Anyone interested in serious tax reform should welcome the changes announced by the Labor government to their proposal to apply additional taxation to large superannuation balances.
The original proposal, known as the Division 296 tax, had two serious flaws which have been heavily criticised since first proposed. The first flaw was the failure to index the $3 million balance threshold to which the tax would apply. Failure to index the threshold meant that a tax, which initially would only apply to about 100,000 people, could apply to as many as 500,000 people by 2050.
The second flaw was the proposal to tax unrealised gains. Taxation of unrealised gains represents a major departure from the taxation of capital gains in the rest of the tax system. For those with large, illiquid portfolios, this tax could have forced individuals to sell assets when it wouldn’t otherwise make sense for them to do so.
The initial proposal to tax unrealised gains made some sense. For industry and retail superannuation funds, where most people have their retirement savings, it is nearly impossible to attribute realised gains to individuals. Taxing individuals based upon the growth in their balance (a mix of income and unrealised capital gains) was a practical and actionable proposal for these funds.
Many of the very large balances that are being targeted by the Division 296 tax are held in self-managed superannuation funds (SMSFs). Some of these SMSFs are composed of property portfolios, businesses or farms. These investments often have low rates of annual income, but their owners plan on selling these assets once in retirement, realising the potentially large capital gains. Taxing unrealised gains was designed to generate more tax revenue from these funds as the value of the assets grew rather than waiting for the revenue when the gains are realised.
The government also had a legitimate concern that SMSFs were being used as tax minimisation vehicles for running businesses rather than vehicles to build assets to generate an income in retirement. Taxing unrealised gains provides a strong discouragement to hold such assets in superannuation.
But the taxing of unrealised gains was always a ham-fisted approach to solving this problem.
Twenty years ago, superannuation was under-taxed and offered a vehicle for wealthy individuals to generate large tax savings if they were willing to postpone consumption until retirement. Over the past two decades, a variety of legislation has chipped away at these concessions.
Notably, individuals can only put $30,000 per year into superannuation at the concessionally taxed rates, and the Division 293 tax ensures that individuals with more than $250,000 a year of annual income pay an additional 15¢ on their concessional superannuation contributions. The transfer balance cap of $2 million results in high wealth individuals paying tax on superannuation earnings even after they move into pension phase.
All of this means that superannuation tax, on average, is now about right. Some of the very large superannuation balances in SMSFs are a legacy issue from an era where super was much less taxed than now.
If Division 296 tax is implemented with the proposed changes, superannuation will still be a tax-preferred investment vehicle for those generating income through their superannuation funds.
For those who are hoping to realise large capital gains, Division 296 will make it less desirable to save through superannuation. The existing 10 per cent tax, plus the additional 15 per cent, means that capital gains will be taxed at 25 per cent. This is higher than the capital gains tax that an individual at the top marginal tax rate would pay on a capital gain outside of superannuation (23.5 per cent because of the capital gains tax discount).
The government is also proposing an additional 10 per cent tax on balances above $10 million.
These two features of the proposed system send a very strong signal that the government does not want people holding large, illiquid assets in superannuation with an expectation of realising a large capital gain in the future. Over time, this will make it less likely that people hold businesses or property in superannuation and that might be desirable.
There are two reasons why these changes make me optimistic about tax reform.
The first is about the process. Tax reform is hard. Tax reform is complicated. For governments of any stripe to think that they can just throw out a proposal and have universal agreement is unrealistic. Even with the best intentions, it is hard to design changes to the tax system that do not have unintended consequences. A debate about reform proposals, followed by concrete action to improve those proposals, is the right process.
As clumsy as it might seem, we need more of this. Government needs to highlight the problems it is seeking to solve. It needs to propose solutions in the form of concrete actions. It then has to weather the stormy and chaotic criticism of those proposals and take on the best of the criticisms to generate better proposals. This has what has happened with Division 296.
The second reason that I am optimistic about tax reform is that these changes to Division 296 will see it get passed. The last 20-plus years of Australian policy debate has seen a litany of tax reform proposals that get shot down and then die without another sound. Then no one ever dares suggest reform of that type, or of any type, again. Every tax reform proposal that gets killed makes the next tax reform proposal less likely.
I’m not suggesting that we should have bad tax reform for the sake of momentum, but sensible tax reform, grounded in compromise and sound debate, paves the way for more reform.
First published at the Australian Financial Review on Monday 13 October 2025.
Recent Comments