This Selected Issues paper examines the role that government policy in Australia plays in influencing household saving, both directly through its own saving and the structure of the tax, social security and welfare systems, and indirectly through the influence of the policy environment on factors that affect saving such as economic growth. The determinants of household saving in a sample of 21 OECD (Organization for Economic Cooperation and Development) countries are also investigated, using both cross-section and panel estimation techniques.

Abstract

This Selected Issues paper examines the role that government policy in Australia plays in influencing household saving, both directly through its own saving and the structure of the tax, social security and welfare systems, and indirectly through the influence of the policy environment on factors that affect saving such as economic growth. The determinants of household saving in a sample of 21 OECD (Organization for Economic Cooperation and Development) countries are also investigated, using both cross-section and panel estimation techniques.

II. Labor Market Issues in Australia1

A. Summary

1. Since the mid-1970s, the unemployment rate in Australia has shown a strong upward trend, rising from around 2 percent to its present level of 8½ percent at end-1996. This increase has largely been due to structural factors, with estimates suggesting that the steady state unemployment rate has increased to around 7½–8½ percent at present.

2. The government has responded to these trends by implementing a broad range of labor market policies, aimed at creating a more flexible labor market. Since the late 1980s, it has been decentralizing the industrial relations system, allowing increasing scope for firms and employees to tailor wage and employment conditions to their particular circumstances, rather than have them dictated by central “awards.” The government has pursued active labor market policies to increase search skills and address skill mismatches, and has reformed welfare policies to improve the incentives for the jobless to seek employment.

3. This paper reviews recent developments in the labor market in Australia and presents some estimates of the steady-state unemployment rate (Section B). It then discusses the government’s policies with regard to the labor market (Section C). Finally, it assesses the outlook for unemployment, and discusses the implications of the 1996 reforms for labor market performance (Section D).

B. Unemployment in Australia

Trends in unemployment

4. For the past three decades, unemployment in Australia has shown a strong upward trend (Chart II.1). From the mid-1970s to the early 1980s, the unemployment rate increased from 2 percent to 6 percent and stabilized temporarily at this level; during the 1980s, it never fell sustainably below 6 percent and reached as high as 10 percent; and, in the 1990s, it has not been less than 8 percent and reached over 11 percent.

CHART II.1
CHART II.1

AUSTRALIA: Unemployment Rate, 1966767–1996/97

(Seasonally Adjusted)

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A002

Source: Australian Bureau of Statistics.

5. Long-term unemployment has also followed an upward trend (Chart II.2). However, unlike in other OECD countries, there does not appear to be a “hysteresis” problem. To the contrary, regression analysis provides no evidence that the trend increase in the rate of long-term unemployment has been independent of changes in the rate of unemployment.2 Rates of long-term unemployment for males and females have followed a similar path, although the rate for males has increased relative to the female rate.

CHART II.2
CHART II.2

AUSTRALIA: Unemployment and Long-Term Unemployment, 1978/79-1996/97

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A002

Source: Australian Bureau of Statistics.

6. As for trends in employment, three distinct subperiods can be identified. During the 1970s, employment growth was low, far below the rate of expansion of the labor force (Tables II.1 and II.2). However, this was not the case during the expansion of the 1980s: from 1983/2 to 1989/4, employment growth averaged 3½ percent per annum—one of the highest rates in the OECD.3 This robust growth stemmed from a rapid increase in female participation rates, especially among the more educated groups. These workers were of a quality that even though many demanded part-time employment, which was relatively expensive for employers, they quickly found jobs as they entered the labor force and, consequently, female employment grew by S percent per annum (Chart II.3). Nevertheless, reflecting this simultaneous expansion of labor supply and demand, the unemployment rate at the end of the 1980s was much the same as it had been at the start of the decade.

Table II.1.

Australia: Changes in Rate of Unemployment, 1974/2–1996/3

(Seasonally adjusted)

article image
Sources: 1974–1977—ABS, Labour Force Australia Historical Summary 1966 to 1984, catalogue no. 6204.0; 1978–1995—ABS, Labour Force Australia 1978–1995, catalogue no. 6204.0; 1996—ABS, Labour Force Australia, August 1996, catalogue no. 6203.0.
Table II.2.

Australia: Percentage Changes in Employment and Labor Force 1974/2–1996/3

(Seasonally adjusted)

article image
Sources: 1974–1977—ABS, Labour Force Australia Historical Summary 1966 to 1984, catalogueno. 6204.0; 1978–1995—ABS, Labour Force Australia 1978–1995, catalogue no. 6204.0; 1996—ABS, Labour Force Australia, August 1996, catalogue no. 6203.0.Note: FTE=full-time employment; E = employment; and LF = labor force.
CHART II.3
CHART II.3

AUSTRALIA: Employment and Labor Force, 1966/67-1996/97

(Seasonally Adjusted)

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A002

Source: Australian Bureau of Statistics.

7. The expansion of the 1990s has been different again. For the first two years of the recovery, employment actually declined, and the unemployment rate rose from 8 percent in 1990/91 to 11 percent by 1992/3. Subsequently, employment began to expand, but at a slower pace than in the 1980s, averaging less than 3 percent per annum from 1993/3 to 1996/3. As before, the fastest growing subcategories were female and part-time employment, which increased at rates of 3½ percent and 4½ percent, respectively.4

8. Other important trends and features of the labor market are:

  • sectoral developments have been important factors behind the divergent trends in male and female employment. Declining employment in manufacturing industry has predominantly affected males, while females have taken advantage of the rapid expansion of employment in the finance and community service sectors (Table II.3).

  • rates of unemployment are highest, and display the greatest cyclical sensitivity, in the manufacturing, construction, and wholesale/retail trade sectors (Chart II.4).5 The same sectors account for a disproportionate share of total unemployment—52 percent of unemployed persons compared to 42 percent of total employment in August 1996.

  • the unemployed are disproportionately from younger age groups. In August 1996, 36 percent of unemployed persons were aged 15–24, even though they only accounted for about 20 percent of persons in the labor force.

  • unemployed persons are disproportionately from groups with lower levels of educational attainment. In February 1994, persons who had not completed high school accounted for 46 percent of unemployed persons, but for only 34 percent of the labor force.

Table II.3.

Australia: Changes in Employment by Industry 1974/2–1996/3

article image
Source: ABS, Labour Force Australia, May 1974, August 1978, May 1981, May 1983, November 1989, August 1993, August 1996, catalogue no. 6203.0.Note: Changes to the method of industry classification in 1993 (from ASIC to ANZSIC) mean that it is not possible to maintain connsistnt industry groups from 1974 to 1996. For purposes of comparison, data for both methods of industry classification are reported for the time period 1989/4 to 1993/3.
CHART II.4
CHART II.4

AUSTRALIA: Unemployment by Industry, 1978-96

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A002

Source: Australian Bureau of Statistics.1/ Totals do not add to unity because those who have never been employed or who have not worked full-time for at least two weeks in the previous two years are included in the definition of labor force, but are not allocated to an industry.2/ Includes Personal Services, Education, Health and Community Services, Cultural and Recreation Services.

Steady-state rate of unemployment

9. The steady-state rate of unemployment abstracts from short-term cyclical fluctuations and hence provides policymakers with an estimate of the “permanent” or long-run rate of unemployment.6 A number of recent studies have estimated this rate for Australia, and they show a fairly consistent picture (Chart II. 5; a detailed summary of the methodology for each study is presented in Table 2 in the Annex). A break in the steady-state rate of unemployment is found around 1973 or 1974, with the rate rising from around 1 to 2 percent in the early 1970s to around 5 to 6 percent in the early 1980s. There is also evidence of a 1 percentage point increase during the 1980s, and a further 1 to 1½ percentage point increase from the late 1980s to the present, bringing the current steady-state unemployment rate to around 7½ to 8½ percent. Estimates of the NAIRU from the Treasury Macroeconomic Model of the Australian economy (TRYM) currently place it around 7½ percent.7

CHART II.5
CHART II.5

AUSTRALIA: Estimates of the Steady-State Rate of Unemployment, 1970-1996

(In percent)

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A002

Sources: Studies listed and staff estimates.

10. Elmeskov (1993) develops estimates of the NAIRU for a range of OECD countries by first estimating a NAIRU series from a simple Phillips curve specification, and then applying a Hodrick-Prescott filter to smooth the series.8 This approach gives a much larger increase in the steady-state rate during the 1970s, thereby estimating the level at 7½ percent in 1981, compared to estimates of about 6 percent in other studies (with the exception of Huay and Groenewold (1992)). After 1981, the estimated change is consistent with other studies, but the difference in level persists; the estimate for August 1996 is 9¼ percent.

Causes of changes in the steady-state rate of unemployment

11. There appear to be four proximate causes of the rising steady-state rate of unemployment in Australia:

  • growth in real unit labor costs. The contractionary phases starting in 1974/2, 1981/2, and 1989/4 were each preceded by significant increases in real unit labor costs, of as much as 8 percent (Chart II.6, top panel). In these episodes, especially in the 1970s and early 1980s, critical roles were played by commodity prices and the industrial relations system. Whenever there was an upswing in prices and wages in the booming sectors, the award system would ensure that similar wage increases were granted in other sectors, thereby raising the economy-wide wage level.9 Then, when commodity prices fell and profit margins were squeezed, the wage floor imposed by the award system forced firms to adjust by reducing employment.10

  • increases in the level of unemployment benefits. The unemployment benefit replacement rate (defined as unemployment benefit payments as a proportion of average earnings) for each family type increased substantially between 1971/72 and 1974/75, but has been relatively stable since (Chart II.7). This tended to reduce the rate of outflow from unemployment (Gregory and Paterson, 1980), and Trivedi and Kapuscinski (1985).

  • slower GDP growth. Around 1973/74, GDP growth exhibited a structural break, and employment growth shifted down in parallel. As in many other OECD countries, the growth slowdown reflected a deceleration in productivity, the underlying cause of which remains unclear.

  • structural change. Increases in unemployment have also coincided with accelerations in structural change, as measured by the inter-industry variation in employment growth (Chart II.7). In large part, this link has been forged by the industrial relations system. After tariff barriers started to be lowered in the 1980s, competitive pressures intensified, yet awards constrained firms’ ability to reduce costs by adjusting wages or adopting productivity-enhancing changes in work conditions. As a result, firms responded by reducing employment. At the same time, rates of exit from unemployment declined because the shift in the pattern of production toward Australia’s comparative advantage created mismatches between the skills of the unemployed (who had lost jobs primarily in manufacturing) and those demanded by employers (primarily in services).

CHART II.6
CHART II.6

AUSTRALIA: Determinants of Employment Growth, 1971/72-1995/96

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A002

Source: Australian Bureau of Statistics.
CHART II.7
CHART II.7

AUSTRALIA: Determinants of Employment Growth, 1971/72–1992/93

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A002

Sources: Australian Bureau of Statistics; and Bradbury, Ross and Doyle (1991).

12. The relative importance of these factors has varied over time. For the period from the mid-1970s to early 1980s, studies generally conclude that the primary factor was an increase in real unit labor costs, accounting for about 2½ of the 4 percentage point rise in the steady-state rate of unemployment (see Valentine, 1993; Pissarides, 1991; and Huay and Groenewold, 1992). Increases in unemployment benefits also appear to have been important, accounting for about 1 percentage point of the increase (Pissarides, 1991; and Huay and Groenewold, 1992). In contrast, the decline in output growth explains only perhaps a half a percentage point of the increase (Pissarides, 1991), while the impact of structural change was small (Trivedi and Baker, 1985; and Dao, 1993).

13. For the period since 1981, studies have generally found that the key factors explaining the increase in the steady-state rate of unemployment have been changes in real unit labor costs, the rate of growth of GDP, and structural change. Labor costs appear to have again been important (accounting for perhaps half of the 1 percent increase over the 1980s), while changes in GDP growth have been critical from the late 1980s onwards (see Huay and Groenewold, 1992; Dao, 1993; and Phipps and Sheen, 1995). As for structural change, Fahrer and Heath (1992) show that skill mismatches caused exit rates from unemployment for males to decline during the 1980s, while there also appears to have been a decrease in the search effectiveness of the unemployed (Fahrer and Pease, 1993).11 Meanwhile, changes in the replacement rate do not seem to have been an important explanatory variable during this period.

C. Labor Market Policies

14. Since the late 1980s, substantial policy changes have been made in an effort to combat the root causes of the trend rise in unemployment. The industrial relations system has been modified, active labor market programs have been pursued, and welfare programs reformed—all with the aim of improving labor market flexibility and strengthening work incentives.

Industrial relations

15. Since the late 1980s, the government has been attempting to shift the focus of industrial relations away from the award system and toward direct negotiations between employers and employees at the enterprise level. In this way, wage and employment conditions can be tailored more closely to the circumstances facing individual firms.

Comparison of Australia with Other OECD Countries

Comparisons between the rates of unemployment in Australia and in various OECD countries are presented in Chart II.8. The unemployment rate in Australia generally moved closely with the OECD average, until the most recent contractionary period, when it moved several points above the norm. This gap has been narrowed, but not closed, during the current expansion. Among individual OECD countries, the United Kingdom and Canada’s unemployment paths have been similar to that in Australia, while those of Japan and the United States have displayed less cyclical variation and no trend increase. New Zealand experienced a longer contractionary period than Australia in the late 1980s and early 1990s, during which the unemployment rate increased from 4 to 5 percentage points below the rate in Australia in the mid-1980s to the same levels in the early 1990s. However, by end-1996, the unemployment rate in New Zealand was once again below that in Australia (by about 2½ percentage points).

Recent studies have examined the causes of differences in unemployment between countries. Jackman and others (1996) undertake a regression analysis of determinants of rates of unemployment in 20 OECD countries during 1983–88 and 1989–94. They found that the relatively high rate of unemployment in Australia is explained by the longer duration of unemployment benefits, lower expenditure on labor market programs, and the higher level of union coverage in Australia than most other countries. Ball (1996) examined causes of changes in the NAIRU for 20 OECD countries between 1980 and 1990. He finds that the main explanation for cross-country differences is the interactions between the extent of disinflation experienced in each country and the duration of its unemployment benefits. For example, although the United States experienced strong disinflation during the 1980s, this did not cause an increase in the NAIRU due to the six-month limit on unemployment benefits; on the other hand, the increase in the NAIRU in Australia despite continued high inflation is explained by the unlimited duration of unemployment benefits.

CHART II.8
CHART II.8

AUSTRALIA: International Comparison by Unemployment Rates, 1972-95

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A002

Source: Australia Bureau of Statistics; and OECD.

16. While some progress was made prior to the enactment of the Industrial Relations Reform Act 1993 (IRRA), this legislation significantly advanced the move toward enterprise bargaining. The Act provided for two types of enterprise agreements—certified agreements (CAs) for the union sector and enterprise flexibility agreements (EFAs) for the lightly-unionized or non-union sectors—which henceforth were to provide the main avenues for employees to achieve wage increases. The role of the Australian Industrial Relations Commission (AIRC) would be reduced to guarding the interests of employees who were not able to gain enterprise bargain wage increases, by providing general “safety net” increases in award wages. Nonetheless, awards retained a crucial role in anchoring the system, since enterprise bargains were to be screened to ensure that they contained “no disadvantage” for employees, compared to the relevant awards. This reform has been successful in encouraging the spread of enterprise bargaining. Currently, one-third of the workforce have their wages determined by formal enterprise agreements, one-third by informal unregistered agreements, and one-third by awards.12 The bulk of enterprise agreements are in the manufacturing, construction, and health and community service sectors.

17. A number of trends in federal enterprise agreements are apparent. First, the rate of commencements is increasing, rising from 1605 in 1994 to 3277 in 1995. Second, genuine enterprise-level agreements appear to becoming more common, with the number of multi-site agreements falling to 37 percent in 1995, from 80 percent just one year earlier. Third, nearly all registered agreements have been CAs, agreed between employers and unions. The nonunion EFAs have not been extensively used (only 104 such agreements commenced operation in 1995), largely because the small business sector has preferred to rely on informal arrangements, rather than have the AIRC register agreements—and subject them to a “no disadvantage” test, during which unions have the right to object. In addition, some businesses have viewed EFAs as unnecessarily complicated, while others have simply been ineligible, since they are neither incorporated nor a respondent to a federal award (DIR, 1996a).

18. Despite this progress in shifting toward enterprise bargaining, there are a number of features of the industrial relations system that still constrain labor market flexibility in important ways. These include:

  • the central role of the award system: although enterprise bargaining has increased, the award system still remains the linchpin of the industrial relations system. Indeed, surveys indicate that most registered agreements are not true comprehensive agreements, but merely “add-ons” to existing awards, typically increasing wages over the basic award rate. This situation persists largely because awards represent the benchmark against which the “no-disadvantage” test is conducted. As a result, awards continue to dictate many working arrangements, entrench wage relativities, and limit wage dispersion.13

  • the excessive involvement of third parties: even if an enterprise bargain is agreed between an employer and employees, it must still undergo screening, during which it may be opposed by unions not represented in the firm, and ultimately rejected by the AIRC.

  • lack of labor market competition: unions have had a monopoly status in representing employees in certain areas of the economy; employees could not register new unions (for example, at the enterprise level) if they could “conveniently belong” to an existing one. This has contributed to the prevalence of “pattern bargaining,” under which unions would first negotiate agreements with key employers, and then use these as the standard for all firms in the sector, regardless of their individual circumstances.

  • excessive employment protection provisions: the unfair dismissal provisions, which were strengthened in the IRRA, are complex and emphasize procedure, rather than outcomes. This has reduced the incentives for firms to hire workers because it is difficult to subsequently dismiss them.

19. Recognizing that further reform was needed, the government legislated in late 1996 the Workplace Relations Act in late 1996, which will help further facilitate direct workplace relations between employers and employees. However, because it does not hold a majority in the Senate, the government had to compromise on some important aspects of its original legislation to ensure parliamentary passage. The main changes to existing industrial relations legislation are:

  • the federal award system will be simplified, to promote its ongoing transformation into a true safety net. The number of employment conditions in awards will be confined to 20 allowable matters, with all other matters being determined at the enterprise or workplace level (although, in exceptional circumstances, the AIRC may be allowed to arbitrate on nonallowable matters).14 Conditions relating to the proportion of different types of employees (for example, full-time and part-time), and to the minimum or maximum hours for part-time employees will no longer be included in awards. Henceforth, the AIRC’s role will focus on the setting of minimum wages and conditions in awards. Although available for conciliation in all matters, the AIRC will generally not be able to arbitrate claims above the minimum safety net. The mechanism by which safety net wages and conditions are to be adjusted was not prescribed in legislation, although a set of guidelines will be specified.

  • new arrangements for CAs will be introduced, to give more flexibility to parties negotiating formal enterprise agreements. Scope will be provided for CAs to be made with employees in workplaces without union members, and a union must have a member in an enterprise to be able to make an agreement or participate in negotiations at that enterprise. Multi-employer agreements will require the consent of all employers and a majority of employees covered by the agreement. Most important, CAs will be subject to a new global “no disadvantage” test, which will allow trade-offs in wages and working conditions, as long as the agreements, when considered as a whole, are no less favorable to an employee than the relevant awards.

  • Australian Workplace Agreements (AWAs) will be introduced, to replace the unsuccessful EFAs, as a framework for the non-union sector. These agreements will reduce third-party involvement, notably because AWAs will be lodged with an Employment Advocate, which will refer them to the AIRC only if it has serious doubts that they satisfy the global no disadvantage test. In addition, uninvited union involvement will be excluded, although employees can nominate a bargaining agent to act on their behalf.15

  • union competition will be increased as the exclusive jurisdiction principle will be repealed, and existing organizations will be able to oppose the registration of a new union only on the basis that the applicant’s members could “more conveniently belong” to the existing union.16 Provisions will be made to allow registration of enterprise unions and the disamal-gamation of federally registered unions. The minimum registration size for a union will be lowered from 100 to 50. Individual employees will have freedom of association to join or not join a union of their choice. Preference clauses will be outlawed, and compulsory unionism or discrimination based on membership or nonmembership of a union will be prohibited.

  • the unfair dismissal legislation will be reinterpreted. The AIRC will base its decisions on the principle of a “fair go all around” (fair to both employers and employees).

20. In recent years the state governments have introduced similar reforms to their industrial relations systems. The most dramatic shift occurred in Victoria, where the introduction of the Employee Relations Act in 1993 abolished compulsory arbitration and existing awards in the state, but retained a legislated safety net of minimum terms and conditions but retained a legislated safety net of minimum terms and conditions. However, under an agreement made in November 1996, the Victorian government will cede its industrial relations powers to the federal government. Workers not already covered under the federal system will remain in the state system until their existing agreements expire, at which time no new state agreements will be allowed and employees will have the option of shifting to federal awards. Other states have tended to retain compulsory arbitration and the award system, but have provided employers and employees with the opportunity to negotiate enterprise agreements. However, the spread of formal enterprise agreements in state jurisdictions has generally lagged behind that at the federal level.

Industrial Relations Reform: A Comparison of Australia and New Zealand

The industrial relations systems in Australia and New Zealand have traditionally shared many common features: compulsory arbitration of industrial disputes; a registration process which ensured recognition of unions before industrial tribunals and which prevented competition for members of existing unions; and a system of awards setting out wages and terms of employment.

During the 1980s, both Australia and New Zealand made tentative steps toward deregulating their labor markets. In 1987, the Labour government in New Zealand introduced the Labor Relations Act which attempted to rationalize the number of trade unions and required the formation of single bargaining units to negotiate at the enterprise level with large employers (O’Neill, 1994). However, the most dramatic shift, and one that caused a significant divergence between the industrial relations systems in Australia and New Zealand, came with the introduction of the Employment Contracts Act (ECA) in 1991.

Under the Act, industry-wide awards were abolished and replaced with a system of free contracting aimed at facilitating agreements negotiated directly between employees and employers at the enterprise level. Individual employees can either negotiate their own contracts or appoint a bargaining agent to represent them. Contracts must meet minimum standards relating to conditions such as wages, leave, hours of work, and grievance procedures. These changes effectively brought to an end compulsory unionism, closed shops, and preference arrangements for union members. Although the ECA does not mention unions, legislative protection for their role is provided under the Trade Unions Act.

A comparison of the ECA and the Workplace Relations Bill reveals that the wage bargaining system in New Zealand still differs from that in Australia. Most importantly, Australia still maintains a system of awards and a significant role for trade unions in wage bargaining. Conditions of employment specified in awards constitute the minimum standards to be met by new workplace agreements, and awards are also to be the mechanism for implementing safety net wage increases for workers unable to gain wage increases through enterprise bargaining. Trade unions are guaranteed a role in bargaining over Certified Agreements provided a member requests their involvement. However, provisions in the Workplace Relations Bill relating to the registration process for unions, and outlawing of preferences clauses and closed shops, will move Australia’s system closer to that in New Zealand

Labor market programs

21. A number of important policy changes have also been made to labor market programs. In 1994, the “Working Nation” statement announced initiatives to improve equity and efficiency by providing the unemployed—especially the long-term unemployed—with the necessary skills and experience to effectively compete for jobs. To achieve this, labor market expenditure was increased significantly, especially on wage subsidy schemes and direct job creation. Subsequently, an official evaluation of these efforts was conducted, which led the government to introduce some major modifications in the context of 1996/97 budget.

22. The main elements of the “Working Nation” program were:

  • the Job Compact, which was to provide a 6–12 month job placement, primarily in the private sector, to all those who had been on unemployment allowance for 18 months or more. The job placements were to be achieved through a number of different programs, including wage subsidies for the duration of the placement period (the JobStart program); job creation schemes (JobSkills and New Work Opportunities); and self employment schemes.

  • a case management strategy, to help disadvantaged job seekers.

  • the Youth Training Initiative, to provide intensive assistance to those under the age of 18 years.

  • an expanded traineeship system, including a national training wage, at lower than standard award levels, so as to facilitate entry into the labor force.

23. The success of these programs, however, has been limited. In 1996, an official evaluation reached the following conclusions about the programs:

  • Job Compact. A job guarantee for all disadvantaged job seekers did not appear to have been appropriate, as the program had relatively limited success in assisting the long-term unemployed into ongoing employment. While the objective of targeting assistance to this group was largely achieved (Table II.4), the actual number of long-term unemployed did not decline, partly because the reduction in assistance to persons unemployed for between 12 to 18 months increased the rate of inflow into the group unemployed for more than 18 months. In addition, considerable variation existed in the effectiveness of the different types of programs. The wage subsidy and business start-up programs proved to be relatively successful. However, a decline in JobStart (wage subsidy) placements occurred, because employers perceived that the shift in the program’s emphasis led to a decrease in the quality of its labor. The decline in JobStart placements, in turn, meant that a higher-than-expected proportion of job seekers were placed in programs such as Job Skills, which were costly and less likely to lead to unsubsidized employment outcomes (Tables II.5 and II.6). Finally, the average cost of achieving unsubsidized employment outcomes for Job Compact clients was also greater than expected.

  • Case Management. While case management appears to have delivered a more personalized service and had some deterrent effect against nonjob seekers registering as unemployed, it does not appear to have achieved better matching of unemployed persons to training programs and employment.

  • Traineeships. These have proved attractive to both employers and trainees, but more emphasis needs to be placed on the quality of the training.

Table II.4.

Australia: Labor Market Program Participants by Duration of Unemployment, 1992–93 to 1995–96

(February)

article image
Source: DEETYA (1996a, p. 44).
Table II.5.

Australia: Positive Post-Program Outcomes by Labor Market Program—Percent of Program Participants in Unsubsidized Employment Three Months After Completion of Program

article image
Source: a) 1992–93—Department of Employment, Education and Training, Annual Report 1992–93, p. 120 (Canberra, AGPS); b) Job Compact—DEETYA (1996a, p. 53).Note: 1992–93 relates to persons who participated in the programs in the 12 months prior to March 31, 1993; and Job Compact relates to persons who participated in the programs in the 12 months prior to February 1996.
Table II.6.

Australia: Cost of Labor Market Programs

(in Australian dollars)

article image
Source: EPAC (1996a, p. 92).Notes: a) JOBSTART, JOBTRAIN, Skillshare, and Job Clubs costs are calculated for 1991–92, and LEAP and JOBSKILLS costs are calculated for 1992–93; b) an unsubsidized employment outcome is defined to occur where a program participant is in unsubsidized employment three months after completion of the program.

24. Based on this evaluation, the new government announced significant reforms to labor market programs in the 1996/97 budget. The reforms aim to improve the efficiency with which labor market services are provided by introducing more competition into the market, and by more effectively fashioning the type of assistance to the individual needs of the client.

25. Under the changes, a Service Delivery Agency (SDA) will provide a single point of contact for those seeking to obtain employment and income support services. The SDA will provide self-help job search facilities, and screen clients for eligibility for further assistance. Those eligible will be referred to Employment Placement Enterprises (EPEs), which will provide three main types of service: labor exchange services, job search assistance, and intensive employment assistance. The last will involve a personal assessment of the client’s needs, followed by a maximum of 12 months of customized assistance such as training and job search assistance.

26. Eligibility for labor exchange services and intensive employment assistance will be restricted—in particular to job seekers unemployed for 12 months or more and who are in receipt of a benefit from the Department of Social Security. Eligibility for intensive employment assistance will also be on a “capacity to benefit” basis—low priority will be accorded to job seekers assessed as likely to find employment without assistance, job seekers who have already received substantial case management assistance, and job seekers who are unlikely to obtain employment even with intensive assistance.

27. Perhaps most significant, a competitive employment placement market will be introduced—the first in the OECD. The public Commonwealth Employment Service will be corporatized, and required to compete with private and community-based EPEs to provide employment placement services. Payments to EPEs will be based on outcomes; to count as an outcome, jobs or training places will need to substantially reduce or end payments to a job seeker for at least six months. Payments will generally be higher for achieving primary outcomes, defined as a placement into full-time employment or traineeships. Payment structures will also be designed to provide incentives for EPEs to attempt to place those clients with the greatest degree of labor market disadvantage.

The welfare system

28. The design of the welfare system is generally believed to have contributed to persistent unemployment and has therefore been subject to a number of reforms. These have not changed, however, the basic features of the system. In particular—and unlike in most other OECD countries—unemployment benefits are not related to past earnings, but are based on current need (“means tested”), and benefits are paid for an indefinite period.17 Both factors play a role in discouraging the unemployed from engaging in intensive job search efforts.

29. The fact that benefits are independent of previous earnings implies that for persons in the bottom 10 percent of the earnings distribution, the replacement rate is close to or even exceeds 100 percent for some household groups (Table II.7).18 However, using data on annual earnings from employment in 1989/90, Whhitock (1994) found that only about 11 percent of families in Australia had replacement rates of more than 80 percent; and only about 3 percent had rates of 100 percent or more. Sole parents and couples with children had the highest replacement rates, and single persons and couples without children had the lowest rates.19

Table II.7.

Australia: Replacement Rates—Unemployment and Other Benefits as Percentage of After-Tax Income of Wage-Earner, 1995

article image
Source: EPAC (1996a, Table 4.3).

30. The fact that unemployment benefits are paid indefinitely (whereas in most OECD countries unemployed persons receive social assistance or welfare benefits which are lower than unemployment benefits after they have exhausted their unemployment benefit entitlements) results in the Australian system being relatively more generous for the long-term unemployed (Table II.8). Even though average replacement rates in Australia are lower than the OECD average in the early part of unemployment, they are higher for persons who are unemployed for 60 months or more.20

Table II.8.

Australia: Unemployment Benefit Replacement Rates for Single-Earner Households—At Average Earnings For Production Workers, 1995

article image
Source: OECD (1996, Table 2.1).Note: Gross replacement rates are before tax. Net replacement rates are after tax and other benefits.

31. The unemployment benefit program in Australia is not only means tested but also activity tested. The program is therefore generally believed to produce a high degree of targeting of benefits to persons who are not in employment, but who are actively seeking it (Weatherley, 1993).21 The activity test has been tightened recently, which appears to have had a significant negative effect on inflows to unemployment. In the 1996/97 budget, a further tightening of the activity test has been announced with unemployed persons being required to complete a job-seeker diary which details job search activity (DSS, 1996).

32. To reduce the incentives to voluntarily enter unemployment, the waiting period for receipt of unemployment benefits for persons with liquid assets was extended in the 1996/97 budget from four to 13 weeks, and the definition of liquid assets expanded to include long-service leave and sick leave payments as well as annual leave payments (DSS, 1996). The increased frequency of reviews of unemployed persons’ incomes, and matching with information from the Australian Taxation Office, has also allowed more accurate assessment of benefit levels (Committee on Employment Opportunities, 1993).

D. Prospects for Unemployment

33. This section assesses the outlook for unemployment, using two different approaches. First, a simple two-equation model of the labor market is estimated, to examine the effects of alternative economic growth rates on the rate of unemployment in the four-year period between 1997 and 2000. Second, the recent policy initiatives are analyzed, to obtain a qualitative assessment of their likely impact on unemployment.

34. The simulation exercise shows that even if economic growth improves, this factor alone is unlikely to achieve the sought-after decisive reduction in steady-state unemployment. This result is based on a model with equations for: (i) changes in the employment/population ratio as a function of changes in real GDP and (ii) changes in the labor force participation rate as a function of changes in the employment/population ratio. Each equation is estimated for persons, males, and females.22 The equations are then used with forecasts of economic growth to predict the employment/population ratio, labor force participation ratio, and rate of unemployment for 2000.23

35. Three scenarios are considered—a low-growth scenario of 2½ percent per annum, a medium (or consensus) growth scenario of 3½ percent, and a high-growth scenario of 4½ percent. In the medium-growth scenario, little change in the unemployment rate is predicted by 2000 (Table II.9). In the low-growth scenario, the unemployment rate increases by about 1 percent to 9½ percent by 2000, while in the high growth scenario, the unemployment rate declines by about 1½ percent.

Table II.9.

Australia: Predicted Employment, Labor Force, Unemployment Outcomes, June Quarter 2000

article image

36. These results imply that a structural improvement in unemployment will depend heavily on the success of the reform efforts in creating a more flexible and efficient labor market. Of course, the impact of the recent policy initiatives is difficult to assess, even on a qualitative basis. Beneficial effects on labor market efficiency, and hence on employment, should follow from the greater decentralization of industrial relations, because it will be possible to match wage growth more closely with to labor productivity changes at the enterprise level. Decentralization will also allow firms to adjust relative wages, so as to better match supply and demand for skills; already, there is evidence that wage bargaining under the IRRA has increased the dispersion in wage outcomes between enterprises and industries relative to the Accord period of the 1980s (Pragnell and others, 1996). Furthermore, the narrowing of the scope of awards to 20 allowable matters, which can be traded off under a new global “no disadvantage” test, will increase the ability of firms to restructure working conditions, and thereby improve productivity.

37. At the same time, the Workplace Relations Act does not represent a fundamental break from the past. The award system will retain its central role: direct negotiations between employers and employees will still be constrained by the conditions of awards and the rulings of the AIRC. Similarly, the role of unions remains largely unaltered. As a result, it is unclear how much decentralization will really follow from the reform. For example, it remains unclear whether wage changes in different enterprises have been decoupled, or whether “pattern bargaining” will continue. Similarly, to the extent that wage increases will be determined by the bargaining strength of unions, rather than changes in labor productivity, efficiency gains may be limited. Moreover, doubts have been expressed about the whether the introduction of AWAs will be sufficient to encourage the non-union sector to take up enterprise bargaining; to the contrary, the simplification of awards may encourage employers to remain within the award stream (Sloan, 1996).

38. The recently announced changes to labor market programs place more emphasis on improving the efficiency of matching processes in the labor market, and less on job creation. With regard to matching processes, changes such as the introduction of a competitive market for provision of employment placement services, expansion of the case management system, and limits on the duration of assistance for participants in labor market programs should have positive incentive effects—both for service providers and for unemployed persons—and therefore improve labor market matching. However, a key issue will be establishing payment systems which ensure that service providers have an incentive to provide placement services and assistance to all unemployed persons. If payment rates are set so that service providers seek exclusively to provide services for short-term unemployed persons, then the absence of training or employment assistance for long-term unemployed may worsen structural imbalances and the potential efficiency of labor market matching. Reductions in expenditure on labor market programs which involve job creation may cause some increase in unemployment relative to the Job Compact period. However, this effect is unlikely to be significant.

39. Policy changes also begun to address the effects of social security benefits on work incentives. The measures to tighten the activity test and to extend the liquid assets waiting period, together with changes to monitoring of compliance for labor market program participants, should improve work incentives, but the size of this effect is difficult to predict.

40. In sum, since the 1980s, Australia has traveled quite some distance along the road toward a flexible, decentralized, labor market. Over time, these reforms should enable the steady-state rate of unemployment to fall. However, further reform is likely to prove necessary, if unemployment is to be decisively reduced.

ANNEX

Table 1.

Australia: Regression Analysis of Relation Between Rate of Long-Term Unemployment and Rate of Unemployment, 1979/1 to 1996/2

Dependent Variable:

Change in Rate of Long-Term Unemployment

article image
Note: Δ(Rate of UE) is the first-difference of the rate of unemployment; standard errors are in parentheses.
Table 2.

Australia: Studies of the Steady-State Rate of Unemployment

article image
Table 3.

Australia: Regression Model for Employment/Population and Labor Force Participation Rate, 1978/1–1996/2

article image
Note: Standard errors in parentheses.

References

  • Ball, L., 1996, ‘Disinflation and the NAIRU,’ Working Paper No. 5520, National Bureau of Economic Research.

  • Bodman, P., 1995, ‘Estimating Frictional Unemployment for Australia and Its States Using a Stochastic Frontier Approach,’ Discussion Paper No. 169, Department of Economics, University of Queensland.

    • Search Google Scholar
    • Export Citation
  • Bradbury, B., R. Ross, and J. Doyle, 1991, ‘Unemployment Benefit Replacement Rates,’ Policy Research Paper No. 60, Department of Social Security.

    • Search Google Scholar
    • Export Citation
  • Catanzaratti, J. and G. Sullivan, 1993, ‘Industrial Relations Legislation in 1992,’ Journal of Industrial Relations, 35, 11041.

  • Committee on Employment Opportunities, 1993, Restoring Full Employment—A Discussion Paper (Canberra, AGPS).

  • Commonwealth of Australia, 1994, Working Nation—Policies and Programs (Canberra, AGPS).

  • Commonwealth of Australia, 1996, Budget Statements 1996–97 Budget Paper No. 1 (Canberra, AGPS).

  • Dao, D., 1993, ‘A Model of Unemployment’, pages 15562 in Committee on Employment Opportunities, Restoring Full Employment - Background Papers (Canberra, AGPS).

    • Search Google Scholar
    • Export Citation
  • Department of Employment, Education, Training and Youth Affairs, 1996a, Working Nation: Evaluation of the Employment, Education and Training Elements (Canberra, AGPS).

    • Search Google Scholar
    • Export Citation
  • Department of Employment, Education, Training and Youth Affairs, 1996b, Reforming Employment Assistance (Canberra, AGPS).

  • Department of Industrial Relations, 1995, 1994 Annual Report Enterprise Bargaining in Australia (Canberra, AGPS).

  • Department of Industrial Relations, 1996a, 1995 Annual Report Enterprise Bargaining in Australia (Canberra, AGPS).

  • Department of Industrial Relations, 1996b, ‘The Reform of Workplace Relations—Legislation Guide,’ mimeo.

  • Department of Industrial Relations, 1996c, ‘Agreement Between the Commonwealth Government and the Australian Democrats on the Workplace Relations Bill,’ mimeo.

    • Search Google Scholar
    • Export Citation
  • Department of Social Security, 1996, ‘1996-97 Budget: What’s New? What’s Different?,’ mimeo.

  • Economic Planning and Advisory Committee, 1992, Unemployment in Australia (Canberra, AGPS).

  • Economic Planning and Advisory Committee, 1996a, Future Labor Market Issues for Australia (Canberra, AGPS).

  • Economic Planning and Advisory Committee, 1996b, The Changing Australian Labor Market (Canberra, AGPS).

  • Elmeskov, J., 1993, ‘High and Persistent Unemployment: Assessment of the Problem and Its Causes,’ Working Paper No. 132, Economics Department, OECD.

    • Search Google Scholar
    • Export Citation
  • Fahrer, J. and A. Heath, 1992, ‘The Evolution of Employment and Unemployment in Australia,’ Research Discussion Paper No. 9215, Reserve Bank of Australia.

    • Search Google Scholar
    • Export Citation
  • Fahrer, J. and A. Pease, 1993, ‘The Unemployment-Vacancy Relationship in Australia,’ Australian Economic Review, 4th quarter, 4357.

    • Search Google Scholar
    • Export Citation
  • Goodridge, S., D. Harding, and P. Lloyd, 1995, ‘The Long-Term Growth in Unemployment,’ Working Paper 2/95, IAESR, University of Melbourne.

    • Search Google Scholar
    • Export Citation
  • Green, R., 1993, ‘Wages Policy and Wage Determination in 1992,’ Journal of Industrial Relations, 35, 14256.

  • Gregory, R., 1986, ‘Wages Policy and Unemployment in Australia,’ Economica, 53, Supplement, 5374.

  • Gregory, R. and P. Paterson, 1980, ‘The Impact of Unemployment Benefit Payments on the Level and Composition of Unemployment in Australia,’ Discussion Paper No. 11, Centre for Economic Policy Research, Australian National University.

    • Search Google Scholar
    • Export Citation
  • Groenewold, N. and L. Taylor, 1992, ‘Insider Power as a Source of Hysterisis in Unemployment: Tests with Australian Data,’ Economic Record, 68, pg.5764.

    • Search Google Scholar
    • Export Citation
  • Huay, O. and N. Groenewold, 1992, ‘The Causes of Unemployment in Australia,’ Australian Economic Papers, 31, pp. 7793.

  • Jackman, R., S. Nickell, and R. Layard, 1996, “Combatting Unemployment: Is Flexibility Enough?,’ Discussion Paper No. 293, Centre for Economic Performance, London School of Economics.

    • Search Google Scholar
    • Export Citation
  • Jaeger, A., 1994, ‘Mechanical Detrending by Hodrick-Prescott Filtering: A Note,’ Empirical Economics, 19, pp.493500.

  • Layard, R., S. Nickell, and R. Jackman, 1991, Unemployment (Oxford, Oxford University Press).

  • Lewis, P. and D. Spiers, 1990, ‘Six Years of the Accord: An Assessment,’ Journal of Industrial Relations, 32, pp.23758.

  • Mitchell, W., 1992, ‘Wages Policy and Wage Determination in 1991,’ Journal of Industrial Relations, 34, pp. 15361.

  • Norris, K., 1989, ‘Wages Policy and Wage Determination in 1988,’ Journal of Industrial Relations, 31, pp.11117.

  • OECD, 1996, Employment Outlook (Paris, OECD).

  • O’Neill, S., 1994, ‘Labor Market Deregulation: The New Zealand Experience,’ pages 13–56 in Department of the Parliamentary Library, Labor Market Papers (Canberra, AGPS).

    • Search Google Scholar
    • Export Citation
  • Phipps, A. and J. Sheen, 1995, ‘Macroeconomic Policy and Employment Growth in Australia,’ Australian Economic Review, 1st quarter, pp. 86104.

    • Search Google Scholar
    • Export Citation
  • Piggott, J. and B. Chapman, 1995, ‘Costing the Job Compact,’ Economic Record, 71, pp.31328.

  • Pissarides, C., 1991, ‘Real Wages and Unemployment in Australia, Economica, 58, pp.3555.

  • Polette, J., 1995, ‘Distribution of Effective Marginal Tax Rates Across the Australian Labor Force,’ Discussion Paper No. 6, NATSEM, University of Canberra.

    • Search Google Scholar
    • Export Citation
  • Pragnell, B., T. O’Loughlin, S. O’Keefe, and J. Buchanan, 1996, ‘Wages Policy and Wage Determination in 1995,’ Journal of Industrial Relations, 38, pp.96113.

    • Search Google Scholar
    • Export Citation
  • Reitano, R., 1995, ‘Legislative Change in 1994,’ Journal of Industrial Relations, 36, pp.8494.

  • Sloan, J., 1996, ‘Labor Market Reform in Australia,’ paper presented to Conference of Economists, Canberra.

  • Stevens, G. and R. Robertson, 1993, ‘The Australian Labor Market 1967–1992: A Brief International Perspective,’ Discussion Paper No. 292, Centre for Economic Policy Research, Australian National University.

    • Search Google Scholar
    • Export Citation
  • Stretton, A. and B. Chapman, 1990, ‘An Analysis of Australian Labor Market Programs,’ Discussion Paper No. 247, Centre for Economic Policy Research, Australian National University.

    • Search Google Scholar
    • Export Citation
  • Trivedi, P. and G. Baker, 1985, ‘Equilibrium Unemployment in Australia: Concepts and Measurement,’ Economic Record, 61, pp.62943.

    • Search Google Scholar
    • Export Citation
  • Trivedi, P. and C. Kapuscinski, 1985, ‘Determinants of Inflow into Unemployment and the Probability of Leaving Unemployment,’ pp. 15188 in P. Volker (ed.) The Structure and Duration of Unemployment in Australia (Canberra, AGPS).

    • Search Google Scholar
    • Export Citation
  • Valentine, T., 1993, ‘The Sources of Unemployment: A Simple Econometric Analysis,’ Economic Papers, 12, pp. 120.

  • Watts, M. and W. Mitchell, 1991, ‘Alleged Instability of the Okun’s Law Relationship in Australia: An Empirical Analysis,’ Applied Economics, 23, pp. 182938.

    • Search Google Scholar
    • Export Citation
  • Weatherley, R., Compliance Policies in Social Security,’ mimeo, Urban Research Program, RSSS, Australian National University.

  • Whitlock, B., 1994, ‘Does the Social Security Income Support System Remove the Incentive to Work?,’ Discussion Paper No. 303, Centre for Economic Policy Research, Australian National University.

    • Search Google Scholar
    • Export Citation
  • Wooden, M., 1996, ‘Hidden Unemployment and Underemployment, Report commissioned by Department of Employment, Education, Training, and Youth Affairs

    • Search Google Scholar
    • Export Citation
1

Prepared by Dr. Jeff Borland, Centre for Economic Policy Research, Australian National University, who was a consultant to the IMF during October-December 1996.

2

A regression model for the first-difference of the rate of long-term unemployment was estimated (results are reported in Table 1 in the Annex). To select the preferred specification of the regression equation the first-difference of the rate of unemployment and four lags of this variable were initially included as explanatory variables. Sequential omission of insignificant variables yielded the equation forms specified in the table. All equations were estimated using OLS. The main point of interest is that in none of the preferred equations—for persons, males or females—is the constant term significant. Also see Economic Planning Advisory Council, 1996a, p. 131.

3

Quarter j is defined as a turning point if changes in the rate of unemployment in the next three quarters (from j to j+1, j+1 to j+2, and j+2 to j+3) are in the opposite direction to the change in the rate of unemployment from the preceding quarters (from j-1 to j). This gives turning points at 1974/2, 1978/3, 1981/2, 1983/2, 1989/4, and 1993/3.

4

Over the past four quarters, there has been a flattening in employment growth as female employment has begun to grow much more slowly than in the 1980s expansion (from August 1995 to August 1996, female employment increased by 45,000 whereas in the comparable period in the 1980s expansion—May 1985 to May 1986—it increased by 194,300). The main source of this divergence appears to be the lower rates of growth in employment in traditional female industries such as finance, business and property, and health, education and community service. However, growth in female participation has also been lower than in the 1980s expansion.

5

Industry of last job is recorded in the Australian Bureau of Statistics (ABS) Labour Force Survey for all unemployed persons who had worked full-time for at least two weeks in the previous two years. No industry status is recorded for unemployed persons who have never been in employment, or who have not worked full-time for at least two weeks in the previous two years. To calculate rates of unemployment by industry, the “labor force” in an industry is defined as employed persons plus unemployed persons whose last job was in that industry and who satisfy the condition of working full-time for at least two weeks in the previous two years.

6

The term “steady-state rate of unemployment” is used here to encompass several different concepts, including the nonaccelerating inflation rate of unemployment (NAIRU), the natural rate of unemployment, and the steady-state rate of unemployment derived from an unemployment-vacancy (UV) curve or structural models of the rate of unemployment. Each of these approaches has different theoretical underpinnings, but all produce steady-state estimates of the unemployment rate from regression analysis. To illustrate the theoretical underpinnings of a model of the rate of unemployment, consider the example of the Layard-Nickell structural model estimated in Huay and Groenewold (1992). In that model the economy is represented in a system of three equations for employment, price determination, and real wages. Estimates of the steady-state rate of unemployment are derived from the equation for employment together with information on exogenousty determined labor supply. Hence, in this framework changes in the steady-state rate of unemployment occur due to changes in factors which are included in the employment equation—for example, real wages, expected aggregate demand, and the level of unemployment benefits.

7

Based on a rolling regression of the TRYM wage equation from 1974 to 1994 (see Downes and Stacey, 1996). Estimates of the NAIRU are subject to a good deal of uncertainty. The 95 percent confidence level around the Treasury NAIRU estimate is 5.2 to 9.6 percent.

8

The first step is to use a simple Phillips Curve specification: Δ(p’)=a(u - u*) where p’ is the rate of inflation, u the rate of unemployment, u* the steady-state rate of unemployment, a < 0, and Δ the first difference operator, to derive a series for the NAIRU. In the Phillips Curve equation there are two unknowns (a and u*). However, by using the Phillips Curve relation for the preceding period, it is possible to substitute out for a, and obtain a solution for u*. The second step uses the Hodrick-Prescott filter to remove the cyclical component from the NAIRU series. The Hodrick-Prescott filter determines a trend component series {gt, t = 1, …, T) from a time-series {yt} by minimizing:

Σt=1T(ytgt)2+μΣt=2T-1[(gt+1gt)(gtgt-1)]2

where μ is a fixed parameter set to 1600 for this study as is usual for quarterly data (Jaeger, 1994).

9

The award system sets a whole range of minimum wage rates based on skills and duties, a comprehensive package of minimum employment entitlements, and also places restrictions on how work can be organized, including, in some cases, directions on who can perform certain types of job. Traditionally, maintaining wage relativities has been an explicit objective of this system. See Appendix 1 of “Australia—Recent Economic Developments” (SM/94/17; 1/18/94) and Appendix 1 of “Australiar—Background Material” (SM/95/33; 2/10/95) for a more extensive discussion of the award system, the Accord between the Labour government and the unions, and the initial moves toward a more decentralized labor market. Further information on the industrial relations system is provided in Section C.

10

In particular, during the 1970s, wages were effectively indexed to prices, preventing real wages from adjusting after Australia’s export prices fell and oil import prices increased.

11

Defined as the average of exit rates from unemployment weighted across groups with different lengths of unemployment spells.

12

Enterprise agreements which are not registered are defined as “…collective written or verbal agreements which have not been lodged with any industrial tribunal or registrar” (DIR, 1996a).

13

While wage dispersion between the highest and lowest paid (the 10th and 90th percentiles) is higher than in many European and Scandinavian countries, the differential in the middle (the 50th and 90th percentiles) is relatively low, thereby discouraging skill acquisition (Blau and Kahn, 1996). These features may be explained by the award system, which covers the bulk of occupations, but excludes managerial positions and some lower paid occupations.

14

These allowable matters are: classification of employees and skill-based career paths; ordinary time hours of work; rates of pay; piece rates and bonuses; annual leave and leave loading; long service leave; sick leave; other types of leave (for example, maternity or paternity leave); public holidays; allowances; loadings for overtime or shift work; penalty rates; redundancy pay and notice of termination; stand-down provisions; dispute settlement procedures; jury service; type of employment; superannuation; pay and conditions for outworkers; and provisions incidental to the allowable matters and necessary for the effective operation of the award.

15

The main differences between AWAs and CAs are that: AWAs can be collective or individual but collective AWAs require approval by each employee subject to the agreement, whereas CAs are collective and apply to all employees in an enterprise subject to majority approval; and the EA will be responsible for the no disadvantage test for AWAs, but may refer agreements to the AIRC for judgement, whereas the AIRC will have responsibility for the no disadvantage test for CAs.

16

The practical difference for the development of union structures of this change is difficult to judge. While, it seems to provide scope for significant changes, in particular for the development of enterprise-level unions, its effect will likely depend on how the new legislation is interpreted by the courts, and on whether the industrial relations environment provide sufficient incentives for workers and employers to break away from existing unions to create new bargaining units.

17

See Chapter I of this volume for a discussion of the structure of the Australian welfare system.

18

The median replacement rate is, however, well below 100 percent. In Australia, persons receiving unemployment benefits are also eligible for other noncash benefits. In addition, the calculations of the replacement ratio do not take into account the costs of work. The OECD (1996, p.41) has estimated that for a one-earner couple with two children in Australia the gain in income from full-time employment at average production worker earnings would be 41 percent not taking account of childcare costs, but negative 25 percent taking account of childcare costs and negative 5 percent taking account of childcare costs and benefits for childcare.

19

The means-testing of benefits also results in high effective marginal tax rates (EMTRs) for some groups. Whitlock (1994) estimates that single unemployed persons could face EMTRs of 120 percent, while for unemployed couples EMTRs of 108 percent existed. Polette (1995) using simulated data on income for 1994 finds that 62 percent of unemployed persons faced an EMTR of zero percent, and only 14 percent faced an EMTR of 80 percent or greater. In order to face a zero EMTR an unemployed person must be earning either zero income or less income from labor market activity than the threshold level ($45 in 1995 for recipients of allowances). Hence, the results of Polette’s study show that most unemployed persons have very low levels of labor market earnings. The findings from Polette’s study suggest either that unemployed persons face significant demand constraints in seeking part-time employment, or that the disincentive effects from EMTRs are causing them to choose not to seek part-time employment.

20

As for different household groups, compared with the OECD average, replacement rates in Australia are relatively higher for couples with children than for couples with no children, and higher for couples than for single persons (Table 9).

21

The activity test specifies that unemployed persons must satisfy the Department of Social Security that they have actively sought work over a two-week period. Failure to meet the activity test results in the cessation of unemployment benefits for a specified time period (which depends on the number of previous breaches).

22

The results from estimation of the model are reported in Table 3 in the Annex. The equations are estimated using OLS on quarterly seasonally adjusted data for the period from 1979/1–1979/2 to 1996/1–1996/2. The employment/population equation was initially estimated using the change in GDP and four lags as explanatory variables. Sequential omission of the final lagged term if insignificant yields the preferred forms reported. The labor force participation equation was initially estimated with the employment/population rate and an interaction between the employment/population rate and a dummy variable for time periods in which the change in the employment/population rate was greater than zero as explanatory variables. The interaction term variable is intended to capture asymmetries in the response of the labor force participation-rate to positive and negative changes in the employment/population rate. The interaction term was insignificant in the labor force participation rate regressions for males and females, but significant in the regression for persons.

23

It is important to note that other factors apart from those specified in the model are likely to affect the employment/population rate and labor force participation rate. Changes in the employment/population rate are also likely to depend over the long run on changes in the age composition of the population, and over the medium run on real labor costs and on technological change. Changes in labor force participation will also depend on changes in the age composition of the population over the long run, and in the medium run may also depend on various factors which affect participation of particular sub-groups—for example, participation of younger persons is likely to depend on the rate of expansion of education and on incentives for undertaking further education; participation of older persons will be affected by provisions for superannuation benefits; and participation of married females will be affected by benefits which are designed to allow females with children to remain at home as careers without being disadvantaged relative to labor market participants. (For a further review of factors which affect labor supply and labor demand see EPAC, 1996b).

Australia: Selected Issues
Author: International Monetary Fund
  • View in gallery

    AUSTRALIA: Unemployment Rate, 1966767–1996/97

    (Seasonally Adjusted)

  • View in gallery

    AUSTRALIA: Unemployment and Long-Term Unemployment, 1978/79-1996/97

  • View in gallery

    AUSTRALIA: Employment and Labor Force, 1966/67-1996/97

    (Seasonally Adjusted)

  • View in gallery

    AUSTRALIA: Unemployment by Industry, 1978-96

  • View in gallery

    AUSTRALIA: Estimates of the Steady-State Rate of Unemployment, 1970-1996

    (In percent)

  • View in gallery

    AUSTRALIA: Determinants of Employment Growth, 1971/72-1995/96

  • View in gallery

    AUSTRALIA: Determinants of Employment Growth, 1971/72–1992/93

  • View in gallery

    AUSTRALIA: International Comparison by Unemployment Rates, 1972-95