Using official data from the Australian Bureau of Economic Statistics and a formal growth accounting framework, this paper shows that the rapid accumulation of information processing and communication technology (ICT) capital over the last two decades in Australia has played a significant role in explaining the impressive, structural acceleration of labor productivity. The following statistical data are also included: household income, expenditure and savings, labor market, fiscal indicators, credit aggregates, capital and financial account, external assets and liabilities, export by commodity group, and so on.


Using official data from the Australian Bureau of Economic Statistics and a formal growth accounting framework, this paper shows that the rapid accumulation of information processing and communication technology (ICT) capital over the last two decades in Australia has played a significant role in explaining the impressive, structural acceleration of labor productivity. The following statistical data are also included: household income, expenditure and savings, labor market, fiscal indicators, credit aggregates, capital and financial account, external assets and liabilities, export by commodity group, and so on.

IV. Welfare Reform—The State of Play and Challenges Ahead1

A. Introduction

1. In December 2000, the Australian Government made its initial response to the Final Report of the Reference Group on Welfare Reform (the “McClure Report”),2 thereby culminating a two year process that bas brought welfare reform near the top of the domestic policy agenda. The response signaled a substantial additional budgetary investment in enhancing and complementing Australia’s social safety net, noting that welfare reform would be among the highest funding priorities in the 2001/02 Budget. Given the support of the major recommendations of the McClure report across a broad spectrum of civil society, Australia appears set to embark on a comprehensive, long-term reform of its welfare system. In doing so, it would join a growing number of OECD countries including the US, UK and Canada—although (he approach in Australia looks likely to remain distinct.

2. The increased interest in welfare reform in Australia stems in large part from the costs—both social and economic—of the persistence of welfare dependency. Key indicators of welfare use (other than those relating to the receipt of unemployment benefits) have showed little or no improvement in 1990s in the face of a prolonged and vigorous economic expansion. In particular: (i) the proportion of the workforce age group receiving some form of income support has risen sharply since the 1960s (much more so than the unemployment rate) while the proportion of jobless of workforce age has actually fallen; (ii) one seventh of workforce age population receives at least 90 percent of its income from government transfers; (iii) public spending on income support has continued to rise as a percentage of GDP; (iv) jobless families are becoming an increasingly prevalent characteristic of the economic landscape (as are two-earner families); and (v) research has shown that that joblessness and welfare dependency of adults can adversely impact the life chances of children, suggesting that social and economic disadvantage could persist across generations.

3. The main drivers of these outcomes are economic and demographic factors, as well as government policies. The most important economic change over the past 30 years has been the increase in the unemployment rate, notwithstanding its steady decline in the current business cycle. The increase in part-time and casual unemployment has also contributed, albeit to a lesser extent. The demographic factors with the most impact have been the increase in lone parenthood, the decline in fertility rates and the aging of the baby boom cohort. Changes in government policy—including expanding the coverage and types of income support payments as well those measures aimed at improving targeting through tightening the income and asset tests—have led to an increase in the complexity of the system and have exacerbated existing disincentives for labor force participation.

4. This paper provides an overview of the current state of play regarding welfare reform in Australia and points to key issues and challenges in the period ahead. It is not meant to be a comprehensive study, but rather aims to highlight the main issues as they relate to the economic—rather than the social—dimensions of welfare reform. From this perspective, the key problem is how to provide incentives for and policies which support labor force participation while continuing to offer an adequate safety net and containing budgetary costs—subject to the social and political constraints in the Australian context. Solving this problem is likely to require a gradual, iterative process, drawing on observation of what actually works in practice.

5. The remainder of the paper is structured as follows. Section B looks at the stylized facts of the Australian welfare system and trends in the labor market and welfare use. Section C provides a cross-country (OECD) view of trends in social spending and welfare reform efforts. Section D looks at the need for reform in the Australian welfare system and the key recommendations of the McClure report. Section E presents the Government’s initial response to the McClure Report. Section F looks at the expected macroeconomic effects of welfare reform, and Section G concludes with the challenges ahead.

B. Stylized Facts of the Australian Welfare System3

The Nuts and Bolts of the System

6. Australia’s welfare system (called the “social security” system) is a national system funded out of general revenue. It provides income support payments to over 5 million adults (defined as age 15 and over) out of a total population of 15 million. The main objective of the system is poverty relief rather than income maintenance—welfare payments are designed to provide an adequate income to eligible people. Payments are based on a person’s family situation, income and other factors related to need—not on previous earnings. Also, payments are made subject to the claimant satisfying the relevant eligibility requirements, but are not time-limited. Importantly, the system is highly targeted, and features both income and asset tests.

7. There are three broad classes of income support payments, which comprise (in descending order of generosity): pensions; allowances; and special benefits. Details are provided in Boxes IV.1 and IV.2. In theory, pensions are intended to cover situations requiring long-term income support such as old age, severe disability or widowhood. Allowances, on the other hand, are intended for people with short-term needs such as the unemployed and the sick. Pension payments therefore give more weight to long-term income maintenance while allowances are structured on the expectation that the individual will return to work. As a whole, the system provides: (i) non-activity tested-entitlements for various groups to subsist outside the labor force (most on pensions, although some are on allowances); (ii) allowances for the unemployed and for students; and (iii) a safety net for others in need.

Summary of Eligibility Criteria and Activity Requirements, Main Workforce-Age Income Support Payments

article image

Summary of Payment Rates and Income Tests, Main Income Support Payments for People of Workforce Age

article image

8. Extensive use of means testing in Australia has resulted in a welfare system that costs less and is more redistributive relative to other industrial countries. The consensus is that the system has been very effective in eliminating poverty and providing a strong safety net.4 Income tests, which are applied on a fortnightly basis, involve:

  • a free area, or the amount of money a person can earn without any reduction in welfare payments received, and

  • one or more taper rates, or the proportion of welfare payments withdrawn for each additional dollar of private income earned above the free area.

Free areas and taper rates serve to encourage welfare recipients to earn private income, for example through part time or casual work. In general, pensions have more generous income tests than allowances since recipients are presumed to require longer-term income support The same logic applies to asset tests. In general, pensions have tapered asset tests, while allowances have “sudden death” cut-out levels, above which no income support is available.

9. A basic element of the Australian approach to welfare is the notion of “mutual obligation.”5 This notion can be interpreted as a view that a contribution from all sectors of society—government, business, communities and individuals—is required for a well-functioning welfare system (or, conversely, that requiring action from only part of the broader society will lead to sub-optimal results). Government and the community have a “mutual obligation” to provide income support to eligible people as well as to provide opportunities for them to participate and improve their situation. Individuals are expected to take up such opportunities—for example, people receiving unemployment benefits are required to actively look for work and to undertake activities such as training that will improve their employability. Business has the responsibility to generate wealth and employment, including for those on income support.

Trends in the Labor Market

10. The most prominent trend in the Australian labor market over the past three decades has been the rise in the unemployment rate. While unemployment as a percentage of the workforce age population was steady at around 1 percent from the mid-1960s to the early 1970s, it rose sharply thereafter, and averaged 6½ percent in the 1990s. (By way of comparison, the unemployment rate measured in terms of the labor force averaged 8⅔ percent in the 1990s.)

11. Accompanying the increase in unemployment has been an increase in its duration and a change in its distribution across families. The mean duration of unemployment rose steadily from less than ten weeks in the mid-1970s to over 50 weeks by the early 1990s, where it has remained since. When the period between employment is relatively short, there is a lower likelihood that agents will need to resort to the income support system and job-related skills will remain fresh. On the other hand, longer spells of unemployment both increase the likelihood of using the social safety net and decrease connectedness with the workforce, thereby increasing the risk of persistent joblessness. Importantly, Australian families have also become increasingly divided between two-job households and zero-job households, which has been a strong contributor to the rise in welfare usage. Over the 15 years ending in 1997, the number of families with one working adult fell, while those with two or zero working adults rose.

12. Another important development over the past few decades has been the increase in labor force participation of women and, relatedly, the growth of part-time employment. While male participation fell from 91 percent in the mid-1960s to 81 percent in the late 1990s, female participation rose from 41 percent to 64 percent Reflecting the higher incidence of part-time work among females, the proportion of part-time workers rose from under 10 percent in the mid-1960s to over one-quarter in the late 1990s, It should be noted that to the extent that part-time employment reflects worker preferences, and to the extent that the interaction between part-time employment with the welfare system is non-distortionary, the rise in part-time employment is not in itself a negative outcome.


Mean Duration of Unemployment

Citation: IMF Staff Country Reports 2001, 055; 10.5089/9781451802030.002.A004


Change in the Number Workforce-Age Income Unite by Number of Employed Adults, 1962 to 1996-97

Citation: IMF Staff Country Reports 2001, 055; 10.5089/9781451802030.002.A004

13. Australia also has a significant number of underemployed workers and people marginally attached to the labor force. The underemployed are defined as those part-time workers who would prefer to work longer hours plus full-time employed persons who work less than 35 hours per week due to economic reasons (e.g., if there was insufficient work to allow them to work more hours). The vast majority of underemployed are in the former category. Underemployment is higher for men than for women. Between the 1985 and late 2000, the proportion of men in this category rose from 23 percent to 36 percent, while women rose from 15 percent to 21 percent. The marginally attached are those who would like a job but are either not available for work or not actively seeking work. In September 1999 there were 883,000 marginally attached, and the number of such people is currently at its lowest level since late 1996. The two main groups with marginal attachment are women aged 25-44 engaged in home duties, and students. Over 40 percent of this group received government cash payments as their main source of income.

14. Finally, the Australian labor market is characterized by a high rate of turnover relative to the rest of the OECD. The occupations with the highest turnover—blue collar and lower-skilled white collar positions—are the types most likely to claim welfare benefits. As will be argued below, the high rate of turnover may play a valuable role in absorbing an increase in labor supply resulting from welfare reform.

Trends in Welfare use

15. Reflecting in part the labor market trends noted above, the proportion of the workforce age population receiving income support has risen sharply over the past three decades from under 4 percent in 1965 to about 22½ percent in 1998—a much higher jump than in the proportion unemployed. While it remains the case that not all jobless people receive income support payments, it is quite clear that the increase in the proportion of the workforce on income support has risen sharply despite a fall in the overall proportion of joblessness. This has been particularly true in the 1990s, where use of income support as a proportion of the workforce age population has not changed appreciably since the end of the 1990-91 recession, while unemployment has dropped steadily.


Jobless Workforce-Age Population (1985-98)

Citation: IMF Staff Country Reports 2001, 055; 10.5089/9781451802030.002.A004

16. In addition to the rise in unemployment and its duration, several reasons explain the increase in welfare use. First, the proportion of couples with no paid work has increased, mostly due to rising joblessness among men. Second, there have been sustained increases in the rate of lone parenthood—these families are much more likely to be reliant on income support Third, the number of people combining part-time work and income support has grown sharply. Finally, the aging of the population has contributed to the rise in income support usage reflecting that this cohort has a higher propensity to use the welfare system (see below). Also relevant are the individualization of benefit eligibility in 19956 and a number of policy changes aimed at promoting part-time and casual work, the income from which are often complemented by support from the welfare system.

17. Far from being a homogeneous group (which was more the case in the past), the welfare system covers a wide variety of workforce-age people. In 1965, over 40 percent of income support recipients were classified as disabled or sick, more than twice as large as any other group. By 1998, however, the distribution was more balanced with one-third classified as unemployed, 20 percent disabled or sick, 17 percent as partner or child carers, and 14 percent each for lone parents and students.

18. With the rise in the proportion of the workforce claiming income support payments, the fiscal costs have risen sharply. According to Australian Bureau of Statistics data, Commonwealth expenditure on social security (excluding old age payments) was about 2½ percent of GDP from the mid-1960s before jumping to around 4 percent of GDP in the late 1970s, jumping again to 5 percent of GDP in the mid-1980, and again to 6 percent of GDP in the early 1990s, where it stands at present7


Commonwealth Spending on Income Support (as a percentage of GDP)

Citation: IMF Staff Country Reports 2001, 055; 10.5089/9781451802030.002.A004

19. Income units with no adult in paid employment are much more likely to be on income support than those with working adults. Moreover, the uneven distribution of jobs across families noted above also has intergenerational effects. Children from families receiving income support are two and one-half times more likely to end up on income support themselves than children from families that do not receive income support. Moreover, in Australia joblessness affects a larger proportion of families with children than in most other industrial countries, which results in greater reliance on government support for this group compared to singles and couples without children (Oxley, et al., 1999).

20. Reflecting the trends outlined above, welfare dependency—defined as receiving over 90 percent of income from government cash payments—continues to increase. In 1986, some 12 percent of income units were welfare dependent, rising to over 14 percent in 1996-97. The incidence of dependency is much higher among single persons than couples. Dependency is notably high for lone parents, at over 40 percent Also, couples with children are less likely to be dependent on welfare than couples without children—the reverse is true for single parents. Dependency was also high among the elderly reflecting an earlier, more general expectation of relying on old-age pension in retirement It should be noted that longterm reliance on income support is entirely appropriate for certain segments of the workforce age population, including people with serious disabilities that prevent them from taking paid employment and people who provide care for the disabled and elderly. The concern in Australia is that a large and growing proportion of the welfare dependent population is made up of potentially employable adults who would normally be expected to be engaged to some extent in the labor market.

Wockforce-age Income Units with at Least 90 Percent of Income from Government Cash Payments, 1986 and 1996-97

article image

C. Cross-Country Comparisons

21. This section takes a brief look at social spending across the OECD and the emerging consensus in the OECD of “what works” as regards labor market and social policies.8 Table IV.1 shows data for gross cash social expenditure—defined as labor market plus non-health, non-retirement public expenditure—across a number of OECD countries in 1995.9 Australia is ranked toward the low end of the spectrum—although it is in the cluster of “Anglo-countries” with which it is often grouped. This result holds up even if one excludes unemployment benefits, which would tend to bias the numbers up or down depending on the relative position of countries in the business cycle. As noted above, Australia’s tightly targeted income support system has been able to achieve its objectives in a relatively efficient manner as measured by total spending.

Table IV.1.

Gross Total Cash Social Expenditures in Selected OECD Countries, 1995 1/

(As a percentage of GDP)

article image
Source: Adela (1999).

Total social expenditure comprises cash benefits (shown here) plus health and social services.

22. In recent years OECD countries have shown a greater willingness to innovate in the area of social policy, and a consensus on a “New Social Policy Agenda” has begun to emerge (OECD, 1999). In short, the consensus suggests that there are three pillars in an effective social assistance system: cash assistance; social services and labor market services. Importantly, these three pillars are seen as interrelated—experience shows that progress in welfare reform cannot be made by advancing on only one front. Other factors in the new consensus are that: (i) while there is no trend toward either increasing or reducing the level of cash assistance, there is strong movement toward making benefits dependent on labor market participation including for, after an increasingly short period, lone parents; and (ii) sufficient attention needs to be paid to the fixed costs of labor market reinsertion.

23. A key focus in recent social policy discussions has been on “making work pay.” In general, this involves avoiding unemployment traps (where the benefits from unemployment and related benefits approach or exceed post-tax income from work) and poverty traps (where rapid withdrawal of benefits leads to effective marginal tax rates, EMTRs, approaching or exceeding 100 percent). These traps can be avoided through lowering benefits on the one hand and/or lowering taper rates or providing other types of in-work benefits on the other hand.10 Also included in this class are participation benefits and other schemes that seek to address the issue of the fixed costs of participation.

24. One of the better known and most often analyzed examples of in-work benefits is the Earned Income Tax Credit (EITC) in the US (see, e.g., Disney, 2000 and OECD, 1999). The principle behind the EITC is to lower the EMTR for a person taking up work by providing for an employment contingent tax credit. The trade-off for introducing incentives along the lines of the EITC is that the benefit net is widened to include higher income workers, thus increasing the fiscal cost of the system. For example, the EITC costs ⅓ percent of GDP per year and the UK’s Working Family Tax Credit costs about ⅔ percent of GDP year while there is no presumption that the costs would be different in Australia, it should be noted that the existence of taper rates and free areas in the benefit structure provides the tools to replicate the incentive effects of the EITC without introducing any new program into what is already arguably a fairly complex system.

25. Another, related, area of recent attention by the OECD has been to distinguish active versus passive labor market measures. Active labor market policies and measures are those aimed at improving the access of the unemployed to the labor market and jobs, job-related skills and improving the functioning of the labor market Passive policies and measures would include unemployment and related social benefits, and early retirement benefits. The consensus above suggests that active measures would be more effective in achieving social policy goals.

26. Overall, Australia’s welfare system is distinguished from those of other OECD countries in a number of ways. As noted above, the income support system in Australia aims at providing an adequate level of income and benefits are not time limited. Also, participation requirements are relatively weak for many of those on income support programs other than unemployment. However, while these benefits are not work contingent, they do serve—through the taper rates and free areas—to subsidize part-time and low paid full time employment, which are seen as important objectives in Australia. While the means tests have led to a highly targeted and cost effective system, they have resulted in high EMTRs relative to the OECD with the attendant work disincentives.

D. The McClure Report

27. Reflecting concern at the increase in welfare dependency as well as the other trends noted above, in September 1999 the Government announced its intention to review the welfare system, and appointed a Reference Group—chaired by Mr. Patrick McClure of Mission Australia (a country-wide community service organization)—to consult with the community and provide advice to the Government. The Group was charged with developing options for changing the income support system and the provision of associated services with the overall aim of preventing and reducing welfare dependency. The Group issued an Interim Report in March 2000 and, after receiving feedback from a broad spectrum of civil society, issued a Final Report in July 2000.

28. The McClure Report noted that the social support system had not responded adequately to the social, demographic and structural changes in the economy. Importantly, the ongoing long expansion (now in its tenth year) has not altered the negative trends outlined above—with a notable exception of unemployment benefits—implying that something more than strong economic growth is required.

29. The McClure Report recognized that the social support system in Australia has its origins in a different environment, one characterized by low, short-term unemployment and the “nuclear family”; i.e., one where income support was more of an exceptional circumstance. As such, the system may be failing those individuals it was designed to help. Four main shortcomings of the current social support system were identified: (i) service delivery is fragmented and not adequately focused on participation; (ii) there is an overly complex and rigid categorical array of benefits; (iii) there are inadequate incentives/rewards for some forms of participation/work, reflecting in particular high EMTRs; and (iv) the system does not provide enough recognition of participation.

30. As already noted, the flipside of Australia’s highly targeted and income tested (and, thus, relatively cost effective) welfare system is high EMTRs that arise when the tax and transfer systems interact Specifically, the high EMTRs arise because of the impact of the taper rates in the transfer system combined with the marginal income tax rates in the tax system. When several benefits are withdrawn over the same income range, the EMTR can exceed 100 percent, implying mat after tax income actually falls as hours worked increase. EMTRs exceeding 80 percent are not uncommon, with obvious effects on workforce participation. The text charts (which are taken from the Interim Report) illustrate the incidence of EMTR for two hypothetical income support recipients: a single person on disability support, and a woman with two children whose husband earns $A 500 per week. High EMTRs can also occur as a consequence of the “bunching” of part-time income (e.g., for seasonal or other intermittent workers), resulting in a diminution of benefits, which are assessed on fortnightly earnings rather than earnings over a longer period. This is the case where the average fortnightly wage would imply a certain level of benefits, but the actual benefits received are much less owing to the existence of free areas and increasing taper rates in the periods when work is actually undertaken.


Single Disability Support Pensioner

Citation: IMF Staff Country Reports 2001, 055; 10.5089/9781451802030.002.A004


Couple: One Child Aged 15, One Child Aged 17 Husband Earns $500 Per Week

Citation: IMF Staff Country Reports 2001, 055; 10.5089/9781451802030.002.A004

31. Some progress in improving work incentives was made in conjunction with the introduction of the New Tax System in July 2000. In addition to lowering marginal income tax rates, family assistance was simplified and the related income and asset tests were eased. Also, the taper rate for the new Family Tax Benefit was lowered relative to the arrangements it replaced. The new benefits also increased the assistance for child care, thereby lowering the costs of participation. Overall, EMTRs were lowered for a range of incomes, incentives were improved for low-income families to participate in the labor force, and benefits were increased to families, making them better off when moving to full-time employment.

32. It has also been recognized that the sheer complexity of the income support system may lead to uncertainties about benefits and the associated work incentives—so-called “model uncertainty.” This complexity stems in large part from policy changes over the past few decades, where the scope of me welfare system was broadened reflecting an increase in the number of objectives and programs as the social environment on which the system was based changed (e.g., with the rise in the number of lone parents). Available research shows that the perceived return to working is actually less than the actual return. The effect of this complexity—like that of high EMTRs—is to reduce participation in the labor force.

33. The “costs of participation” have also been identified by welfare recipients in the context of the work on the McClure report as a barrier to workforce entry. These relate to the use of formal childcare, transportation costs and start-up costs relating to work or study. Transport costs were particularly relevant for those in rural areas and those with physical disabilities.

34. The Report recommended the creation of a “Participation Support System” built on five, inter-related pillars: individualized service delivery; a simple and responsive income support structure; incentives and financial assistance; mutual obligation; and social partnerships—building community capacity. Below is a sample of proposals from each group. The complete list of recommendations is presented in Annex IV.1.

  • Individualized service delivery: Design systems for monitoring outcomes from interventions; develop a better means of assessing the capacity of people with disabilities to participate in employment and other activities.

  • A simple and responsive income support structure. Reconceptualize existing income support payments as “participation support payments;” integrate existing payments for mature age people into a single payment.

  • Incentives and financial assistance. Introduce a transition bank (see below); develop research bases for, e.g., analyzing optimal withdrawal rates and means testing structures.

  • Mutual obligation. Implement (phase in) participation models for parents with school aged children; implement a framework for mature aged jobless people which requires some form of (economic) participation.

  • Social partnerships—building community capacity. Further develop and expand community capacity building in rural and remote areas; develop strategies to foster the growth of micro-businesses.

35. The Report also took the view that the welfare reform process would take time, and divided its recommendations into “initial steps” and medium-to longer-term measures. The distinction was made in light of the need for evaluation of the initial steps and, in some cases, the need for more research to ensure that implementation is guided by good information and sound analysis. Some elements of the package may also require the build—up and maintenance of political momentum.

E. The Government’s Response

36. The Government set out its “initial steps” in response to the McClure Report in December 2000. It agreed with the thrust of the Report and its objectives, noting that the income support system had become too rigid and passive, and that it did not encourage participation. It also agreed that the issue of welfare reform is complex, and that the process will take several years to complete.

37. The Government does not see welfare reform as a cost-cutting exercise—on the contrary, it is expected that implementation will require a substantial upfront investment. Indeed, the objective is to prevent the exclusion of large segments of the population from economic and social participation. In the longer term, it is expected that higher participation rate and the corresponding reduction in the welfare rolls will increase output and lower the public cost for social services (as a percentage of GDP).

38. The objectives of welfare reform—all geared toward reaching the goal of minimizing social and economic exclusion—are threefold: (i) a significant reduction in the incidence of jobless families and jobless households; (ii) a significant reduction in the proportion of the working age population that needs to rely heavily on income support; and (iii) stronger communities mat generate more opportunities for social and economic participation. It is expected that the achievement of these objectives would be bolstered by policies designed to support economic and employment growth and to avoid recessions, which have tended to have long-lasting effects on unemployment and the reliance on income support.

39. As to specific recommendations of the McClure Report, the Government will make detailed announcements in the 2001/02 Budget. In the area of incentives the Government has indicated that it will move to develop a transition bank and improve publicity to existing incentives and financial assistance to help meet the costs of participation. The transition bank aims to give casual and intermittent part-time workers the same income support levels as those part-time workers with more uniform work schedules—see Annex IV.3 for details. As noted above, the fortnightly measurement period combined with the existence of free areas and taper rates per fortnight imply that casual part-time workers receive less income support than part-time workers with equal, but more regular, fortnightly income. It is the intention of the publicity program to address more fully the barriers to participation.

40. The Government is also disposed to increase gradually participation requirements for parents with school-aged children, including for sole parents. Australia is one of the few OECD countries that allow sole parents to receive income support until their youngest child turns age 16 without having to look for work. Societal attitudes toward this group are changing in favor of a participation requirement, although a part-time requirement polls much better than full-time requirement in available surveys (Eardley, 2000), In response, the Government has indicated a preference for a gradual approach: recipients with the youngest child 6-12 years will be required to attend an annual interview to discuss training and participation plans, while those with the youngest child 13-15 must develop a participation plan for part-time activity.

41. The Government has also committed itself to ongoing consultation with key stakeholders. An important subject for this process will be the modalities of implementing the more complex medium-term measures in the McClure Report. These will include research on the optimum withdrawal rate and means testing structure, the optimum approach to improving in-work benefits, and the appropriate balance between full-time and part-time work.

F. Macroeconomic Effects

42. The main macroeconomic effect of successful welfare reform will be an increase in labor supply resulting from improved work incentives. There are several channels for this effect on work incentives, each of which would be expected to have a different response rate (or elasticity). Thus, the scale and pace of these effects is, in the end, an empirical issue.

43. An obvious follow-on issue to a welfare-reform induced increase in the supply of labor is whether there would be a commensurate increase in the demand for labor. If not, men the new labor force participants would simply contribute to a rise in the unemployment rate. However, there are grounds for optimism (Dawkins, 2000).

  • The increase in participation should lead to a more efficient labor market (assuming that new participants have the requisite skill mix), implying a lower number of voluntary unemployed—and less wage pressure—for a given level of vacancies.

  • The high turnover in the Australian labor market suggests relatively favorable job opportunities for those entering the labor market (compared to a labor market with lower turnover). Although these positions may not be particularly secure or entail sufficient hours, such jobs would be an improvement for those who have been excluded from the labor market.

  • Any increase in employment from welfare reform leading to higher aggregate wages would, in turn, increase the demand for goods and services, thereby increasing the demand for labor.

  • The effects of recent and prospective reforms in the labor market would reinforce outcomes arising from welfare reforms. In particular, wage flexibility and decentralized bargaining would be complementary to an (appropriately skilled) increase in labor force participation.

44. A successful welfare reform effort would also be expected to provide more flexibility to macroeconomic policy making. Specifically, there would be more scope for expansionary policies without generating wage pressures. The benefits of labor market reforms to date in this area have been evident in the mature phase of the current expansion.

45. Another positive macroeconomic effect of successful welfare reform would be a reduction over the medium- and longer-term in the percentage of GDP devoted to income support payments. This would create scope for a re-allocation of future spending, which would seem particularly desirable in view of the projected effects on the public purse stemming from the aging of the population. A welfare reform induced increase in the participation rate could also serve to mitigate the effects on the budget of population aging.

46. Finally, the success of welfare reform should not be measured in terms of a reduction in the unemployment rate. As noted above, the channels through which labor demand could increase are complex, suggesting that the effect of welfare reform on the unemployment rate is ambiguous.11 In light of the arguments presented in this paper, better measures of success would be the employment to population ratio and the number of long-term jobless. Any drop in the unemployment rate could therefore be seen as a bonus from welfare reform.

G. Challenges Ahead

47. Overall, the main challenge ahead in implementing welfare reform will be to achieve an appropriate balance among the competing economic, social and political objectives and constraints while making sufficient progress in all of the three pillars identified in the recent work of the OECD to ensure that the reform effort as a whole is viable.

48. Relative to other OECD countries, Australia has placed a priority on targeting income assistance as a way to reconcile the importance of maintaining an adequate safety net while keeping a lid on costs. However, this has had a largely negative effect on work incentives. The economic objective of the welfare reform effort is to shift the balance toward improved work incentives. This effort must face two difficult questions: (i) at what speed to make the transition and (ii) what tools to employ to achieve the desired results. The social and political constraints placed on the process argue for a gradual transition and an incremental use of tools (e.g., improving in-work benefits building on what already is in place)—this is in fact the approach being taken.

49. How well will this strategy work? An approach that is too gradual runs the risk of a tepid labor supply response, little reduction in welfare dependency and a loss of political momentum for reform. An aggressive increase in in-work and related benefits may improve incentives for work, but at the price of an unacceptably large expansion in the fiscal costs. An overly strong push toward participation requirements could risk a political backlash. A balanced approach would seem to require the following:

  • Increasing participation requirements relatively aggressively for those who society sees as being able to work (e.g., the young, and parents of school aged children) and in a relatively cautious way for those who have strong constraints on labor force participation (e.g., the physically disabled).

  • An early shift in indexation from wages to prices for a sufficiently broad class of benefits. This would allow for the approximate preservation of benefits to wages over the short term, and a reduction of this ratio over time.

  • Continued research into the tax-transfer nexus and a willingness to innovate and experiment with measures to foster participation.12 Resulting policy initiatives should be evidence based.

  • Continued progress in labor market reforms needed to increase the efficiency and flexibility of the labor market and help absorb the increase in labor market participation stemming from welfare reform,

50. Finally, the continuation of sound macroeconomic policies and structural reforms must underpin any successful welfare reform effort Healthy growth in output and employment with an efficient structural base would improve incentives for labor force participation and provide for a relatively smooth transition to employment for those making the transition to work. It would also serve to bolster political support for increased short-term spending on welfare reform aimed at increasing the supply of labor.


  • Adema, Willem, 1999, Net Social Expenditure, OECD Labour Market and Social Policy - Occasional Papers No. 39 (Paris: Organization for Economic Co-operation and Development)

    • Search Google Scholar
    • Export Citation
  • Creedy, John, 1999, “Take-Up of Means-Tested Benefits with Labour Supply Variations,Melbourne Institute Working Paper No. 14/99 (May)

    • Search Google Scholar
    • Export Citation
  • Dawkins, Peter, 1998, “A Plan to Cut Unemployment in Australia: An Elaboration on the ‘Five Economists’ Letter to the Prime Minister, 28th October 1998,” (Melbourne: Melbourne Institute of Applied Economic and Social Research)

    • Search Google Scholar
    • Export Citation
  • Dawkins, Peter, 2000, “Special Topic: Labour Market Issues in Welfare Reform,Quarterly Bulletin, Version 1.00 (Melbourne: Melbourne Institute of Applied Economic and Social Research)

    • Search Google Scholar
    • Export Citation
  • Department of Family & Community Services, 2000, Technical and Other Appendices, Interim Report of the Reference Group on Welfare Reform (March) (Canberra)

    • Search Google Scholar
    • Export Citation
  • Department of Family & Community Services, 2000, Participation Support for a More Equitable Society, Interim Report of the Reference Group on Welfare Reform (July) (Canberra)

    • Search Google Scholar
    • Export Citation
  • Department of Family & Community Services, 2000, Government Response—Welfare Reform, Interim Report of the Reference Group on Welfare Reform (December) (Canberra)

    • Search Google Scholar
    • Export Citation
  • Disney, Richard, 2000, “The Impact of Tax and Welfare Policies on Employment and Unemployment in OECD Countries,IMF Working Paper 00/164 (Washington: International Monetary Fund)

    • Search Google Scholar
    • Export Citation
  • Eardley, Tony, 2000, “Sole Parents and Welfare Dependency,Social Policy Research Centre, Newsletter No. 76 (May)

  • Martin, John P., 1998, What Works Among Active Labour Market Policies: Evidence From OECD Countries’ Experiences, OECD Labour Market and Social Policy - Occasional Papers No. 35 (Paris: Organization for Economic Co-operation and Development)

    • Search Google Scholar
    • Export Citation
  • Organization for Economic Co-operation and Development, 1999, Implementing the OECD Jobs Strategy: Assessing Performance and Policy

  • Organization for Economic Co-operation and Development, 1999, A Caring World: The New Social Policy Agenda

  • Oxley, H., Dang T.-T., Förster, M. & Pellizzari, M., 1999, Income Inequalities and Poverty Among Children and Households with Children in Selected OECD Countries: Trends and Determinants, Paper for the LIS/DGV Conference “Child Well-Being in Rich and Transition Countries: Are Children in Growing Danger of Social Exclusion?, Luxembourg, 30 September2 October

    • Search Google Scholar
    • Export Citation

Annex IV.1

Detailed Recommendations of the McClure Report

article image
article image
article image
article image

Annex IV.2

Taper Rates and Labor Supply

This appendix presents a simple framework for analyzing the effect of an increase in taper rates—the rate at which welfare benefits are withdrawn for increases in income—on labor supply.1 Under a scaled down, but plausible, benefit structure and Cobb-Douglass preferences, it is shown that this effect is negative for all reasonable parameter values.

Assumed Tax and Benefit Structures

On the tax side, assume that there is a single marginal income tax rate, t, that applies to all gross income above some threshold, a.2 If all income comes from earnings, y, after tax income is:


Otherwise, t = T = 0. On the benefit side, assume a single program that pays a means tested benefit, B, with taper rate, s ε (0,1)—i.e, between 0 and 100 percent—and threshold a > 0 (identical to the income tax threshold, for simplicity). Thus,


and B = 0 otherwise. Moreover, assume that there is a fixed cost, k, of applying for the means tested benefit, which can be regarded as a non-refundable cost of applying for B.

Budget Constraint

The budget constraint in this model is piece-wise continuous, consisting of three sections, corresponding to: (i) those who earn above the threshold, pay tax and receive no welfare benefits; (ii) those who earn below the threshold but do not receive benefits owing to the existence of the fixed cost k of obtain such benefits; and (iii) those who earn less than the threshold and receive welfare benefits.

Suppose that net of tax income and consumption are synonymous (i.e., that we are in a oneperiod world) and that the price index is normalized to unity. For those who work and pay tax the budget constraint is c = y - T (y) or


For those who have y < a but do not claim any welfare benefit the budget constraint is simply


Finally, for those who have y < a and claim welfare benefits the budget constraint is c = y + B(y)- k, or:

Utility Maximization

This section will derive the labor supply decision consistent with utility maximizing behavior. While, three different cases are possible (as noted above) as to the take up or eligibility of welfare benefits, the only case of interest for the present exercise is those who work, qualify for welfare and claim benefits. Individuals are assumed to maximize their utility over consumption and leisure according to the following Cobb-Douglas function:

U(c, h)=cαh(1α)(6)

where h represents leisure, and the time endowment has been normalized to unity; i.e. labor supply is (1-h). Labor is the only endowment—there are no assumed asset stocks to run down, meaning that labor is the only source of income (other than welfare benefits). There is no bequest motive. If the wage, w, is taken as given and agents live for only one period, then all earnings and benefits are consumed


It is straightforward to substitute (7) into (6). The first order condition yields, with a little manipulation:


If h*< 1, then the agent will find it optimal to work (and will receive welfare benefits as long as wh* < a). To ensure an interior solution it must be the case that


From (9), it is easy to note that the smaller is the taper rate, s, ceteris paribus, the more likely that an individual would supply a positive amount of labor. The figure below illustrates over which constellations of parameters (9) would hold.3


Wage Thresholds for Various Taper Rates (Area above threshold implies positive labor supply)

Citation: IMF Staff Country Reports 2001, 055; 10.5089/9781451802030.002.A004

Leisure/Labor Supply Effects

We are now in a position to ascertain the effect of changes in the taper rate for welfare benefits on leisure chosen and labor supplied. Differentiating (8) with respect to s yields:


Which is positive as long as a > k; i.e., if the threshold at which welfare benefits end is greater than the fixed cost of applying for such benefits. Since a violation of this condition would imply that no one would claim any welfare benefits, we can conclude that a always exceeds k and that the effect on labor supply (leisure) of increases in the taper rate is always negative (positive).

For completeness, the analysis is taken one step further, and we now look at the rate of increase in leisure/labor supply from changes in the taper rate. Differentiating again,


With the requirement that a > k in order to induce anyone to claim welfare benefits, (11) shows that changes in the taper rate s increase leisure (decrease labor supply) at an increasing (decreasing) rate.

Annex IV.3

A Transition Bank

One of the main measures contained in the Government’s initial response to the McClure Report was the creation of a “transition bank.”1 The idea behind a transition bank is to allow those income support recipients with seasonal, casual or other intermittent types of work to be compared similarly under the income test with those who earn an equivalent amount on a regular basis. Under the current system, with its “free area” and “taper rates” implemented on fortnightly time intervals,2 income support recipients with intermittent work face a much larger reduction in benefits man those with regular, part-time work, with the attending disincentives for labor force participation. This is illustrated below with a series of examples for hypothetical individuals who receive the Newstart (unemployment) Allowance.

Simple Examples

Consider two individuals, X and Y, who are both single and work part time. X works part time on a regular basis, earning $A 100 per fortnight or $A 1,300 over a six month period, while Y works intermittently, earning the same amount over the same period. Y’s fortnightly earnings are assumed to range from zero to $A 450. Both X and Y qualify for the Newstart Allowance. The free area for benefits under Newstart Allowance is $A 62 per fortnight, with a taper (or benefit withdrawal) rate of 50 percent for fortnightly income between $A 62 and $A 142, and 70 percent for income over $A 142.

The analysis is undertaken by comparing benefit reductions. Table 1 shows the benefit reductions for person X. Given a free area of $A 62, person X has “non-free income” (i.e., used to calculate a reduction in benefits) of $A 38 each fortnight. Since X’s fortnightly non-tree earnings fall within the 50 taper rate for non-tree income, X faces a benefit reduction of $A 19 per fortnight On a cumulative basis, person X loses $A 247 over the 13-week period.

The situation is quite different for person Y, who earns the same amount as person X over the 13-week period, but with higher fortnightly earnings variability. Person Y’s benefit reductions over the period total $A 646, over two-and-one-half times the reductions faced by person X (Table 2). These reduction range from zero (in weeks where income is zero or inside the free area) to $A256, which is equivalent to 57 percent of that week’s income.

The intuition for these differences is straightforward. With the more variable profile, person Y’s earnings are more penalized by the income test, which is administered on a fortnightly basis. Alternatively, person Y is losing” the use of the free area in those weeks where income is zero or negligible, while person X gains the full benefits of the tree area. Indeed, while 62 percent of person X’s earnings are in the “free area,” only 23 percent of person Y’s earnings enjoy this status. Moreover, over half of person Y’s earnings are subject to the 70 percent taper rate, which kicks in at $A 142 per fortnight, while none of person X’s earnings face this taper rate. From a macroeconomic perspective, the high average and effective marginal tax rates negatively affect person Y’s labor supply.3

Solution: A Transition Bank

Any strategy to ameliorate this problem would need to move away from applying the income test on a purely fortnightly basis, which has the effect, as shown above, of penalizing those income support recipients who have highly variable fortnightly earnings. The intention of a transition bank is to effect such a change. It seeks to achieve this objective by allowing income recipients to build up a stock of unused free area “credits” over time. This stock can then be used as needed as an offset against any fortnightly income earned subsequently in order to increase the amount covered by the free area and lower or eliminate future benefit reduction. Person Y’s earning schedule and benefit reductions using a transition bank are shown in Table 4.

As can be seen from the Table, person Y’s benefit reductions fall by more than 50 percent when using a transition bank (although they remain about 20 percent higher than person X’s) compared to the example above, which would serve to eliminate most of the negative labor supply effects arising from the divergent treatment of similar total income streams. Person Y’s high benefit losses in fortnights 6 and 11 under the no transition bank regime have largely been offset by using balances built up in the transition bank. Moreover, since person Y has used up all of the cumulative free area over the 13-week period, the percentage of income not subject to a taper rate is the same as for person X (see Table 3).

Table IV.3.1.

Person X: Benefit Reduction under Newstart Allowance

article image
Table IV.3.2.

Person Y: Benefit Reduction under Newstart Allowance

article image
Table IV.3.3.

Summary of Earning Subject to Various Taper Rates

article image
Table IV.3.4.

Person Y: Benefit Reduction under Newstart Allowance with Transition Bank

article image


Table 1.

Australia: Selected National Accounts Aggregates at 1998/99 Prices, 1995–2000 1/

article image
Source: Australian Bureau of Statistics.

Quarterly data are seasonally adjusted.

Includes real estate transfer expenses.

Includes livestock and intangible fixed assets.

Contribution, to GDP growth, at annual rates.

Table 2.

Australia: Sectoral Components of Gross Domestic Product at 1998/99 Prices, 1995–2000 1/

article image
Source: Australian Bureau of Statistics.

Quarterly data are seasonally adjusted.

Includes defense.

Table 3.

Australia: Household Income, Expenditure and Savings, 1995–2000 1/

article image
Source: Australian Bureau of Statistics.

Quarterly data are seasonally adjusted.

Includes investment income of insurance enterprises and superammuation funds attributable to policyholders and imputed interest on government unfunded superammuation arrangements.

Refers to unincorporated enterprises owned by households in which the owners or members of the same household contribute unpaid labor

Includes workers’ compensation.

Unincorporated enterprises owned by households.

Table 4.

Australia: Saving and Investment Balances, 1995–2000 1/

(In percent of GDP)

article image
Source: Australian Bureau of Statistics.

Quarterly data are seasonally adjusted.

National accounts basis, as measured by the authorities.