Printer Friendly

Pressing towards full employment? The persistence of underemployment in Australia.

The economic boom in Australia since the early 1990s has been accompanied by a steady increase in jobs and a reduction in official unemployment rates. This has sparked some excited commentary. Responding to the release by the Australian Bureau of Statistics (ABS) of the July 2007 labour force estimates, which cited an unemployment rate of 4.3 percent, the then-federal Treasurer, Peter Costello, announced that "we are pressing towards full employment" (The Australian, 9 August 2007). He was loyally echoed a week later by the Secretary of the Treasury, who pointed to the same figure and declared that "today the Australian economy is as close to full employment as it has been for more than 30 years" (The Australian, 16 August 2007).

This article examines an aspect of labour market restructuring that has been neglected in the flurry of recent commentary. Most attention has been narrowly focused on unemployment, but this article turns to the somewhat separate phenomenon of underemployment. The latter term can have different meanings in labour market analysis, but it is used here in its most familiar sense as 'time-related underemployment', broadly understood as 'insufficient hours of paid work' amongst workers classified as currently employed.

The article begins by presenting official labour force data that outline the extent of underemployment and its pattern of growth in Australia. Both unemployment and underemployment surged upwards in the last recession. However, in contrast to the pattern of unemployment, underemployment failed to recede during the subsequent period of economic boom and even increased slightly. The underemployment rate has more than tripled since thirty years ago, reaching a point where it now surpasses the unemployment rate. Moreover, the current level of underemployment in Australia seems high in comparison with the other advanced capitalist societies grouped together in the Organisation for Economic Co-operation and Development (OECD) (ILO, 2007). In short, underemployment has become a significant phenomenon, demanding careful analysis.

Though pushed aside in current political discussion, underemployment is recognised as an important issue by many labour market researchers, who have sketched out its main features (Watson et al, 2003; Wilkins, 2004, 2006, 2007). This article covers similar points of description, but it also seeks to push on into less familiar territory by exploring the dynamics behind the increase in underemployment, looking at how underemployment is influenced by cyclical movements in the economy and also by more fundamental structural factors. It shows the contribution of industry conditions and employer practices in producing part-time jobs that sustain underemployment, using the example of retail sales assistants and contract cleaners. It also draws out some implications of the analysis, stressing that the persistence of high levels of underemployment undermines exaggerated claims of labour market success. Finally, it poses important questions about the implications of economic growth and the quality of jobs in Australian labour markets.

Describing Underemployment

Unemployment and Underemployment

Researchers have known for a long time that the headline unemployment rate, which measures the unemployed as a percentage of the labour force, is a poor indicator of the efficiency of labour markets in utilising labour resources (De Neubourg, 1988; OECD, 1995; Mitchell and Carlson, 2000; Watson, 2000; Denniss, 2001). More than twenty-five years ago, Stricker and Sheehan (1981:26) pointed out that the Australian measure of the unemployment rate was 'totally inadequate'. Thus it does not capture the many persons who want to take up hours of paid work (and may respond to opportunities if these become available) but who do not meet the criteria to be included in the count of the unemployed. Instead, these persons are counted either as 'not in the labour force' or even as 'employed'. Stricker and Sheehan (1981) concentrated on the former group, but this article focuses on the latter group of persons, who have been included amongst the 'employed'. In the standard ABS classification people are counted as employed if they have worked (in a job or business or on a farm) for as little as one hour in the reference week (ABS, 2006b). Many part-time workers are content with their hours, and can therefore be regarded as 'fully employed', but other part-time workers state that they want more hours of paid work. These workers may be part-time employed, but they also feel themselves at least partly unemployed. They may be counted as employed, but they can also be regarded as underemployed.

In order to get a better idea of labour utilisation, the standard count of the unemployed needs to be supplemented by more comprehensive measures (Mitchell and Carlson, 2000; Watson, 2000; Denniss, 2001; ABS, 2001a, 2004). Advances have been made in Australia in developing such measures. The ABS has introduced two supplementary measures--a 'labour force underutilisation rate' and an 'extended labour force underutilisation rate'. The first includes the underemployed as well as the unemployed, while the second adds in certain groups that can be regarded as 'marginally attached to the labour force' (ABS, 2006b). Other research groups have developed similar measures. For example the Centre of Full Employment and Equity (CofFEE) at the University of Newcastle publishes regular information on labour underutilisation, drawn from ABS data, including an indicator of underemployment (CofFEE, 2007).

The central problem in using the unemployment rate as an indicator of labour market health derives from its flexible meaning. As O'Donnell (2003, 2005) points out, unemployment is best seen as a contingent notion, which emerged in the post-World War 2 period in association with particular forms of labour-management practices and labour regulation. The unemployment rate has the comforting appearance of a hard number, but its precise meaning has always varied according to the social and economic context. This point is most obvious if we compare labour force statistics across different nations. Many countries currently boast a low official unemployment rate (ILO, 2007), but this can mean markedly different things, and it is not necessarily an indicator of a healthy labour market. Indeed, in many countries a low unemployment rate is linked with a widespread 'informalisation' of economic activity. In the absence of a social welfare net and in the absence of decent job opportunities, many people are compelled to engage in informal activities that can earn them a small income. They are counted as employed, but they are in reality severely underemployed, and the society as a whole can be seen as marked by a chronic problem of labour underutilisation (Lee, McCann and Messenger, 2007:55-60).

If we confine our attention to OECD countries, the problem of interpretation is eased, but it is still pertinent. The unemployment rate was a good enough indicator of labour utilisation and labour market health in past decades, when male and female workers were limited to specific patterns of workforce participation, generally in the form of fulltime employment (from the end of education and training till marriage or retirement), supported by relatively comprehensive social security. But the measure has lost its potency as labour markets in OECD countries have changed, producing more fractured patterns of participation at different stages of the life course and more varied work schedules. Education and social security systems have also altered, with the result that contemporary labour markets in such countries are characterised by multiple and often-repeated transitions (Schmid, 1995; Schmid and Schomann, 2003; Watson et al, 2003). Underemployment and marginal attachment have become more important, though in patterns that vary widely across nations according to the nature of the employment and social security regime (MacDonald, 2002). In countries such as the United States, researchers warn of a process of informalisation that draws at least parts of the employment structure closer to a Third World model (Freeman, 2007). In short, even in OECD countries, a narrow focus on the unemployment rate is less and less justifiable. It risks missing much of what is happening in contemporary labour markets, including in particular the changing employment patterns of women.

The current unemployment rate in Australia has declined and it is now below the OECD average, after many years sitting comfortably above the OECD average. This may be grounds for cautious congratulation, but it would be premature to send out invitations to a party. The unemployment rate remains lower in several other countries, for example Norway, the Netherlands and New Zealand. Most important, we cannot be confident of the precise meaning of this low unemployment rate, unless we can examine what is happening with groups such as the underemployed.

Sources of Data on Underemployment

Individual countries use different approaches to measuring underemployment in their national statistics, some of which are better than others. The International Conference of Labour Statisticians adopted a resolution at their 1998 Conference, proposing a new standard for measuring time-related underemployment (Hussmanns, 2007; ABS, 2006b). However, little progress has been made so far in implementing the recommendation and most countries continue their traditional approaches (ILO, 2007).

The main source of data on underemployment in Australia is the monthly ABS Labour Force Survey. The ABS produces data for two measures that can be called 'underemployment': 1. a measure of 'part-time workers who want more hours'; and 2. a more recently developed measure, often simply called 'time-related underemployment', which has been designed to be consistent with the ILO recommendation (ABS, 2006b). The two categories overlap. The first is straightforward. In effect it is based on a simple division of part-time workers into two groups those who state a preference for more hours and those who state a preference for the same number of hours.1 The second category of underemployment is slightly more complicated. Roughly summarised, it starts, like the first measure, with part-time workers who want more hours. But it diverges in two respects: first, it includes some full-time workers (those who are usually full-time but are working less than thirty-five hours in the reference week for 'economic' reasons); and, second, it does not count all part-time workers who want more hours but only those who were available for more hours (in rough parallel to the measures of unemployment) (ABS, 2006b). This paper refers to these two measures as 'underemployment (1)' and 'underemployment2'. (2)

Data for 'part-time workers who want more hours' (underemployment1) have been collected on a monthly or quarterly basis as part of the Labour Force Survey and are available since 1978 (ABS, 2006x, 2008x). Data on 'time-related underemployment' (underemployment2) have been produced from a supplementary survey (the Underemployed Workers Survey) that has been conducted in September each year since 1994 (see ABS, 2007 and various issues). This supplementary survey includes some information on the group of 'part-time workers who want more hours' (underemployment1) as well as a wide range of information couched in terms of underemployment (2).

The ABS measures are preferable to most alternatives, but they are still far from perfect. As can be seen, both versions of underemployment refer to weekly hours. They capture what could be called regular underemployment, but they miss the phenomena of underemployment over a longer period that might be based in irregular hours or indeed irregular employment. Moreover, like all labour market indicators, the precise meaning can vary depending on the broader social and economic circumstances. Indeed this is particularly true for underemployment, which is couched in terms of individual workers' desires or preferences, and is therefore difficult to compare across different circumstances, because of the variation in the many background factors that shape such statements of preference.

Some recent analyses of underemployment (Watson et al, 2003:37-41; Wilkins, 2004, 2006, 2007) use Wave 1 of a new longitudinal survey, the Household, Income and Labour Dynamics in Australia (HILDA) survey, as an alternative source of data. This allows a measure that closely resembles underemployment (1). Though HILDA data are useful for some purposes, I use ABS data in this paper, primarily because they allow a description of the long-term trends and pattern of growth of underemployment.

Trends and Extent

Figure 1 compares unemployment trends with underemployment trends in Australia since 1978, using ABS data for the two versions of underemployment (underemployment1 and underemployment2). It shows underemployment1 as a proportion of the labour force for August each year since 1978, and it shows underemployment2 as a proportion of the labour force for September each year since 1994. The two are fairly close for the period since 1994 (though the latter is always lower than the former and the gap between the two has widened somewhat in recent years).


The story of unemployment is a familiar one, rising with each recession and then falling back in subsequent periods of recovery. Since the early 1990s, the unemployment rate has fallen steadily. But the story of underemployment is less familiar. Figure 1 illustrates important several points. First, it indicates that the underemployment trend diverges from the unemployment trend, influenced by cyclical trends but showing a reluctance to recede after the jumps linked to recessions. This is particularly true of the most recent period. The 'underemployment1 ' rate jumped sharply during the recession of the early 1990s, but in contrast to the trend for the unemployment rate it has not subsequently dropped away. Instead, it has plateaued, and indeed it now sits at a level (6.2 percent) slightly higher than the peak in 1993 (6.0 percent). The 'underemployment2' rate shows a similar flat pattern (though with some recent signs of retreat). This suggests that the cyclical influences affecting employment and unemployment have been supplemented by separate structural influences.

Second, Figure 1 indicates that the trajectory of growth has reached a point where the number of underemployed persons now clearly surpasses the number of unemployed persons. In August 2007, 'underemployedl' persons were 676,700 while unemployed persons were 447,700 (ABS, 2008x, 2008b). For 'underemployment1 ' the transition to numerical superiority occurred in 2001, and the advantage has grown stronger in the subsequent years. (3)

This high level in Australia stands out in cross-national comparison. High levels of underemployment are characteristic of developing countries, but they are less common in OECD countries outside of recessionary periods when both unemployment and underemployment could be expected to rise. In the 2005 data presented in a recent edition of Key Indicators of the Labour Market (ILO, 2007), several OECD countries show evidence of a similar trend of growth in underemployment. But Australia recorded the highest level (7.5 percent of the workforce), far ahead of the levels recorded for other leading OECD countries such as Canada (4.7 percent), Japan (4.7 percent), New Zealand (3.7 percent), and Sweden (3.2 percent). There are important limitations on comparability because of the different institutional frameworks that shape worker perceptions and because of differences amongst nations in definitions and methods of data collection (ILO, 2007). (4) Nevertheless, the gap is large and cannot be fully explained away as a result of measurement peculiarities. The data underline the impression that Australia has a significant problem of underemployment. It can be seen as a country where many workers suffer a 'working hours deficit' (Lee, 2004:46).

Underemployment primarily concerns part-time workers, and it can be seen in the context of the growth of part-time jobs. Part-time jobs have grown strongly since 1978, and indeed Australia boasts a particularly high part-time share for both males and females (OECD, 2007:261). As Lee notes (2004:49), countries with a relatively high incidence of part-time jobs tend to have a greater reliance on 'marginal' part-time jobs. A substantial proportion of the part-time jobs found in Australia are characterised by short 'marginal' hours of less than twenty per week (Lee, 2004:47-49; Lee, McCann and Messenger, 2007:55-58).

This high level of part-time work can be seen as a fertile environment in which underemployment can flourish. Of course, not all part-time workers want more hours. In August 2007 just less than one in four (22.7 percent) part-time workers stated that they wanted more hours. This reminds us, on the one hand, that underemployment in Australia--as measured at one point in time--is a minority phenomenon and that it is outnumbered by the part-time workers who state that they prefer not to work more hours. On the other hand, we can also note that the proportion of part-time workers who want more hours was just 12.9 percent in 1978. It increased steadily to reach a peak of 27.9 percent in 1993, in the immediate aftermath of the last recession, but it has fallen back slightly since then to its present figure of 22.7 percent (ABS, 2006x, 2008a). This suggests that the overall growth of underemployment is promoted by the growth in the part-time workforce but that increased underemployment is not a simple by-product of increased part-time work. (5)


Previous analyses draw attention to the characteristics of the jobs held by underemployed workers. The latter are predominantly employees in less skilled occupations (Watson et al, 2003; Wilkins, 2006). There is a concentration in service sector industries.

Table 1 shows that the extent of underemployment differs according to industry. Three industry groupings, all located in the service sector, have a particularly high underemployment rate ('underemployed1' workers as a proportion of total employed in that industry). They are accommodation, cafes and restaurants (17.6 percent), retail (13.6 percent) and cultural and recreation services (12.1 percent). It is true that these are also the industry groupings with high proportions for all part-time workers. But high proportions of part-time workers do not always involve high proportions of underemployed workers. For example, health and community services comprises many part-time workers, but the proportion that is underemployed is lower, leading to a lower 'underemployment1 ' rate overall.

Previous analyses also draw attention to the personal characteristics of underemployed workers (Watson et al, 2003; Wilkins, 2006). Underemployment tends to be concentrated in younger age groups (ABS, 2007 and various issues). But perhaps the most striking pattern is according to sex (Table 2). Underemployment is more of a problem for women than for men. Although male part-time workers are more likely to state a preference for more hours, their numbers are swamped by the many female part-time workers who want more hours. As a result, most underemployed workers (63.2 percent) are female. Almost ten percent of all female workers are part-time workers who want more hours, whereas this is true for less than five percent of male workers.

In light of the significance of gender divisions, it is useful to present the unemployment and underemployment trends for men and women separately (Figures 2 and 3). The patterns of change are similar in both cases, but the impact is different for men and women. For men, unemployment has been the dominant experience for most of the period, and underemployment1 has only recently become more important than unemployment. For women, in contrast, underemployment1 was already as widespread as unemployment in 1989 and it had clearly overhauled the unemployment rate by the mid-1990s.




Little research has been undertaken on the dynamics underlying the growth of underemployment in Australia (Wilkins, 2007:265). Too often underemployment has been assimilated with unemployment, and it has been assumed that both obey the same cyclical laws. Similarly, the research on underemployment from other OECD countries is still only sketchy and exploratory (Comte, 2005; Forssell and Jonsson, 2005). This section offers a few preliminary remarks, with the aim of defining a starting point for an analysis in Australia.

The rich literature on the dynamics behind the growth of part-time employment in general offers a few clues. This literature points to a complex mix of factors that have promoted increases in part-time jobs in many OECD countries over the past decades. Part-time jobs are most frequently found in the service sector, and structural shifts favouring service sector jobs can be seen as helping to boost the proportion of part-time jobs in a country. However, shift-share analyses both internationally (Walwei, 1998) and for Australia (Dawkins and Norris, 1995; de Ruyter and Burgess, 2000) suggest that restructuring effects make only a minor contribution to part-time growth and that the growing intensity of part-time employment across all sectors is more important. This analysis points to the role of supply-side factors such as changing household strategies, demand-side factors such as changing employer labour-use practices, and alterations in the institutional framework around labour regulation, tax and social security.

It is widely acknowledged that the growth of specific segments of part-time work may need specific explanations. For example, the growth of 'marginal' part-time work is often linked closely to industry conditions and changing employer practices (within a specific context that may include thresholds in tax and social security rules as well as new patterns of combining education and paid work amongst full-time students) (Lee, 2004:47-49). In this perspective, employer practices are seen as fostering jobs with very short hours. Recent studies in the United Kingdom refer to the emergence in several industries of 'fragmented time systems' in which "the agreed period of time is becoming fragmented into shorter, discontinuous periods and is being scheduled across the week or the year to match the requirements of employers" (Rubery, Ward and Grimshaw, 2006:123; see 2005). The researchers argue that these systems are anchored in the efforts of employers to achieve cost savings by removing paid on-the-job inactivity, that is, by intensifying work and closing up the pores in the working day. Though not confined to part-time work, such fragmented time systems tend to swell the ranks of part-time workers and alter their wages and conditions.

Explanations for underemployment can draw on these analyses. However, the analytic task remains formidable, because we are dealing with a segment of part-time workers defined in terms of their preferences. As several scholars note (Fagan, 2001; Bielenski, Bosch and Wagner, 2002; Lee and McCann, 2006), preferences around part-time work can be strongly influenced by contextual factors such as economic incentives, institutional barriers and sociological/ psychological factors (social norms), and these factors can in turn vary in complex ways according to circumstance. As a result preferences can be elusive and difficult to interpret.

Nevertheless, it is possible to make a few points. The issue is not so much why workers should want more hours but rather why they are impeded from obtaining the hours they want. Because underemployment directly signals deficient control over hours, the explanation will inevitably be drawn towards broad factors that constrain worker choice. (6) This includes cyclical influences, whereby poor labour market conditions push people into unsatisfactory jobs. But, as noted above, cyclical factors seem to have driven unemployment rates down and left underemployment rates high, and it is therefore necessary to bring in additional factors. As the analysis of the emergence of short hours' work hints, changing employer practices, especially in parts of the service sector, are likely to be influential (Comte, 2005).

Examples of how specific employment patterns and forms of job design can lead to underemployment can be found in recent Australian case-studies of shop assistants and checkout operators in retail (Campbell and Chalmers, 2008) and contract cleaners (Peeters and Campbell, 2008). Retail is an expanding industry, with significant concentration of capital in areas such as food retailing and recent experience of large-scale restructuring and innovation (Price, 2004). As in other advanced capitalist societies, large employers in retail have turned to part-time work as the most cost-effective way of matching labour to fluctuations in customer flows (Jany-Catrice and Lehndorff, 2005). Part-time employment offers several cost advantages, the most powerful of which is the ability to schedule part-time hours to save on paying for idle time. This advantage becomes increasingly important as working hours and opening hours are decoupled. Part-time employees can be used not only for daily peaks but also for periods during non-social hours such as evening and night work and weekend work. In this management approach, work is fragmented by time and customer flow, and labour inputs above a minimum complement are 'increasingly a variable factor of production ... purchased almost on an "as needed" or "just-in-time" basis' (Walsh, 1990:526). As a result, part-time jobs are configured in 'all shapes and sizes' and are constantly reconfigured according to employer perceptions of changing demand and the need for labour cost discipline.

Contract cleaning is also an expanding industry, with a core of large firms and a long tail of small firms. Because labour costs are the key item in expenses and there are few capital barriers to entry, new firms are constantly forming, eager to secure contracts from the owners or managers of large premises. Private contracting firms, both old and new, are continually seeking improvements in unit labour costs in order to achieve an advantage over their rivals in tendering (Ryan and Herod, 2006). Apart from efforts to lower standards by circumventing regulatory provisions for wages and conditions, the main thrust of employer cost-cutting is towards the intensification of work. As contracts turn over, firms seek to cut the total hours devoted to the job (generally by cutting the number of team members, while preserving the set number of hours for each individual worker). This approach builds on features of the work organisation, in which working hours of employees are tied to specific 'jobs' and are generally short. From the point of view of the worker, it implies a constant pressure on working hours that are already compressed. The majority of employees are part-time, often juggling several different 'jobs' with the one contractor, but they are still obliged to be constantly scrambling for work.

Both industries present examples of 'fragmented time systems', in which short hours under the control of the employer lead to widespread underemployment (see also Walsh, 2007: 164-165). As Table 1 indicates, retail is the industry group with the largest number of 'underemployedl' persons--202,200 in August 2006 (28.1 percent of all underemployed persons in Australia). The proportion of 'underemployedl' workers in the total retail workforce is high, and it has shown a pattern of slow growth in the period since 1994, increasing from 11 percent to 13.6 percent (ABS, 2006c). Similarly, in a recent survey of retail workers in Melbourne, around 40 percent of all part-time employees stated that they wanted more weekly hours (Campbell and Chalmers, 2008). Good data for cleaning workers are harder to obtain. However, a recent survey amongst cleaners in the Central Business District of Melbourne indicated that almost half felt underemployed (LHMU, 2007), and interviews with employees of contract cleaning firms in Melbourne and Adelaide (Peeters and Campbell, 2008) confirmed that underemployment was often a central theme in worker experience.

In both industries employer practices produce short hours and underemployment. But there are differences in the extent to which this is a specific objective of employers. On the one hand, underemployment seems to be actively welcomed and actively pursued by employers in retail. What is called 'passive flexibility' (Lehndorff and Voss-Dahm, 2005:291-2, 304-5), whereby workers are available for additional hours and shifts, often at short notice, according to employer needs, can be useful to retail employers. It allows them to replace absences, respond to unexpected fluctuations in demand, and fill difficult shifts more easily. Such 'temporal availability' depends on several conditions, but it is clearly facilitated when the workers in question want to work more hours, that is, when they are underemployed. In cleaning, on the other hand, 'passive flexibility' is not as important. The fragmentation in jobs is anchored in structural features of the industry. The short hours are sustained because they offer a convenient platform for the process of intensification that employers have adopted as the main vehicle for cost-cutting and competitiveness. The fact that workers desire more hours may encourage compliance with ever-changing demands and strict workloads, and as such underemployment may be appreciated by employers. Nevertheless, in this case it appears as more of an incidental by-product of structural conditions and practices aimed at other objectives.

Why are employees in these industries vulnerable to short and fractured hours that entail underemployment? One answer can be framed in terms of powerlessness in the face of strong and determined employers. Sales assistants and cleaners possess few marketable skills, and they can be easily replaced by other workers looking for part-time work. Both gender and ethnicity can contribute to marginalisation. Employers can draw on groups such as full-time students, who more readily tolerate short and fragmented hours. Though unions are present in some parts of each industry, they have difficulty in asserting themselves in collective bargaining. One decisive factor is the weakness of protective regulation around working-time in Australia (Lee, 2004; see Campbell, 2007: 5157). Certainly in the two industries discussed above, there are few obstacles to the establishment of fragmented schedules. In cleaning, the minimum payments for part-time employees prescribed by the relevant award, generally two hours per day (three hours on Sunday), represent only a minor obstacle. In retail, awards and agreements contain more complex provisions, including penalty payments for work during nonsocial periods, but nothing that would constitute a major barrier to fragmentation.

Vulnerability to underemployment and lack of control over working hours could perhaps be expected in the case of casual employees, who have few rights and entitlements under regulation. But weak protective regulation in these industries extends vulnerability to many part-time permanent employees as well. In contrast to casuals, part-time permanent employees possess certain basic rights to paid leave entitlements and regular hours and regular rosters. But this does not seem sufficient to provide protection against employer demands. In retail, employers often make the decision about the number of hours and the rosters given to part-time permanent employees, and the hours are often significantly less than desired by the employee (Campbell and Chalmers, 2008). In cleaning, as we have seen, the number of hours is seen as prescribed by the specific contract or 'job', and permanent status does not necessarily lead to any enhanced rights to acquire several jobs.

The service sector offers a favourable environment for underemployment. However, other industries with low skill workers, constrained trade unions, and weak protective regulation can also exhibit widespread underemployment. Such conditions allow fragmented time systems to be generalised within specific industries and then across the boundaries into other industries, thereby providing a platform for widespread underemployment.


Although there have been advances in employment growth during the economic boom of the last sixteen years, talk of full employment is clearly misleading. As this article demonstrates, the growth and persistence of underemployment suggests that labour markets in Australia are still marked by extensive labour underutilisation. Underemployment may yet recede somewhat, if labour market conditions continue to improve and if this flows through to into an enhanced capacity for workers to resist employer pressure on hours. However, it is does not appear that cyclical factors can be relied on to solve the problem.

Underemployment remains a puzzling issue. It represents one of the crucial contradictions of the present period of economic buoyancy (Bramble, 2004). Though comparative data are slippery, underemployment seems to be more prominent in Australia than elsewhere in the OECD. In spite of the good economic conditions, it has held up and even increased while the unemployment rate has steadily declined. Analyses of this pattern of growth, as well as the concentration of underemployed part-time workers in specific industries and occupations, suggest that underemployment is a distinct labour market phenomenon, only weakly related to changes in unemployment. This presents a challenge for explanation. This article responds to the challenge by stressing industry conditions and dominant employer strategies and practices (within an institutional framework marked by weak working-time regulation). However, this is only a starting point for explanation.

The analysis in this article questions the conventional discussion of tight labour markets. But it also points beyond 'labour market slack' to the broader topic of 'labour market well-being' (MacDonald, 2002). The discussion of underemployment spans the crucial issue of job quality as well as the issue of job quantity. The fact that part-time workers judge the number of hours in their jobs as insufficient can be seen as one indicator of poor quality in part-time jobs, readily connected to other indicators such as low wages and lack of control over hours and schedules (Chalmers, Campbell and Charlesworth, 2005). Underemployment can be seen in this light as one indicator of a broader problem that stems from the specific way in which the economic boom has unfolded in Australia.

The broader discussion of job quality inevitably leads back to the effect of federal government employment policies. These policies, inspired by neoliberalism and vigorously pursued by the previous Coalition government, weakened protective regulation and encouraged employers to circumvent labour standards, thereby helping to fragment wages and conditions and to spread precariousness within certain parts of the workforce. They may not have played any role in contributing to growth in the quantity of jobs, but these policies seem to have had a major negative impact on the quality of many jobs in Australia, both full-time and part-time (Watson et al, 2003). The task of cleaning up this unfortunate legacy constitutes a major challenge for labour market policy.

The research for this paper was supported under the Australian Research Council's Discovery Project funding scheme (DP0449771).

Thanks to Warren Friddell, who helped in chasing up the ABS data, to Jenny Chalmers and Sara Charlesworth, who are involved in the broader project, and to participants at the AIRAANZ conference in February 2007 and the meeting of the International Working Party on Labour Market Segmentation in July 2007, who offered comments on earlier versions of the paper. The author would also like to thank Inger Jonsson and Johanna Forssell, who helped to focus his interest in the topic of underemployment.


ABS [Australian Bureau of Statistics] (2001a) 'Special Article--Unemployment and Supplementary Measures of Underutilised Labour', Labour Force Australia, Cat. no. 6203.0.

ABS (2001b) Information Paper: Questionnaires Used in the Labour Force Survey, Cat. no. 6232.0.

ABS (2004) 'Feature Article--Labour Underutilisation', Australian Labour Market Statistics, Cat. no. 6105.0, July.

ABS (2006a) Labour Force, Australia, Detailed Quarterly, data cube: SuperTABLE, Cat. No. 6291.0.55.003, E01 Feb78.

ABS (2006b) Labour Statistics: Concepts, Sources and Methods, August, Cat. No. 6102.0.55.001.

ABS (2006c) data provided on request.

ABS (2007) 'Updated Volume Measures of Labour Underutilisation', Australian Labour Market Statistics, Cat. No. 6105.0, April.

ABS (2007 and various issues) Underemployed Workers Australia, September, Cat. No. 6265.0.

ABS (2008a) Labour Force, Australia, Detailed Quarterly, data cube: SuperTABLE, Cat. No. 6291.0.55.003, E01 May01, viewed 23 January 2008.

ABS (2008b) Labour Force Australia, 'Table 03. Labour force status by sex', Time series spreadsheet, Cat. No. 6202.55.001, viewed 23 January 2008.

Bielenski, H., Bosch, G. and Wagner, A. (2002) Working time preferences in sixteen European countries, Dublin, European Foundation for the Improvement of Living and Working Conditions.

Bramble, T. (2004) 'Contradictions in Australia's "Miracle Economy"', Journal of Australian Political Economy, No. 54, pp. 5-31.

Campbell, I. (2007) 'Long Working Hours in Australia: Working-Time Regulation and Employer Pressures', The Economic and Labour Relations Review 17 (2), 37-68.

Campbell, I. and Chalmers, J. (2008) 'Job Quality and Part-time Work in the Retail Industry: An Australian Case Study', International Journal of Human Resource Management 19 (3), 487-500.

Chalmers, J., Campbell, I. and Charlesworth, S. (2005) 'Part-time Work and Caring Responsibilities in Australia: Towards an Assessment of Job Quality', Labour and Industry 15 (3), 41-66.

CofFEE [Centre of Full Employment and Equity] (2007) 'CLMI 2: Underemployment', CofFEE Labour Market Indicators, May, Newcastle, CofFEE.

Comte, M. (2005) 'Le sous-emploi: pourquoi?' Travail et Emploi no. 103, 95-111.

Dawkins, P. and Norris, K. (1995) 'The growth of part-time employment in Australia and the United Kingdom, 1978-93', Labour Economics and Productivity 7(1), 1-27.

De Neubourg, C. (1988) Unemployment, Labour Slack and Labour Market Accounting, Amsterdam, North-Holland.

de Ruyter, A. and Burgess, J. (2000) 'Part-time employment in Australia: Evidence for globalization?', International Journal of Manpower 21(6), 452-463.

Denniss, R. (2001) Measuring Employment in the 21st Century: New measures of underemployment and overwork, Discussion Paper no. 36, Canberra: The Australia Institute.

Eurostat (2007) Eurostat Metadata in SDDS format: Summary Methodology, online at sdds/en/lfsi_series_sm.htm, accessed 17 August 2007.

Fagan, C. (2001) 'Time, Money and the Gender Order: Work Orientations and Working-Time Preferences in Britain', Gender, Work and Organization 8(3), 239-266.

Forssell, J. and Jonsson, I. (2005) Part-time unemployment in a European comparative perspective--with special reference to the situation in Sweden, paper for an Arbetslivsinstitutet workshop, Visby, Sweden, December 7-8.

Freeman, R. (2007) 'The challenges to global labour', New Statesman, 10 September, 6-7.

Hussmanns, R. (2007) 'Measurement of employment, unemployment and underemployment-Current international standards and issues in their application', Bulletin of Labour Statistics 2007-1, online at: http:/ bureau/stat/download/articles/2007-1.pdf, accessed 25 January 2008.

ILO [International Labour Office] (2007) Key Indicators of the Labour Market, 5th edition, Geneva, ILO.

Jany-Catrice, F. and Lehndorff, S. (2005) 'Work organization and the importance of labour markets in the European retail trade', in G. Bosch and S. Lehndorff (eds) Working in the Service Sector: A tale from different worlds, London, Routledge, 211-36.

Lee, S. (2004) 'Working-hour gaps: trends and issues', in J. Messenger (ed.), Working Time and Workers' Preferences in Industrialized Countries: Finding the balance, London, Routledge, 29-59.

Lee, S. and McCann, D. (2006) 'Working time capability: Towards realizing individual choice', in J.-Y. Boulin, M. Lallement, J. Messenger and F. Michon (eds) Decent working time: New trends, new issues, Geneva, ILO, 65-91.

Lee, S., McCann, D. and Messenger, J. (2007) Working Time Around the World: Trends in working hours, laws and policies in a global comparative perspective, London, Routledge.

Lehndorff, S. and Voss-Dahm, D. (2005) 'The delegation of uncertainty: flexibility and the role of the market in service work', in G. Bosch and S. Lehndorff (eds) Working in the Service Sector: A tale from different worlds, London, Routledge, 289-315.

LHMU (2007) Preliminary report on survey findings, 22 August, Melbourne, LHMU.

MacDonald, M. (2002) Underemployment as a Measure of Labour Market Well-Being: Measurement and Meaning Across Countries, paper for a conference on the Development of a new Cross-National Architecture for Labour Market Statistics, Bellagio, September 2327, online at, accessed 25 January 2008.

Mitchell, W. and Carlson, E. (2000) Beyond the unemployment rate--labour underutilisation and underemployment in Australia and the USA. Working Paper No. 0006, Newcastle, Centre of Full Employment and Equity (CofFEE), University of Newcastle.

O'Donnell, A. (2003) 'Inventing Unemployment: Labour Market Regulation and the Establishment of the Commonwealth Employment Service', Federal Law Review 31(2), 343-372.

O'Donnell, A. (2005) Re-inventing Unemployment: Welfare Reform as Labour Market Regulation, Working Paper no. 36, Centre for Employment and Labour Relations Law, University of Melbourne.

OECD [Organisation for Economic Co-operation and Development] (1995) 'Supplementary Measures of Labour Market Slack: An analysis of Discouraged and Involuntary Part-time Workers', ch. 2 in OECD Employment Outlook, Paris, OECD, 43-97.

OECD (2007) OECD Employment Outlook, Paris, OECD.

Peeters, M. and Campbell, I. (2008) 'Low Pay and Working Time: The Case of Contract Cleaners', in P. Stanton and S. Young eds., Workers, Corporations, and Community, Proceedings of the 22nd Conference of AIRAANZ, volume 2, Melbourne, AIRAANZ, 296-305.

Price, R. (2004) Checking out Supermarket Labour Usage: The Nature of Labour Usage and Employment Relations Consequences in a Food Retail Firm in Australia, unpublished PhD thesis, Department of Industrial Relations: Griffith University, Brisbane.

Rubery, J., Ward, K. and Grimshaw, D. (2005) 'The Changing Employment Relationship and the Implications for Quality Part-time Work', Labour and Industry 15(3): 7-28.

Rubery, J., Ward, K. and Grimshaw, D. (2006) 'Time, work and pay: understanding the new relationships', in J.-Y. Boulin, M. Lallement, J. Messenger and F. Michon (eds) Decent working time: New trends, new issues, Geneva, ILO, 123-151.

Ryan, S. and Herod, A. (2006) 'Restructuring the Architecture of State Regulation in the Australian and Aotearoa/ New Zealand Cleaning Industries and the Growth of Precarious Employment', Antipode 38(3): 486-507.

Schmid, G. (1995) 'Is Full Employment Still Possible? Transitional Labour Markets as a New Strategy of Labour Market Policy', Economic and Industrial Democracy 16, 429-456.

Schmid, G. and Sch6mann, K. eds., (2003) The Concept of Transitional Labour Markets and Some Policy Conclusions: The State of the Art, Working Paper, TLM.NET, Amsterdam.

Stricker, P. and Sheehan, P. (1981) Hidden Unemployment: The Australian experience, Melbourne, Institute of Applied Economic and Social Research, University of Melbourne.

The Australian (2007a) 'Australia nearing "full employment"', The Australian 9 August.

The Australian (2007b) 'How to sustain our record low unemployment rates', The Australian, 16 August.

Walsh, J. (2007) 'Experiencing Part-Time Work: Temporal Tensions, Social Relations and the Work-Family Interface', British Journal of Industrial Relations 45 (1), 155-177.

Walsh, T. (1990) 'Flexible Labour Utilisation in the Private Service Sector', Work, Employment and Society 4(4): 517-30.

Walwei, U. (1998) 'Are part-time jobs better than no jobs?' in J. O'Reilly and C. Fagan eds., Part-time Prospects: An International Comparison of Part-time Work in Europe, North America and the Pacific Rim, London, Roudedge, 96-115.

Watson, I. (2000) 'Beyond the Unemployment Rate: Building a Set of Indices to Measure the Health of the Labour Market', Australian Bulletin of Labour 26(3): 175-190.

Watson, L, Buchanan, J., Campbell, I. and Briggs, C. (2003) Fragmented Futures: New Challenges in Working Life. Sydney: Federation Press.

Wilkins, R. (2004) The Extent and Consequences of Underemployment in Australia. Working Paper no. 16/04, Melbourne, Melbourne Institute of Applied Economic and Social Research.

Wilkins, R. (2006) 'Personal and Job Characteristics Associated with Underemployment', Australian Journal of Labour Economics 9 (4), 371-393.

Wilkins, R. (2007) 'The Consequences of Underemployment for the Underemployed', Journal of Industrial Relations 49 (2): 247-275.

Wooden, M. and Drago, R. (2007) The Changing Distribution of Working Hours in Australia, Working Paper no. 19/07, Melbourne Institute of Applied Economic and Social Research, University of Melbourne.

(1) Part-time workers are defined as employed persons "who usually work less than 35 hours a week (in all jobs), and either did so during the reference week, or were not at work during the reference week" (ABS, 2006b). To establish the estimate of 'part-time workers who want more hours', part-time workers are asked whether they would prefer to work more hours than they usually work (ABS, 2001b).

(2) We focus just on the main contemporary Australian measures. Another measure appears in the official statistics for most European countries--it is used in the European Labour Force Survey (ELFS)--as well as in other countries such as Canada and the United States (ILO, 2007). This measure counts only those part-time workers who report involuntary reasons for working fewer than full-time hours, and it is usually known as 'involuntary part-time work'. The range of reasons counted as 'involuntary' can vary, but 'involuntary' part-time workers are often defined as those who state that they work part-time because they cannot find full-time work (Eurostat, 2007), while those who declare any other reason are counted as 'voluntary' part-time workers. This is the narrowest and least plausible measure of underemployment. A somewhat similar measure was used in Australia from 1964 to 1977, but it was subsequently dropped (Stricker and Sheehan, 1981:28-29). Though no real equivalent is currently used in Australia, it is possible to obtain an approximation by calculating the number of part-time workers who state that they would prefer to work full-time. According to 2006 data for Australia, slightly more than half (56.8 percent) of all part-time workers who wanted more hours stated that they wanted to work full-time hours. This was equivalent to 3.2 percent of the workforce (ABS, 2007 and various issues).

(3) We refer here just to the headcount measure. Though underemployed workers want a significant number of extra hours of work, the volume of extra hours desired is generally less than the volume of hours desired by the unemployed. As a result, on an hours-adjusted or volume measure, underemployment may be seen as less important than unemployment. The ABS has recently developed 'experimental volume measures' of labour underutilisation. In the latest estimate (ABS, 2007), underemployment is given a value of 2.1 percent compared to 3.8 percent for unemployment. The indicator of underemployment used by CofFEE (2007) is also a volume measure, where the number of underemployed workers is converted to a full-time equivalent.

(4) For example, Australia uses a broad definition of part-time workers (with 35 hours as the weekly cut-off) and uses the broader definition of underemployment, in comparison with some countries that just use a notion of 'involuntary part-time'. Both factors will tend to boost the estimate of underemployment for Australia compared to estimates in certain other countries.

(5) A simple shift-share calculation suggests that about one third of the growth in underemployment as a proportion of all employment from 1978 to 2006 can be associated with the growth in part-time employment as a proportion of all employment. The other two thirds is linked to other causes.

(6) This does not, of course, imply that choice is completely denied. The sharp polarity of 'voluntary' and 'involuntary' is not appropriate in the case of wage labour, where (constrained) choices often come into play. For example, to accept (and remain in) a job, despite insufficient hours, implies a degree of choice. Moreover, a statement that working hours are insufficient often implies that the worker is seeking alternatives and that s/he is likely to exercise a decision to alter the situation in the medium term. The three main paths forward are: 1. find another job with more hours; 2. find more hours in the same job; 3. find a second job to add to the current job. We could expect many workers to choose to take up one of these options in the medium term and to thereby 'solve' their immediate problem of underemployment. Of course, 'solutions' such as multiple job holding may not last long and may come with their own problems of control over hours. We can note here that longitudinal data on workers' careers (Wooden and Drago, 2007) suggest that the individual worker experience of underemployment does not usually last more than a year. However, the worker's experience should not be confused with the underlying structure, which continues to generate part-time jobs that are experienced by a minority of workers as underemployment.

Iain Campbell is at the Centre for Applied Social Research, RMIT University, Melbourne
Table 1: Underemployed1 Workers by Selected
Industries, Australia, August 2006

 Part-time workers who
 workers who prefer not to
 want more hours work more hours
Industry ('000) ('000)

Manufacturing 32.7 99.0
Construction 38.8 99.1
Wholesale trade
Retail trade 14.4 62.9
Accomm., cafes & 202.2 517.2
restaurants 84.6 165.0
Transport & storage 23.0 58.7
Communication services 7.7 19.6
Finance & insurance 9.6 62.0
Property & bus. services 68.2 260.1
Govt. admin. & defence 13.9 64.5
Education 57.0 207.5
Health & community services 89.1 388.9
Cultural & recreation services 31.8 78.7
Personal & other services 31.7 77.0
All industries 719.6 2236.7

 oyment1 as
 proportion of Total
 total employed employed
Industry (%) ('000)

Manufacturing 3.1 1056.2
Construction 4.3 892.8
Wholesale trade
Retail trade 3.0 482.4
Accomm., cafes & 13.6 1490.6
restaurants 17.6 481.2
Transport & storage 5.0 465.0
Communication services 4.3 178.7
Finance & insurance 2.5 377.6
Property & bus. services 5.5 1245.8
Govt. admin. & defence 2.7 507.0
Education 8.1 704.0
Health & community services 8.3 1067.2
Cultural & recreation services 12.1 263.3
Personal & other services 8.2 388.8
All industries 7.1 10172.0

Source: ABS data available on request; ABS 2008b.

Table 2: Underemployed1 Workers by Sex,
August 2007 ('000)

 Part-time workers who
 workers who prefer not to
 want more work more Total Total
 hours hours part-time employed

Male 248.8 617.3 866.1 5749.7
Female 427.9 1683.9 2111.9 4686.1
Total 676.7 2301.2 2978.0 10435.8

Source: ABS 2008a; 2008b.
COPYRIGHT 2008 Australian Political Economy Movement
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Campbell, Iain
Publication:Journal of Australian Political Economy
Article Type:Essay
Geographic Code:8AUST
Date:Jun 1, 2008
Previous Article:The institutional intractability of recession.
Next Article:Western Australia's boom economy: insights from three studies.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters