Andrew Podger. A fair, effective and sustainable retirement incomes system.

May 23, 2015

Fairness, Opportunity and Security
Policy series edited by Michael Keating and John Menadue.

In his introduction to this series, Ken Henry said he could not recall a poorer quality debate, on almost any issue, than what we have had in Australia in recent times. Ian Marsh, in his contribution, advocated pursuing bi(multi)partisanship opportunities as far as possible.

Sadly, Henry’s comment seems most apt when it comes to retirement incomes policy, and Marsh’s call seems a long way off after the Prime Minister and Treasurer ruled out a comprehensive review of the policy after the recent Budget. This is despite the Treasurer recently saying in relation to taxation that all options were on the table, and the Opposition indicating a willingness to work with the Government. In addition, there has been some excellent work in academia over recent years, and a quality report from David Murray’s Financial Services Inquiry which highlights the importance of drawing all the threads of the retirement income system together. We can only hope some others in the Government can find a way to allow proper discussion and wide engagement on this critical issue for everyone.

Careful analysis of Australia’s retirement income system would reveal it has considerable strengths, but also some serious weaknesses and challenges, most of which have not been addressed by the Government’s Budget proposals – either this year or last year.

Such an analysis requires, first, some agreement on the objectives of the system. David Murray’s Financial Services Inquiry made an important contribution in its simple admonition to articulate in legislation the objectives of the superannuation system, the primary one being, ‘to provide income in retirement to substitute or supplement the age pension’. Such a focus would avoid debate being hijacked by those promoting housing investment, or infrastructure financing, or broader wealth accumulation and so on.

The wider retirement incomes system in fact has two objectives:

  • The alleviation of poverty amongst the aged (addressed mainly by the age pension); and
  • The maintenance of income and living standards at and through retirement (addressed mainly by superannuation).

These core objectives are complemented by general principles such as value for money and sustainable cost, simplicity and understandability, and stability and certainty.

Australia’s ‘multi-pillared’ system (to use the language of the World Bank) has considerable strengths. Its ‘foundation pillar’, the age pension financed by general revenue, addresses poverty alleviation reasonably effectively and efficiently. The level of the pension is slightly below the OECD benchmark for poverty (50% of median income) so the headline poverty rate amongst our elderly is quite high (35% compared to an OECD average of 12.5%0, but the severity of poverty is lower (the average gap being 12.4% compared to the OECD average of 18.4%). With significant increases in the pension over the last decade and more, and with increasing numbers having superannuation as well as the pension, our main underachievement against the first objective concerns those fully reliant on the pension who are in private rental accommodation whose after housing costs are much higher than those who own their own home or are in public housing. The case for increasing rental assistance is strong.

Perhaps our system’s greatest strength comes from our emphasis on ‘pillar two’ mandated contributions and ‘pillar three’ tax-encouraged voluntary savings. These pillars are mostly fully funded instead of a ‘pillar one’ national superannuation scheme with unfunded promised benefits as is common in Europe and North America. In theory at least, our approach imposes less risk on governments and future taxpayers, and hence offers greater intergenerational equity.

At the current mandated contribution rate of 9.5%, most people will accumulate superannuation savings which, with some age pension, will be able to deliver at least 70% net income replacement in retirement after 35 years of contribution. The rates are higher at low income levels because pension eligibility is higher. This suggests we have the mandated contribution rate about right already if one accepts the international standard of adequate income maintenance of between 70 and 80%. Raising it further would only force people on low incomes to save more when their needs are greater in order to improve retirement incomes that are already sufficient. Most people on or above median earnings are already contributing more than the mandated amount taking advantage of the incentives available and, on average, it seems they also are likely to have sufficient accumulated savings to achieve 75% replacement rates.

The problem is that these income replacement rates are only potentially available. That our system does not in fact deliver them and ensure they last everyone’s full life is perhaps its greatest weakness. It certainly contrasts with every national superannuation scheme in other countries, and indeed with our own age pension.

We allow people too much freedom to take benefits in the form of lump sums. This can leave people with insufficient funds for their later retirement years and make them overly reliant on the age pension. Evidence gathered by the FSI suggests that this is not as yet a major concern but it could become one.

Of more concern according to the FSI is that too many people are trying to manage longevity risk on their own. To do this, they are holding back consumption from their accumulated savings so as not to run out of savings before they die. The result is lower consumption (and a lower standard of living than their accumulated savings suggest they should be able to have in retirement), and much larger bequests to the next generation than they would have planned (and much more than the system was intended to provide). Also, some still live to a very old age and run the risk of running out of savings.

There is also capacity to exploit the tax concessions to accumulate wealth including for planned transfer to the next generation rather than genuine retirement purposes.

Our system needs products that deliver retirement streams and provide insurance against the risk of longevity, and for policies which promote the take up of these. The FSI proposed requiring superannuation funds to offer their members a ‘comprehensive retirement income product’ which would include a longevity insurance element. It hoped these products would become the default retirement benefit products which most will take up, and thereby also addressing in part some of the ‘market failures’ such as adverse selection. This may not be sufficient and, eventually, consideration may need to be given to a mandated approach and to complementary measures to address market failure such as the options identified in the Henry Report including the issue of longevity bonds and the sale of annuities by government to supplement the age pension. These might be more likely to make lifetime annuity products available and limit capacity for people to use superannuation tax concessions for purposes other than retirement income.

The structure of our system makes us much better prepared for demographic and economic changes, but we still have serious cost challenges. The 2015 Intergenerational Report projects the cost of age pensions will grow from 2.9% GDP to 3.6% over 40 years unless the legislation changes. Health and aged care costs are projected to grow further. These increases will need to be managed and, if possible, curbed while ensuring the programs still deliver what the community needs and prefers.

What the IGR did not report was the cost of superannuation tax concessions which are growing faster than the age pension and are concentrated on those on high incomes. We should not however exaggerate the scale of these.

Treasury estimates of close to $30 billion are based on a ‘comprehensive income tax’ benchmark or TTE approach (taxing contributions and fund earnings as income and exempting the benefits). This may apply to your bank account but it is clearly excessive as it eats into the real level of savings. Last year Treasury presented estimates of the tax expenditures if a ‘comprehensive consumption tax’ or TEE benchmark was used. This suggested the costs of superannuation concessions are around $12 billion. But even that is arguably more than the revenue forgone that might be reaped if we agree the purpose of superannuation is to spread lifetime incomes to maintain living standards in retirement. That would suggest an EET approach, the orthodox approach used elsewhere but way too hard for us now given policies of the last 25 years. I have not seen any estimate of our tax expenditures on this basis, but they would be much lower as few retired people would have large amounts of other income so the tax rate would be much lower than their marginal rate when making contributions.

Given it is not feasible now to replace the current regime with an EET one, the question is what tax arrangement might most closely replicate an EET one, containing the costs and ensuring tax equity. I suspect the Henry Report approach would get pretty near to it by allowing a 20 percentage deduction from contributor’s marginal tax rate when setting the contributions tax. In practical terms, this would mean applying a 30% contributions tax for all those with incomes at the top marginal tax rate and no change for the vast majority of contributors; Henry also proposed a flat 7.5% tax on fund earnings at both the accumulation and drawdown phases.

The Government is right to draw attention to the costs of the age pension even if our challenges are small compared to those facing many others. But we also need to be realistic and to consider carefully how the pension will fit with superannuation as our population ages and the transition to retirement shifts and varies.

Australia has already been remarkably successful in reducing eligibility for pensions amongst women under 65, and has legislated to increase the age pension age to 67. When considering possible further increases consideration needs to be given to the implications for those with limited capacity to continue work, and the savings actually generated by such a change. The savings may be modest given the falling numbers of full-rate pensioners and the increasing proportion of these already on welfare before transferring to the pension. The Government’s proposal to increase the age to 70 in the 2030s was designed to maintain the ratio of working years to retirement years: that has some attractions but we need to look more carefully at the effects of the increase to 67 first and review whether the overall impact of a further increase would be acceptable.

The Government proposed last year to change the pension index to the CPI rather than AWOTE. That was always far too tough, reducing relativities with community incomes very substantially if continued for a lengthy period. But as Minister Morrison suggested in February, there is a case for modifying the current AWOTE approach which will over time increase the pension relative to community incomes. Using the CPI for automatic increases then having independent reviews to make adjustments for community income changes every two or three years would in fact be very sensible, and could form the basis for a uniform approach to indexation of all welfare payments. The welfare lobby might like to reconsider its opposition to any change in pension indexation arrangements.

Tightening the means test offers another way of achieving savings but it is important to recall that the original intention of the superannuation reforms was to allow most retired workers to supplement age pensions not to fully replace them. We have already seen a drop in the proportion of the aged on full-rate pensions from around 60% to 50% and this is projected to drop to 30%. The proportion on part-rate pensions however is increasing, so the forecast involves only a modest reduction in the total pensioner population.

To achieve a much greater reduction would require radical changes which could have adverse implications. The income and assets levels at which pension eligibility ceases are of course a function of the level of the pension and the means test withdrawal rates. The income test withdrawal rate has already been increased to 50%: a higher rate could affect incentives to continue part-time work. The Government has proposed an increase in the assets test withdrawal rate but few (including in the welfare sector) seem to realise this involves an effective wealth tax of 7.8% removing incentives to improve assessable assets above the threshold, contrasting sharply with superannuation tax arrangements intended to encourage saving.

More sensible suggestions include Henry’s proposal for a single merged income test which converts assets into appropriately deemed income: this would not radically change the numbers eligible for some pension, but would provide a more coherent effect on incentives to work and save. Another is to include the home in the assets test beyond some threshold, allowing people to continue to receive the pension but requiring repayment from their estate through a reverse mortgage arrangement. But in all likelihood over half our retired population will continue to receive some age pension, and that should not be regarded as bad so long as the system as a whole is delivering adequate incomes efficiently and at an affordable cost.

All this goes to demonstrate how a bi-partisan review of our retirement income system could build on its strengths, make it more effective and sustainable, and give people full confidence as they plan for their retirement years.

The demographic changes now underway should not be presented as a crisis; they represent a triumph of increased life expectancy and years of health living at older ages. They provide new opportunities for people to contribute to society and their families and communities as they transition from full-time employment. Our retirement income system can provide the security people need against poverty and reduced living standards while offering the flexibility for people to manage this new transition to retirement in the way they want.

Andrew Podger, Professor of Public Policy, Australian National University. He was previously the Public Service Commissioner and Secretary of the Departments of Health and Ageing, Housing and Regional Development, and Administrative Services.

Share and Enjoy !

Subscribe to John Menadue's Newsletter
Subscribe to John Menadue's Newsletter


Thank you for subscribing!