Increasing employment participation in Australia and how to finance it.

Abstract

This paper considers the optimal combination of policies that maximise the opportunities for paid employment while also being acceptable in terms of the income distribution effects. While the official unemployment rate has remained relatively low over recent years, the growth in full time employment has been 'sluggish'. The officially unemployed comprise only part of the picture of labour supply. In addition we need to take into account those unemployed who have disengaged from the labour market (who might be enticed back under the right conditions) and those who are underemployed-typically part-time workers working less hours than desired. Adding these three groups together gives a more realistic account of the potential labour supply of those wanting more work (estimated at 11 per cent). These three groups have in common low educational levels or lack of opportunity at their current skill level and barriers to employment, indicating a role for government in terms of employment generation and facilitation.

Introduction

Australia has now experienced thirteen years of continuous economic growth at an average annual rate of 3.5 per cent. This makes Australia one of the fastest growing developed market economies over an extended period. Yet the increase in full-time employment has been sluggish, averaging an annual rate of increase of slightly less than 1 per cent between 1991 and 2004. The official rate of unemployment is still around 5 per cent, compared to just over 1 per cent in the mid-1960s, and male employment participation has fallen by about 16 percentage points since then.

So, although economic management in Australia has been relatively successful for more than a decade, it is not altogether surprising that some people have questioned whether economic growth will be sufficient to ensure paid employment for all who want it. It is this line of thinking that has led to proposals in favour of the development of so-called transitional labour markets, which would allow people to participate in gainful non-market activities that will preserve and enhance their future employability. For example, the Transitions and Risk conference descriptor (Centre for Public Policy 2004, p. 1), suggests that '(t)he notion of transitional labour markets reflects the reality that ongoing employment for all who seek it is difficult to attain'. In addition, as Schmid (1998, p. 3) has noted, our highly individualistic society has been associated with increasing female participation in paid employment, but women naturally also want more flexible labour markets that are better able to respond to changing individual tastes and preferences over their life-cycle.

It is this combination of circumstances that underpins Schmid's (1998, p. 4) assertion that 'the goal of a more employment intensive growth can only be achieved by a working time reduction of those who have a job, and this, realistically, implies some kind of income redistribution' [my emphasis]. This paper, however, questions the first part of this assertion that the desired increase in employment is impossible without a working time reduction. On the other hand it is accepted that the goal of employment for all who want it cannot be achieved without a combination of policies that will very likely alter the distribution of income. For example, the experience with the then Labor Government's Accord with the trade unions during the 1980s, when employment increased rapidly relative to economic growth, suggests that if people were prepared to restrain their wages then we could come much closer to employing all who want at least some paid employment. But will many people be prepared to make such wage sacrifices today? Instead it seems likely that many families are working more hours than their parents did precisely because they are having difficulty in meeting their material expectations, which have apparently risen faster than their means. Accordingly the focus of this paper is on the optimum combination of policies that will maximise the opportunities for paid employment while also having the best chance of acceptance in terms of their income distribution effects.

The first part of the paper begins by considering the extent of the restructuring of the Australian labour market, and what further changes might be anticipated, including the impact of an ageing population. The aim is to assess the extent of unemployment, withdrawals from the active labour market, and underemployment in Australia today. This assessment provides the basis for a rough measure of the potential employment, and by how much labour demand would need to increase in order to reasonably satisfy all demands for work that could be expected in a fully employed economy. The second part of the paper then considers the appropriate policy measures to achieve this desired level of employment and how these measures might be financed. The critical issue will be to find an acceptable approach that has the best chance of resolving the conflicting claims affecting the distribution of income. But failure to achieve this resolution could risk Australia becoming a less cohesive and egalitarian society where there is a sharper dividing line between the 'haves' and the 'have-nots'.

The Nature and Extent of Labour Market Restructuring in Australia

The first major structural change in the Australian labour market over the last 30 to 40 years has been a substantial decline in male employment participation, which has been matched by an increase in female employment participation. Since 1966, male workforce participation has fallen from 84 per cent to 72 per cent while female participation has increased from 40 per cent to 55 per cent (Table 1). Combining the two sexes, the overall workforce participation rate for persons aged from 15 to 64 has increased from 60 per cent to 64 per cent since 1966. However, if allowance is made for the four percentage point increase in unemployment, the overall rate of employment participation is now the same as it was nearly forty years ago.

The second major change has been in the occupational composition of the workforce. Over the last thirty years, Australia has created many skilled professional jobs, but there has been negligible growth in manual jobs of the traditional 'blue collar' kind for males (Table 2). Indeed, it is most probable that the number of full-time male blue-collar jobs probably fell over the last twenty years, and we know that it did during the 1990s. It is this decline in 'blue collar' employment, from almost two-thirds of total male employment in 1971 to less than half by 2000, that appears to be the immediate cause of the decline in male employment participation, already described.

The third major structural change in the workforce has been the relatively rapid growth in part-time employment, which has tended to offset the first two changes described. The share of part-time employment has nearly doubled from 15.5 per cent of total employment in 1978 to 28.5 per cent today (ABS: 6202.0 and ABS electronic database for earlier years). Australia now has the second highest rate of part-time employment in the OCED. Indeed, unlike many other countries, the relatively rapid growth of part-time employment has effectively allowed Australia to embrace a considerable measure of job-sharing on a largely voluntary basis for both employers and employees. Many married women, in particular, prefer part-time employment as it helps to balance their employment with their family responsibilities, and part-time employment now accounts for nearly half of total female employment.

Bringing these three major changes together, there has been on balance a significant shortfall in the demand for labour. Thus, the present supply of labour, and even more so the potential supply, is outstripping the current demand. This imbalance between labour demand and supply has especially affected male employment, and is the main reason why male employment participation has fallen so much in the last forty years.

Rough estimates of the available labour supply and the potential supply in a fully-employed economy are provided in Table 3, where the following three groups are identified as either wanting or potentially being available for more paid employment:

* People who are employed part-time and who would like to work more hours

* Unemployed people who are searching for work

* People who are marginally attached to the workforce, as they want to work and they would be available to start work in the next four weeks. In addition, this third group includes an allowance for those men on disability support pensions or who have taken early retirement and who could be induced back into employment by changes in government policy.

The estimates are explained in more detail in the Data Appendix (at the end of this paper) but, unfortunately for some of these groups, the estimated additional hours potentially available are really only 'educated guesstimates', which could be improved upon with better information. Nevertheless the estimates of potential employment are believed to be sufficiently robust to indicate broad orders of magnitude and to assess relative priorities. Some key assumptions are first, that a fully-employed labour force would still have some frictional unemployment. Second, after having regard to the present activities of different groups of marginally attached males and females, not all these people are assumed to actually be available for work, and different allowances were made for the proportion of each component sub-group who could actually obtain employment. In addition, in the case of marginally attached males, it is assumed that their potential labour supply could be further increased in response to new government policies directed towards reducing the number of disability pensioners and early retirees. Third, the average number of additional hours of work wanted has been obtained from ABS labour force surveys for those people already in the active labour force, and has been inferred from the characteristics of the component groups in the marginally attached labour force.

What stands out in Table 3 is that, relative to total employment of 9.6 million persons in September 2003, as many as 1.7 million persons wanted or could be reasonably induced to have wanted more work. That would represent an increase of another 18 per cent. Excluding those people who are already employed, although underemployed, the number of extra people who could have been employed falls to 1.2 million people, representing an increase in total employment of 12 per cent. However, the additional hours that would be worked by most of these people who want more employment would be less than a full-time working week. Thus the potential addition to the labour supply measured in terms of hours worked would be of the order of 11 per cent. (1)

A matching increase in the demand for labour of 11 per cent in an economy which in significant respects is already running at close to full capacity is clearly a major challenge. Over time the task will become easier, not least because with an ageing society the demand for labour will tend to rise faster than its supply. But an increase in the number of hours worked of as much as 11 per cent would be more than sufficient to accommodate the expected demands of an ageing society. Also this increase in demand from an ageing society is still some way off and will only slowly take up the present excess in the potential labour supply. There is therefore a strong argument for government action now to improve people's job chances. Quite apart from the benefits to the individuals concerned by generating this extra employment, there would be considerable benefits to society as a whole, including a significant reduction in the projected future fiscal demands.

Key Characteristics of Those People Who Want More Work

The starting point in considering how to ensure that everybody is able to obtain the extra hours of employment they desire is to better understand the nature of the disadvantages and/or perverse incentives that people wanting more work face when they compete in today's labour market. Of course, the target groups are not completely homogeneous, but they do present sufficient common features that can provide a basis for policy.

Prime-age Males and Females

The first and probably the largest group of people whose employment potential is not currently being fully realised are prime-age males and females, aged between their early twenties and mid-fifties. The common characteristic shared by the three groups of people who have been identified as wanting more paid employment--unemployed persons, part-time workers, and people who are marginally attached to the workforce--is that their skills and education levels all seem to be much lower than for fully-employed persons. First, in 2003 the highest level of education attained by 50 per cent of unemployed males and 47 per cent of unemployed females did not extend to completion of secondary school (ABS: 6222.0 Table 3). Second, among those part-time workers who want to work more hours, about 60 per cent of the males and 55 per cent of the females consider that they lack the necessary skills and/or that there are insufficient vacancies for people with their skill levels (ABS: 6265.0 Table 7). Third, the Australian Government (2004) has shown that almost all the decline in workforce participation for those males aged 25 to 55 is accounted for by men who have no post-school qualifications, and in most cases left school at year g or even earlier. And although female workforce participation has increased for all educational levels over the last twenty-five years, it continues to be about 30 per cent lower for women who have no post-school qualifications compared to those women who have. Fourth, this decline in employment participation by low-skilled people has been associated with a decline in the number of 'blue-collar' jobs which used to employ most unskilled people. In sum, it seems highly likely therefore that the major reason why most people aged between 25 and 55 cannot get as much work as they want or might want is because they lack the necessary skills to be fully competitive in today's labour market.

In addition, it is possible that some of these people, who have low education and skill levels, and consequently low pay expectations, may then feel that they have insufficient incentive to pursue extra work. The disability support pension (DSP) for a couple is as much as 88 per cent of the 'minimum' wage that a low-skilled person could expect to earn in employment. In a situation where these people are finding it difficult to obtain work they may therefore lose the incentive to actively seek employment when they can survive in not much worse circumstances on the DSP. Certainly by 2001 the proportion of the active population receiving the DSP had doubled compared to twenty years earlier, and it is hard to believe that this increase only reflects a commensurate decline in health status.

Other possible disincentives to actively pursue employment can be brought about by overlapping means tests for different forms of income and other support and the interaction of the tax system with the social security system. Very high effective marginal tax rates for members of families with children on low incomes can then result. This disincentive may be a factor leading to the situation where as many as one-sixth of children live in households where no-one is engaged in paid employment.

Finally, one other factor that may affect the incentives for employment are the extra costs associated with employment. Interestingly, however, the ABS surveys do not report difficulties in obtaining child care as a major reason why women, in particular, experience difficulty in obtaining the hours of work that they would like. However, Toohey (2005) points out that the costs of child care can substantially reduce return to work for families, and especially for women, with children.

Mature-age Males and Females

Many of the difficulties facing unskilled prime-age workers in finding sufficient employment also affect mature-age people, aged from their mid fifties to late sixties, or in some cases even older. In particular, Australian Government (2004, p. 6) found that 'Thirty per cent of 50- to 65-year-olds who retire in Australia do so because of illness or disability'. This has led to a situation where nearly 30 per cent of 50- to 64-year-olds receive income support, and very few who claim income support after the age of 50 ever come off benefits (Australian Government 2004, p. 8).

But it should equally be noted that more than half of male retirees aged from 45 to 64 declare that their 'early' retirement was involuntary, and many of the other 'voluntary' retirements may have received some encouragement from employers whose firms were restructuring. These mature-age people who are seeking employment often lack not only the skills to be competitive but they report that the attitude of employers that they are too old is also a problem. When mature-age workers become unemployed they are much more likely to be unemployed for a long time, (2) and as they become discouraged they are more likely to drop out of the workforce altogether than people in younger age groups. Consequently, many of these mature-age people are living on benefits because they perceive that they have no other choice, while others are using up their redundancy pay or accessing their typically limited superannuation entitlements early.

On the other hand, early retirement can be an attractive proposition for many skilled professionals on higher incomes, who have sufficient superannuation to provide an acceptable living standard. At this stage, however, most mature-age employees have not accumulated sufficient amounts in their superannuation accounts to finance their early retirement. But this situation can be expected to change over time as superannuation entitlements accumulate; although less so for those in casual and low paid work.

Youth

Most young people who would like but cannot obtain full-time work are not students. Again, these unemployed youths are typically early school-leavers with low skills. However, although youth unemployment rates appear to be high, school retention rates have increased greatly in the last two decades. As a result, in 2001 only 3 per cent of people aged 25 finished their education at year 9 or less compared with 30 per cent at age 60 Population Census (ABS 2001). This means, first, that a declining proportion of all youth is experiencing unemployment. Second, the increasing proportion of young people who have higher qualifications and post-school skills can be expected to reduce the risk of today's youth later experiencing periodic unemployment over their lifetimes.

It is clearly important to try to keep youth engaged in continuing their studies. Improvements in fostering school-to-work transition are an obvious policy response, which can help to maintain young people's interest in learning as well as increasing their immediate employability.

Policies to Increase Paid Employment

As noted in the introduction, economic growth is necessary but not sufficient to realise the potential supply of employment. Indeed the demand for goods and services is currently running ahead of supply in Australia. Consequently, our current account on the balance of payments is in deficit, and shortages of skilled labour are being reported. Faster expansion of aggregate demand in these circumstances could not be sustained, notwithstanding that potentially there is considerable excess labour supply. Instead a fully-employed economy requires an increase in the employment content of Australian output, while still maintaining sufficient productivity and real income growth to be acceptable to the majority of Australians who appear to be reasonably satisfied by present policies.

It has been shown elsewhere that the loss of blue-collar jobs described above is not due to the opening up of the Australian economy (see Keating 2000, pp. 101-5; Keating 2004, pp. 122-3). In brief, this research shows that unskilled jobs have been lost in all industries whether or not they are directly affected by international trade, and that 'the fall in employment associated with increased low-wage imports was outweighed by the increase in jobs associated with greater export' (Murtough, Pearson and Wreford 1998, p. xi). Instead, technology has been the key factor leading to a loss of mainly unskilled jobs, which in turn is a principal reason for the potentially excess labour supply. But it is difficult for governments to stand in the way of technological progress, and very likely it would do more harm than good. Consequently, an increase in the number of jobs relative to output must come by reducing the cost of employment.

Relative Wage Rates and Taxation

One obvious way to increase employment relative to output would be to restrain the increase in wage rates relative to output growth. This wage restraint would have its greatest impact on labour demand if the wage rates for unskilled labour fell relative to other wage rates so that the labour that is in excess supply became relatively cheaper. The counter-argument, however, is that the price elasticity of demand for this unskilled labour is low--possibly substantially less than one. In that case, the number of extra jobs generated would be relatively small, while the distribution or earnings would then become more unequal. The USA provides an example of this type of policy approach where there has been considerable success in generating jobs, including unskilled jobs, but the distribution of earnings is more unequal than in any other developed OECD country.

In addition, in Australia the level of income support limits the scope for reducing low wage rates in practice. The basic pension rate for a couple already represents 88 per cent of the minimum wage rate, and the costs of working would absorb much of the presently remaining gap. Consequently, a proposal some years ago by a group of 'five economists' to 'freeze' the living wage for four years also involved the introduction of an earnings tax credit so that take-home pay for a family with children would be maintained in real terms (Dawkins and Keating 2003). These five economists estimated that this freeze in minimum wages could translate into average real wages being about 2.5 to 3 per cent less than they otherwise would have been. This wage restraint was then estimated to increase employment by 2.5 to 3 per cent.

Economic modelling suggested that this strategy based on a wage-tax trade-off would pay for itself over about a decade or even less, because of the extra jobs and taxation revenue generated. But initially this policy initiative would have impacted negatively on the government's budget and possibly have required an income transfer from those who are employed to those who want more work. Overall the modelling suggested that there would be an increase in the disposable incomes accruing to households in deciles 2 to 7 of the income distribution and a decline in the share of the upper income deciles.

Whatever view is taken of using a wage-tax trade-off to increase employment, clearly it would be a much more ambitious task to increase the number of hours worked by as much as 11 per cent, compared with the 2.5 to 3 per cent envisaged by the 'five economists'. In particular, the initial fiscal impact would be more negative, and/or it would last over a longer period.

Education and Training

An alternative way of reducing the cost of labour is to make it more productive relative to its price. A classic way of achieving this increase in labour productivity has been through education and training. Moreover in an economy like Australia's, that is short of skilled labour, increasing the supply of skills in this way can allow both employment and output to grow faster.

To some extent the labour market can be expected to adjust to accommodate the projected shortage of skills. The school retention rate has risen substantially in the last twenty or so years, which will lead to older workers becoming more highly educated, and having better employment prospects in the future. The ageing of the population will also tend to increase labour scarcity and that will encourage employers to improve their training efforts. But projections suggest that these market responses to what is an impending increase in skill shortages will not on their own meet all the expected increase in the demand for skilled labour over the next forty years (Karmel and Woods 2004). Moreover the expected continuation of a high rate of technological change and the demands of a knowledge economy mean that members of the workforce will constantly need to refresh and upgrade their qualifications. So, for these various reasons governments will have to support an increase in education and training.

The advantages and disadvantages of different approaches to education and training will be discussed in other papers for this conference. Here it is noted that a wide spectrum of possibilities are available but, in terms of promoting both efficiency and social equity, perhaps the most pressing need for reform is to improve the opportunities for those people who left school early and who have low capabilities and low skills. Based on the earlier figuring in this paper, it is arguable that almost 20 per cent of the active population are unable to get as much work as they want and would benefit from some re-skilling. Furthermore, this number is consistent with separate estimates that around 20 per cent of the adult population continue to have very poor literacy skills (Australian Government 2004, p. 4).

The cost of developing the capacities and skills of those people who are least competitive in obtaining employment will be high, running into billions of dollars over a period of years. The size of the effort required can usefully be compared with the period immediately following the World War Two, when a large proportion of the then Australian workforce was being demobilised and there was a widespread fear that this would prompt a return of the unemployment experienced in the 1930s. At that time the Australian Government responded with a massive program of retraining and university education.

A critical issue will be who should pay? In the past most training for people without post-school qualifications has largely been provided by their employers, at the employer's cost. But this approach to financing training biases its nature in favour of very job-specific training; whereas in a world of constantly changing technology, people need more generic skills that enable them to change jobs and to engage in continuous learning. For these reasons, it would be better to rely less on employers for training low-skilled people in the future, with more of the costs being recovered from the immediate beneficiary or the government.

At present, students who complete their schooling and then go on to higher education continue to be heavily subsidised and can obtain a government loan on very favourable terms to finance their fees. By contrast, those students attending TAFE pay up-front fees with no loans, while early school leavers who do no further study pay taxes which subsidise those who continue to study. A more equitable arrangement would be for everybody to have a learning entitlement that could help finance their post-school education and training. Some will seek to use that entitlement immediately after they graduate from school as they continue their studies. Others will postpone further study after leaving school and will go directly into the work force, but they should be able to draw down on this learning entitlement later as they wish.

In addition, it is possible that this learning entitlement will prove to be insufficient to cover the total costs, and especially the living costs of people with families who return later in their lives for further education and training. Older people could perhaps then be permitted to draw down on their superannuation accounts to help finance their further education, possibly by way of a loan with a HECS type of system to ensure its subsequent repayment.

Labour Market Programs and Job Search Assistance

Labour market programs are in a sense an alternative form of training to improve the beneficiary's capabilities. Often the programs embody some form of education and training. Even when the programs are limited to a subsidy for employment the intention is that the work experience will improve the person's morale, work habits and capabilities. The main argument against this form of assistance has been concerns about the programs' cost effectiveness. But for some people who cannot find employment the programs are reasonably cost-effective (Stromback and Dockery 2000; DEWR 2003, pp. 2, 21-3).

Job search assistance can be a relatively low-cost way of improving people's skills and the flow of information. In particular, a recent pilot program of intensified assistance directed to people on the disability support pension (DSP) placed more than a third of the participants in employment. Of these new employees, 70 per cent had been receiving the DSP for more than two years, and over half had received the pension for over five years. Most DSP recipients want to work--90 per cent according to a recent survey--but over 80 per cent of the participants in the pilot program had not previously accessed the available assistance. The pilot program has gathered evidence of 'significant disincentives and widespread ignorance inhibiting DSP recipients' take-up of work opportunities' (DEWR 2004, p. 4). Major disincentives that discourage DSP recipients from pursuing employment opportunities appear to be the fear of losing the pension and/or associated concessions, and insufficient guarantees of a return to pension if a recipient leaves work for reasons other than their disability.

Removal of these disincentives seems unlikely to cost much, if anything. Indeed their removal might even result in budget savings, because experience suggests that otherwise it will be very difficult to reduce the number of disability pensioners, despite their preference for work. Thus a combination of increased assistance to improve the skills and restore the capabilities of DSP recipients, while still providing ready access to the social security safety net, could substantially reduce expenditure on the DSP. As a rough estimate, if much of the increase in the rate of take-up of the DSP by males aged 35 and over since the early 1980s was reversed, then it is estimated that it would be possible to employ another 122,000 men. Most of these men would only be available for part-time work, but they could be employed in additional jobs as they would have become sufficiently productive to justify their wage rates.

Improving Other Incentives for Employment

Government policy reforms have focused on the one hand on improving the incentives for people to pursue employment, and on the other hand the penalties for not pursuing employment. There has been understandable concern about the potential disincentives from the very high effective marginal tax rates facing low-income families with dependent children that were referred to earlier. This has been the main reason for a number of government initiatives in the last decade that have sought to reduce the magnitude of such disincentives. So far, however, the impact on employment by the people most affected has been very modest, and it seems likely that the lack of employment opportunities has been more important than a lack of incentives for most families.

Overall, it seems safe to conclude that further significant reductions in effective marginal tax rates would require radical reform of social security, possibly removing or absorbing some benefits or abolishing their means tests. Fundamentally government faces a difficult choice between:

* Tight targeting with high disincentives but lower budget costs, or

* Consolidating or removing benefits to remove overlapping means tests so as to improve employment incentives, with the cost depending on the specific proposals, (3) or

* Relaxing the present targeting and incurring even higher budget costs.

Each different strategy will impact on the distribution of government assistance while the extent of the impact on employment may be only modest. In these circumstances it is not surprising that the government has made so little progress in implementing some of the key recommendations by the McClure Report on welfare reform (FACS 2000). Moreover, it is quite likely that other policies that improve the demand for unskilled labour would be more cost-effective. This is because politically acceptable ways of reducing effective marginal tax rates (EMTRs) are likely to involve less targeting of assistance and will therefore be very expensive.

Reducing Incentives for Early Retirement

At present, people can retire as early as age 55 and access their superannuation contributions to support them until they are eligible for the age pension. In 1997 the government legislated to increase the preservation age for superannuation from 55 to 60 by 2024. But that change is still some way off, and to some extent the option of early retirement will still remain.

The incentive for early retirement would be greatly reduced if future superannuation savings could only attract their present generous tax treatment if all or most of each person's entitlement had to be drawn down as an income stream, actuarily determined to continue over the remainder of the person's expected life. This would

probably allow many people to partially retire early, but most early retirees would probably find that they needed to continue part-time work in order to maintain a satisfactory living standard. Again there would be no cost to the Budget and probably savings over time.

Conclusion

The critical problem facing Australia is that there is a structural mismatch between the labour supply and the demand for labour. (4) There is a shortage of skilled labour and a substantial excess supply of people with low education and skill levels. Potentially the number of hours worked in a fully-employed economy could be expanded by 10-11 per cent. But any attempt to expand aggregate demand faster without remedying the present structural imbalance in the labour market is doomed to failure.

Equally, attempts by government to enforce job-sharing are unlikely to work. To a significant extent the expansion of part-time work has permitted the extension of job-sharing on a voluntary basis. But those people who want more work for the most part do not have the skills necessary to share the jobs where labour is in short supply. Moreover, there is little evidence that people want to share their jobs by reducing their working hours and taking home less pay. Instead the people with less education and skills often want to work more hours rather than less so as to increase their incomes.

The best way forward is to improve the match between the supply of labour and labour demand. In principle the present imbalance could be removed by altering relative wage rates, but at a risk of greater inequality unless there is more fiscal support for low-wage households through a wage-tax trade-off. The priority for future policy is therefore to focus on improving and extending education and training policies. The cost will be substantial and will need to be shared between the government and the individuals concerned. Results will also take time, but in the longer run more education and training offers the possibility of a more egalitarian society by improving the productivity of low-wage people and thus levelling up rather than down. Moreover, an approach to income redistribution along these lines will increase the total amount of income. It is therefore less likely to take money away from people and will probably prove most acceptable to society as a whole.

Data Appendix

Table 2

Data for employment by occupation was obtained from various issues of the ABS Labour Force, Australia, Catalogue number 6203.0. There have, however, been numerous changes to the ABS occupational classification over the years. For the period from 1989 to 2000 the data were adjusted to ensure consistency as described in Keating (2003). For the earlier periods, 1971 to 1986 and 1987 to 1989, the changes in the occupational classification mean that it is not possible to link one period with the next. The employment in the different occupations in each of the sub-periods was however allocated to the broad occupational groupings--managers and professionals, clerks and services, and blue-collar workers (shown in Table 2 by the author).

The series for each of these broad occupational groupings is therefore not strictly comparable between the different sub-periods. Nevertheless it is considered to portray the major changes in the occupational structure of male employment over the three decades shown.

Table 3

The number of people already employed who want to work more hours and the average number of extra hours that they wished to work was obtained from The Year Book of Australia 2004, Labour Feature Article--Labour Underutilisation (ABS 6105.0) and Underemployed Workers (ABS 6265.0). Both these sources reported for September 2003. Similar data were also available in ABS 6105.0 for unemployed persons, but the number of unemployed persons who could be employed at any one time was adjusted to allow for some residual frictional unemployment in a fully-employed economy equivalent to an unemployment rate of 1.5 per cent.

In addition, the ABS considers persons to be marginally attached to the labour force:

* If they want to work and are actively looking for work but are not available to start work in the reference week, or

* If they want to work and are not actively looking for work but are available to start work within four weeks.

The inclusion of these marginally attached persons in a measure of potential labour supply is consistent with the finding by Gray, Heath and Hunter (2002, p. 32) 'that there are few differences between the dynamic behaviour of the marginally attached and the unemployed. ... The marginally attached have a similar attachment to the labour force as the unemployed, which is an argument for including both groups in assessing effective labour supply'.

The numbers of such marginally attached males and females were obtained from 'Persons not in the labour force' (ABS 6220.0) for September 2003. The number of these marginally attached persons who would actually be able to undertake employment was then obtained after a downwards adjustment to allow for the probability that not all of these marginally attached people would in fact be able to find employment at any one time. In this case, the size of the adjustment was greater than the analogous adjustment made to allow for continuing frictional unemployment among those people who are presently recorded as unemployed. This adjustment for continuing frictional unemployment among marginally attached people was also greater for females than for males, with the greatest adjustment being for females with children.

In addition, the number of marginally attached males was increased to allow for an estimate of the number of men who could potentially become attached and find employment if policy changes were successful in assisting disabled pensioners back into employment and in reducing the extent of early retirement. This estimate assumed first, that over time the proportion of males aged 35-54 and 55-64 on DSP could potentially be reduced by 40 per cent and one-third respectively, which would still leave that proportion somewhat higher than it was 20 years ago. (5) The number of these male recipients of DSP in each age group was obtained directly from the Department of Family and Community Services. Second, it was assumed that the proportion of males aged 35-54 and 55-64 who retire completely prior to age 65 could potentially be reduced by 45 per cent. These two potential additions to the male workforce would increase male workforce participation by males aged 35-54 by 4.4 percentage points, and by 27 percentage points for males aged 55-69.

However, some of these males recipients of DSP and early retirees would probably have been counted among the number reported in the ABS survey as marginally attached to the workforce. The total number of marginally attached males who are retired and who have an injury or an illness is published in ABS 6220.0. Using these figures rough estimates were then made of how many of these men were aged 35-44 and 55-64 and how many might have been double counted. A downwards adjustment was then made using these estimates in order to derive the total of marginally and potentially attached males shown in Table 3. In sum, these estimated net additions to the potential male workforce would still leave male participation for both age groups significantly below where it was at the beginning of the 1970s.

The average number of hours worked by these marginally attached and potential additions to the workforce was based on a weighted average of the hours worked by full- and part-time workers, where the weights reflected the proportions of marginally attached male and female persons wanting full- and part-time work. These estimates were then further adjusted to reflect the age and marital status of particular groups of marginally attached workers.

References

Australian Bureau of Statistics (2001), Census of Population and Housing: Selected Social and Housing Characteristics, Australia, cat. no. 2015.0, ABS.

Australian Bureau of Statistics (2004), Year Book of Australia, cat. no. 6105.0, ABS.

Australian Bureau of Statistics (various years), Labour Force Australia, cat. no. 6202.0 and cat. no. 6203.0, ABS.

Australian Bureau of Statistics (2003), Persons Not in the Labour Force, Australia, cat. no. 6220.0, ABS.

Australian Bureau of Statistics (2003), Job Search Experience, cat. no. 6222.0, ABS.

Australian Bureau of Statistics (2003), Underemployed Workers, Australia, cat. no. 6265.0, ABS.

Australian Bureau of Statistics (2002, 2003), Australian Labour Market Statistics, cat. no. 6105.0, ABS.

Australian Government (2003), Budget Paper No. 1 Fiscal and Economic Outlook, Commonwealth of Australia, Canberra.

Australian Government (2004), Australia's Demographic Challenges, Commonwealth of Australia, Canberra.

Argy, F. (2005), 'An analysis of joblessness in Australia', Economic Papers, vol. 24, no. 1, pp. 75-96.

Centre for Public Policy (2005), Transitions and Risk: New Directions in Social Policy. Available at: www.public-policy.unimelb.edu.au/Conferences2005, accessed February 2005.

Dawkins, P. and Keating, M. (2003), 'The five economists' plan', summarised in Dawkins, P. and Kelly, P. (eds), Hard Heads, Soft Hearts: A New Reform Agenda for Australia, Allen and Unwin, Sydney.

Department of Employment and Workplace Relations (DEWR) (2003), Good Jobs or Bad Jobs: An Australian Policy Perspective, Australian Government, Canberra.

Department of Employment and Workplace Relations (DEWR) (2004), 'JobAble. Job Network Disability Support Pension Pilot: Interim Evaluation Report', Available at: http://www.JobAble.gov.au/home/DSP_Pilot_Interim_Evaluation. asp.

Department of Family and Community Services (FACS) (July 2000), Participation Support for a More Equitable Society--Final Report of the Reference Group on Welfare Reform, Department of Family and Community Services, Canberra.

Gray, M., Heath, A. and Hunter, B. (2002), 'An exploration of marginal attachment to the Australian labour market' Research Discussion Paper, Research discussion paper, RDP 2002-07, Reserve Bank of Australia, Canberra.

Karmel, T. and Woods, D. (2004), Lifelong learning and older workers, National Centre for Vocational Education Research, Adelaide.

Keating, M. (2000), 'Realising people's potential: labour markets', in Davis, G. and Keating, M. (eds), The Future of Governance: Policy Choices, Allen and Unwin, Sydney.

Keating, M. (2003), 'The labour market and inequality', Australian Economic Review, vol. 36, no. 4, pp. 374-96.

Keating, M. (2004), Who Rules: How government retains control of a privatised economy, The Federation Press, Sydney.

Murtough, G., Pearson, K. and Wreford, P. (1998), Trade Liberalisation and Earnings Distribution in Australia, Industry Commission Staff Research Paper, AGPS, Canberra.

Schmid, G. (1998), 'Transitional labour markets: A new European employment strategy', Discussion Paper, FS I 98-206, Wissenschaftszentrum Berlin fur Sozialforschung.

Stromback, T. and Dockery, A. (2000), Labour Market Programs, Unemployment and Employment Hazards, ABS Occasional Paper, Catalogue number 6293.0.00.002, ABS, Canberra.

Toohey, M. (2005), 'The effectiveness of child care benefit at improving returns to work for women, with children', 9th IFS Conference, 9-11 February.

Ziguras, S., Considine, M., Hancock, L. and Howe, B. (2004), Social Policy for the 21st Century: The Role of Transitional Labour Markets. Background Paper, March, Centre for Public Policy, Melbourne.

Endnotes

(1) Argy (2005) has suggested that labour underutilisation in Australia is equivalent to about 9 per cent of the total number of hours potentially available. From the limited information provided by Argy, the small difference between his and the present estimate of 11 per cent seems to reflect: 1. T[begin strikethrough]t[end strikethrough]he different denominators used, with Argy using the workforce as his denominator whereas this paper uses employment as the denominator; 2. T[begin strikethrough]t[end strikethrough]he more detailed estimates in the present paper of the hours that underemployed people would be willing to work; and 3. A[begin strikethrough]a[end strikethrough]n allowance in the present paper for the effect of possible policy changes on the number of people seeking work.

(2) In 2002 half of unemployed 55-64-year--olds had been unemployed for more than 12 months (ABS Catalogue number 6105.0).

(3) The Labor Party's Tax and Family Policy at the last election incorporated a proposal along these lines.

(4) Argy (2005) similarly finds that this structural mismatch is the overwhelming reason for unemployment and underemployment in Australia today. [??check formatting of footnote]

(5) As heavy manual labour is less common than it was 20 years ago, and will become even less common in the future, it could be argued that there is even more scope to reduce the rate of take-up of the DSP.

Michael Keating, Australian National University 

Table 1: Workforce Participation (%) 1966 to 2003

            1966     2003

Male         84       72
Female       40       55
Persons      60       64

Source: Australian Government (2003, pp. 4-8)

Table 2: Changing Structure of Male Employment (%) 1971 to
2000

                                             Managers and    Clerks and
                                             Professionals    Services

1971 share of employment                         18.0           19.0
1971-1986 average annual employment change
p/a                                               2.6            1.4
1987 share of employment                         32.0           16.0
1987-1989 average annual employment change
p/a                                               3.8            1.8
1989 share of employment                         34.0           16.0
1989-2000 average annual employment change
p/a                                               1.9            0
2000 share of employment                         39.0           15.0

                                              Blue-Collar
                                                 Group

1971 share of employment                         63.0
1971-1986 average annual employment change
p/a                                               0
1987 share of employment                         52.0
1987-1989 average annual employment change
p/a                                               3.2
1989 share of employment                         50.0
1989-2000 average annual employment change
p/a                                              -0.5
2000 share of employment                         46.0

Source: See Data Appendix.

Table 3: Potential Additions to Employment, September 2003

                                                    Number
                                                 wanting/able
                                                 to undertake
                                   Number       more employment
Employment Group                   (000)             (000)

Males already employed             5,313.7            230.5

Females already employed           4,256.9            336.8

Unemployed males                     321.0            240.1

Unemployed females                   275.2            206.4

Marginally attached females          572.0            317.6

Marginally & and potentially         449.0            413.2
attached males

Total                             11,187.8          1,744.6

                                  Average
                                 number of      Potential total
                                 additional      of additional
Employment Group                   hours          hours (000)

Males already employed              17.1             3,942

Females already employed            14.3             4,816

Unemployed males                    34.0             8,163

Unemployed females                  28.0             5,779

Marginally attached females         20.0             6,380

Marginally & and potentially        21.0             8,429
attached males

Total                               21.5            37,509

Source: See Data Appendix.