Treasurer Jim Chalmers identified three priority areas for tax reform he said had attracted support from his roundtable participants:

• A fair go for working people, including in intergenerational equity terms,
• An affordable, responsible way to incentivise business investment, and
• A simpler, more sustainable tax system to fund the services people need.

It is hard to disagree, but much of the material before the roundtable on intergenerational equity exaggerated the problem and pointed participants, and the public, in the wrong direction.

Let me highlight three misleading aspects of the advice given to the roundtable.

Superannuation tax concessions

First, the Treasury paper on budget sustainability and tax reform repeats its longstanding but misleading advice that superannuation receives huge tax expenditures and that these are heavily skewed towards people on high incomes (highlighted in Chart 10).

The chart is taken from the 2024-25 Tax Expenditures and Insights Statement. In footnote 1 on page 2 of that Statement, however, Treasury mentions that:

“Previous publications have explored alternative benchmarks for tax expenditures relating to the taxation of superannuation, alcohol and owner-occupied housing.”

Yes, indeed. In the 2017 Statement, Treasury included an alternative benchmark for taxing superannuation, one much nearer to the approach most commonly used across the OECD where superannuation withdrawals are fully taxed, but contributions and fund earnings are fully exempt, or EET (the alternative benchmark used in the Treasury statement fully taxed contributions and fully exempted fund earnings and withdrawals, or TEE).

This TEE alternative produced a radically different picture of the “tax expenditures” from that based on the benchmark Treasury insists on using, of taxing both contributions and earnings in full and exempting withdrawals, or TTE. In particular, some $20 billion of “tax expenditures” from exempting fund earnings became a $10 billion negative tax expenditure. If repeated today, that turnaround would be larger. The claimed total tax expenditures essentially disappear under this alternative benchmark – they would certainly do so under the more standard EET benchmark. Moreover, most of the picture of the concessions skewed to the rich would disappear.

This is not just a theoretical point. I am dead certain that no-one at the roundtable would suggest that superannuation fund earnings should be taxed at people’s full marginal tax rates as the Treasury preferred benchmark assumes.

Treasury has not repeated such a presentation using alternative benchmarks since 2017. It should do so in every annual statement. It should also review its preference for the TTE benchmark which no country does, or ever would, use for long-term savings.

Falling amounts of tax paid by the elderly

Second, a lot of attention was given to the Treasury paper’s claim that the income tax burden on Australians aged 70 and over has declined from about 30% in the 1990s to only 16% in 2022-23.

That claim is based in part on identifying nearly all the taxes on superannuation contributions and earnings as a burden on younger people. But they are not.

Those taxes may provide revenue from employers and funds while the employees are working, but they do not impact the disposable income of those workers at all. They are taken from the contributions by employers which the workers are not allowed to access.

What they do impact is the disposable incomes of the workers when they retire.

So, while it is true that no tax is payable on superannuation withdrawals (or earnings) after age 60, contributing to the apparent reduction in the income tax paid by older Australians, that is because the older Australians have already paid taxes on their contributions and earnings and are having their disposable incomes reduced as a result.

We have chosen a much more complicated way to tax superannuation than most other OECD countries. Had we pursued the more common approach of taxing at the back end rather than the front, the picture of significant reductions in tax by the elderly would be significantly different.

I suspect that some of the apparent reduction in the tax burden on the aged since the 1990s also reflects the shift among public sector workers from defined benefit schemes to defined contribution schemes. The declining number of retirees on defined benefit pensions are taxed in the standard OECD way at the back end (though with a 10% rebate that is hard to justify) while the increasing numbers with defined contribution superannuation would have paid tax at the front end.

The social safety net for older Australians

Third, the paper from the ANU’s Tax and Transfers Policy Institute claims that:

“As Australians have become wealthier, we might expect government spending on social safety nets for older Australians to fall. Instead, we have seen these programs grow in real, per person terms.”

The problem is that the paper includes health and aged care as part of the “social safety net”. Had it used a more widely accepted definition, the result would be very different.

Indeed, the 2023-24 Intergenerational Report revealed that the proportion of people of age pension age and over receiving an age or service pension was already under 70% (it was about 80% in 2010) and is projected to fall to around 55% by 2062-63.

Moreover, the proportion of these receiving the maximum rate is also projected to continue to fall substantially. Unlike almost every other OECD country, government spending on retirement incomes is projected to continue to fall as a percentage of GDP despite our ageing population.

The ANU paper authors are correct, of course, in identifying expenditure pressures from services including not only health and aged care, but also the NDIS and childcare, and the need to control those pressures and to fund the expenditures involved fairly and efficiently. But it exaggerates the intergenerational dimension involved.

Future policy directions

Fortunately, one of the ANU authors, Bob Breunig, has been reported as warning against a focus on increasing taxes on superannuation, despite the urgings of the Grattan Institute, suggesting instead that the government help retired people to spend their superannuation.

It is the absence of good advice and good retirement products that is causing too many retirees to leave more of their savings unspent than they consciously intend, as they struggle to manage the market, inflation and longevity risks they face. The resulting transfer of unused savings to the next generation is a genuine intergenerational equity problem, increasing wealth disparity among future generations.

Here is a sensible agenda item, consistent with Chalmer’s third priority area, the need to make the system simpler. Apart from ensuring access to good advice, the government should offer retirees the option to buy government annuities. It should also simplify the pension means test — including adding the value of the home above a high threshold — to encourage retirees to use their wealth in rational ways that give them security and the living standards they want.

This is not to downplay other reforms such as the indexation of the personal income tax scale and increased reliance on indirect taxes, but to ensure we avoid unfairly penalising the aged and instead complete the Keating retirement income reforms to deliver adequate and secure incomes in retirement.

A slightly edited version of this article appeared in AFR, 27 August and The Mandarin, 28 August 2025

The views expressed in this article may or may not reflect those of Pearls and Irritations.

Bob Breunig and I, together with John Piggott from UNSW, co-authored a chapter ( 5. Retirement Incomes: Increasing Inequity, Not Costs, across Generations Is the Intergenerational Problem) on retirement incomes in a book on the Intergenerational Report in 2023, which made similar points.