Pulling the wool over the eyes of those being fleeced

Labor and the unions want the levy to rise to 15 per cent or more, which would increase the quantum of tax concessions while reducing outlays on the Age Pension: fiscal shock and horror.

Many commentators assert that heavier taxation of super is clearly justified because the Treasury projects that by 2050 superannuation tax concessions will cost the government more than it spends on the Age Pension. Jim Chalmers has now joined the chorus, announcing yet another tax grab at super. However, that chorus’s budget comparison is not a robust reason to increase the tax rate.

One rationale for the introduction of the superannuation guarantee levy in 1992 was to reduce the future fiscal burden of the Age Pension. If the SGL had stayed at 3 or 4 per cent, then undoubtedly in 2050 pensions would cost Treasury much more than it lost through tax concessions on super. However, Labor and the unions want the levy to rise to 15 per cent or more, which would increase the quantum of tax concessions while reducing outlays on the Age Pension: fiscal shock and horror.

It is surely paradoxical to force people to accumulate super funds to reduce pension outlays, then complain that super’s fiscal cost (which rises as forced savings accumulate) will eventually exceed the fiscal cost of the (correspondingly reduced) pension outlays.

Just as those claims are flawed, so are contentions about the extent of super tax concessions. The Grattan Institute notes that a worker earning $100,000 pays $23,000 in income tax, while retirees pay no tax on anything earned by their super accounts. Grattan therefore wants a tax imposed on those earnings.

However, putting aside bad luck, today’s workers will become tomorrow’s retirees and, having paid superannuation taxes along the way, they would quite rightly consider it insulting to be told they are freeloading on the income tax-paying worker.

The Treasurer wants the legislated purpose of super to be “to preserve savings to deliver income for a dignified retirement, alongside government support, in an equitable and sustainable way”. But taxes on super greatly affect how readily Australians can secure that “dignified retirement”.

A scenario: a middle-income earner makes a steady inflation-adjusted stream of SGL contributions for 25 years, with rates of inflation and earnings as in the 2020 Callaghan review of retirement income; that is, 2 and 5 per cent. Let us assume if all this were tax-free, the retiree, like the worker, would have $100,000 to spend.

But each year he pays 15 per cent tax on contributions and fund earnings, which reduces his retirement income to $75,000; his cumulative tax burden is $25,000, which far from being less than the worker’s, is more – and this is for the lower concessional rates.

Notice that the super tax burden is greater than 15 per cent of the tax-free $100,000. This is the effect of compound interest: $100 tax paid in year one reduces the tax-paid amount in year 25 by more than does $100 tax paid in year 25. Because the tax paid at the beginning and on the way through reduces the amount that can grow at a compound rate in subsequent years, it has a cumulative effect that makes the effective tax rate much larger than the headline tax rate.

Now consider the same savings scenario, except this time for a better-off taxpayer on the top marginal rate who falls under division 293 and pays 30 per cent on contributions and 15 per cent on fund earnings. Tax-free, she would have four times $100,000 to spend in retirement; but after super taxes her actual amount is four times $62,000. This retiree’s effective tax burden is therefore four times $38,000. Again, this is as a result of the compounding effect of the year-on-year tax wedge. Those tax rates don’t look especially generous.

In fact, an ordinary taxpayer earning income of four times $100,000 would pay income tax of four times $40,000. When properly calculated, the much-touted concession to the better-off retiree therefore amounts to all of four times $2000.

Indeed, according to the OECD, Australia’s actual tax rates on private retirement savings put us around the middle of the international pack and not – as Grattan’s claims imply – at its lower end. Moreover, the OECD estimates of those tax rates don’t capture the totality of our hypothetical savers’ fiscal contributions: the higher their SGL retirement income, the less they would receive by way of the Age Pension. And Labor’s proposal of raising to 30 per cent the tax on earnings from super balances over $3m will further boost these already hardly generous tax rates.

However, in calculating the value of the tax concessions to super, Treasury’s tax benchmarks and variations statement takes no account of those realities. It calculates the difference between the ordinary income tax rates and the 15 per cent “headline” super tax rates rather than calculating the difference between ordinary income tax rates and the effective super tax rates. But were those ordinary income tax rates actually applied to super they would result in extremely high effective tax rates on retirement savings.

Our middle-income worker, for example, who would, tax-free, have had $100,000 to spend would be left with only $51,000.

As for the better-off saver, she would pay $62,000 in tax for each $38,000 of after-tax retirement income, a ratio of tax paid to after-tax income of 163 per cent.

Imposing the ordinary income tax rates that Treasury uses as a benchmark on super would almost certainly cause substantial economic and social costs. But those costs are inevitable when steeply progressive income tax rates such as ours are applied to retirement savings that accumulate across long periods.

That is why almost all OECD countries have less punitive concessional arrangements for savings for retirement: and Ken Henry, in his tax review, recommended exactly that for Australia.

In the end, Treasury’s estimates of superannuation tax concessions are designed to show how much income tax revenue is notionally forgone in a single financial year from the tax treatment of that year’s super contributions and earnings. In contrast with the OECD’s, they are calculated as if the savings were there for one year only – which means they have no relevance whatsoever to determining the effective tax burden on savings that are long term.

This is why the tax benchmarks and variations statement that Chalmers has just released is deeply flawed. Until the analysis is put on a sensible basis, it will be little more than an exercise in pulling the wool over the eyes of those who are being fleeced.

Jonathan Pincus is visiting professor at Adelaide University and author of Superannuation tax concessions are overestimated, Tax and Transfer Policy Institute WP 20/2021.