orienting their policies towards the retention of older workers but at the time of writing it was difficult to assess the impact of policy interventions because of their recency (but see also Blondal and Scarpetta 1998 below, who model the potential impacts). Taylor et al's third key theme--and possibly their most profound contribution--is that much recent policy making has proceeded on the basis of an incomplete understanding of older workers' orientation to work and how retirement behaviour may change in response to policies aimed at delaying retirement (also Taylor 2001, 2004). This is different to simple evaluation of the impact of the policies, which are often based on simplistic 'black board economics' cost-benefit foundations and directed at individual actors (Henken and van Dalen 2002).
In general, and as is increasingly well-recognised, both 'push' and 'pull' factors enter the decision to retire: it is seldom a uni-dimensional decision. Economically-oriented factors revolve around the widely-acknowledged incentive and disincentive aspects of public and private pension schemes and their interaction with specific labour market policies. The dissipation of retirement benefits for those continuing to work beyond the normal retirement age has been quantified by many. Gruber and Wise (2005), for example, examine the biases for early retirement and against retirement beyond the normal age in nine European countries, Japan and the US, and find that the penalties imposed by existing social security provisions are often severe. The implicit tax rate on retirement incomes past the normal age is at its lowest in Japan but stands at around 90 per cent in the Netherlands. The universal findings from this study indicate a strong correspondence between the age at which benefits are available and departure from the labour market (see also Kune 2003).
The retirement decision-making environment is much broader, however, than purely economic factors. Among their key findings, Taylor et al. (2000) also point to the role of employer attitudes and behaviour in contributing to early retirement, finding that generally positive attitudes towards older employees are not always met with actions to retain them; and to differences in individual orientation to work in terms of a range of psycho-social, workplace, health, gender and socioeconomic factors. In Taylor et al's study the importance of the latter dimension was unclear, with conflicting evidence emerging both across and within the countries studied--hence their concerns with the incompleteness of knowledge as a basis for policy development.
This is an important dimension with which to engage. As Henkens and van Dalen (2002) argue, there are at least four assumptions underlying retirement policy reform that may be questioned. These are that financial incentives work, that the intrinsic rewards of work are not as important as financial factors, that the retirement decision is an individual decision, and that the different routes to retirement (such as via a disability pension) are unrelated. In each case these authors proffer plausible challenges to convention. They cite a range of studies which show that financial incentives explain only a small part of long run retirement trends; that intrinsic factors such as job satisfaction play an important part in the decision-making process; that the decision is strongly influenced by the conjugal (and other family) relationship; and that unemployment and disability schemes do provide important alternative routes to early retirement.
Velladics, Henkens and van Dalen (2005, Table 5) also identify a sizeable gap between the average age of exit (actual retirement age) from the labour force, preferred retirement age and expected retirement age, across 11 Eastern and Western European countries around 2001. The preferred age of retirement was typically the lowest (low to mid-50s), above which was the actual age of retirement (mid-late 50s), and above that again, the expected or intended age of retirement (late 50s-early 60s). These data give a valuable longitudinal illustration of the process of retiring, with few achieving their 'preference' for an early exit but equally few being able to work as long as they expected or perhaps 'needed' to. (3)
Notably many older workers who lose their jobs do not retire as such, but come to realise that they have retired when they fail to get another job. For many, this process renders the concept of the retirement 'decision' moot. The insight is important. Taylor et al. (2000, p. 2) propose that longitudinal studies would be preferable for studying retirement behaviour because they would enable better exploration of the retirement decision-making process; in particular how individuals weigh up the various factors when making the decision. They also draw attention to the fact that several important items are often omitted from studies on retirement behaviour, such as the role of gender, ethnicity, disability, socio-economic group and, in some cases, even age (see Williamson and McNamara 2001 on the interrelated role of age, race, gender and retirement in the US). Arguably, individual orientations towards work and retirement will differ not only by these discrete characteristics, but also by their particular combination, so that their absence from such data collections will constrain the ability of researchers to undertake more comprehensive analyses, particularly regression.
Incorporating a broad range of such variables, an international review and econometric analysis of 16 OECD countries by Blondal and Scarpetta (1998) offers important insights into the push and pull-and mutually-compensating-effects of a wide range of government-mandated pension and other income support schemes. (4) These authors also confirm that across the 1990s old-age pension systems in most OECD countries made it financially unattractive to work after the age of 55, and that these disincentives had increased markedly across the decades. They show that the variation in older worker participation rates across these countries and time can be explained by the various features of old-age pension systems, including their respective replacement rates, age of entitlement and pension accrual rates (5)--in addition to differing structural conditions in local labour markets (note that Australia was not included in this study, and that Australia's pension and income support system differs markedly--see for example Castles 1994). They caution, however, that many of the recent revisions to these systems designed to increase participation after age 55 will have little effect, because in most cases they will continue to be countered by a range of de facto early-retirement programs, such as unemployment and disability schemes. The latter, they argue, simply provide alternative routes to early retirement (see also Henkens and van Dalen 2002; Taylor 2004; Gruber and Wise 2005).
While acknowledging the possible negative political implications of restricting access to unemployment and disability income support schemes by older workers, Blondal and Scarpetta (1998, pp. 45-46) conclude that co-ordination of the reforms of the different age-pension and non-employment benefit schemes is imperative. Modelling the effects of policy simulations that generate actuarially-neutral systems (where there is no incentive to retire early and no disincentive to continued participation) indicates that such systems would increase the participation rates of those over the age of 55--albeit more in some countries than others. But they also acknowledge the participation-enhancing effects of higher educational attainment, service sector employment and self-employment, each of which is anticipated to increase among the older population in coming years and may well generate a 'natural' increase in participation.
The foregoing arguments are given some qualitative context by the AARP International Retirement Security Survey (AARP 2005) which, while focussing on retirement income security, also explored attitudes to and behaviour surrounding retirement issues across 10 OECD countries. (6) Implicitly agreeing with the idea of extended participation at older ages, less than one in three AARP respondents (31 per cent) expect to retire without some form of ongoing participation, with proportions ranging from 14 per cent in the United States to 47 per cent in Italy (that is, at 14 per cent the U.S. has the lowest proportion planning to stop working completely at retirement, while Italy at 47 per cent has the highest). Most respondents planning to continue some form of work desire to either work part-time (19 per cent), alternate between work and leisure (16 per cent), or do volunteer work (17 per cent) (AARP 2005, p.17).
Only four in ten (37 per cent) of these respondents, however, rate the need for money as the primary motivating factor for continued participation. Instead, the main reasons are a need to stay connected (45 per cent) and enjoyment of work (43 per cent). These proportions are consistent across currently employed people, those who are not currently working, and current retirees (AARP 2005, p. 18). They are also consistent across those who are married and those with incomes in excess of US $40K (AARP 2005, p. 85). The findings thus challenge the convention that pension reform should be the primary focus of policy revision; they certainly indicate a concomitant need for work-place reform.
It is also important to note that nearly one in five (17 per cent) of AARP respondents had given no thought at all to retirement, while only one in four rated as good the level of information they had in order to have a happy and successful retirement (AARP 2005, pp. 9-10). Australians were among the most optimistic about retirement of respondents from the 10 OECD countries. But they were also among the least likely to have thought very much about retirement, or to have given it any thought at all (AARP 2005, pp. 30, 33).These deficits, along with the widespread lack of confidence regarding the capacity of governments to provide adequate pensions in the future, imply an unmet need for public education campaigns about retirement policies and planning.
The United Kingdom
Focussing on the UK labour market only, Humphrey, Costigan, Pickering, Stratford and Barnes (2003) incorporated a broad range of social variables in their study on the continued labour market participation of 50-69 year olds. (7) They found that health status played a major role, with one-fifth having to retire or leave a job because of poor health--either their own or that of a family member (caring responsibilities had affected the paid work of 20 per cent of respondents). Ill-health was also cited by 50 per cent of those not working as a reason for not seeking future employment. Relatedly, Humphrey et al. found that people appear to redefine their personal status as they move past the official State Pension Age (SPA), with a significant drop in the proportion of both men and women who previously identified as long-term sick or disabled. Since health problems and disabilities tend to become more prevalent with age, this trend is counter-intuitive, and must therefore reflect personal redefinition.
Other main correlates of not working and not seeking future employment were financial security (22 per cent) and caring responsibilities (12 per cent). In general, men in this study were more likely than women to have retired early, as were those with tertiary qualifications, access to private superannuation, and previous employment as managers, professionals or associate professionals.
By contrast, self-employment was identified as a significant factor in delayed and transitional shifts to retirement, with semi-retired respondents disproportionately likely to be self employed. This was somewhat different to the situation for semiretired employees, who were much less likely than other employees to have a permanent contract. At the same time, most of the semi-retired (around 89 per cent) were working part-time and expressed greater satisfaction with their hours and jobs in general than did those working full-time.
Perceived retirement age was given an interesting construction by Humphrey et al., who argue that the setting of a clearly defined age of access for a state pension (SPA) establishes (or reinforces) a cultural expectation of retiring at or close to that age. In this study over half of those who had fully retired had done so before reaching SPA. While of those who retired at SPA around half claimed that it was what they had always expected to do, the other half claimed that it represented their first opportunity to access the pension. As was noted in several other studies reported here, however, detailed knowledge about pensions and their features was also relatively low among respondents who were still working.
Several of the foregoing findings are supported in an analysis of secondary data on those still working after reaching official pension age. McKay and Smeaton (n.d.) (8) found that a high proportion of those still working beyond pension age were self employed, although these authors interpret this as reflecting the phenomenon of later retirement among the self-employed per se, as opposed to any significant post-pension-age transition to self employment. Part-time employment and fewer work hours featured highly among these delayed retirees, as did job satisfaction, working in smaller firms, continuous employment, and a particular wish to retire at the same time as a partner. Persons holding mortgages were more likely (than tenants) to be working post-pension age, as were those wishing to improve their financial circumstances for retirement, and/or not dependent on an age-pension. Marital status was also a significant predictor of working beyond pension age, with single or divorced women particularly likely to continue working. Those employed in the distribution, hotels, restaurants and 'other service' occupations were more likely to work on than those in the manufacturing and construction sectors.
A growing range of studies focussing on specific occupational and industrial groups add further to this picture. Surveying 2000 general practitioners in Britain, Sibbald, Bojke and Gravelle (2003) (9) identify an interaction between declining job satisfaction, increasing age, and a dramatically changed working environment (in terms of changes in the organisation and governance of general practice) as the principle cause of labour market exit by doctors. Specifically, they see doctor discontent as related to difficulties in adapting to the changed working environment, and propose that because this--or its manifestation as job dissatisfaction--is amenable to policy intervention, attention to this factor could increase the supply of doctors (Sibbald et al. 2003, p. 3). They also acknowledge that intentions do not always translate into behaviour. They note, however, that even if only half of the number of doctors they surveyed actually retire when they say intend to, this would still be a concern because of the current shortage of doctors. Ironically, the health status of doctors was an omitted variable in this study.
The findings concur with those of Eagles et al. (2005) who, in a UK study of the retirement intentions of consultant psychiatrists, (10) note the growing number of unfilled consultant positions alongside the fact that 45 per cent of survey respondents plan to retire over the next decade. Survey respondents gave an average intended retirement age of 58 years, with female consultants planning to retire earlier than males. Respondents indicated that they might be induced to retire later if they were able to work part time without any detrimental effects to their pension (67 per cent agreeing with the statement), if they experienced a reduction in workload (63 per cent), could have sabbatical/career break opportunities (49 per cent), a pay increase (40 per cent), increased staffing levels (39 per cent) and a change of job content (31 per cent). Only 7 per cent said that none of these inducements would cause them to delay retirement.
In a study examining the medical professions in Australia, Schofield and Beard (2005, p. 80) similarly conclude that retirement among ageing clinicians and nurses will result in significant workforce shortages within the next five years. Analysing secondary data, (11) this study was also unable to ascertain the health status and many other characteristics of its 'respondents'. It did, however, show that a large proportion of Australian GPs work beyond the traditional retirement age of 65, while nurses retire at younger ages than doctors. An interesting trend was that all GP cohorts worked fewer hours in 2001 than in 1986, and that Generation X GPs worked fewer hours than did their baby boomer counterparts when they were at the same age. The findings are posited by Schofield and Beard (2005, p. 82) most likely to reflect changing income replacement rates, flexibility (or lack thereof) of working arrangements and the health status of medical practitioners. Certainly the trends by age and generation resonate with the findings of Sibbald et al. (2003) that they may reflect organisational changes in the work environment. They may also have a cohort or generational dimension per se.
Examining the Australian situation in general, Shacklock and Brunetto (2005, p. 741) (12) similarly propose that three main factors contribute to older Australians' decisions about retirement: health considerations, financial considerations and organisational policies and practices (which generally translate into employment opportunities or lack thereof). In addition to the role of increasing age as a strong predictor of retirement, their findings emphasise the intrinsic factors of job satisfaction and autonomy (especially control over 'choice' about when to retire) in encouraging or discouraging delayed exit from employment. Integral to, and essentially dominating these factors, however, were flexibility in organisational practice and employee knowledge about the organisation's policies and practices, such as the facility to shift to part-time work. In sum, organisational factors were listed by respondents as of greater import to continued participation (31 per cent of respondents) than either intrinsic factors or financial considerations (25 per cent each). Despite their strong presence in other studies (e.g. ABS 1997, 2000, 2006; Wolcott 1999, p. 78; HILDA 2004; and Warren 2006, which found that ill health or injury was the primary reason for early retirement among Australians), health considerations were of less significance in this particular study's findings.
The finding appears anomalous. Analysing the Household, Income and Labour Dynamics in Australia (HILDA) survey Wave 3 data, (13) Warren (2006, p. 3) found that ill health was the primary reason for present Australian retirees' withdrawal from the labour force, although this differed slightly by sex and age (see also ABS 1998, 2006). (14) For women, family and lifestyle reasons were the main cause of withdrawal (whether partial or full), with ill health second, while for men ill health was the primary cause for full withdrawal and family and lifestyle for partial withdrawal. For women aged between 45 and 64 years, however, ill health was the most common reason for fully retiring, while for women aged 65+ years the main reason was family and lifestyle (Warren 2006, pp. 17-18). For both men and women, ill health as a reason for retiring also reduced with age, suggesting that work stresses and demands have a particularly strong impact during peoples' 40s and 50s, and that those with the financial capacity to retire at those ages may therefore do so. Certainly regression analysis showed that for women, higher superannuation savings correlate with earlier retirement age, although the effect is minor (Warren 2006, p. 3).
An early retirement--age and/or a desire to retire early--was also in clear evidence across all Australian studies reviewed:
* In 1997 the average age at which retirees had retired from full time work was 48 years (58 for males and 41 for females). Seventy-seven per cent of males and 97 per cent of females had retired before age 65, with most males retiring at age 55-64 and females at less than 45 (ABS 1998, pp. 3-4). The main reason for withdrawal was ill-health or injury (over half of those who retired before age 55) and retrenchment (19 per cent of those aged 55-59 years). For those retiring at 65-69 years, the main reason was having reached the appropriate age (ABS 2000).
* Wolcott (1998, p. 4, 1999) shows that just over half of the men in her survey had retired before the traditional compulsory age of retirement (65 years), and that just on half of all men and 63 per cent of all women had been happy to retire when they did. A further small proportion (11 per cent of men and 6 per cent of women) would have liked to retire earlier.
* Delpachitra and Beal (2002, p. 7) (15) found that 61 per cent of their yet-to-retire respondents planned to retire between ages 55 and 60, and a further 27 per cent by 65, meaning that 88 per cent did not expect to work beyond age 65. But they also found that workers adjust their planned retirement age as they live through the years leading up to retirement (2002, p. 3). Using the psychological concept 'locus of control' (personal ability to control), they found that retirement plans premised on factors such as having accumulated sufficient funds are often defeated by external factors such as ill health, inability to find work, lack of job satisfaction and the need to care for others (see also Smith 2000 on OECD countries (16). Significant differences in reasons for retiring were also given by those already retired and those yet to retire, indicating the possible presence of cohort effects.
* The Australian Public Service (APS) Management Advisory Committee (2003) found that around 23 per cent of the APS workforce is likely to retire over the next five years. Some of this is due to the '54/11 effect' the temporary effect of a specific superannuation scheme in which it is advantageous for personnel in the scheme to resign just before their 55th birthdays. Most, however, is due to the ageing of the baby boomer cohort, higher levels of recruitment of highly skilled people, who tend to be older than their lower-skilled entry-level counterparts, and declines in the employment of young people and graduate retention. Thus the effect is deeper than the 54/11 problem alone would imply.
* The estimated intended average age of retirement for the HILDA Wave 1 sample (run in 2002) was 60.5 years, with females planning to retire earlier than males (59.3 and 61.4 respectively) and partnered females planning to retire earlier than partnered males (58.6 and 61.3) (Knox 2003, pp. 6-7). (17)
* Wave 3 of the HILDA survey (run in 2003) found that 77 per cent of the Australian workforce over the age of 45 expects to retire by age 65, and that a greater proportion (81 per cent) would choose to retire earlier if they could (HILDA 2004, pp. 21). Approximately 34 per cent of employed men and 41 per cent of employed women desire to retire permanently by or prior to their 60th birthdays (see also McAlister, Lindenmayer and McLean 2005, pp. 15-18).
* The ANOP study (2004) (18) found a mean retirement age of 59 years.
* A NATSEM (2005, p. 20) study found that of respondents aged 45-64, one-third were already retired and almost three-quarters of the remainder would like to be.
* The AARP (2005. pp. 40-41) survey found that Australians (along with Canadians) planned to retire at earlier ages than their counterparts in nine other OECD countries.
* Shacklock and Brunetto (2005, p. 750) found that (in 2005) one-quarter of their sample planned to retire before age 55, while overall 61 per cent planned to retire at or before 60.
* the Australian Bureau of Statistics' (2006) found that the average age of retirement among all retirees aged 45 and over in Australia in 2004-05 was 52 (58 for males and 47 for females), while among recent retirees it was 60 (61.5 for males and 58.0 for females). (19) This is somewhat older than evidenced in the ABS' 1997 survey (ABS 1998), which found an average retirement age of 48 years (58 for males and 41 for females) but discontinuities in the data and attrition rates due to mortality make it difficult to be sure. In 2004-05, 80 percent of surviving male retirees had retired by age 60, as had 94 per cent of females. Among both retired men and women whose last job was less than 20 years ago, the main reason for ceasing that job was 'reaching retirement age/being eligible to receive superannuation or the pension' (34 per cent). Other main reasons were 'sickness, injury or ill health' (26 per cent) and 'being retrenched, dismissed or no work available' (11 per cent). Among those still to retire, the decision was expected to be influenced by personal health or physical abilities (40 per cent), the achievement of financial security (36 per cent) or reaching the eligible age for an old age or service pension (15 per cent). Average intended retirement age was 62 years (63 for men and 61 for women).
relatively to their international counterparts, but also that average retirement age and possibly intended retirement age--may be going up. This, however, may be true only for females. The ABS data, for example, show no change in the average retirement age for males between 1997 and 2004-05, the increase in the total retirement age (from 48 to 52 years over the period) being driven solely by the increase in the female retirement age.
Nevertheless minor increases in the participation rates of older Australians of both sexes have recently been recorded, albeit most for females (NATSEM 2005, pp. 8-9; Colebatch cited in Healey 2002, p. 32). Recent monthly labour force statistics also confirm these trends (ABS June 2006). Figure 1 shows the shift by birth cohort, indicating a small upward movement at age 60-64 for the cohort born 1939-43 (currently aged 62-66 years and in the process of retiring). For males, participation at this age is not yet back to that experienced by the cohort born 1914-18 (which retired around 1980), while it is substantially exceeded for females. These are very interesting trends, because they indicate a possible reversal beginning in the late 1990s, which would pre-date both the recent government interventions and general coverage of related issues in the media. A similar shift has been recorded in the United States, with a break in the trend to early retirement being temporally located in the 1980s (Quinn 1999 in van Dalen and Henkens 2002, p. 227). Van Dalen and Henkens (2002, p. 228) also note a small rise in participation rates of older workers in The Netherlands, although they put it down to favourable business cycle conditions rather than a although they put it down to favourable business cycle conditions rather than a reversal of early retirement per se.
[FIGURE 1 OMITTED]
Whether or not these trends will be maintained--or, in the case of Australian females, participation levels reach those projected by the Productivity Commission (2004)--is a question that only time can answer. In the interim, we can get an indication of the answer by reviewing the thoughts or intimations of Australians about working longer.
Are Older Australians Interested in Working Longer?
On the whole, Australian retirees adjust well to retirement, with most claiming to enjoy being retired and to have about the same (52 per cent) standards of living as previously, or better (21 per cent) (HILDA 2004, p. 21). These proportions may also be increasing, with Australian Seniors Finance (2005) (20) citing these variables at 60 and 30 per cent respectively. Retired Australians also disproportionately feel that people do not respect them less than when they were working, that their self-esteem and self-confidence are high, that they feel less stress than when working, and that they do not expect to become bored with retirement (see also Wolcott 1998, pp. 11-12 (21); McAllister, Lindenmayer and Mclean 2005). The overall message is that most enjoy their freedom. These are important findings, because they tell against the idea of mass delayed retirement.
For a sizeable minority, however, retirement does not appear to be a totally fulfilling experience, and it is from this group that we may expect to glean particularly useful insights into the propensity for delayed retirement. The recency of data collection on this topic means that little research has yet been published, but there are growing indications that ultimately the level of satisfaction with retirement correlates most strongly with the level and source of retirement income. The most recent ABS survey of retirement intentions (2006), for example, found that over 200,000 people who had previously retired had either returned to work or were looking for work, most claiming financial need (46 per cent), or boredom/the need for something to do (35 per cent).
What literature exists on the topic also provides evidence of a strong orientation to work among many older Australians. A 1996 survey of older Australians (22) found that 60 per cent of older working women and 70 per cent of their male counterparts agreed that they would prefer to have a paid job even if they had a reasonable income without work, while 45 per cent of women and 80 per cent of men said that they would be unhappy if they were not in paid work (Wolcott 1999, p. 78). For these people it was the intrinsic value of paid work that was the issue, not the money per se. Of those aged 50-70 years who had already retired, 39 per cent of men and 31 per cent of women also indicated they would have preferred to delay their exit from paid employment. The patterns reflect the above-noted discrepancy between intended and actual retirement age, with the latter occurring somewhat earlier than the former (also Wolcott 1998, pp. 3-5)., The study found that those still in employment, perhaps yet to succumb to this discrepancy, appeared to be planning to remain in employment until later ages than those who had already retired: only 12 per cent of currently employed men and women over the age of 50 were thinking of retiring before age 60.
Responses to a MatureAge Workers Issues Paper put out by the National Seniors Association (23) (NSA Australia n.d.)identify a related issue, that of discouraged workers and long term unemployment among the growing numbers of divorced people in their fifties. Many of these people are finding themselves with young (second) families or unexpectedly single and financially insecure. In response to the question whether NSA members wanted to work longer, feedback was mixed. On the one hand, members cautioned against enforced upward shifts in the age of access to pensions, believing that the choice should be personal and informed; also that health issues often militate against longer participation. On the other, many acknowledged that retirement is not an option for all, that many of those forced on to the dole or age pension at an early age would have preferred to keep working, and that for many the retirement-driven loss of mental stimulation was real. There was also consensus on the need for continued access to training and a change in corporate attitudes to mature workers (see also Shacklock and Brunetto 2005, p. 741, and Horin cited in Healey 2003, p. 30).
Surveying the West Australian Public Service in 2000, the Ministry of the Premier and Cabinet (2001) found widespread support for casual or contract employment in the Public Service after retirement, with 78 per cent of respondents indicating that they would consider this option if it were available. Reduction of working hours leading up to retirement and reduced responsibility were also favourably cited, although it is notable that these features have been available across the Public Sector for some years. The Ministry (p. 49) also acknowledges the important issue of the gap between intended retirement age and actual retirement age, citing evidence that many employees retire at a younger age than they indicate when asked their intention.
A retirement intentions survey administered to the ACT Public Service in 2002 (CPA 2002) (24) similarly found that while 21 per cent of respondents intended to retire over the following five years (a retirement rate treble that of the previous five years), 65 per cent agreed with the statement that they would find transitional arrangements such as reduced hours (53 per cent) and working from home (25 per cent) attractive, and would be likely to defer retirement if these were available (CPA 2002, pp. 3-7). Only 35 per cent agreed with the statement that none of these arrangements would cause them to defer retirement. Notably there is much diversity within this population, with some occupational groups experiencing much higher rates of intended exit than others (e.g. teachers 54 per cent; executives 50 per cent).
The HILDA survey adds to this picture of retirement-prone and non-retirement-prone groups. Private sector males plan to retire slightly later than those in the public sector (61.9 compared with 59.7 years--a difference probably attributable to the relatively generous provisions of public service superannuation schemes (Knox 2003, p. 8). (25) University educated persons plan to retire later than others, as do those who own their home as opposed to renting. Many differences exist by sex, however For example, the higher the expected payout on retirement for women, the later the intended retirement age for women; the opposite being true for men.
Citing a survey run for the government in 2000 (26), Knox (2003, p. 17) also reports that approximately 65 per cent of Australian males and 62 per cent of females who had retired before the age of 60 had done so involuntarily (inclusive of illness, caring responsibilities etc). Of these, 24 per cent would have preferred to be still working, indicating as above that for this group actual retirement age is lower than intended retirement age, with the gap being greater for women. Of those who had retired voluntarily, 54 per cent of males and 43 per cent of females had done so 'because they could afford to', at the same time indicating that there are many reasons for retirement other than financial ability, especially for females. Knox argues, however that since more than half who had retired voluntarily did so when they could afford to, the size of the sum remains relevant and any reduction in it could delay retirement (Knox 2005, p. 19). Consequently he argues that should the low returns on superannuation that were extant around the time (2000) persist over a long period, actual retirement age could even exceed intended retirement age, while an increase in returns could have the opposite effect. It is plausible that the former is one reason behind the increases in labour force participation shown above in Figure 1, as the periods coincide.
Although the proportions are lower, the HILDA survey Wave 3 data, collected in 2003, similarly confirm that not all current retirees consider their retirement to have been truly voluntary, with little more than half (54 per cent) seeing it as so (HILDA 2004, p. 21). Thirty-six per cent felt they had been pressured into retirement, with the remaining ten per cent citing a mixture of both. Pressure to retire [early] came had come from employers (41 per cent) and medical advice (42 per cent), while 17 per cent (mainly women) had experienced pressure from their partners.
Studies using the Wave 3 data have also drawn attention to the complex patterns of part- or semi-retirement that exist and the variety of meanings that people attach to the concepts, with people in similar circumstances sometimes having different perceptions of whether or not they were retired (McAlister, Lindenmayer and McLean 2005, p. 3; see also Warren 2006). While the majority of Wave 3 respondent retirees had retired by their early 60s (58 per cent), a further 16 per cent deemed themselves to be only partly retired (HILDA 2004). For this group the status meant working fewer hours (46 per cent), working only casually or occasionally (42 per cent), taking a less demanding job (23 per cent) or doing voluntary or charity work (17 per cent) (HILDA 2004; see also ABS 1998, 2000).
The distinction between working fewer hours or partially retiring--transitioning to retirement--is particularly pertinent for women. Merkes (2003) (27) examined women's perceptions about working beyond age 65, and found that many are open to the possibility but wish to do so at a different pace. Respondents stressed the importance of a work-life balance, and reported an equal interest in volunteer work as in formal employment. Women with high qualification levels and in high status occupations were more likely to consider later retirement than those in low status jobs, although for many the lack of sufficient financial support to retire was a key deterring factor (see also Olsberg 2002, 2004; Clare 2004; Warren 2006).
Summary and Conclusion
A broad literature identifies the following factors as contributing to the early retirement 'decision', although the findings are not without contradiction; notably a number of writers draw attention to the lack of multivariate analyses that would assess the relative importance of the various factors along with the extent to which they interact, and thus potentially resolve these contradictions:
* current age, which determines both access to pensions and/or the level of superannuation at retirement, and contributes to perceptions about when it is appropriate to retire;
* pension and superannuation scheme availability and attributes (which can either 'push' people into a longer working life or 'pull' them into retirement). This includes knowledge about pension scheme characteristics, which appears lower for those still working than for the retired;
* pension wealth/implicit tax on age-pension or superannuation (high pension wealth and/or high implicit tax contribute to early retirement);
* de facto early-retirement schemes, such as unemployment and disability pensions, that facilitate an alternative path to early retirement;
* positive organisational policies, attitudes to and practices regarding older workers, and flexible working arrangements, particularly the availability of part-time work, which contribute to delayed retirement (employee knowledge of these is very important);
* job satisfaction and the psychosocial aspects of work (being valued, feeling that the work is of value, having a supportive supervisor, a high level of autonomy, challenging/interesting work), all of which contribute to delayed retirement;
* health status (poor health of either self or spouse hastens retirement; however health status appears to be re-defined as people move toward retirement, indicating the impact of work stress rather than ill-health per se);
* employment and labour force status. Self-employed men tend to retire later, self-employed women earlier, than employees; those with access to part-time work are more likely to work longer; and current employment increases the probability of delayed retirement;
* educational status (the tertiary educated tend to work longer);
* occupation (unclear--in some studies, managers, professionals, semiprofessionals are likely to retire earlier than those employed in manufacturing and construction sectors; in other cases--e.g. doctors--they work longer);
* housing status (those with their own home or a mortgage tend to retire later than those who rent);
* indebtedness (higher debt equates with later retirement);
* sex (females tend to retire earlier than males, but not in all studies; females with insufficient superannuation tend to work longer than females with sufficient super, but not in all studies);
* partnered status/dependents/caring responsibilities (including divorce/ second families), which both constrain and contribute to delayed retirement (constrain when family responsibilities have to take priority; delay when compounded with indebtedness; men with older children retire earlier; single people, especially women, retire later than married people; spouses tend to retire together);
* health insurance (lack of it may cause people to delay retirement);
* autonomy as regards choice about when to retire (later retirement when locus of control is high and not compromised by external factors, such as displacement).
Reflecting the last point, many studies allude to the fact that people adjust their retirement plans as they transit the retirement age zone, typically because of external factors beyond their control. Involuntary retirement appears particularly high (40-65 per cent of retirees), with most studies consistently reporting that 10-40 per cent of these people would prefer to be still working. A resulting gap between intended and actual retirement age appears to be consistent across most studies which have sought this information, with actual retirement occurring 1-3 years earlier than intended retirement, especially for women. But actual retirement also occurs later than the stated desired age of retirement (where this is asked), typically by 1-3 years. This 'desired-actual-intended' sequence indicates an elasticity of around 6 years that, on the one hand, could see retirement age move slightly upwards, from actual to intended, but, on the other, slide back from actual to desired, should the conditions that facilitate either sequence of events be met. Importantly, McAlister, Lindenmayer and McLean (2005, p. 20) make the salient point that although a minor increase in age at retirement appears to have recently occurred in Australia, this should not be confused with an upward shift in the age at which Australians desire to retire.
Also of some significance is that, almost universally, financial security or 'more money' is rarely given as the priority factor that would ensure continued participation (although this may change once retirement becomes a reality). By contrast, it would appear that people wishing to work longer broadly desire a working environment that is the opposite to what they presently have: a 'small firm' in which they and their work are individually valued, in which they are not physically or psychologically stretched (or stressed) beyond reason, where there is flexibility regarding working hours and leave arrangements, and the needs of family can be more readily accommodated. Ironically, the family-friendly workplace currently being mooted as essential for the labour force participation of young families (especially women) may be equally as or even more important as retirement approaches; contrary to popular perception, older workers often have multiple layers of family responsibility: ageing parents, partners, middle-aged children and grandchildren. The extent to which such conflicting pressures may be entwined in the 'health concerns' (or work stress) also consistently recorded as a key reason for early retirement, raises the question as to the role of the neo-liberal workplace itself in driving the trend.
Indeed, as Henkens and van Dalen (2000) argue, current approaches to 'the problem' miss many of the main issues. This is because much of the logic surrounding the development of incentives to continued participation for older workers relates to individual motivations, while in reality the retirement decision is most often a joint decision and/or imposed by external circumstance. The decision to leave work is strongly shaped within the conjugal relationship (Henkens and van Solinge 2002), while organisational policy arrangements surrounding such factors as joint pension access and entitlement rules play a significant role, as do other structural factors over and above implicit tax rates on pension and superannuation entitlements, such as rules relating to the governance and management of the workplace in general. While involuntary retirement is widespread, voluntary redundancies are also implicated in the early retirement trend (Patrickson and Clark 2001). Another part of 'the problem', then, would appear to be the neo-liberal framework within which retirement and workplace policies are conceptualised.
Contradictions also abound in the rationale behind the concept of private provision for retirement and its impact on delayed retirement. Studies almost universally show that it is those with access to private superannuation--and the highest levels of it--who are most likely to take early retirement. While increasing coverage will reduce pressure on the public purse, and this is of course the motivation for encouraging an increase in the retirement age, paradoxically increased coverage may contribute to early retirement unless other issues, such as the family-friendly or regulation-unfriendly workplace, are attended to (Perry 2001; Perry and Freeland 2002).
Further, Sheen (2001) argues that much of the debate about the ageing workforce and retirement involves significant gender bias. Specifically, she argues that it reflects the collapse of full-time, permanent employment in male-oriented work such as manufacturing and middle management, and its impact on males; and that, relatedly, the idea of retirement and early retirement is based on male experiences of full-time, full-year, permanent employment. By contrast, a continuing increase in female participation is clear, especially at the older ages (refer back to Figure 1), with women now comprising 45 per cent of the Australian labour force, up from 35 per cent in the 1970s. It is extremely difficult even to calculate an average retirement age for women, with so many leaving and re-entering the workforce for family reasons across their life cycle (Olsberg 2004). Instead, the issue relating to delayed retirement for women is not low or declining participation, but lack of equity in their ability to build superannuation resources (Perry 2001; Olsberg 2004; Clare 2004; Warren 2006). (28)
Viewed from these perspectives, recent policy initiatives designed to increase the age of retirement may have some distance to go. In Australia they appear to take the early retirement 'decision' as opposed to involuntary or imposed retirement at face value, and are almost entirely directed at gender-neutral financial factors. (29) Specifically, they include pension bonuses for those who will eventually receive a government pension but choose to work beyond age 65; the ability to access private superannuation while continuing to work, once the preservation age (30) has been reached (for example, 55 years in the case of people born before 1960); the ability to have superannuation income streams turned on and off if/when reentering the labour force on a temporary basis; and from July 2007 an historic shift to tax-free superannuation after age 60 and a vastly simplified superannuation and taxation system. (31) The measures also include the 2004 introduction of an Age Discrimination Act (NATSEM 2005, p. 20) (32) which has the objective of ensuring the immutable right of older workers to work.
Notably, the initiatives directed at earlier access to private superannuation represent an important departure from earlier plans, which had been moving towards raising the preservation age from 55 to 60 (Ministry of Premier and Cabinet 2002, p. 41). Arguably there is now a tasty 'carrot' whereby these funds will be tax-free after age 60, and this may reduce the numbers accessing them at age 55. At the same time, it is likely that the moves will add to, rather than reduce, the contradictions already in place, permitting semi-retirement to drift downwards towards desired retirement age, instead of upwards towards intended age. Writing somewhat ahead of these policy changes, Perry (2001, p. 39) argues that in the context of trying to encourage delayed retirement, the ability of those with private superannuation to access it from age 55 under whatever circumstances sends a confusing message because it both sets and reinforces the normative example of a socially sanctioned retirement age.
The policies as they are presently formulated would also appear to do nothing to address specifically the situation of almost half (45 per cent) of the labour force--women whose greater demands from family responsibilities will continue to expose them to weaker lifetime attachment to the labour force; nor are they likely to do much for those many older workers who suffer involuntary displacement--or accept a voluntary redundancy and then find themselves 'retired'. Information on how these policies, circumstances and other ideas about delayed retirement actually accord with the views and interests of older Australian workers (especially women, whose increasing participation may well be behind the recent small increase in the retirement age), and what other incentives and initiatives might be considered, is therefore urgently needed: the phenomenon of early retirement may have had very little to do with individual intentions.
Moreover the concept of retirement remains in need of critique. If a discourse analysis were applied to the literature reviewed herein it would generate two overwhelming conclusions. One is that preservation ages and their associated policies create a discourse that determines, normalises and reinforces an appropriate age (or stage) of retirement. Discourse--the language and concepts we use assigns--meanings and causes to phenomena that both shape the way we think about them and delimit the way we conceive of and react to them as a social policy issue: retirement is definitely one such phenomenon. The other is that for most people retirement is a process, not an act, and therefore policies that treat it as an act are likely to deliver that outcome: an abrupt departure from the labour force, rather than a slow transition. In this sense Australia's present approach may be on the right track, albeit still a narrowly-defined track based on a rather gender-blind reading of the issues.
AARP (2005), International Retirement Security Survey, HarrisInteractive, www. harrisinteractive.com (15th January 2006).
ANOP (2004), Community Attitudes to Saving for Retirement: Overview of Main Findings and Implications, ANOP Research Services.
Australian Bureau of Statistics (1998), Retirement and Retirement Intentions Australia 1992 and 1997, Catalogue No. 6238.0.
Australian Bureau of Statistics (2000), Australian Social Trends 2000, Catalogue No. 4102.0.
Australian Bureau of Statistics (2005, 2006), Labour force status by social marital status, age and sex, Table 01, Labour Force, Australia, Detailed--Electronic Delivery, Catalogue No. 6291.0.55.001.
Australian Bureau of Statistics (2006), Retirement and Retirement Intentions Australia, August 2004 to July 2005, Catalogue No. 6238.0.
Australian National Seniors Association Employment for mature age workers issues (no date), www.nationalseniors.com.au/Policies%20New/ mature%20Age%20Workers.htm (15th January 2006).
Australian Seniors Finance (2005), Attitudes of Senior Australians, Study No. 14457, Sweeney Research Consultants, Melbourne and Sydney
Blondal, S. and S. Scarpetta (1998), The Retirement Decision in OECD Countries, Economics Department Working Papers No. 202, Organisation for Economic Co-operation and Development, www.oecd.org/eco/eco (15th January 2006).
Borland, J. (2005), Transitions to Retirement: A Review, Melbourne Institute Working Paper No. 3/05, Department of Economics and Melbourne Institute of Applied Economic and Social Research, University of Melbourne,.
Castles, F. (1994), The Wage Earners' Welfare State Revisited: Refurbishing the Established Model of Australian Social Protection, 1983-1993, Discussion Paper No. 39, Australian National University Graduate Program in Public Policy.
Clare, R. (2004), 'Why Can't a Woman be More Like a Man: Gender Differences in Retirement Savings', paper presented to the ASFA 2004 National Conference and Super Expo Super: Saving 4 the Nation, Adelaide Convention Centre 10-12 November.
Commissioner for Public Administration (2002), Retirement Intentions Survey for the ACT Public Service, ACT PS Human Resources Council.
Delpachitra, S. and Beal, D. (2002), 'Factors Influencing Planning for Retirement', Economic Papers, vol. 21(3), pp. 1-13.
Duval, R. (2003), The Retirement Effects of Old-age Pension and Early Retirement schemes in OECD countries, Economics Department Working Papers No. 370, OECD Publishing, doi:10.1787/308728704511.
Eagles, J.M. et al. (2005), 'Retirement Intentions of Consultant Psychiatrists', Psychiatric Bulletin vol. 29, pp. 374-376.
Gruber J and Wise D (2005), Social Security and Retirement around the World, University of Chicago Press.
Healey, J. (2003), 'Employment and Retirement', Chapter 3 in Healey, J. (ed), Our Ageing World, Sydney, The Spinney Press, pp. 29-40.
Henkens, K. (1999), 'The Decision to Leave Work', Demos, Netherlands Interdisciplinary Demographic Institute, vol. 15, (www.nidi/ral/public/demos/ dm99epc3.html) (15th January 2006).
Henkens, K. and van Dalen (2002), P., Early Retirement Systems and Behaviour in an International Perspective, Netherlands Interdisciplinary Demographic Institute (NIDI), The Hague, unpublished paper provided by authors.
Henkens, K. and van Solinge, H., 'Spousal Influences on the Decision to Retire', International Journal of Sociology, vol. 32(2), pp. 55-74.
HILDA (2004, 2005), Living in Australia, HILDA Survey Annual Report, Melbourne Institute of Applied Economic and Social Research, University of Melbourne.
Humphrey, A., Costigan, P., Pickering, K., Stratford, N., and Barnes, M. (2003), Factors Affecting the Labour Market Participation of Older Workers, Research Report 200, Department for Work and Pensions, www.dwp.gov.uk/asd/(15th January 2006).
Knox, G. (2003), 'Retirement Intentions of Mature Age Workers', paper presented to the Australian Social Policy Conference, 9-11 July, University of New South Wales.
Kune, J.B. (2003), On Global Ageing. Old-Age Income Systems in the EU and Other Major Parts of the World, Contributions to Economics, Physica-Verlag in association with the Netherlands Foundation for Pension Studies, The Hague, Netherlands.
Management Advisory Committee (2003), Organisational Renewal, Commonwealth of Australia.
McAllister, D., Lindenmayer, P., and Mclean, P. (2005), Three Dimensions of Retirement--Aspirations, Expectations and Outcomes, Paper prepared for the HILDA Conference 29-30 September 2005, Melbourne.
McKay, S. and Smeaton, D. (n.d.), Working after State Pension Age: Quantitative Analysis (summary), Research Report No. 182, Department for Work and Pensions, http://www.dwp.gov.uk/asd/asd5/182summ.pdf, (28th January 2006).
Merkes, M. (2003), A Working Future for Australian Women of the Baby Boom Generation? Work after Age 65:Women's Perceptions and the Policy Implications of an Ageing Female Labour Force, LaTrobe University, http://member.melbpc.org.au/~monika/vcoss_congress_2003.ppt (28th January 2006).
Ministry of the Premier and Cabinet--WA (2001), Workforce Beyond 2000, Public Sector Management Division.
NATSEM (2005), May the Labour Force be with You, AMP:NATSEM Income and Wealth Report, Issue 12, University of Canberra.
Olsberg, D. (2002), Women, Ageing and Retirement Challenges and Long-term Strategies, a submission to the Standing Committee on Ageing Inquiry into Long Term Strategies to Address the Ageing of the Australian Population, University of New South Wales Research Centre on Ageing and Retirement, http://www.aph.gov.au/house/committee/ageing/strategies/subs/sub84.pdf (15th January 2006).
Olsberg, D. (2004), 'Women and Superannuation: Still Ms ... ing Out', Australian Journal of Political Economy, vol. 53, pp. 161-178.
Productivity Commission (2004), Economic Implications of an Ageing Australia, Draft Research Report, Productivity Commission Canberra.
Perry, J. (2001), 'The New Moral Economy of Retirement', Growth Vol. 49, pp. 33-40.
Perry, J. and Freeland, J. (2002), Too Young To Go. A Review of Good Practice in the Employment of Mature Age Workers, Report Commissioned by the New South Wales Committee on Ageing.
Phillipson, C. (2004), 'Work and Retirement Transitions: Changing Sociological and Social Policy Contexts', Social Policy and Society, vol. 3, pp. 155-162.
Productivity Commission (2004), Economic Implications of an Ageing Australia. Draft Research Report, Canberra, Productivity Commission.
Schofield, D. and Beard J.R. (2005), 'Baby Boomer Doctors and Nurses: Demographic Change and Transitions to Retirement', Medical Journal of Australia, vol. 183, pp. 80-83.
Shacklock, K., and Brunetto, Y. (2005), 'Employees' Perceptions of the Factors Affecting their Decisions to Retire', International Journal of Organisational Behaviour, vol. 10, pp. 740-756.
Sheen, V. (2001), Older Women in Australia's Workforce, paper presented to Australian Women Speak, Inaugural National Women's Conference of the Status of Women, National Convention Centre, Canberra.
Sibbald, B., Bojke, C., and Gravelle H. (2003), 'National Survey of Job Satisfaction and Retirement Intentions among General Practitioners in England', British Medical Journal, vol. 326: 22- doi:10.1136/bmj.326.7379.22, pp. 1-4, 4th January, www.bmj.com (15th January 2006)
Smith, T.W. (2000), A Cross-national Comparison on Attitudes towards Work by Age and Labour Force Status, National Opinion Research Centre, University of Chicago, http://www.oecd.org/dataoecd/52/60/2535801.pdf
Taylor, P. (2001), Reversing the Early Retirement Trend, Labour Market Trends, vol. 109, pp. 217-221.
Taylor, P. (2004), Age and Work: International Perspectives, Social Policy and Society, vol. 3, pp.163-170.
Taylor, P., Tillsley, C., Beausoleil, J., Wilson, R., and Walker, A. (2000), Factors Affecting Retirement Behaviour: A Literature Review, Research Brief No. 236, http://www.dfee.gov.uk/research/ (15th January 2006).
United Nations (2000), Replacement Migration, Population Division, Department of Economic and Social Affairs: United Nations Secretariat.
Van Dalen, H.P. and Henkens, K. (2002), 'Early-retirement Reform: Can It and Will It Work?', Ageing and Society, vol. 22, pp. 209-231.
Velladics, K., Henkens, K., and van Dalen, H.P (2005), Do Welfare States Produce Different Individual Policy Preferences?: Opinions on Pension Reforms in Eastern and Western Europe, Working Paper, Nederlands Interdisciplinary Demographic Institute, The Hague.
Warren, D. (2006), Aspects of Retirement for Older Women, Melbourne Institute of Applied Economic and Social Research http://ofw.facs.gov.au/downloads/ pdfs/Aspect_of_Retirement%20_report_final.pdf (10th August 2006).
Williamson, J.B. and McNamara, T.K. (2001), Why Some Workers Remain in the Labour Force beyond the Typical Age of Retirement, Working Paper 69, Centre for Retirement Research at Boston College.
Wolcott, I. (1998), Families in Later life. Dimensions of Retirement, Working Paper No. 14. Australian Institute of Family Studies
Wolcott, I. (1999), 'Older Workers, Families and Public Policies', Family Matters No. 53, pp. 77-81.
World Economic Forum (2004), 'Living Happily Ever After: The Economic Implications of Aging Societies', Executive Summary of a Report to the World Economic Forum Pension Readiness Initiative developed in partnership with Watson Wyatt Worldwide.
(1) This paper was supported under the Australian Research Council's Discovery Grant scheme (Grant No. DP0557922; Will Older Workers Change their Retirement Plans in Line with the Government's Calls for later Retirement?)
(2) Retirement intentions, attitudes and behaviours are best studied longitudinally. The HILDA survey, begun in 2002 and now run over four consecutive years, is slowly providing appropriate data, but published studies as yet remain few due to this recency. Prior toBefore the HILDA survey, Australia had only two sources of information on retirement transitions (Norris and Bradbury 2001 and Commonwealth Department of Family and Community Services 2003, cited in Warren 2006). Notably eEven the HILDA survey included a comprehensive module on retirement in only one of its four data collection waves (Wave 3). The most recent Australian Bureau of Statistics retirement survey (2005) complements and expands its earlier collection from 1998, but even there the range of questions remains limited.
(3) However ilt is also important to reflect on the self-evident point that actual retirement pertains to those who have already retired, while intentions pertain to those yet to retire. These are different groups of people, typically older and younger (respectively), so that the differences may reflect cohort effects.
(4) Note that occupational pensions negotiated between employers and employees were not covered by this study, only government pensions (NB. Australia was not included in this analysis).
(5) Blondal and Scarpetta (1998: 7) provide a clear description of the 'problem': 'An extra year of work implies foregoing one year of old-age pensions and often paying pension contributions for an additional year, with little or no increase in ultimate pensions after retirement. The drop in old-age pension-wealth, i.e. the discounted value of future pension streams minus pension contributions, often amounts to 50 to 80 per cent of gross income, and this implicit tax is likely to strongly discourage work beyond the pensionable age.'
(6) Telephone survey of just over 4,000 respondents aged 30-65 years in 10 countries (G7 plus Sweden, Australia and the Netherlands); approximately 400 respondents per country.
(7) Survey of 2,800 respondents aged 50-69 years carried out by the (UK) National Centre for Social Research in conjunction with the Institute for Fiscal Studies.
(8) Cross-sectional analysis of the (UK) Labour Force Survey, Family Resources Survey, and the British Household Panel Study
(9) National postal surveys of 1,949 general practitioner principals in 1998 and 2001 (of whom 790 were surveyed in 1998 and 1,159 in 2001).
(10) A survey of 387 consultant psychiatrist members of the Scottish Division of the Royal College of Psychiatrists, working in Scotland in 2003.
(11) Secondary analysis of previously unpublished Census data for the years 1986, 1991, 1996 and 2001.
(12) Mixed quantitative and qualitative survey of 97 respondents emailed to two public sector organisations. The sample was heavily biased towards males (81%per cent) and those aged 41+ years (78 %per cent).
(13) The Household, Income and Labour Dynamics in Australia (HILDA) survey, which began in 2002, has now collected four waves of data on the retirement intentions and experiences of Australians of various ages above 45 years, although only Wave 3 included a comprehensive module on the topic of retirement. The Wave 3 survey interviewed 5,756 people aged 45+ years. As yet there are relatively few papers of relevance to this review drawing on these data.
(14) These questions were asked only of people aged 45+ years who had retired before 1990.
(15) Random postal survey of 5000 residents of one suburb each in Brisbane, Ipswich, and Toowoomba, and rural residential Queensland carried out in 2001. 717 responses were received of which usable (complete) forms were received from 313 non-retired and 184 retired persons.
(16) Draws on data from the International Social Survey Program (ISSP) 1997 Work Orientation Study (covers 8 OECD countries but excludes Australia)
(17) Draws on the Wave 1 survey which interviewed 2,428 people in the 45-54 year age range, of whom 1,870 (77 per cent) were currently in paid employment.
(18) A telephone survey of 755 Australians aged 30-69 years conducted in May/June 2004. This survey compared the experiences and savings attitudes of retired and yet-to-retire respondents, with a focus on baby boomers.
(19) The age differs for all retirees and recent retirees (retirements over the past 5 five years) because the former includes only those who are surviving; it is more likely to include those who retired at younger ages (e.g. 40) in any given year than those who retired at older ages (e.g. 65) in that same year.
(20) The study (carried out by Sweeney Market Research Consultants) comprised of two parts--a qualitative component, consisting of 12 focus groups, conducted in Melbourne, Sydney, the Sunshine Coast and Forster, and a telephone survey of 600 home owners aged 65+ years living in Melbourne, Sydney, Brisbane, Newcastle, Sunshine Coast and Bendigo.
(21) The study reports on data from the 1996 Australian Institute of Family Studies' Later Life Families Study (a random national telephone survey); the sample comprised 721 respondents aged 50-70 years, within which 34 per cent of men and 47 per cent of women identified as retired.
(22) The study draws on data from the 1996 wave of The Australian Family Life Course Survey (a random national telephone survey conducted by the Australian Institute of family Studies); the sample comprised 721 respondents aged 50-70 years, 49 per cent of which identified as in the labour force.
(23) Feedback to the paper was conveyed by 13 zone committees representing 188,000 members.
(24) Postal survey to 5,362 ACT PS employees aged 45+ years who were covered by the Public Sector management Act (representing 48 per cent of the ACT PS workforce); the response rate was 51 per cent.
(25) Wave 1
(26) The Workforce Circumstances and Retirement Survey (2000) undertaken by Tthe Wallis Consulting Group for the Department of Family and Community Services.
(27) The findings are based on three studies used in Merkes' PhD research: one focus group, one computer-mediated communication, and quantitative data from the Healthy Retirement project
(28) It is somewhat ironic that an article on retirement and retirement intentions as recently as 2000 (ABS Australian Social Trends 2000) focussed on the experience of men because, despite recent increases in female participation, 'women were more likely to have had a working history of part-time and causal employment'.
(29) Australia's new Transition To Retirement regulations announced on 4 April 4th 2005 and implemented on 1 July 1st 2005 (http://www.superchoice.gov.au/employees/general/retirement. asp).
(30) The age at which people are entitled to access their private superannuation.
(31) The 2006 Federal Budget included arguably sweeping changes to become effective from July 2007. After age 60 the measures include tax-free superannuation and/or lump sum payments where the superannuation has already been taxed at input, and the abolishment abolition of the compulsory payment of superannuation benefits upon reaching a certain age. For the self-employed, previous age-based tax deductible contribution limits have also been removed up to age 75. See http://www.budget.gov.au/2006-07/overview2/html/ overview_01.htm
Natalie Jackson * Maggie Walter * Bruce Felmingham ** and Anna Spinaze ***
* School of Sociology and Social Work, University of Tasmania
** School of Economics, University of Tasmania
*** University Department of Rural Health, University of Tasmania
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||CONTRIBUTED ARTICLE|
|Publication:||Australian Bulletin of Labour|
|Date:||Dec 1, 2006|
|Previous Article:||Creating decent jobs: the role of regulation in facilitating transitions between employment and unemployment in the labour market.|
|Next Article:||The changing contours of labour disputes and conflict resolution in Australia: towards at post-arbitral measurement framework.|