Good jobs, bad jobs?: Understanding the quality of aged care jobs, and why it matters.

Martin, Bill

Introduction

There can be little doubt that paid care work is here to stay. Overall, paid care workers certainly number in the hundreds of thousands. Accurate numbers are hard to come by, but the best recent estimate indicates that more than 110,000 Australians have paid jobs that focus around providing direct care in Australia's nursing homes and hostels (Richardson and Martin 2004, see also Meagher and Healy 2005). And there is every reason to believe that the forces that have produced this army of paid carers will ensure that it grows in coming years. Generally, more and more care will continue to be 'outsourced' from families, as household structures change, patterns of participation in paid work evolve, and expectations shift (see Breunan's contribution to this special issue). Our population will continue to age creating growing need for aged care workers, women's participation in paid work will continue to rise (if more slowly than in the past), and more care of the ill and disabled will be provided by paid carers as expectations of family provision decline (see Hugo's contribution to this special issue). In short, it is a safe bet that paid care work will offer significant and growing employment opportunities for many years to come.

Popular accounts of work in the future often emphasise jobs and work that are associated with the 'new economy'. They focus on 1T and other 'knowledge' work, or the importance of 'creative' workers with a similar emphasis on high technology work in dynamic, rapidly evolving organizations. Yet rising employment in paid care is an equal part of our 'new economy'. Saskia Sassen emphasised nearly 20 years ago that the growth of business service jobs (a form of 'knowledge work') goes hand in hand with rising demand for personal services, including some forms of paid care (1988). She also saw this as a source of rising inequality, as employment tended to polarize between well paid, challenging jobs in and around growing global companies, and poorly paid jobs with limited future in the personal service sector. There is ample scope for developing these ideas to understand the complex relations between the glamorous new jobs so often emphasized by popular accounts of our work future, and the growing realm of paid care work whose attractions may be far more ambivalent. But a first step is to understand what paid care work is like as work. How does it compare with other kinds of work? What are its advantages and disadvantages, its costs and rewards?

In recent years, researchers have certainly begun to focus on people's experiences working as paid carers (e.g., Harrington Meyer 2000). Unsurprisingly, how we think about paid care has been heavily influenced by the movement of caring from the unpaid private realm of the family to the realm of paid work and the labour market. Comparing the provision of paid care with that of unpaid care, usually based around family relationships, has certainly produced important insights about the nature and experience of paid care work, whether this comparison is made explicitly or implicitly. However, measuring paid care work against other paid work (rather than other care work) is also important, especially in fully understanding what the growing opportunities in the area mean for workers. There is already considerable research along these lines, especially outside Australia. Much has been concerned with the quality of care jobs, particularly their relative pay, and some with the more subjective experience of paid care work (for a partial review, see England 2005).

How paid care jobs compare with other jobs is important from both the 'demand' and 'supply' sides, so to speak. If these jobs are unattractive, for whatever reasons, we can expect difficulties in filling them in the future. Indeed, some paid care jobs, particularly those highly feminised ones requiring few formal qualifications, are often seen as 'bad jobs'--they are poorly paid, have poor security, offer few training opportunities or opportunities for advancement, and involve hard, dirty or demeaning work. But these characteristics are also problematic from the 'supply' side. We cannot be content with arrangements that condemn large numbers of people to jobs that are 'bad', even if the other alternatives available to them are not much more attractive. Do paid care jobs actually have the characteristics of 'bad' jobs? And how do workers experience these jobs, do their objective characteristics determine the experience of doing them? What implications do variations in workplace experience and the characteristics of jobs have for workers' inclination to stay in their jobs, and thus for the future supply of paid care workers?

In this paper, I examine these issues by focusing on the direct care workforce in Australian residential aged care facilities (nursing homes and hostels). The aged care workforce is an important component of the overall care workforce. It sharply raises many of the dilemmas and issues that populate paid care work: Much aged care work has always been poorly paid, most carers are women, the work can be physically demanding and unappealing, career opportunities are often limited, many facilities are under strong pressure to control or reduce costs. At the same time, at least some workers find the experience of actually caring to be rewarding and fulfilling, and they appear to continue to do this work even though they might get other work that pays more or has better conditions.

The care workforce in Australian nursing homes and hostels is one of the few paid care workforces on which good data is now available. In 2003, the National Institute of Labour Studies at Flinders University was commissioned by the Commonwealth Department of Health and Ageing to survey all Australian aged care facilities and a sample of direct care workers employed by them. In all, we received responses from about 1,800 facilities and over 6,000 workers (for details of the surveys, see Richardson and Martin 2004). Both surveys were fielded in mailback form and yielded good response rates (just over 60% and 40% respectively for the facilities and employees surveys). In this paper, I use data from the survey of employees. This survey provided information about a range of employment experiences and job characteristics, both objective and subjective.

Are Direct Care Jobs in Aged Care Facilities 'Bad' Jobs?

The idea that the economy produces 'good' and 'bad' jobs originally gained currency in the early 1970s, largely through the work of a group of American institutional economists (e.g., Bluestone 1970, Doeringer and Piore 1971). They argued that the American labour market was bifurcated between a core of 'good' jobs and a secondary segment of 'bad' jobs. Good jobs offered high wages and good conditions (notably health insurance in the U.S.), employment security, promotion prospects and a range of employment protections secured through legislation and unionization. Bad jobs offered none of these advantages. In subsequent years the image of a simple binary split between a primary and secondary labour market was largely rejected. However, the rise of 'non-standard' forms of employment in all the Anglophone economies over the past 20 years has reinvigorated interest in the quality of jobs, and analysis of the relation between bad jobs and non-standard employment has proven a useful way of approaching the issues (e.g., Kalleberg et al. 2000, McGovern et al. 2004).

Many current analysts of work in Australia believe that the quality of jobs is a concern, and may be declining (e.g., Watson et al. 2003). At the least, measuring the quality of jobs against the standard of the 'bad' job is a useful heuristic for assessing their quality. To assess the quality of aged care jobs, I compare key characteristics of the jobs of workers in this industry with those in the wider economy. The characteristics on which I focus are: pay, type of employment contract (casual vs. permanent), hours of work, job tenure and job security, and training opportunities. In the classic formulations, 'bad' jobs are characterized by poor quality on all of these measures.

To assess the relative pay of aged care workers, I compare their actual pay with the pay they might expect in the wider economy, given their 'human capital'. As in much of the remaining analysis in this paper, I focus on the two occupational groups that make up the lion's share of direct care workers in aged care facilities, personal care workers and nurses. Personal care workers (hereafter 'personal carers' or 'PCs'), also known as 'assistants in nursing' and 'personal care assistants', have become an increasingly important component of the aged care workforce in recent years. Although their predominance varies somewhat from State to State, they make up nearly 60% of all direct carers in Australian residential aged care facilities, with nurses composing about 35%, and allied health workers making up the remainder (Richardson and Martin 2004: 19). Personal carers are not usually required to have any qualifications in aged care, though most acquire at least the Certificate III in Aged Care. In contrast, of course, nurses have professional qualifications, whether they are RNls (registered nurses) or RN2s (enrolled nurses).

Since PCs and nurses are quite distinct in their level of qualification, the work they perform and their typical career paths, I examine them separately throughout this paper. In analyzing each group's pay, their expected earnings are calculated on the basis of standard wage equations estimated on appropriate sub-populations from the HILDA survey, a high quality national survey (see Watson and Wooden 2004). Thus, for example, I compared PCs earnings from the survey of aged care workers with those of women with Certificate IV qualifications or less from HILDA. This is an appropriate comparison group for PCs. I compare nurses observed earnings with predicted earnings based on a model for all women workers. (1)

At first blush, these analyses suggest that aged care workers are not particularly poorly paid, at least compared to the earnings they might expect elsewhere in the economy (Table 1). PCs earn about what we might expect if their earnings reflected their human capital in the way that earnings reflect human capital for other women with relatively limited formal qualifications. (2) If anything, it appears that nurses in aged care facilities earn more than we might expect on the basis of how human capital affects earnings across the whole workforce. However, women continue to suffer significant earnings penalties in Australia (ABS 2005), and it is plausible that they are particularly susceptible to being confined to 'bad' jobs. In other words, the predicted earnings of PCs based on the experience of other similar women may not be a useful test of whether their earnings are low since they reflect the fact that women are far more likely than men to be in 'bad' jobs. To test this idea, I provide estimates of PCs' predicted earnings if they were rewarded for their human capital attributes in the same way as men with relatively low levels of formal education. (3) These estimates suggest that, in such an imaginary world, PCs might earn nearly 30% more than they actually do. By contrast, if nurses suffered a wage penalty because they are women, it was much smaller. On these measures, then, PCs' earnings are low, given their age and qualifications. But they are probably not dramatically lower than those of other similarly qualified women in other jobs. In this, their wages probably reflect a general tendency for women with low levels of formal education to suffer very low wages. In contrast, the argument that nurses in aged care facilities are poorly paid is more difficult to sustain.

On other measures of whether aged care jobs are 'bad' jobs, the picture is generally more consistently positive. Table 2 shows a range of such indicators. Casualisation of employment is a widely remarked feature of the Australian labour market in recent years, and is often taken as a prime component of decline in the quality of jobs (e.g., Watson et al. 2003). The aged care workforce displays much lower levels of casual employment than any comparable segment of the whole Australian workforce. For example, about 18% of PCs are employed casually compared to around 37% of all women with certificate IV education or less, 31% of all women and 21% of all men. Other measures of employment security suggest that aged care workers are as well off as most other workers--they have similar (PCs) or greater (nurses) average job tenure, and, though their satisfaction with their employment security is a little lower than is generally the case in the Australian workforce, the vast majority are happy on this score. PCs do show a much lower level of full-time work than is usual across the economy, but it is unlikely that this straightforwardly reflects a lack of choice since a significant majority are happy with their hours of work, or want to work less (Martin and Richardson 2004: 234).

'Bad' jobs are often characterized as offering few training opportunities. However, training participation is, if anything, higher than expected amongst aged care workers, particularly PCs. Once we take account of their older average age, PCs have quite high rates of participation in formal training, and aged care nurses' rates are at least comparable to those of all Australian workers. Indeed, some 15% of PCs and 8% of nurses aged 3044 were currently undertaking aged care qualifications in 2003, and 9% of PCs and 6% of nurses aged 45 and over were doing so. Training opportunities are usually seen as an indicator of job quality in that they provide routes to higher pay within jobs, or to career progression. Training in the aged care sector does appear to have this effect amongst nurses. However, PCs achieve no wage benefits from gaining the typical qualification for their work, in line with the experience of all women workers with lower levels of formal qualification. (4) Moreover, they do not gain access to straightforward career paths, for example into aged care nursing, by achieving these qualifications. This is not to say that this training is not a real advantage for these workers, but its effects are less than might be expected.

Aged care jobs are certainly not unequivocally 'bad'. They offer security and training opportunities that are comparable with much of the rest of the labour market. Indeed, for the lesser qualified workers, those who work as PCs, levels of casual employment and opportunities for gaining formal qualifications appear to be significantly better than for many of their peers. On the other hand, there is little doubt that these workers are not well paid. But they are not alone in this; their low pay largely reflects the fact that women with limited formal qualifications suffer significant wage penalties compared to similarly qualified men. The heavily gender segmented labour markets for people with limited formal qualifications continue to sustain significant pay inequities between men and women. Personal carers, like many of their peers, suffer the wage consequences. By contrast, Nurses' wages appear to be more or less in line with those in other parts of the economy, though, in fact, aged care nurses generally have been paid at lower rates than other nurses. (5) In general, nurses have been able to use their claims to professional status and their professional organization to establish pay arrangements that put them on a par with other similar professions. Combined with the other features of their jobs, this means that aged care nurses are certainly not in 'bad' jobs.

Retaining Aged Care Workers

Turnover rates amongst aged care workers are quite high. Data from the 2003 survey of aged care facilities indicated that nearly 20% of nurses and just over a quarter of PCs need to be replaced each year (Richardson and Martin 2004: 31). (6) If aged care jobs are not 'bad' jobs, as the previous sections suggests, then an important issue in understanding both the experience of aged care work and the challenges managers of aged care facilities face in retaining workers, is what affects the likelihood that workers will leave their jobs. Because the survey of aged care workers was of a sample currently employed in aged care facilities, it is not possible to directly assess these factors. However, the survey did collect data on whether workers expected to be working in the aged care sector 3 years from the time at which they were surveyed. (7) About 72% of nurses and 75% of PCs said that they expected to continue working in the sector for at least 3 years. These expectations could be affected by four sets of factors. First, if employees do not find their employment arrangements amenable, they may leave a job or industry. Relevant issues here would include shift characteristics, whether workers are able to get the hours of work they prefer, the type of employment contract available, and wages. Secondly, workers may leave jobs if their experience of the work means that they are unable to achieve the non-wage rewards they hope for. Some of the key factors here might be aspects of subjective work experience, including job satisfaction. In addition, much research suggests that paid care workers' positive experience of their work is largely through the rewards they receive from actually doing caring work. So that the extent of opportunities to do this work, and to do it satisfactorily, may be important aspects of the experience of work for such workers. Third employees' personal characteristics may affect their expectations about staying in the aged care sector, partly because some characteristics may make them more likely to leave paid work altogether. Personal characteristics also connote human capital which will affect the alternative opportunities open to them. Finally, some organizational characteristics of facilities may make work more or less attractive to workers, and also indicate the extent of alternative employment opportunities (e.g., the size of facilities or where they are located).

Table 3 shows the results of estimating a logistic regression model predicting whether workers expect to be in the aged care sector 3 years after they were surveyed. First, the experience of work has a large impact on expectations about staying in the sector for both groups. In particular, job satisfaction is a major factor in explaining such expectations. On the 60 point job satisfaction scale used here, each 1 point increase in job satisfaction produces a 5-6% increase in the odds that a worker expects to remain in the aged care sector over 3 years (controlling for other variables in the model). Adding only the job satisfaction variable to a model containing all employment arrangements, personal characteristics, and facility characteristics raises the Nagelkerke [R.sup.2] from 0.13 to 0.26 for nurses and from 0.14 to 0.21 for PCs (analysis not shown in Table 3). Second, there are strong indications that where nurses are able to use their professional skills--acting out their professional identity, as it were--they are more likely to expect to continue in the aged care sector. Thus, those who say they make more use of their skills in their current jobs, and those who work in facilities with only high care beds (which demand greater use of their professional skills), are more likely to expect to remain in the sector. (8) However, these factors do not affect PCs' expectations, suggesting that they do not have the same professional identities as nurses. Indeed, PCs' expectations about remaining in the sector are affected by variables that may measure the general conduciveness of workplaces in fairly routine jobs: PCs are less likely to expect to remain in the sector the longer they have been with their current employer, the more they feel under pressure to work harder, and the larger the facility. None of these factors affects nurses' expectations.

The employment arrangement variables measured here have a very limited effect, with working regular night or evening shifts making respondents less likely to expect to continue in the sector compared to working other shifts. However, neither being employed casually nor wage levels have any effect. Personal characteristics were included in this analysis primarily as controls, and generally have limited effects on expectations. It is notable, however, that older PCs and those with more limited secondary education are more likely to expect to remain in the sector. The age effect may indicate that older workers find this work particularly conducive. Alternatively, it may arise because aged care work is something of a last resort for workers with few alternatives, as the education effect suggests.

In this analysis, aged care workers' experience of their work looms as the major explanation for their expectations about whether they will stay in the sector, with job satisfaction being a very powerful predictor of intentions. But what determines their job satisfaction?

Job Satisfaction Amongst Aged Care Workers

Standard job satisfaction questions offer a useful window on how workers feel about their jobs. They provide indications of whether workers see their jobs as 'bad' ones or not, and are also known to predict whether workers will stay in their jobs (Clark et al. 1998). Australian workers, like those elsewhere, generally express quite high levels of overall satisfaction with their jobs, and with most specific aspects of them (Martin and Pixley 2005). In doing so, they appear to be making the judgement that the jobs they hold are as good as they might reasonably hope for. Those expressing high job satisfaction are not necessarily saying that they love their work, but they are at least indicating that they do not reasonably believe they could get a job that was much better. (9)

Aged care workers generally express quite high levels of job satisfaction with most aspects of their jobs, though their satisfaction appears to be consistently a little lower than for comparable sectors of the whole Australian workforce (Table 4). The exception is pay, where aged care workers express low satisfaction, whether they are PCs or nurses. This is a striking result, especially in view of the fact that neither group appears to suffer major wage penalties compared to comparable women in other jobs. It suggests that aged care workers may be attaching a meaning to their pay that is beyond simply what they might plausibly earn in other jobs open to them. (10) Indeed, this very low satisfaction with pay suggests that we should be cautious in drawing strong conclusions about the somewhat lower job satisfaction of aged care workers on other aspects of their jobs. Very high pay dissatisfaction may have a suppressing effect on the expression of satisfaction with other aspects of people's jobs.

Leaving aside their feelings about pay, then, aged care workers probably feel virtually as positive about their jobs as the average comparable Australian worker. This is quite encouraging news in terms of the morale of the aged care workforce, especially given the effect of job satisfaction on expectations to stay in the aged care sector. To understand the determinants of job satisfaction amongst aged care workers, I constructed a conventional index of job satisfaction by adding workers rating of their satisfaction using an 11 point scale (highly dissatisfied to highly satisfied) on the 6 aspects of their jobs shown in Table 4. This variable was used in the analysis of expectations about staying in the aged care sector in Table 3. The scale ranges from 0 (highly dissatisfied with all aspects of the job) to 60 (highly satisfied with all aspects of the job). I estimated regression models predicting scores on this job satisfaction scale using a similar set of variables to that used for expectations. The results are shown in Table 5.

Work experience is the overwhelming determinant of job satisfaction amongst aged care workers. One set of work experience factors is common to many studies of job satisfaction, and relates to general workplace conditions. Thus, job satisfaction is substantially higher if workers believe that they use 'many' of their skills on the job or if they feel that they have some significant autonomy in their jobs ('freedom to decide how to do their job'), and it is lower if they feel under pressure to work harder. The skill usage and job autonomy measures probably indicate whether employees feel that they are able to perform their jobs in appropriately 'professional' ways. These are usually thought of as related to 'intrinsic' job satisfaction. Workers who feel under pressure to work harder are most likely experiencing some aspect of the work intensification that is a feature of many contemporary workplaces.

A second set of factors relates more directly to the distinctive nature of the care work that these workers do. Thus, how much of their workplace time employees spend with residents matters, though the effect is more important amongst Nurses than PCs. For Nurses, there is a gradient of effect: the more time Nurses spend in direct care, the higher their job satisfaction. For PCs, the key issue appears to be that if they spend less than two-thirds of their time in direct care work, their job satisfaction will be lowered. These differences between Nurses and PCs undoubtedly arise from the different job expectations of Nurses and PCs: PCs expect to spend most of their time in direct care and are unhappy as soon as this expectation is not met; Nurses may expect that direct care is a smaller part of their job, but the smaller it becomes the lower their job satisfaction. Perhaps most strikingly, even beyond this more objective measure of how much time workers spend with residents, their perception of whether they are able to 'spend enough time with each resident' substantially affects job satisfaction.

Finally, several variables affecting job satisfaction probably indicate whether workplace arrangements are conducive to other aspects of workers' lives. Such issues are likely to be particularly important in a workforce that is primarily female and part-time. For example, preferring to work different hours (either more hours or less) reduces job satisfaction quite substantially for both nurses and PCs, reducing their job satisfaction score by 4-5 points on our 60 point scale. Moreover, preferring to work a different shift has a similar effect on nurses' job satisfaction (though not PCs'). Objective employment arrangements have only small effects for both groups once these preferences are taken into account: for Nurses being on an irregular shift actually increases job satisfaction, and for PCs being on a casual contract decreases it. (11)

These results have important implications. First, all of the most important predictors of aged care workers' job satisfaction are determined primarily by how work is organized in aged care facilities, and are therefore largely under the control of facility managers. How managers allocate hours and shifts will determine whether workers' preferences are met. How they organize the daily performance of the care work will largely determine how much time workers are able to spend with residents, whether they use their skills in their jobs, whether they experience some job autonomy, and whether workers experience the pressures of work intensification. This is to say that managers in aged care facilities have substantial scope to improve or hamper the morale of their workforces. Moreover, our earlier analysis indicates that these factors also affect whether workers expect to remain in the aged care sector, sometimes having their effects directly and sometimes through job satisfaction. While cost pressures undoubtedly do have the capacity to derail even the most well intentioned management, our results indicate that many aged care facilities in Australia were able to manage these issues so as to produce high levels of job satisfaction amongst workers.

The second implication is that the 'intrinsic' job satisfaction experienced by aged care workers contains two distinct elements. First, there is a component related to the 'professional' aspects of their work. Many of these workers clearly see themselves as skilled at the jobs they do, and being able to use their skills in an appropriately responsible and autonomous work situation enhances their job satisfaction. In this respect, they are like other workers with vocationally specific skills that are certified and widely recognized. Secondly, however, there are distinctive aspects of aged care work that affect workers' job satisfaction. For both nurses and PCs the opportunities they have, and feel they have, for directly caring for residents have substantial effects on their job satisfaction. And this is true even taking into account all of the other factors that affect job satisfaction, including the opportunities to act 'professionally'. This result is consistent with a range of research that emphasizes the importance of 'intrinsic' motivations amongst paid care workers (England 2005). However, in demonstrating the distinctiveness of the care related component of this motivation, it raises again the vexed question of how exactly this motivation arises. In particular, it reminds us that almost all aged carers are women (around 95%), and highlights again the fraught issue of how such a heavily gender coded form of work as caring, with such correspondingly gendered forms of intrinsic motivation, can be organized equitably for both workers and those receiving care (see Meagher 2006).

Conclusion

Focusing on the quality of care work and jobs in Australian nursing homes and facilities leads us well beyond any simple notion that they may be 'good' or 'bad' jobs. Some aspects of the jobs are, at the least, far from desirable: the pay of PCs is undeniably low, and significantly less than they would earn were they not women in a women's occupation. But other aspects of the working conditions are not so emblematic of bad jobs: job security appears to be generally on a par with other parts of the labour market, most are employed permanently, and the opportunities for training appear to be good. On the other hand, career paths for PCs are certainly limited. Moreover, aged care workers seem to be almost as satisfied with their jobs as other workers, except on the issue of pay.

One important consequence of the quality of jobs is, of course, its implications for people's willingness to undertake them - the problem of labour supply. I have shown that most of the factors that seem to affect aged care workers' willingness to remain in the industry are in the control of management. A key issue is the effects work arrangements can have on job satisfaction. Job satisfaction is important in its own right as a component of job quality, and has important effects on workers' expectations about remaining in the sector. Workplace experience is the major determinant of variation in aged care workers' job satisfaction. Some issues are common to most workplaces (shift and hours arrangements, opportunities for job autonomy and to use skills, etc.). However, there is also something distinctive about the experience of caring work that has substantial effects on job satisfaction. This is a widely noted phenomenon, often being discussed in terms of the emotional bonds that develop between carets and those they care for (e.g., Meagher 2006, England 2005).

Understanding paid care as work involves seeing both its similarities to other work, and its distinctiveness. Indeed, I have argued that its distinctive aspects are as important as those it shares with other work, both for the quality of workers' experience at work and for ensuring adequate supply of labour for the increasing demand for paid carets. Although there is already considerable research on paid care work, its distinctive character as work remains a fertile ground for future investigation. To the extent that it becomes better understood, the capacity of managers in aged care facilities and similar institutions to arrange work practices so as to maintain and enhance job satisfaction will be increased. And this will enrich the lives of workers, and improve the prospects that they will continue to work as paid caters.

Acknowledgements

The data used in this paper comes from research commissioned by the Commonwealth Department of Health and Ageing. Sue Richardson led that research, and I am very grateful for her insightful analysis of it, for advice and discussions about the work reported here, and for her encouragement to use the data in this paper. Megan Moskos discussed many of the ideas in this paper with me. I am also grateful for the work Josh Healy and Megan Moskos undertook to compare the aged care workforce with the Australian workforce, which provided the impetus for some of the comparisons in this paper.

References

Australian Bureau of Statistics (2004) Labour Mobility, Cat. No. 6209.0, Canberra: Australian Bureau of Statistics.

Australian Bureau of Statistics (2005)Australian Social Trends, 2005, Cat. No. 4102.0 Canberra: Australian Bureau of Statistics.

Bluestone, B. (1970) 'The tripartite economy: Labor markets and the working poor,' Poverty and Human Resources Abstracts 5, 15-35.

Clark, A.E., Georgellis, Y. & Sanfrey, P. (1998) 'Job satisfaction, wage changes and quits: Evidence from Germany,' Research in Labor Economics, 17, 95-122.

Doeringer, P.B. & Piore, M.J. (1971) Internal Labor Markets and Manpower Analysis Lexington, Ma: DC Heath.

England. P. (2005) 'Emerging theories of care work,' Annual Review of Sociology, 31, 381-99.

Harrington Meyer, M. (2000) (ed.) Care Work: Gender, Class, and the Welfare State. New York: Routledge.

Hodson, R. (2001) Dignity at Work Cambridge: Cambridge University Press.

Kalleberg, A., Reskin, B.F. & Hudson, K. (2000) 'Bad jobs in America: Standard and non-standard employment relations and job quality in the United States,' American Sociological Review, 65 (2), 256-278.

Lopez, S.H. (2006) 'Culture change management in long-term care: A shop-floor view,' Politics and Society, 34 (1), 55-79.

Martin, B. & Pixley, J. (2005) 'How Australians feel about their work,' in R. Gibson, G. Meagher, S. Wilson (eds.), Australian Social Attitudes: the First Report, UNSW Press, 2005.

McGovern, P., Smeaton, D. & Hill, S. (2004) 'Bad jobs in Britain: Nonstandard employment and job quality,' Work and Occupations, 31 (2), 225-249.

Meagher, G. (2006) 'What can we expect from paid carers?' Politics and Society, 34 (1), 33-53

Meagher, G. & Healy, K. (2005) Who Cares, Volume. 1: A Profile of Care Workers in Australia's Community Service Industries, Strawberry Hills, NSW: ACOSS.

Richardson, S. & Martin, B. (2004) The Care of Older Australians: A Picture of the Residential Aged Care Workforce, National Institute of Labour Studies, Flinders University, Adelaide.

Rose, M. (2003) 'Good deal, bad deal? Job satisfaction in occupations,' Work, Employment and Society, 17 (3), 503-530.

Sassen, S. (1988) The Mobility of Labor and Capital: A Study in International Investment and Labor Flows, Cambridge: Cambridge University Press.

Sennett, R. (1998) The Corrosion of Character New York: W.W. Norton & Co.

Watson, I., Buchanan, J., Campbell, I. & Briggs, C. (2003) Fragmented Futures: New Challenges in Work Life Annandale, NSW: Federation Press.

Watson, N. & Wooden, M. (2004), 'The HILDA survey: A summary,' Australian Journal of Labour Economics, 7, 117-124.

Endnotes

(1) The restriction of the comparison population for PCs is made because the key issue in this analysis is whether aged care jobs are worse than other jobs available to workers with similar human capital, and limited formal education effectively excludes PCs from jobs requiring significant post-secondary education. In contrast, nurses, with significant post-secondary qualifications, are able to access jobs at all levels of the labour market, so that it is appropriate to compare them to all women.

(2) The HILDA data on which these wage equations are estimated was collected a year earlier (2002) than the data for aged care workers (2003). Given wage inflation, it is arguable that these results actually indicate that PCs did earn a little less than might be expected. Table 1 does show the effect of adjusting for wage growth using the observed increase in hourly ordinary time earnings for the health and community services sector between 2002 and 2003.

(3) There is not a good measure of experience in our wage equation. I use current tenure and age as proxies for experience, and these will be better approximations for men than women (due to women's tendency to leave the paid labour force for periods). This will tend to overestimate women's experience, and therefore inflate our estimates of their expected earnings. However, this problem is unlikely to explain all of the large difference in expected earnings using the male and female regressions estimates.

(4) The wage equation for all Australian women with certificate IV or lower education, on which the predicted wage for PCs in Table 1 is based, shows that post-secondary education has no effect on wages for this group. Similarly, estimating wage equations for PCs on our aged care workforce survey shows that gaining the Certificate III in Aged Care has no significant effect on their wages.

(5) The future outlook for wages amongst these workers undoubtedly depends on the overall state of the labour market. Any overall deterioration means that the risk is definitely towards lowering wages, especially in the context of the recent industrial relations law changes.

(6) The overall turnover rate for women in health and community services industries in 2004 was about 21% (ABS 2004). This figure would include some very brief and casual employment.

(7) This item simply asked where workers expected to be working in 3 years, and thus does not distinguish those who expect voluntary and involuntary departure. However, given the state of the aged care labour market, it can be assumed that most who expect to depart will do so voluntarily.

(8) Interestingly, the more that nurses feel they have the skills to do the jobs, the less likely they are to expect to continue working in the sector. This probably indicates that nurses with the strongest professional identities may seek more challenging work outside aged care.

(9) See Rose (2003) for a more extended discussion of the meaning of job satisfaction measures.

(10) Part of the explanation for the very low level of pay satisfaction in our survey of aged care workers may be that it was undertaken during a prolonged campaign for higher wages in the NSW aged care sector. The focus of union arguments was on the claim that aged care workers were poorly paid compared to other similar workers. However, it is implausible that this campaign could fully explain aged care workers' disenchantment with their pay.

(11) In addition, a few of workers' personal characteristics have small effects on job satisfaction. Nurses who were born in non-English speaking countries and those who are the main income earners in their household have somewhat lower job satisfaction, while those in 'excellent' health have higher job satisfaction. Amongst PCs, being older and in better health (very good or excellent) increases their job satisfaction quite substantially. Being the main income earner also has much the same negative effect on job satisfaction as it does amongst Nurses. Finally, PCs in rural facilities and those in larger facilities are more satisfied with their jobs. However, there are no significant effects of facility characteristics on Nurses' job satisfaction.

Table 1. Expected and observed earnings of personal carers and nurses
In Australian nursing homes and hostels, 2003

 Personal carers Nurses (RNs and ENs)

Expected hourly wage--female $16.96 $21.67
regression

Expected hourly wage--male $21.53 $24.52
regression

Observed hourly wage $16.80 $24.01
--female aged care workers ($15.33) ($22.89)

Note: Expected wages are calculated from standard wage regressions
estimated using HILDA data. Regression for personal carers is based on
workers with Cert. IV education or less and uses the following human
capital predictors of wages: age, age squared, tenure, no. of
employees at workplace, years of schooling, years of post-secondary
education, partner status. Regression for nurses is based on all
workers and uses the same predictors as the RC equation plus
non-English speaking background status, and health status (excellent
health vs. other). Observed figures in brackets are deflated by the
rate of wage rise for the community services sector between 2002 and
2003 to allow for the fact that HILDA data were collected in the
former year, and the aged care survey data in the latter.

Table 2. Comparing job characteristics of the aged care workforce to
those of the Australian workforce

 Aged Care Workforce

Job characteristic Personal Nurses
 carers (RNs and
 ENs)

Casual 17.9% 8.1%
Full-time 28.4% 41.2%
Job tenure (mean years) 5.8 8.0
Satisfied with job security 72.2% 74.5%
Currently studying post-sec 13.1% 6.5%
qualification
Currently studying post-sec 15.2% 7.8%
qualification, aged 30-44
Currently studying post-sec 8.9% 6.1%
qualification, aged 45+

 Australian Workforce

Job characteristic Low All All male
 education female workers
 female workers
 employees

Casual 37.0% 30.6% 20.5%
Full-time 41.3% 46.7% 81.6%
Job tenure (mean years) 5.6 5.9 9.0
Satisfied with job security 84.8% 85.4% 83.3%
Currently studying post-sec 17.2% 16.2% 12.4%
qualification
Currently studying post-sec 10.6% 11.7% 8.8%
qualification, aged 30-44
Currently studying post-sec 4.2% 5.9% 3.9%
qualification, aged 45+

Note: Figures for aged care workforce include women only. Figures for
Australian workforce include workers 18 and over.

Table 3. Determinants of Aged Care workers' expectations about working
in aged care sector in 3 years

 Nurses

 B Exp(B)
Employment Arrangements
Shift type (0 = regular day shift)
 Regular eve/night shift -.731 ** .481
 Irregular shift -.064 .938
Casual employee -.589 .555
Wage per hour .015 1.015
Experience of work
Months at current facility .002 1.002
Time spent in direct care (0=more than two thirds)
 < one third .253 1.288
 > one third and < two thirds .102 1.107
Able to spend enough time with each resident .050 1.051
Have the skill I need to do my job -.317 ** .729
Use skills in current job .278 ** 1.321
Have freedom to decide how I do my work .086 1.089
Do not feel under pressure to work harder .011 1.011
Job satisfaction .056 ** 1.058
Personal characteristics
Male -.756 * .470
Age .023 1.023
Born non-English speaking country -.762 * .467
Live with a partner -.075 .928
Main income earner .394 1.483
Self assessed health (0=good, fair or poor)
 Very good health .084 1.088
 Excellent health -.110 .896
No. hrs caring for people in own household .005 * 1.005
Years secondary education -.064 .938
Years post-secondary education -.098 .906
Hold aged care qualification
Currently studying aged care qualification .807 2.240
Facility characteristics
Facility location (0 = metropolitan)
 Regional facility .512 * 1.669
 Rural facility .556 * 1.743
High care places only .666 ** 1.947
Total no. employees .000 1.000
Facility ownership (0 = not-for-profit)
 For-profit facility -.169 .845
 Publicly owned facility -.623 * .536
Constant -2.910 .054
N 724
Nagelkerke R2 .298

 PCs

 B Exp(B)
Employment Arrangements
Shift type (0 = regular day shift)
 Regular eve/night shift -.559 * .572
 Irregular shift -.001 .999
Casual employee -.100 .905
Wage per hour .013 1.013
Experience of work
Months at current facility -.004 ** .996
Time spent in direct care (0=more than two thirds)
 < one third .184 1.202
 > one third and < two thirds -.267 .765
Able to spend enough time with each resident .002 1.002
Have the skill I need to do my job -.111 .895
Use skills in current job .058 1.059
Have freedom to decide how I do my work .065 1.067
Do not feel under pressure to work harder .126 ** 1.134
Job satisfaction .050 ** 1.051
Personal characteristics
Male .387 1.472
Age .035 ** 1.036
Born non-English speaking country -.184 .832
Live with a partner .492 * 1.635
Main income earner .308 1.360
Self assessed health (0=good, fair or poor)
 Very good health -.173 .841
 Excellent health -.378 .685
No. hrs caring for people in own household .003 1.003
Years secondary education -.193 * .825
Years post-secondary education -.119 .888
Hold aged care qualification .465 1.592
Currently studying aged care qualification -.324 .723
Facility characteristics
Facility location (0 = metropolitan)
 Regional facility .434 1.544
 Rural facility .412 1.510
High care places only .098 1.103
Total no. employees -.003 * .997
Facility ownership (0 = not-for-profit)
 For-profit facility .140 1.150
 Publicly owned facility .049 1.050
Constant -2.459 .086
N 784
Nagelkerke R2 .221

* p [less than or equal to] .05

** p [less than or equal to] .01

Note: Table shows results of a logistic regression in which the
dependent variable is whether respondents expect to be working in the
aged care sector 3 years after the survey. 'B' is the logistic
regression coefficient, and Exp(B) is the odds ratio corresponding to
a one unit change in the relevant independent variable (e.g., the odds
of a respondent saying he/she will remain in the aged care industry in
3 years increase by 32% if he/she says that his/her skills are used in
the current job). Nagelkerke's R2 is a logistic regression analogue of
explained variance in OLS regression; it provides an indication of the
net power of all independent variables to explain variation in the
dependant variable.

Table 4. Comparing the aged care workforce's job satisfaction to other
Australians'

 Aged Care Workforce Australian Workforce

Job Personal Nurses Low All All
characteristic carers (RNs education female male
 and female workers workers
 ENs) employees

Security 0% 75% 85% 85% 83%
Pay 26% 29% 72% 72% 73%
Work itself 76% 65% 84% 85% 86%
Hours 71% 77% 78% 77% 77%
Flexibility 70% 66% 81% 80% 77%
Job overall 77% 71% 87% 87% 87%

Note: Figures for aged care workforce include women only. Figures for
Australian workforce include workers 18 and over.

Table 5. Determinants of Nurses' and PCs' job satisfaction

 Nurses

 B

Employment Arrangements
Shift type (0=regular day shift)
 Regular eve/night shift 1.148 .038
 Irregular shift 2.526 ** .092 **
Casual employee .270 .006
Wage per hour .130 .022
Experience of work
Months at current facility .001 .004
Time spent in direct care (0 = more than
two thirds)
 < one third -4.825 ** -.202 **
 > one third and < two thirds -2.993 ** -.122 **
Able to spend enough time with each resident 1.146 ** .155 **
Have the skill I need to do my job -.194 -.014
Use skills in current job 1.507 ** .196 **
Have freedom to decide how I do my work 2.383 ** .330 **
Do not feel under pressure to work harder 1.104 ** .191 **
Prefer different shift -4.141 ** -.170 **
Hours preferences (0 = same as current)
 Prefer workless -4.233 ** -.143 **
 Prefer work more -5.020 ** -.147 **
Personal characteristics
Male -2.254 -.045
Age .033 .025
Born non-English speaking country 2.322 ** .068 *
Live with a partner -.957 -.035
Main income earner -2.034 ** -.086 **
Self assessed health (0=good, fair or poor)
 Very good health -.245 -.010
 Excellent health 2.319 * .084 *
No. hrs caring for people in own household .010 .035
Years secondary education -.430 -.037
Years post-secondary education .414 .037
Currently studying aged care qualification .328 .007
Facility characteristics
Facility location (0 = metropolitan)
 Regional facility .489 .018
 Rural facility -.527 -.020
High care places only -.219 .009
Total no. employees .006 .063
Facility ownership (0 = not-for-profit)
 For-profit facility -.945 -.037
 Publicly owned facility 1.430 .047
Constant 15.756 *
R2 .409
N 733

 PCs

 B

Employment Arrangements
Shift type (0=regular day shift)
 Regular eve/night shift -.393 -.016
 Irregular shift .475 -.022
Casual employee -1.758 ** -.062 **
Wage per hour .089 .041
Experience of work
Months at current facility -.007 -.044
Time spent in direct care (0 = more than
two thirds)
 < one third -2.887 ** -.098 **
 > one third and < two thirds -2.082 ** -.095 **
Able to spend enough time with each resident 1.329 ** .222 **
Have the skill I need to do my job .787 * .081
Use skills in current job .868 ** .104 **
Have freedom to decide how I do my work 1.068 ** .172 **
Do not feel under pressure to work harder .760 ** .147 **
Prefer different shift -.587 -.028
Hours preferences (0 = same as current)
 Prefer workless -4.463 ** -.122 **
 Prefer work more -4.296 ** -.201 **
Personal characteristics
Male -.627 -.016
Age .120 ** .123 **
Born non-English speaking country -1.620 -.056
Live with a partner -1.521 -.070
Main income earner -1.552 * -.075 *
Self assessed health (0=good, fair or poor)
 Very good health 3.137 ** .152 **
 Excellent health 4.028 ** .156 **
No. hrs caring for people in own household .016 .059
Years secondary education -.287 -.030
Years post-secondary education -.497 -.050
Currently studying aged care qualification .380 .011
Facility characteristics
Facility location (0 = metropolitan)
 Regional facility 1.187 .050
 Rural facility 3.132 ** .129 **
High care places only -1.208 -.055
Total no. employees .013 ** .087 **
Facility ownership (0 = not-for-profit)
 For-profit facility .656 .027
 Publicly owned facility .103 .002
Constant 15.713 **
R2 .327
N 783

* p [less than or equal to] .05

** p [less than or equal to] .01

Note: Table shows results of an OLS regression on job satisfaction
scale.

'B' is the unstandardized regression coefficient, '_' is the
standardized regression coefficient.