Superannuation policies must not crimp saving

Benefits of current retirement policy settings are that incomes of both age pensioners and self-funded retirees are growing in real terms, while spending on the age pension is shrinking as a share of GDP.

Revenue from superannuation savings, which are mostly taxed at 15 per cent on both contributions and investment returns over the typical 40 years of working life, is likely to be rising as a share of GDP.

So what’s the problem? The RIR points to “tax expenditures” on superannuation now claimed to be some $40 billion a year and growing at 10 per cent a year, driven by the concessional tax treatment on superannuation investment returns.

These hypothetical costs are projected to outweigh the actual cost of the age pension by the late 2040s. It is this alarming picture that drives the RIR to many of its policy preferences.

Tax expenditures are the gap between revenue actually collected from superannuation and revenue that would be collected under some hypothetical benchmark.

Treasury produces these calculations, but what the RIR sweeps under the carpet is the fact that Treasury has produced two tax expenditure estimates for superannuation – one of $40 billion a year rising at 10 per cent a year, and the other of $7.5 billion a year, roughly steady over time.

The high estimate, favoured by the RIR, benchmarks super saving against the tax treatment of a savings account, and is biased towards current consumption and against saving, especially very long-term saving such as superannuation.

If super were in fact taxed like a saving account as this benchmark imagines, it would not exist.

The low estimate is typically about one-fifth of the high estimate, and is more appropriate for treating savings neutrally.

It uses an expenditure tax benchmark, according to which full tax applies to the income from which savings are sourced but investment returns are not taxed.

This is the benchmark advocated by earlier studies such as the FitzGerald National Saving Report in 1993 and the Australia’s Future Tax System Report of 2009.

By this measure, overall retirement system costs (super plus age pension) would likely not rise as a proportion of GDP, despite population aging and rising retirement incomes.

If superannuation taxes were benchmarked this way, many of the RIR’s policy preferences would fall to the ground.

The RIR also suggests government should aim to induce faster and more complete consumption of super capital and housing equity to prevent retirees’ wealth rising, leading to growing bequests.

But with savers’ equity in housing typically about double their savings in super, no acceleration of super spending is likely to outweigh inflation of housing prices in an era of fiscal and monetary stimulus and asset price inflation.

The RIR claims that most are not drawing down their super savings as fast as it would prefer. But that is based on a selective reading of the research it cites.

When the data properly includes elderly retirees who once had super balances but have completely exhausted them, spending of life savings is significantly higher than the RIR claims.

While the RIR claims most (except some retiring as renters) are already saving more than enough for a retirement that it considers “adequate”, it is unwise and unnecessary for government to set its policies to constrain citizens’ choice of the self-funded retirement living standards they want to work and save towards.

The minimum may not need to be raised beyond the current 9.5 per cent, but that does not mean policies should discourage voluntary contributions above the minimum, as some of the RIR’s preferences would.

Policies to crimp voluntary saving and accelerate retirement spending would create more uncertain retirements and a more fragile economy.