Australia's Future Tax System

Retirement Income Strategic Issues Paper

2. Australia's three-pillar system

2.1 Overall goals of the retirement income system

The Retirement income consultation paper (AFTS 2008) identifies five objectives of a retirement income system, namely:

  • it should be broad and adequate, in that it protects those unable to save against poverty in their old-age and provides the means by which individuals must or can save for their retirement;
  • it should be acceptable to individuals, in that it considers the income needs of individuals both before and after retirement, is equitable and does not bias inappropriately other saving decisions;1
  • it should be robust, in that it deals appropriately with investment, inflation and longevity risk;
  • it should be simple and approachable, in that it allows individuals to make decisions which are in their best interests; and
  • it should be sustainable, in that it is financially sound and detracts as little as possible from economic growth.

These criteria are interconnected and need to be applied in a holistic manner when evaluating the performance of the retirement income system.

The Retirement income consultation paper sought feedback on these objectives. Most submissions support them, either explicitly or implicitly. There is particularly strong support for the proposition that the system should be adequate and sustainable.

Section 4 provides an assessment of the retirement income system against these objectives. It finds that the present three-pillar architecture broadly addresses these objectives in a reasonable, balanced way. The architecture also provides flexibility and sharing of risk to face the challenges of the substantial but uncertain economic and social changes of the 21st century and the impact of demographic change. However, tax concessions are not distributed equitably; there are insufficient products to insure against a person outliving their assets; and the system's complexity affects the individual experience of the retirement income system.

2.2 Australia's three-pillar retirement income system

Over the past 100 years, the retirement income system has evolved from one where the Age Pension primarily fulfilled a poverty alleviation function, to one where a combination of public and private provision meets a broad range of retirement income needs.

The three–pillar structure of the Australian retirement income system — the Age Pension, compulsory saving through the superannuation guarantee and voluntary superannuation saving — is unusual among developed countries, but it has considerable strengths.2 In particular, it provides a system intended to satisfy the minimum needs of all Australians, provides the capacity for individuals to enhance their retirement income, and spreads risks between the public and private sectors in a fiscally responsible way.

Although the three pillars are described as one system, distinctions between the three pillars exist and are reflected in the form of the retirement benefit, the coverage of the three pillars, an individual's exposure to risk and whether a residual payment is made to a person's estate on their death. Table 2.1 summarises these differences.

Table 2.1: Characteristics of Australia's three pillar retirement income system

  Form of benefit Level of benefit How funded Coverage Effective tax rate Coverage of longevity, investment and inflation risk Residual value at death
Age Pension Income Depends on marital status and subject to means tests Current taxpayers Universal(a) for a resident subject to means tests Means tested based on income and assets. Subject to personal income tax with relief through offsets. Yes, payable for life with wage indexation No
Superannuation guarantee An asset which can be converted to income Depends on rate, salary/wage, returns and period in workforce Employer contributions(b) Employees, but upper cap on superannuation guarantee contributions exists. Exemptions also apply. Flat 15 per cent on contributions and 15 per cent on earnings(c) Depends on how people choose to take their benefit(d) Yes
Voluntary superannuation(e) An asset which can be converted to income Depends on amount invested and returns Personal/employer contributions

Government co-contribution if eligible
Work tests apply from age 65 years. The amount that can be contributed is capped. Tax of 0 or 15 per cent depending on type of contribution, and 15 per cent on earnings(c) Depends on how people choose to take their benefit(d) Yes
  1. A person must be a resident for at least 10 years before becoming eligible for the Age Pension. However, there are rules which can provide for a shorter period.
  2. Although these contributions are made by the employer, the incidence is likely to fall on the employee through lower real wages.
  3. Benefits paid from a taxed superannuation fund to a person aged 60 years or older are tax-free. Earnings on assets supporting an income stream are tax exempt. Funds are eligible for imputation credits and a one-third capital gains tax reduction on assets held for at least 12 months.
  4. In principle, people can insure against these risks by purchasing an indexed lifetime annuity. The individual bears more risk with other types of income streams such as allocated pensions.
  5. Significant voluntary saving also occurs outside the superannuation system including owner-occupied housing, other property, financial assets and business assets.

2.3 The fiscal cost of the three-pillar architecture

The budgetary cost of the three-pillar architecture is influenced by a range of key policy parameters including: the level of the Age Pension; means testing of the Age Pension; the ages people can access the Age Pension and their superannuation; the rate of the superannuation guarantee; the taxation and other concessions provided to superannuation contributions; caps on contribution concessions; and the taxation of superannuation earnings.

Governments need to set these parameters on a long term basis to balance the requirements of adequacy and fiscal sustainability. Policy changes to vary any of these parameters will affect the long term work, consumption and saving decisions made by individuals over the course of their lifetime.

Adjustments to settings in the retirement income system should take account of these interactions across the broader system and reflect a long term strategic view about the structure of the system and its fiscal sustainability.

2.4 Retaining the existing three-pillar architecture

The strengths of the existing three-pillar system should be preserved. Not only does this system spread the responsibility and risk of providing retirement incomes in a fiscally sustainable way, it is also a structure that is likely to be durable and relevant across a broad range of economic, demographic and social outcomes.

Retirement arrangements involve very long term planning horizons and there is considerable merit in avoiding inessential large changes.

The Age Pension

The means tested Age Pension is a particular strength of the system. Its purpose is to ensure that all Australians have access to a safety net level of income throughout their retirement that is adequate to provide a reasonable minimum standard of living. It substantially underpins the retirement incomes of most low to middle income earners. It supports people who live longer than expected and exhaust their private savings, and it supports those who have less than average full-time employment due to periods of unemployment, caring responsibilities, working part-time or spending part of their working life overseas. Being taxpayer-funded, the Age Pension provides protection against investment, inflation and longevity risk.

The Pension Review addressed separately the adequacy of the Age Pension.

The superannuation guarantee

The superannuation guarantee was motivated to address a particular consequence of life-cycle 'myopia' — specifically, people not saving adequately for retirement because it is too far in the future for them to adequately 'see', and so make adequate provision for their needs. It is a private saving and asset accumulation vehicle that contributes to the improved wellbeing of employees in retirement. It enables employees to achieve a level of retirement income above that provided by the Age Pension, with the extent of the increase affected, as for all savings, by the means tests. However, it does not apply universally; the individual bears some or all of the investment risk and there is no requirement that accumulated funds be applied to fund a retirement pension. Therefore, the superannuation guarantee has not been designed to support specific income replacement goals and, in the Panel's view, cannot be made to do so.

Even when mature (toward the end of the 2030s), the current rate of compulsory contribution by employers of 9 per cent will not be sufficient on its own to meet everyone's retirement income aspirations. However, together with the Age Pension, the superannuation guarantee is expected to provide the opportunity for people on low to average wages with an average working life of 35 years to have a substantial replacement of their income.

For example, on a set of assumptions detailed in Appendix F, as a consequence of the first and second pillars only, a worker on median income of 75 per cent of average weekly ordinary time earnings (AWOTE)3 might be expected to have a replacement rate4 of about 73 per cent. A worker on AWOTE might be expected to have a replacement rate of about 63 per cent.

Whilst obviously a matter for judgment, the Panel considers that, for most employees on low to middle incomes, the 9 per cent superannuation guarantee rate can provide a reasonable balance between before and after retirement incomes, at least insofar as compulsory arrangements should require. It is, of course, open to any individual to save more. In addition, adoption of the recommendations in this report and those of the Pension Review would have the effect of increasing the replacement rates achieved by the existing 9 per cent rate of compulsory superannuation saving.

Several submissions to the Panel have argued that replacement rates for low to middle income workers should be raised further. The Panel notes, in particular, the proposal in some submissions to increase the compulsory saving rate to 12 per cent (either through increasing the superannuation guarantee, mandatory employee contributions or a 'soft' compulsion arrangement). An increase in compulsory saving would increase potential retirement incomes. However, it would also reduce an employee's pre-retirement income. Low to middle income earners, who are typically unable to offset the impact of increased compulsory saving by reducing voluntary saving, could be expected to experience larger reductions in pre-retirement consumption opportunities than those experienced by higher income earners. For those who undertake other saving it would mandate a greater proportion of that saving in the form of superannuation. An increase in the superannuation guarantee would also have a net cost to government revenue even over the long term (that is, the loss of income tax revenue would not be replaced fully by an increase in superannuation tax collections or a reduction in Age Pension costs).

At higher levels of pre-retirement income, the Age Pension and the superannuation guarantee provide lower replacement rates. However, the retirement incomes of higher income workers provided by the Age Pension and the superannuation guarantee are higher than those of lower income workers. For example, as a consequence of the first and second pillars only, a worker on AWOTE might be expected to have a disposable retirement income equivalent to about $43,000 per year in today's dollars5; whereas, a worker on 2.5 times AWOTE might be expected to have a disposable retirement income equivalent to $60,000 per year. There may well be a case for such a person seeking a higher retirement income, but the case for the government mandating that outcome is much less clear.

On average, employees on incomes of 2.5 times AWOTE contribute around an additional 10 per cent of their income to superannuation under the third pillar (through salary sacrifice arrangements), providing an expected income replacement rate of about 95 per cent when all three pillars are taken into account. Accordingly, the public policy case for insisting on higher levels of compulsory retirement saving for workers on above average incomes is relatively weak.

Some submissions have noted that older workers have not had the benefit of the superannuation guarantee for a full working life and could, as a consequence, have relatively low replacement rates. However, it is much less likely that older workers generally suffer from 'myopia' with respect to saving for retirement. Accordingly, the Panel does not consider that there is a public policy case for increasing the rate of compulsory saving of older workers.

The level of retirement incomes provided by the Age Pension and the superannuation guarantee would be increased further by adopting the recommendations in Section 3 to increase the age at which a person can access their superannuation, the recommendation to increase the Age Pension age, and any increase in pensions arising from the findings of the Pension Review.

The superannuation guarantee, since its inception, has not been applied to business income. The reluctance to extend the superannuation guarantee to small business recognises the diversity in the small business sector, its varying capital, liquidity and investment needs. While small business people and the self-employed should make provision for their retirement, the costs of compulsion may be higher in this sector than for employees. Many small business people have alternative strategies for saving for their retirement, often with different time profiles than those applying to employees. Including small business people would also be administratively difficult and add to the complexity of the system. Accordingly, the Panel recommends against extending the superannuation guarantee to small business people.

However, there can be a fine line between those who are self-employed and those who are performing duties similar to an employee. This issue arises in a number of areas of policy. The review will give further consideration to this issue in its final report, including whether there is scope and need to extend the superannuation guarantee to include with greater clarity and certainty, contractual arrangements that are close in nature to an employer-employee relationship.

Another exclusion from the superannuation guarantee system is for people earning less than $450 per month. Several submissions propose that this threshold be abolished. However, there are significant differences in the type of work people may do and applying the superannuation guarantee from the first dollar of income may not be appropriate or cost-effective in many cases. The Panel is of the view that a simple threshold should continue to apply to ensure that the compliance costs to the employer of providing the contribution are outweighed by the benefits to the employee. On this basis, the current $450 per month threshold provides an appropriate means of balancing these costs and benefits.

Many people have less than an average working life of 35 years of full-time employment due to periods of unemployment, caring responsibilities (in particular, women who may have a number of periods out of the workforce caring for others), working part-time or entering the workforce later than the average worker (for example, migrants). This affects the amount of compulsory superannuation savings they can expect to accumulate by the time they retire.

In the Panel's view, increasing the rate of the superannuation guarantee is an inappropriate way to increase the retirement savings of these groups. Like the self-employed, people who have an intermittent connection with the labour force are likely to prefer more liquid savings than superannuation. Support for people who have experienced broken work patterns should be achieved through the Age Pension.

Voluntary saving for retirement

The third pillar of the retirement income system generally has been seen as the tax-assisted voluntary part of the superannuation system. Generous tax concessions encourage and assist those with saving capacity (including those not subject to the second pillar) to provide for their retirement.

The third pillar can also be viewed more broadly to include other forms of lifetime voluntary savings — owner-occupied housing, other property, financial assets and business assets. People on higher incomes are better able to save voluntarily through the third pillar and, on average, have similar replacement rates to people with lower pre-retirement incomes once these savings are taken into account.

Recommendations

The superannuation guarantee rate should remain at 9 per cent. The Panel has considered carefully submissions proposing an increase in the superannuation guarantee rate. Such an increase could be expected to lift the retirement incomes of most workers. However, the Panel considers the rate of compulsory saving to be adequate. The Age Pension and the 9 per cent superannuation guarantee (when mature) can be expected to provide the opportunity for people on low to average wages with an average working life of 35 years to have a substantial replacement of their income, well above that provided by the Age Pension. This strikes an appropriate balance for most individuals between their consumption opportunities during their working life and compulsory saving for retirement. The Panel considers that more can be done through preservation and other rules to ensure that the 9 per cent contribution rate produces an adequate retirement income for greater numbers of people, and its other recommendations are made partly for this purpose. For higher income workers especially, the third pillar provides an opportunity to access significantly higher income replacement rates.

The superannuation guarantee broadly should continue to cover employees. While those who derive business income should make provision for their retirement during their working lives, the diverse and varying risks and circumstances of business and entrepreneurship argue for allowing full flexibility in their saving and investment decisions. The voluntary superannuation system is available to small business people for contributing to meeting their retirement needs. However, there can be a fine line between those who are self-employed and those who are performing contracted duties similar to an employee. This distinction arises in a number of areas of policy. In its final report, the Panel will consider further how to distinguish the self-employed, including whether the scope of the superannuation guarantee could be extended to include with greater clarity and certainty arrangements that are close in nature to a formal employer-employee relationship. The $450 per month threshold should continue to apply, as the compliance costs to the employer of providing superannuation guarantee contributions to marginally attached workers are outweighed by the benefits to the employee.


1 This paper uses the term 'saving' to refer to the act of adding to an individual's assets, and the term 'savings' to refer to the stock of those assets.

2 The retirement income system in many OECD countries provides a taxpayer or contribution funded retirement income based on a proportion of an individual's pre-retirement income. These countries also provide a minimum retirement income to alleviate poverty for those with a limited working life.

3 AWOTE is currently about $60,000 per year with 75 per cent AWOTE about $45,000.

4 A replacement rate compares an individual's spending power before and after retirement (that is, after tax is paid). For example, a replacement rate of 75 per cent would mean that an individual would be able to spend in a given time period $75 in retirement for each $100 spent before retirement.

5 Note that these projected disposable income estimates refer to a future year when AWOTE (even when discounted to today's dollars) will be considerably higher than the present level of $60,000 because wages are projected to grow faster than prices.