Article 3 - Towards higher retirement incomes for Australians: a history of the Australian retirement income system since Federation


The Australian retirement income system, comprised of a publicly provided age pension, mandatory private superannuation savings, and voluntary private superannuation savings, has come increasingly to be viewed as a model for other countries. The age pension was introduced in 1908 and since that time has retained its character as a modest benefit provided on the basis of need. Although having undergone frequent change, the age pension means test has consistently ensured that the age pension is targeted to those most in need while remaining affordable for the Government. Similarly, the introduction of indexation has helped to maintain the value of the age pension, most notably since the 1970s.

Private superannuation first emerged for a small group of salaried employees in the nineteenth century and spread amongst white-collar employees. After several failed attempts at introducing national superannuation, private superannuation became more widely available in the 1970s through negotiation on its inclusion in industrial awards. This process accelerated under Productivity Award Superannuation, and subsequently under compulsory superannuation through the Superannuation Guarantee. In this way, the maturing superannuation system has become the vehicle for providing higher incomes in retirement for most Australian employees. At the same time, the age pension remains as an essential safety net income, ensuring that all Australians have security in retirement.


The modern Australian retirement income system comprises three elements, which have become known as the three pillars:

  • a publicly provided means tested age pension;
  • mandatory private superannuation saving; and
  • voluntary saving (including voluntary superannuation saving).

These pillars were not established on a systematic basis as part of a grand design. Rather, each pillar emerged and evolved separately.

Nonetheless, since the implementation in 1992 of compulsory superannuation (through the Superannuation Guarantee), the fundamental elements of Australia's retirement income system have come increasingly to be viewed as a model for other countries. This is reflected not least in the World Bank's effective endorsement, in its 1994 report Averting the Old Age Crisis, of Australia's three pillars as the approach which offered the best prospect of simultaneously being fiscally sustainable in an environment where the population is aging, of improving national saving, ensuring intergenerational equity and providing higher incomes in retirement.

This article charts the development of Australia's retirement income system over the last 100 years and highlights some of the key forces that influenced that development. The focus is on the development of the age pension and mandatory private superannuation saving, which, from a retirement income policy perspective, are where Government involvement has been at the forefront.

Aged care in Australia prior to Federation

Prior to Federation, welfare provision for the aged in the Australian colonies was rudimentary and dominated by voluntary charitable organisations. Generally, support was provided in the form of institutionalised care, and was motivated by a desire to alleviate only the worst poverty (Kewley, 1973, p 15). The predominant mode of care for the aged in this period was by family members.

Rapid changes in the colonies' population in the period immediately prior to Federation (including a significant increase in the proportion of the population aged over 65), combined with the effects of economic depression in the 1890s, placed the voluntary system of aged care under strain (Dixon, 1977, p 161). As a result, there was increased community acceptance of the notion that support for the aged was not solely the responsibility of the individual, but a collective responsibility for the whole community (Dixon, 1977, p 4). In this respect, recognition of a role for Government in aged support mirrored developments overseas and followed the introduction of insurance or pension schemes in Germany (1882), Denmark (1891), Belgium (1894) and New Zealand (1898).

The first age pension legislation in Australia

In 1900, New South Wales (NSW) became the first Australian jurisdiction to introduce an age pension, which took the form of a non-contributory flat rate payment funded out of general revenue. A compulsory contributory insurance scheme of the type introduced by Bismarck in Germany (and subsequently in most other continental European nations) was rejected on the grounds that it was not suited to NSW' relatively underdeveloped social infrastructure and small population (Kewley, 1973, p 33) and would leave the unskilled, casual workers, the sick, unemployed and non-working (ie married) women without coverage (Dixon, 1977, p 11).

The New South Wales pension was set at a level which the community could afford (26 per year for a single person) and was means tested against property and income to minimise costs and ensure that assistance was directed towards those genuinely in need (Dixon, 1977, p 12). Nonetheless, the pension was sufficient to keep an individual in modest comfort.

Of the other States, only Victoria and Queensland introduced age pensions prior to the introduction of the age pension by the Commonwealth Government. Both were more basic than the New South Wales scheme. The Victorian age pension, for example, was introduced at the same time as the New South Wales age pension but provided significantly less support than the NSW pension and was subject to a tighter means test (Kewley, 1980, p 9).

The Commonwealth age pension

Federation in 1901 gave the Commonwealth explicit power through the Constitution to legislate for the provision of age pensions. As a concurrent power, at the point at which a Commonwealth age pension was introduced, any age pension legislation introduced previously by the States would be overridden. However, the introduction of the Commonwealth age pension was delayed at first by the operation of section 93 of the Constitution, which required the Commonwealth to hand over any surplus revenue to the States on a monthly basis.

The passage of the Surplus Revenue Act 1908 reversed this situation by permitting the Commonwealth to retain all surplus revenue. Age pensions became the first item to be funded through revenue appropriated under this Act (Kewley, 1973, p 72), and subsequently, with the passage of The Invalid and Old Age Pensions Act 1908 (henceforth the 1908 Act), the rate of the age pension and the basic principles for its provision were established. The age pension rate was set at 26 per year, or 10 shillings per week - the same rate as the New South Wales pension. The basic principles for the provision of the age pension, while having undergone some changes in operation since that time, remain the cornerstone of the age pension today.1

Pension eligibility

The 1908 Act limited eligibility for the age pension according to character and race, age, residency and means.

The character and race provisions of the 1908 Act, which denied the age pension to people who were not of `good character' as well as Aliens (that is, non-residents), non-Australian born Asians, Australian aborigines, Africans, Pacific Islanders and New Zealand Maoris, were progressively removed over the next 60 years.

Under the 1908 Act, age pensions were payable to both men and women at age 65. In 1910, the pension age for women was reduced to 60 on the grounds that women generally became `incapacitated for regular work at an earlier age than men' (Kewley, 1973, p 75), their longer life expectancy not withstanding. These age thresholds remained unchanged until 1993, when provision was made to increase the pension age for wome
n to age 65 over a twenty-year period to bring it into line with the threshold for men.

The 1908 Act initially required an individual to have been continuously resident in Australia for 25 years before they were eligible to claim an age pension.2 However, in 1909, the residency requirement was reduced to twenty years to reflect the residency requirement of the superseded Victorian and Queensland age pensions, thereby avoiding the situation where pensioners in Victoria and Queensland with residency periods of greater than 20 years but less than 25 years would be ineligible for a pension.

Further changes to the residency test occurred in 1962 and twice in the early 1970s, largely in recognition of the needs of post-war migrants (Kewley, 1973, p 302). In 1962, the residency period was reduced to 10 years, while in 1972 the age pension legislation was amended to enable pensions to continue to be payable in overseas countries which made corresponding arrangements for their own pensions to be payable in Australia. This amendment was superseded in 1973, when age pensions granted in Australia could be paid in whatever country the pensioner later chose to live. These measures have subsequently been complemented by an extensive network of bilateral social security agreements to facilitate the transfer of pensions between Australia and other countries.

As noted above, the Commonwealth age pension has, from the outset, been subject to a means test, not only to target the pension to those most in need but also to limit its potential budgetary cost. While the essential nature of the means test (ie the assessment of a pensioner's eligibility for a pension based on their accumulated wealth) has not changed markedly since its introduction, the way in which the means test affects an individual's pension rate has undergone significant change.

The two key aspects of the means test which have been varied are the extent to which income and assets are included in the means test and the method for determining the degree to which the age pension should be removed once the means test thresholds have been exceeded.

The 1908 Act provided for the pension amount to be reduced on a one-for-one basis for every pound by which the net capital value of the pensioner's property exceeded 50. The pensioner's home was excluded from the means test (provided the pensioner lived there permanently and no income was derived from the property) until the pensioner's accumulated assets reached 100. The age pension cut out when the individual's accumulated property exceeded 310. Where both spouses received the age pension, the income and property of each was deemed to be half the total income and property of both.

It is notable that, from the outset, owner-occupied housing was at least partially excluded from the means test. In 1912, the 1908 Act was amended to wholly remove the family home from the means test. The exclusion of owner-occupied housing from the means test is a continuing feature of the Australian age pension system, and has helped to keep age pensioners (who mostly own the homes in which they live) above the poverty line (Castles, 1985, p 96).

The emphasis on self-reliance that underpinned the exclusion of the owner-occupied home from the means test was further reinforced by increases in the income and assets limits in 1946 and 1954. The Government justified the changes in the grounds that the means test was acting as a penalty upon thrift which discouraged private saving for retirement (Kewley, 1973, p 290).

In 1960, the merged means test was introduced through the Social Security Act 1960. Under the merged means test, income and property became interchangeable. The Act provided for a property component, equivalent to 1 for each complete 10 of the value of a pensioner's property above 200, to be added to their annual rate of income. This sum, called `means as assessed', thus could consist entirely of income, entirely of property, or of various combinations of income and property. The rate of pension payable in any case was to be calculated by deducting from the appropriate maximum annual rate of pension the amount by which means as assessed exceeded 182 (which was the same amount as the former permissible income).

In 1969, the merged means test was replaced by the tapered means test, under which only half of the amount by which a pensioner's means as assessed exceeded the permissible income was to be deducted from the pension. The tapered means test, by replacing the previous `one-for-one' means test, significantly increased the incidence of part-rate age pensions.

The means test was repeatedly modified in the 1970s, with the Whitlam Government initially acting to remove it for people aged 75 or over in 1973 and then for people aged 70 to 75 in 1975. The Fraser Government continued the trend in 1976 by replacing the income and assets test with a test on income only (including income from assets). However, the test was subsequently tightened again in 1978 when the income test was partially restored for people aged 70 and over. The means test was tightened further by the Hawke Government in 1985 when the assets test was reintroduced. Under the 1985 changes, either the income or assets test was applied, depending on which test gave a lower pension level.

The last fifteen years have seen several amendments to the scope of the means test, largely in response to the wide range of new financial products on offer to retirees, and to take account of the increased take-up of superannuation, initially under the Accord agreements between the Hawke Labor Government and the Australian Council of Trade Unions (ACTU), and subsequently under the Superannuation Guarantee. The most significant of these have been the introduction of income test deeming rules, the inclusion of allocated pensions in the means test and the introduction of the Deferred Pension Bonus Plan.

In 1990, income test deeming rules were introduced for pension eligibility. Under the new rules, cash and money in accounts were assessed as yielding an interest rate of nil or the actual rate on the first $2000. For amounts over $2000, a rate of 10 per cent was to be assessed, or the actual rate, whichever was the higher. Deeming was extended in the 1995-96 Budget to include financial investments, such as bank, building society and credit union accounts, cash and term deposits. However, the family home remained exempt from the rules. The deeming rates and thresholds, as at 1 July 2000, are 3.5 per cent for the first $31,600 of total financial investments held by a single age pensioner and 5.5 per cent for balances above this amount.

In 1992, allocated pensions (which are commutable) became subject to both the income and assets tests, while in 1998, non-commutable lifetime and life-expectancy annuities that met certain minimum standards became exempt from the assets test in order to encourage greater take-up of retirement income stream products.

In the 1997-1998 Budget the Howard Government introduced the Deferred Pension Bonus Plan, which offered individuals reaching pension age a positive incentive to defer retirement. Under the plan, a person who deferred age pension take-up while continuing to work a minimum of 25 hours per week, accrued a cumulative tax exempt bonus entitlement of 9.4 per cent of his or her basic pension entitlement for each full year of employment past pension qualifying age (maximum 5 years deferral).

The most recent changes to the means test have occurred in conjunction with the introduction of the Goods and Tax (GST) on 1 July 2000. In order to ensure that individuals on the age pension received their fair share of the benefits of tax reform, and to compensate recipients for the cost of living effects associated with the introduction of the GST, the Government provided a 2.5 per cent increase in the income and asset test free areas and a reduction of the income and asset test taper rate from 50 per cent t
o 40 per cent. The Government also increased the maximum rate of the age pension by 4 per cent, increased the maximum Pensioner Tax Rebate and introduced a one-off Aged Persons Savings Bonus. The 4 per cent increase in the age pension comprised of an up front advance of 2 per cent on the cost of living adjustment on 20 March 2001 and a real increase of 2 per cent (ie indexed to movements in the consumer price index).

The age pension rate

As noted above, the 1908 Act set the age pension at a rate of 26 per year, or 10 shillings per week, sufficient to keep an individual in modest comfort. The objective was thus to assist the aged poor, but to do so in a way that would not undermine self-reliance (Dixon, 1977, p 27).

Since the introduction of the age pension, the most significant legislative changes affecting pension rates have been the introduction of indexation in the 1930s, the introduction in the 1960s of different rates of pension for singles and couples and legislating the minimum age pension at a rate of 25 per cent of male average weekly earnings in 1997.

Indexation was first introduced for the age pension by the Financial Relief Act 1932, which reduced the age pension rate in response to the Depression and also introduced an annual review of the rate of pension based on the retail price index for food and groceries. Indexation was subsequently removed and the rate of the age pension was restored to its pre-depression rate by an amending Act in 1937.

Indexation was reintroduced once again in 1940, this time against the price index for household expenditure items in the six capital cities. However, indexation according to this method would have meant the payment of different rates of pension in each State, as the rate of inflation varied in each capital city, and was removed in February 1944 (Kewley, 1973, p 284).

Indexation was introduced for a third time in 1973, when the Whitlam Government introduced a provision indexing the age pension biannually against average weekly earnings. In 1976 the Fraser Government modified this so that the age pension was adjusted against the consumer price index.

In 1963, the Menzies Government introduced different pension rates for single and coupled age pensioners. This measure, which was designed to address the disadvantage faced by single pensioners who were unable to share fixed costs, set the single rate pension at a little under 10 per cent higher than the per person rate payable to a couple ($11.50 per week). However, increases in pension rates since that time have resulted in the standard rate currently being a little under 20 per cent higher than the per person rate payable to a couple.

As noted above, from the outset the overriding goal of the pension was to provide a sufficient level of support to ensure a retired person was able to live in modest comfort. Since the introduction of the 1908 Act, this view has become entrenched. However, until the Whitlam Government came to power in 1972, the benchmark for a minimum age pension had been vague. In December 1972, the then Prime Minister Gough Whitlam promised that the basic pension rate would be raised until it reached 25 per cent of average weekly male earnings. This benchmark was bolstered by a relaxation of the means test and the introduction of indexation in 1973 (see above).

However, it was a further twenty five years before the benchmark announced by Whitlam was enshrined in law by the Howard Government. The Social Security and Veterans' Affairs Legislation Amendment (Male Total Average Weekly Earnings Benchmark) Act 1997 ensured that the full rate age pension for a single adult would be maintained at a minimum rate equal to or greater than 25 per cent of male total average weekly earnings (MTAWE). The Act also set the partnered adult rate of pension at 83 per cent of the single adult rate of pension. Chart 1 outlines movements in the age pension as a share of male total average weekly earnings since 1965.

Chart 1: Single pension rate compared to MTAWE
(MTAWE benchmark basis) 1965 to 2000

Chart 1: Single pension rate compared to MTAWE (MTAWE benchmark basis) 1965 to 2000

Source: Department of Family and Community Services.

It has always been recognised that the age pension was to provide a `safety net' level of income, rather than an income at a level concomitant with an individual's pre-retirement standard of living. Hence, the future of higher retirement incomes for Australians has resided firmly in the realm of self-provision, in particular, through an extension of access to the already established private superannuation system.

Superannuation and earnings-related pensions
in Australia

Occupational superannuation first emerged in Australia in the mid-nineteenth century. Its purpose since then has changed markedly and the period prior to the introduction of mandatory superannuation in 1992 can be clearly divided into three eras. During the first era, spanning the nineteenth century to the 1940s, superannuation provided a select group of salaried employees with an independent retirement income. During the second era, from the 1950s to the 1970s, with the relaxation of means test arrangements for the age pension, superannuation acted to supplement the age pension for mostly white-collar workers. From the 1970s until the introduction of award superannuation in 1986 and subsequently mandatory occupational superannuation in 1992, superannuation was an employment fringe benefit which although more generally available, was still concentrated among professionals, managers and administrators; public sector employees; and the financial sector (Gunasekera and Powlay, 1987, p 3).

By 1974, 32.2 per cent of wage and salary earners were covered by superannuation, comprising 40.8 per cent of male wage and salary earners and only 16.5 per cent of females, with most superannuation assets located in defined benefit accounts.3

Superannuation was, and still is, a preferentially taxed investment vehicle. Changes to the tax arrangements for superannuation have, however, reduced inequity in the arrangements. Tax was first levied on superannuation in 1915. Up until 1983, superannuation contributions by employers were exempt from tax and employee contributions (as contributions from wages and salary) were taxed at the employee's marginal rate and were not taxed again when withdrawn from the fund. Investment earnings of superannuation funds were not taxed when earned, but were taxed when withdrawn from the fund.

Superannuation benefits paid as a pension were fully taxable once the income of the recipient exceeded the income tax threshold. For superannuation benefits paid as a lump sum, 5 per cent of the total lump sum benefit was included as assessable income and taxed at the individual's marginal taxation rate. In effect, this meant that the maximum rate that applied to lump sums was 3 per cent (ie the 60 per cent top marginal rate applied to 5 per cent of the lump sum).

The uneven coverage of superannuation (and associated superannuation tax concessions) provided the catalyst for the push for universal occupational superannuation, which culminated in mandatory occupational superannuation under the Superannuation Guarantee legislation of 1992. The precursors of this legislation lie as far back as 1928, with the first Federal Government inquiry into the possibility of introducing universal contributory superannuation.

The first proposals for national superannuation in Australia

Three Government inquiries into the retirement income system, in 1928, 1938 and 1976, resulted in failed proposals to introduce a universal contributory
national superannuation or insurance scheme in Australia.

In September 1923, the Bruce Government established a Royal Commission to examine the possibility of introducing a comprehensive national insurance scheme for retirement, sickness and disability (Kewley, 1973, p 143). The Royal Commission supported the Government's proposal on the grounds that it offered the prospect of improving significantly the level of security for the aged, sick and disabled. On the basis of the Royal Commission's recommendations, the Bruce Government introduced the National Insurance Bill 1928 into Parliament. The Bill required all individuals, subject to a maximum income limit, aged between 16 and 65 (60 in the case of women) to contribute to national insurance. The scheme would pay benefits for sickness, invalidism, and old age, as well as for widows and orphans (pp 144-145).

However, the Bill was strongly opposed by Friendly Societies, who were concerned that national insurance would make them redundant, and by employers, who were concerned that the compulsory insurance contributions would increase labour costs (Kewley, 1973, p 147). The Bruce Government initially postponed debate on the Bill indefinitely to engage in further consultations with interested groups and the Bill lapsed when the Government was defeated in 1929 (Kewley, 1973, p 149).

A second attempt at introducing similar legislation was made in 1938, largely in response to concerns about the growing cost of the existing non-contributory age pension. The National Health and Pensions Bill retained many of the features of the 1928 Bill but covered a wider range of benefits. The Bill was passed, despite facing similar criticisms to the 1928 Bill, including the fact that it excluded self-employed people, and was due to come into force on 1 January 1939. The introduction of the new scheme was subsequently delayed and then abandoned due to the onset of the Second World War and the increased priority given to defence in the Commonwealth Budget (Kewley, 1973, pp 159-164).

The Hancock Report into national superannuation

The Whitlam Government, after coming to power in 1972, argued the case for a national contributory earnings-related superannuation scheme as a way to improve equity and broaden superannuation coverage. The following year, the National Superannuation Committee of Inquiry was established by the Government under the Chairmanship of Keith Hancock.

The Committee's final report, handed down in 1976 after the election of the Fraser Government, contained both a majority recommendation, supported by Hancock and Richard McCrossin, and a minority recommendation, supported by Kenneth Hedley. The majority scheme was a partially contributory, universal pension system with an earnings-related supplement that would raise pension rates to a minimum of 30 per cent of average weekly earnings. Kenneth Hedley advocated a broadening of existing arrangements through a scheme encompassing a non-contributory flat rate universal pension, a means tested supplement and greater encouragement of voluntary savings through an expansion of occupational superannuation (National Superannuation Committee of Inquiry, 1976).

The majority findings of the Hancock Report ultimately were rejected in 1979 by the Fraser Government, primarily on the grounds that such a scheme would place too great a burden of contributions on low to middle income earners (Social Welfare Policy Secretariat, 1983, p 16).

It is notable that the most enduring legacy of the Hancock Report is the closer resemblance of present day retirement income arrangements to the scheme advocated by Hedley. The work of the Hancock Committee led to other suggestions for reform of the age pension, most prominently from the Life Insurance Federation of Australia (1981, revised in 1983) and in 1982 from Labor Party MP Chris Hurford. Both of these proposals adopted, in broad terms, the features of the Hedley scheme. The significance of all of these proposals for reform was the shift in emphasis for retirement income policy away from poverty alleviation through the minimalist age pension towards income maintenance through contributory superannuation.

Towards a comprehensive contributory superannuation system: productivity award superannuation

While the various proposals for a national superannuation scheme provided some policy background for universal contributory superannuation, the actual impetus for such a system came primarily through the industrial relations arena, beginning in the 1970s (Bateman and Piggott 1996).

For the union movement, occupational superannuation provided a vehicle through which members could obtain deferred wage increases in the form of retirement savings without going outside the bounds of Australia's then centralised wage fixing system. The most prominent push for worker superannuation was mounted by the Federated Storemen and Packers Union (FSPU), which developed a portable (ie industry-wide) accumulation fund for its members. The FSPU subsequently launched a concerted industrial campaign to force employers into contributing to its fund on behalf of their employees.

In 1983, the newly elected Hawke Labor Government expressed support for the principles of employee superannuation and initiated discussions with the Australian Council of Trade Unions (ACTU) on the possibility of broadening access to superannuation throughout the workforce as part of the Government's negotiated Accord with the trade unions. However, superannuation was not included as part of the first Accord.

The Hawke Government's first significant initiative in the area of superannuation was to reform the taxation of superannuation to address concerns that the small proportion of often higher income individuals who received part of their total remuneration in the form of superannuation were enjoying significantly lower effective marginal tax rates than individuals who received all of their remuneration as salary and wages.

The Government's May 1983 Economic Statement announced that for lump sums received at age 55 or later, the first $50,000 would attract a tax rate of 15 per cent, with the remainder to be taxed at a flat rate of 30 per cent. Where lump sums were received at an age below 55 years, a 30 per cent rate of tax would apply to the whole of the taxable component. These arrangements applied only to lump sum amounts relating to service after 1 July 1983. The existing taxation arrangements were preserved for benefits that related to service prior to this date.

Sustained Government support for employee superannuation was motivated not only by a desire to enhance the retirement incomes of the general workforce, but also by macroeconomic pressures. By 1985, the Government was battling inflationary pressures, relatively high real interest rates, an increasing current account deficit, foreign debt and a decline in the value of the Australian dollar. Employee superannuation was attractive to the Government in this economic climate because it offered scope for deferred wage increases and improvements in national saving. The Government was also motivated by a desire to put in place a clear framework for retirement income policy, rather than to allow a patchwork set of ad hoc arrangements to develop.

The process of institutionalising employee superannuation began in September 1985 when, with the support of the Government, the ACTU, as part of its National Wage Case claim with the Conciliation and Arbitration Commission, sought a 3 per cent superannuation contribution by employers to be paid into an industry fund. Subsequently, Accord Mk II negotiated between the Government and the unions stipulated that compensation to employees should be 6 per cent, to keep pace with inflation, to be comprised of a 3 per cent employer superannuation contribution, a 2 per cent wage rise and tax cuts to take effect from 1 September 19
86 (Bateman and Piggott 1996).

In supporting the unions' claim, the Government argued before the Commission that the implementation of the productivity award in the form of superannuation would provide significant benefits not only to employees, but also to employers, who would not incur the additional on-costs associated with a direct wage increase, such as workers' compensation and payroll tax. The Government also argued that productivity award superannuation would mean greater certainty and control over the growth of labour costs than if unions were to pursue their superannuation claims on an individual uncoordinated basis.

The superannuation agreement between the Government and the unions was endorsed by the Conciliation and Arbitration Commission in February 1986, albeit with reservations concerning the implementation process. The Commission announced that it would approve industrial agreements that provided for contributions to approved superannuation funds resulting from wage increases of up to 3 per cent. Ratification of these agreements would give them status equivalent to an award and make them enforceable under the Conciliation and Arbitration Act.

Employer groups, notably the Confederation of Australian Industry, challenged the Commission's decision in the High Court on the grounds that the Commission did not have the jurisdiction to rule on any claim relating to the payment of superannuation benefits as part of an industrial award as superannuation was not an industrial matter within the meaning of section 51 (xxxv) of the Constitution.4 However, on 15 May 1986, the High Court ruled that worker superannuation was an industrial matter and the Commission did have the jurisdiction to arbitrate on superannuation matters, paving the way for the unions to pursue superannuation award claims on behalf of employees (Dabscheck 1989, p 99).

In conjunction with the agreement between the Government and the ACTU on Accord Mk II, which led to the introduction of productivity award superannuation, the Government agreed to tighten the prudential framework for superannuation. A condition of the Conciliation and Arbitration Commission's June 1986 National Wage Case decision to include superannuation within the Commission's wage fixing principles was that new industry superannuation schemes would be required to conform to the Commonwealth's operational standards for superannuation.

Historically, the prudential regulation framework for superannuation rested broadly on the principles of trust law supplemented by prudential controls in the Life Insurance Act, certain aspects of the Corporations Law and the Income Tax Assessment Act 1936 (ITAA36). Prudential issues were primarily the responsibility of superannuation fund trustees, who were personally liable to fund members for the management of the superannuation savings in their care.

In December 1985, the Government released for comment a set of draft operational standards that superannuation funds would be required to meet in order to be eligible for superannuation tax concessions under ITAA36. Subsequently, on 21 December 1987, the Government introduced the Occupational Superannuation Standards Act 1987 (OSSA).

OSSA prescribed operating standards for:

  • the vesting of benefits arising from employer and employee contributions (ie full vesting of an individual's employer and member contributions, plus earnings, less the costs of administration and any death or disability insurance cover);
  • the preservation of superannuation benefits until age 55;
  • greater member involvement in the control of superannuation funds (eg through equal representation of employees and employers on the trustee board of superannuation funds with 200 or more members); and
  • the security of members' benefits (eg requirements that funds seeking taxation concessions must lodge annual returns with the Insurance and Superannuation Commission certifying compliance with relevant provisions in the Income Tax Assessment Act and other operating standards).

As new industrial awards were progressively negotiated according to the guidelines established by the 1986 National Wage Case, there was a rapid increase in superannuation coverage so that in the 4 years after the introduction of award superannuation, superannuation coverage grew from around 40 per cent of employees to 79 per cent. In the private sector, coverage grew from 32 per cent in 1987 to 68 per cent in 1991 (Bateman and Piggott 1996).

As superannuation became more widespread, the cost to revenue of superannuation tax concessions increased. In May 1988, the Hawke Government released its statement Reform of the Taxation of Superannuation, which contained measures to bring forward the payment of superannuation taxation liabilities by introducing a tax on contributions but reducing the tax on benefits. These taxation arrangements remain in place today.

From 1 July 1988, a 15 per cent tax was applied to all employer contributions and deductible contributions. This tax on contributions was levied when the contribution was received by the superannuation fund (by including the relevant contribution in the assessable income of the fund). The 15 per cent tax also applied to deductible contributions to approved deposit funds and income from the superannuation and rollover annuity business of life insurance companies, friendly societies and other registered organisations.

A 15 per cent tax was also applied to the investment income of superannuation funds derived from 1 July 1988, including approved deposit funds and rollover annuity funds. However, the new tax on investment earnings could be offset by the allowance of full imputation credits for franked dividends received from Australian companies and credits for dividend and interest withholding tax on income received from foreign sources. Previously, imputation credits had not been allowable to superannuation funds.

Finally, subject to transitional rules, the tax on lump sum benefits paid from taxed superannuation funds was reduced:

  • the first $60,000 of a lump sum payment (indexed to the rate of growth of average wages) was exempted from tax if the lump sum was preserved to age 55 or later, with the remainder of the payment subject to a tax rate of 15 per cent; and
  • where lump sums were taken prior to age 55, the tax rate was phased-down from 30 to 20 per cent.

To avoid the imposition of a new tax on a retrospective basis, the taxation treatment of the pre-1983 component of retirement benefits and amounts accumulated between 1 July 1983 and 30 June 1988 remained unchanged. Grandfathering of this nature (which was also a feature of the 1983 amendments to the taxation of superannuation) has added to the complexity of superannuation taxation arrangements.

In spite of the rapid growth in superannuation coverage for employees as a result of productivity award superannuation, by 1991, nearly one-third of private sector employees remained uncovered. Moreover, the existing 3 per cent award was too small to provide any significant improvement in retirement incomes for all but the highest paid workers (Borowski, 1991, p 35).

In addition to the gaps in coverage, the Government became aware that several tensions had emerged in the superannuation system following the implementation of award superannuation, largely as a result of the piecemeal implementation process. Firstly, many employees who were entitled to superannuation under the provisions of an award were not paid their entitlement: in some cases because they were confused about the nature of their entitlement, while in others, unions had not established industry funds into which employer contributions could be paid. Secondly, compliance with the terms of the National Wage Case could only be enforced through a case mounted with the Conciliation and Arbitration Commission, which was a v
ery laborious process given the number of awards which fell under the Commission's jurisdiction. Finally, award superannuation did not effectively take account of the significant number of employees who already enjoyed superannuation rights as part of their employment. Any attempt to differentiate between employees who already had superannuation and those who did not within a new or renegotiated award threatened to increase significantly the complexity of the awards themselves.

In 1991, the compliance problems associated with award superannuation prompted the Conciliation and Arbitration Commission, now renamed the Industrial Relations Commission, to reject an application, supported by both the ACTU and the Labor Government, for a further 3 per cent superannuation increment. The claim was rejected because of perceived shortcomings in the implementation and administration of award superannuation, which the Commission felt were evident to the extent that unless addressed, any further development of the system `may be flawed to the point of frustrating its contribution to the achievement of an adequate national retirement incomes system' (Australian Industrial Relations Commission, 1991).

Instead, the Commission recommended that the Government convene a national conference on superannuation involving all relevant parties to consider issues such as non-compliance; the extension of award superannuation to all awards, including state awards; building more flexibility into award-based superannuation; the extension of superannuation to casual and part-time employees; and the role of the Commission in the long-term agenda for ensuring appropriate retirement incomes.

The Government opted not to act on the Commission's recommendations. Instead, in the 1991-92 Budget, the Government announced that it would introduce a mandatory superannuation system through the implementation of the Superannuation Guarantee.

The Superannuation Guarantee

In the Budget delivered on 21 August 1991, the then Treasurer, John Kerin, announced that from 1 July 1992, under a new system to be known as the Superannuation Guarantee (SG), employers would be required to make superannuation contributions on behalf of their employees. Moreover, if the required amount of superannuation support was not provided, the employer would be liable for a non-deductible Superannuation Guarantee Charge, equivalent to the individual employee shortfall in contributions, an interest component and an administrative charge. The Federal Government's taxation powers under the Constitution provided the mechanism for collecting the Charge.

The Government stated that the SG would provide:

  • a major extension of superannuation coverage to employees not currently covered by superannuation;
  • an efficient method of encouraging employers to comply with their obligation to make contributions on behalf of their employees; and
  • an orderly mechanism by which the level of employer superannuation support could be increased over time, consistent with the Government's retirement income policy objectives and the economy's capacity to pay.

In this way, in conjunction with the age pension, the Superannuation Guarantee would help to meet the challenges of Australia's aging population, the significant trend towards earlier retirement and changing community views about what level of retirement income is adequate (Security in Retirement, 1992).

The Budget announcement provided for an initial level of prescribed superannuation support of 5 per cent of Ordinary Time Earnings (OTE) for employees of employers whose annual payroll exceeded $500,000 and 3 per cent of OTE for other employees.5 The Government also announced a program to progressively raise the minimum level of support, to reach 9 per cent of OTE by 2000-01.6 By collecting superannuation contributions from employers, administrative costs were significantly reduced, as the Australian Taxation Office would only be required to oversee the payment of SG contributions from 500,000 employers instead of more than 5 million employees.

The coverage of the system was very broad, utilising wide definitions of employer and employee.7 The SG legislation also provided for very few exemptions, the most prominent being for employees earning less than $450 per month, part-time employees under 18 years of age and employees aged 65 or over. The self-employed were also excluded from the requirement to make mandatory superannuation contributions through the SG. However, the self-employed have been able to access the superannuation system by making voluntary contributions or by rolling over the proceeds from the sale of their business into superannuation at retirement.8

In terms of superannuation coverage, the SG further expanded the gains in coverage made under award superannuation. As noted above, by 1991 coverage had grown to 70 per cent of private sector employees. In the first year of operation of the SG (1992-1993) 80 per cent of public and private employees were covered. By 1999, this had grown to 91 per cent (Table 1).

Table 1: Superannuation Coverage in Australia

Table 1: Superannuation Coverage in Australia

Source: Insurance and Superannuation Commission, Superannuation Bulletin; Australian Prudential Regulation Authority, Superannuation Trends (various years).

Award superannuation and the SG also brought about a switch in the type of superannuation coverage members enjoyed. In 1982-83, 82 per cent of fund members were covered by defined benefit schemes. As most industry funds established to accept productivity award contributions were accumulation funds, award superannuation and, subsequently, compulsory contributions under the SG, completely reversed this trend. Thus, in 1991-92, 73 per cent of members were in accumulation funds, and by 1999-2000 this figure had risen to 86 per cent (Table 2). With expanding superannuation coverage, most working Australians could look forward to higher incomes in retirement.

Table 2: Proportion of members in defined benefits and accumulation funds

Table 2: Proportion of members in defined benefits and accumulation funds

Source: Insurance and Superannuation Commission, Superannuation Bulletin; Australian Prudential Regulation Authority, Superannuation Trends (various years).

The Superannuation Industry (Supervision) Act 1993

With the mandating of superannuation contributions under the SG came an additional responsibility on Government to build confidence in the superannuation system and to ensure that superannuation funds managed the assets in their care to maximise retirement benefits and superannuation savings were adequately preserved for use in retirement. Thus, the introduction of the SG was accompanied by further reform of the prudential rules governing superannuation funds and a tightening of the preservation rules concerning access to superannuation benefits.

These reforms in part reflected the findings of a Senate Select Committee Report on the security of superannuation funds, titled Safeguarding Superannuation and were outlined on 21 October 1992 in the Keating Government's Economic Statement, Strengthening Super Security.

Under the Government's proposal for an enhanced prudential framework, tax concessions would only be granted to schemes which had a responsible entity (generally the trustee) who was a corporation within the meaning of the Constitution, or which had the provision of age pensions within the meaning of the Constitution as its substantial or dominant purpose.

The Government stated that this approach would facilitate m
ore effective supervision of the industry by allowing the introduction of specific penalties directed at particular offending parties, without the need to determine a fund as non-complying (as required under the OSSA legislation).

The Statement also announced a number of other new and improved regulatory measures. These included:

  • clearly defining the basic duties and responsibilities of trustees and ensuring that they have adequate powers to carry out these responsibilities;
  • improved disclosure requirements applying from 1 July 1992 and improved requirements for equal representation on trustee boards for funds with 200 or more members;
  • measures to increase the role performed by auditors and actuaries, including improved audit standards for superannuation funds, requirements for disclosure where a fund has not acted to address previously notified concerns and/or any actual breach by trustees of OSSA requirements; and
  • the introduction of more direct enforcement powers and improved audit resources for the Insurance and Superannuation Commission.9

These measures were outlined in a new piece of legislation, the Superannuation Industry (Supervision) Act 1993 (SIS Act), which effectively replaced the OSSA legislation from 1 July 1994.

Further refinement of the SG system

As the SG made superannuation more widely available, in particular to low-income and part-time workers, the impact of fund fees and charges on the small account balances held by some workers became more noticeable. In response, the Government introduced, from 1 July 1995, new member protection rules, which prevented fund fees and charges on accounts with balances of less than $1,000 from exceeding any investment gains in those accounts.

At the same time, the Government also established a small accounts collection system, the Superannuation Holding Accounts Reserve, to accept SG amounts from employers where the employer was unable locate a superannuation fund that would accept small or one-off superannuation contributions. In this way, small contributions could be held outside of the formal superannuation system pending their accumulation into an amount sizeable enough to be sustained by investment earnings in a superannuation fund.

In the 1995-96 Budget, the Keating Government announced reforms designed to enhance superannuation's contribution to Australia's national savings. The reforms, which were contained in the Treasurer Ralph Willis' statement `Saving for our Future', contained two elements, namely Government support for additional employee contributions to be phased-in in conjunction with wage increases negotiated through enterprise agreements or awards; and a matching, means-tested, Government co-contribution. This proposal was first raised by the Government in 1992 in its statement introducing the SG, `Security in Retirement' and was to be funded by diverting foreshadowed tax cuts directly into superannuation. However, these measures were not implemented before the Keating Government lost office in 1996.

In its first Budget delivered in August 1996, the new Howard Government introduced a surcharge on superannuation contributions made by high-income individuals. The aim of the surcharge was to make the taxation of superannuation fairer by reducing the taxation bias in favour of high-income earners. Under the scheme, a surcharge of 15 per cent was imposed on employer superannuation contributions made after 20 August 1996 on behalf of individuals with an income (including deductible superannuation contributions) of $85,000 or more. The surcharge was phased in over an income range of $70,000 to $85,000, to be indexed annually. For the 2000-01 year, the phase-in range for the surcharge is $81,493 to $98,955.

In addition, the Government also announced the introduction of:

  • from 1 July 1997, an 18 per cent rebate for up to $3000 of contributions made by an individual on behalf of his or her low income spouse,10 designed to be of particular benefit to women outside the paid workforce who were not covered by the SG;
  • an extension of the upper age limit at which superannuation contributions could be made from 65 to 70, in recognition of the increased tendency for older Australians to work past retirement age;
  • retirement savings accounts (introduced in May 1997 by the Retirement Savings Accounts Act 1997)11; and
  • an opt-out from the SG for workers with incomes between $450 and $900 a month, who could take the equivalent amount in salary.12

In the 1997-98 Budget the Howard Government announced the introduction of a 15 per cent savings rebate, to apply to the first $3,000 of taxable income from all forms of savings, including undeducted superannuation contributions. The rebate was to be introduced in two stages, with a 7.5 per cent rebate to apply from 1 July 1998 and a full 15 per cent rebate from 1 July 1999. However, the rebate was discontinued with the announcement of the Government's major tax reform package, which provided substantial cuts in personal income tax.

The Coalition Government has also strengthened the preservation rules, so that all superannuation contributions from 1 July 1999, including personal contributions and earnings, generally are preserved until preservation age (currently 55, but progressively increasing to 60). These arrangements replaced measures proposed by the former Labor Government in 1994 to counteract double dipping and excessive leakage from the superannuation system, which had been twice deferred.

The early release of superannuation benefits had previously been allowed under a variety of circumstances, including: retirement due to permanent incapacity; permanent departure from Australia; where a person ceased employment with a preserved benefit in a fund of $500 or less; in cases of severe financial hardship; and on compassionate grounds. The current rules allow for early release subject to the governing rules of the fund: where the person's preserved benefits in the fund are less than $200; if the benefits are taken in the form of a lifetime non-commutable pension or annuity; or if the benefits are restricted non-preserved benefits (which largely arise from undeducted member contributions made prior to 1 July 1999). There is also scope for benefits to be released on compassionate grounds in certain limited circumstances or in cases of severe financial hardship (determined on the basis of strict objective criteria).

A distinguishing characteristic of the Australian superannuation system is the large number of self-managed superannuation funds (SMSFs). In September 2000, there were more than 214,000 such funds holding more than $70 billion in assets (APRA 2000). Self-managed funds were previously called excluded funds and were established to allow the self-employed and small businesses to maintain their own cost-effective superannuation accounts. While they were generally subject to the SIS Act, some requirements were relaxed in recognition of the close relationship between trustees and members (Financial System Inquiry, 1997, p 333).

In 1999, the SIS Act was amended by the Superannuation Legislation Amendment Act (No. 3) 1999 (SLAA3) to establish a new category of small superannuation fund, to be regulated by the Australian Taxation Office. The amendments, which had their origins in the recommendations flowing from the Financial System Inquiry, under the Chairmanship of Stan Wallis, replaced the previous definition of an excluded superannuation fund with a new definition of a self-managed superannuation fund. In addition to continuing to require the fund to have fewer than five members, the definition required all members of the fund to be trustees of the fund or directors of the body corporate trustee. The new regulations also ensured that employees could no
t be a member of their employer's self-managed superannuation fund except where they were relatives of the employer. In general, however, SLAA3 imposed a less onerous prudential regime on SMSFs than SIS did because all members were directly involved in the management of the fund and therefore were better able to protect their own interests.

The Government has also sought to advance two further policies announced prior to the election in 1996, namely the distribution of superannuation on separation or divorce, and choice of superannuation fund. The Family Law Legislation Amendment (Superannuation) Bill 2000 was introduced into Parliament on 13 April 2000 and aims to allow separating couples to split the value of their superannuation by agreement, or by Court order.

The choice of funds legislation, which aims to increase competition and efficiency in the superannuation sector by giving employees choice as to where their superannuation contributions should be paid, was first introduced into Parliament on 4 December 1997 as part of the Taxation Laws Amendment Bill (No. 7) 1997, which became the Taxation Laws Amendment Bill (No. 3) 1998 (TLAB3). During the Senate debate of TLAB3, the Government removed choice of funds from that Bill and on 28 May 1998, reintroduced it as Superannuation Legislation Amendment (Choice of Funds) Bill 1998. This Bill expired when Parliament was dissolved following the calling of the election, and was reintroduced into the House of Representatives on 12 November 1998. This Bill is currently awaiting passage in the Senate.


The origins of the present age pension and superannuation systems lie in the period prior to Federation. Both systems have evolved in the century since Federation to meet the changing needs and aspirations of Australian retirees, and the system has come to be viewed as a model for other countries. In the case of the age pension, this has largely been without any substantial change in its character as a modest benefit provided on the basis of need. Indeed, since its introduction, the age pension's capacity to deliver security in retirement for those in need has been improved, most notably through the introduction of indexation and the enshrinement in law of the single rate age pension at 25 per cent of MTAWE. The age pension has proved to be an enduring, effective and affordable instrument for providing a modest safety net retirement income for the aged.

Over this same period, the superannuation system has undergone major transformation since beginning as a benefit available to a small group of salaried employees. After spreading gradually among white-collar employees, superannuation became more widely available in the 1970s through negotiation on its inclusion in industrial awards. This process accelerated under Productivity Award Superannuation and, subsequently, under compulsory superannuation through the Superannuation Guarantee. Today, superannuation is available to more than 90 per cent of employees.

The maturing superannuation system, comprising both mandatory and voluntary superannuation savings, has become an important vehicle for providing higher incomes in retirement for most Australian employees, including the self-employed. At the same time, the age pension remains as an essential safety net income ensuring that all retirees are able to have security in retirement. Australia's three-pillar system, at the end of the first century of Federation, is truly a system for all Australians.


Australian Industrial Relations Commission (1991) `National Wage Case Decision April 1991' Commonwealth Arbitration Reports 1991/4: pp 204-286.

APRA (2000) Superannuation Trends Australian Prudential Regulation Authority, Sydney.

Bateman, Hazel and John Piggott (1996) `Private Pensions in OECD Countries: Australia' University of New South Wales, Sydney.

Borowski, Allan (1991) `The Economics and politics of Retirement Incomes Policy in Australia' International Social Security Review Vol. 44 (1/2): 27-40.

Castles, Francis G. (1985) The Working Class and Welfare: Reflections on the Political Development of the Welfare State in Australia and New Zealand Allen and Unwin, Sydney.

Dabscheck, Braham (1989) Australian Industrial Relations in the 1980s Oxford University Press, Melbourne.

Dixon, John (1977) Australia's Policy Towards the Aged: 1890-1972 Canberra College of Advanced Education, Canberra.

Financial System Inquiry (1997) Financial System Inquiry Final Report Australian Government Publishing Service, Canberra.

Gunasekera, Michelle and John Poway (1987) `Occupational Superannuation Arrangements in Australia' Social Security Review Department of Social Security, Canberra.

Kewley, T.H. (1973) Social Security in Australia, 1900-72 Sydney University Press, Sydney.

Kewley, T.H. (1980) Australian Social Security Today: Major developments from 1900 to 1978 Sydney University press, Sydney.

National Superannuation Committee of Inquiry (1976) A National Superannuation Scheme for Australia: Final Report of the National Superannuation Committee of Inquiry Australian Government Publishing Service, Canberra.

`Security in Retirement: Planning for Tomorrow Today' (1992) Australian Government Publishing Service, Canberra.

Social Welfare Policy Secretariat (1983) `Approaches to National Superannuation' Social Welfare Policy Secretariat, Canberra.

World Bank (1994) Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth The World Bank, Washington.

1 It should be noted that an age service pension, which is essentially akin to the age pension, has been available to war veterans since World War I (Kewley, 1980, p. 183). Administered by the Repatriation Commission (supported by the Department of Veterans' Affairs) the age service pension is payable to eligible veterans, their partners, and widows and widowers 5 years earlier than the age pension, recognising that the effects of war may be intangible and result in premature aging and loss of earning power. The Repatriation Commission was formed on 1 July 1920 under the Australian Soldiers' Repatriation Act 1920. In 1986, this Act and several other related Acts were replaced by the Veterans' Entitlements Act 1986, which continued the existence of the Repatriation Commission.

2 Continuous residence was not to be interrupted by occasional absences not exceeding one-tenth of the total period of residence.

3 Under defined benefit schemes, the end benefit is calculated according to a fixed formula which takes into account pre-retirement salary and the number of years of fund membership. In 1982-83 an Australian Bureau of Statistics survey found that 82 per cent of fund members were covered by defined benefit schemes.

4 Section 51 (xxxv) of the Constitution limits the Commonwealth's jurisdiction to conciliation and arbitration for the prevention and settlement of industrial disputes extending beyond the limits of any one state.

5 The minimum level of superannuation support was to be determined as a percentage of the employee's notional earnings base and was subject to transitional arrangements. In the implementation phase, an employee's earnings base depended on whether the employer was providing superannuation on 20 August 1991. Employers who were contributing to a scheme on behalf of their employees on that date were permitted to use the earnings base that applied to those contributions under the SG. Employers who were not contributing on that date were required to use an earnings base specified under an industrial award or an earnings base not less than OTE. The maximum earnings base for contributions was se
t at $80,000 (indexed annually to AWOTE). In 2000-01, the maximum contribution base for SG contributions is $26,300 for each quarterly period, or $105,200 per year.

6 This timetable was subsequently amended so that the SG phase-in began at 3 per cent contributions for employers with a payroll of $1 million or less, and at 4 per cent contributions for employers with a payroll greater than $1 million. In addition, the transitional period was extended so that mandated contributions would reach a maximum of 9 per cent in 2002-03.

7 The Superannuation Guarantee Administration Act 1992 contained special provisions to extend the definition of employer and employee beyond their ordinary meanings to include company board members, contractors, parliamentary representatives, performing artists and professional sportspeople, Commonwealth, State and Territory office holders and local government councillors.

8 Treasury's Retirement Income Modelling Unit estimates that the participation rate of the self-employed in the superannuation system has hovered between 25 per cent and 30 per cent throughout the 1990s.

9 The Insurance and Superannuation Commission was replaced by the Australian Prudential Regulation Authority on 1 July 1998.

10 The maximum rebate is available to a spouse with an annual assessable income of $10,800. The maximum rebate phases out on a dollar for dollar basis so that it is no longer available when the spouse earns $13,800 or more per annum.

11 Retirement savings accounts were designed to be a low-cost, flexible alternative to superannuation accounts which could be offered by banks, credit unions, building societies and life offices.

12 The SG opt-out is yet to be implemented.