Introduction
The Albanese Labor government’s new tax geared to individuals’ total superannuation balances was unveiled in March 2023, failed to pass the parliament before the May 2025 election, and is to be reintroduced to parliament by the re-elected government with the intention that it take effect from 1 July 2025.
This new tax is only now receiving the critical attention it should have received long ago. The critics may have believed the legislation would be amended or never passed — but have now realised the election result gives the government a much better chance to do what it wants.
The tax is many things and there are many reasons to object to it. I argued two years ago that the very idea of such a tax was based on a misconception — that superannuation is under-taxed at least for some people — and should be shelved.[1] But that argument seems to have been lost, with most of those opposing the tax not objecting to the principle of increased taxation for larger balances but instead to three key design features.
These are: the absence of indexation; the taxation of unrealised capital gains; and the treatment of defined benefit schemes.
The government is doubtless happy to have diverted attention from the concept of such a tax to selected design issues. In this critique I confine myself to these technicalities, while maintaining the view that this tax should not proceed in any form. That case has been made again, with admirable clarity, by Henry Ergas very recently.[2]
The new tax in a nutshell
As the new tax is not widely understood, it is best to start with a sketch of what it is, simplifying as much as possible something that is very complex.
There is a trigger and a tax base.
The trigger determines whether the tax applies to you in any year. It is the sum of your superannuation balances. If this exceeds $3 million at any 30 June from 2026 onwards, then you are subject to the tax.[3]
If the trigger is activated, the tax base is the change in your total balance since the preceding 30 June, minus any fresh contributions to super in that year and plus any withdrawals you have made. This change in the balance is defined as the earnings of the fund(s). If it is a positive amount, then the excess above $3 million determines the proportion of the earnings subject to the new tax.
If your balance declines from year to year you have made a loss, but the government does not subsidise your loss — you only get to carry it forward to set against any future gains.
There are many complications, but that is it in a nutshell.
Before we go any further, there are some important things to understand about this scheme:
- The change in your total balance from year to year can come from anything that adds to the balance: for example, interest, dividends, rent, realised capital gains and unrealised capital gains. They are all blended, and the Tax Office will not need to know the composition to work out the new tax liability. Thus, despite all the (justified) criticism, it is not only a tax on unrealised gains, although it is partly that.
- In this calculation, unrealised losses on individual assets within a portfolio are automatically offset against any other taxable income, whether capital in nature or not. It is only if there is an overall loss that the carry-forward provision comes into play.
- Although the tax is reported as doubling the existing earnings tax rate from 15% to 30%, the new 15% is really a different tax because it applies to a different tax base. As we shall explain below, it is more than doubling the tax payable.
Now we come to the three contentious issues: indexation; unrealised capital gains; and defined benefit pensions.
Indexation
The absence of indexation of the $3 million trigger is the easiest of these points to address. As almost everyone commenting on the tax says: it is obvious that it should be indexed. The government says there are many other thresholds in the tax system — some in the superannuation system — that are not indexed; but so what? Two wrongs do not make a right.
If the $3 million threshold is not indexed, more superannuation members will be caught in its net every year as average member balances grow. The government says that future governments will increase it from time to time, but why leave them the temptation not to? Just as bracket creep plagues the personal income tax system due to the lack of indexation, failure to index the $3 million threshold will create a new source of bracket creep in another part of the tax system.
The only question is what it should be indexed to. Using the consumer price index would preserve the purchasing power of $3 million, but real incomes will increase over the long run and this real growth should be reflected in superannuation benefits. It would be better to index to average wages or to the average total pre-tax return on one type of institutional public superannuation fund, such as the balanced growth type.
If the government fails to do this, it will be clear it sees this tax becoming a major new revenue source over time.
Unrealised capital gains
The new tax is sometimes referred to as if it were only a tax on unrealised gains, but as discussed above it covers much more. While taxation of unrealised gains is just one dimension of the tax, it is the most objectionable and has attracted a storm of opposition.
To repeat from above, the tax base includes anything that causes your balance to increase in a year, whether interest, dividends, rent, realised capital gains or unrealised (‘accrued’) capital gains. The extra taxation of ordinary income, while unwarranted, does not raise the same contentious issues as taxation of accrued capital gains.
How significant accrued gains will be in anyone’s superannuation will vary widely depending on the composition of the portfolio and how actively it is traded. In most balanced portfolios, accrual gains would be a significant component of the total return and would vary from year to year depending on market conditions. But the proportion could be very high in a fund with a dominant, long-held asset such as business premises, active business interests or a farm.
If there is a doctrinal foundation for the new tax, it is the Schanz-Haig-Simons (SHS) comprehensive definition of income, named after the trio of legal and economic theorists who developed it. SHS defines income as the increase in a taxpayer’s economic wealth between two points in time plus the taxpayer’s consumption in that period. This describes the new superannuation tax base, if we equate ‘consumption’ with superannuation withdrawals. SHS would clearly include unrealised capital gains in the income tax base.
However, it is a long bow to claim SHS as a justification for the new tax. The Treasury describes the SHS definition as “the starting point” for defining Australia’s income tax benchmark, but recognises that the actual benchmark “departs from the SHS definition of income in places — for example, it does not include unrealised capital gains”.[4] Indeed, there are so many departures from SHS that it cannot be claimed as the basis for taxation in Australia or any other country. At most, it is sometimes used as a frame of reference for base-broadening tax reform.
As a guide to tax policy, SHS is also contested. The main competition comes from the ‘optimal taxation’ school of thought, which departs from the notion of uniform taxation of a comprehensive base.
Rather than any rigorous theoretical framework, the inclusion of unrealised gains would appear to be based either on a ‘Trojan horse’ strategy to introduce the concept in one part of the tax system as a prelude to its wider application — such as to capital gains tax in the personal income tax system — or on administrative convenience given that some public-offer superannuation funds reportedly lack the systems to track members’ taxable income inside their funds.
If it is the latter, it is a case of the tail wagging the dog. Administrative convenience is not a cogent reason for such a dramatic change as taxing unrealised gains for the first time in any part of the tax system. If the government insists on increasing the tax on fund earnings from larger balances, funds should be required to reengineer their systems to produce the necessary information on members’ earnings from their funds. If this means delaying the start date, so be it.
But this is all supposition, as the government has failed to provide an explanation and merely repeats the mantra that it is a ‘modest change’ — which it is not.
Taxation of unrealised gains has one thing going for it and many things against it.
The one thing for it is that it removes the lock-in incentive of a realised capital gains tax. Deferral of the tax liability until realisation confers a tax benefit. Investors hold assets longer than they otherwise would to avoid incurring a capital gains tax liability. This is good for the taxpayer, but it reduces capital market efficiency.
Against that, there are many negatives, including:
- The extra 15% tax on capital gains, whether realised or unrealised, denies the one-third discount of longer-term capital gains (on assets held for more than 12 months) in the current superannuation fund earnings tax, which reduces the headline 15% rate to an effective 10%. It also removes the benefit of tax deferral discussed above, which also reduces the effective rate. For these reasons, the new tax is not only double but more than double the existing 10% tax on longer term gains, and at 25% it is even higher than the top discounted capital gains tax rate of 23.5% in the personal income tax system. It makes no sense that capital gains would be more (even slightly more) heavily taxed in superannuation than outside it.
- There is asymmetric treatment of gains and losses if there is an overall loss for the year. (Losses on individual assets are automatically offset against other capital gains and ordinary income, if enough is available.) As overall losses can only be carried forward to be offset against future gains, if they occur, the compensation for losses may never come or may take a long time.
- Accurate asset valuations will be needed every 30 June. This is not a problem for deposits and listed securities such as shares and units in trusts, but it is a major issue and transaction cost for funds that directly hold real estate, business assets, private equity and works of art. Funds have always been required to value assets annually, but this will have larger consequences under the new tax. Get it wrong and you could be paying too much tax or have the Tax Office chasing you for having paid too little.
- The tax increases the risk and cost of capital committed to small business ventures through superannuation structures. This is not to suggest all such funding goes through superannuation and it is difficult to say how much of it does; but at least at the margin this tax will discourage venture capital and start-up funding. It will skew affected funds’ asset allocation towards low risk/low return/more liquid assets to reduce the accrual of gains and the risk of losses that may never be recouped. This is a distortion against efficient capital allocation.
- It increases the perception of investment risk more broadly in that it raises fears that taxation of unrealised gains, once established in superannuation, will be broadened to other parts of the tax system.
Defined benefit schemes
The new tax is designed for defined contribution schemes (which is what most people now have) but ever since the original announcement the government has maintained that it would apply “broadly equivalent” treatment to defined benefit schemes.
There are various types of defined benefit schemes but the most common is the public sector unfunded pension financed out of current consolidated revenue rather than any past contributions and fund earnings. These schemes have none of the key parameters in the design of the new tax — a fund and earnings. The quest for “broadly equivalent” treatment is like forcing a square peg into a round hole, with predictably ugly consequences.
Pensions from unfunded defined benefit schemes are taxed at the recipient’s marginal tax rate — as high as 47% — subject to a 10% tax offset capped at a pension of $125,000 a year from 1 July 2025. Pensions from a defined contribution (accumulation) scheme are tax-free (subject to the recipient satisfying the conditions of release). But the contributions and earnings were taxed while the fund was accumulating and the earnings continue to be taxed once the pension draw-down starts subject to there being no taxation on the first $2 million balance (the Transfer Balance Cap). That is the essential difference.
In the jargon of superannuation, defined contribution schemes are ttE, whereas unfunded defined benefit schemes are equivalent to EET. Increasing the ‘T’ to achieve “broad equivalence” is a form of double taxation.
Just how the government intends forcing the square peg into the round hole is obscure and complex, but that is really beside the point; which is that the new tax should not apply to unfunded defined benefits other than to include its capital value in the calculation of the recipient’s total balance at 30 June each year. Then, if the recipient has other defined contribution superannuation interests that push their total above $3 million, they will be subject to more tax.
The Prime Minister’s own future pension details have attracted attention in this context. His ultimate pension will depend on how much longer he remains in the PM’s office, but at current parliamentary salary rates, the maximum Prime Minister’s pension would be around $360,000 a year. On that amount, the marginal tax rate would be 47% and the regular annual income tax with the tax offset would be around $122,000 for an average rate of 34%, whereas an equivalent or even larger income stream from an accumulation fund would be subject to zero tax.
The point here is not to invite sympathy for the Prime Minister, but to highlight the fact that many people stand to be severely affected. They have no way to adjust. They cannot take money out of superannuation even if they wanted to. There should be more informed discussion of the relevance of the new tax to defined benefit schemes.
Conclusion
The new tax as designed is anything but modest. It will have large effects, initially on a small — but growing — number of people.
There are likely to be strong behavioural responses, with some taking funds out of super to reduce their balances or giving up on superannuation because of the history of change. The government probably would not care about such responses, but should understand that the revenue yield they expect from the tax is unlikely to be realised as people shift funds to other concessionally taxed structures, including more expensive owner-occupied housing.
The tax should be shelved, or at least subject to review and amendment with its implementation date postponed for at least 12 months.
If, and when, it proceeds, any impediments to people shifting funds out of superannuation where their total balance exceeds $3 million should be waived for 12 months leading up to implementation.
References
[1] Robert Carling, “Superannuation Tax: Why the proposed total balance threshold should be shelved”, Policy Paper No. 51, The Centre for Independent Studies, July 2023.
[2] Henry Ergas, “Super tax rationale is confused, convoluted”, The Australian, 16 May 2025.
[3] Although the intended start date is 1 July 2025, this date only becomes relevant if your total balance is above $3 million on 30 June 2026. Then the tax base is the increase in your balance since 1 July 2025.
[4] The Treasury, “Tax Expenditures and Insights Statement”, December 2024.
