PensionReforms' summary and comments
Saving for retirement is a good idea for those who can afford to do it.  Individuals can build on the state-delivered retirement benefits in ways and amounts that suit their preferences.

But, should the state subsidise employers or individuals to put money aside for retirement?  Because all developed countries subsidise some types of retirement saving, it might seem the answer to that question is 'yes'.

The report suggests that the very generous subsidies built into Australia's tax system (for retirement savings, medical insurance and housing concessions) aren't good value.  The large amount spent on retirement saving ('superannuation' in Australia) is what the report calls a "case study" of unscrutinised state outlays.

Most tax systems let individuals collect 'income' that is put aside for retirement in a tax-favoured way.  Contributions paid by employers are not taxed in full as income; those paid by the individuals themselves are often deductible from income for tax purposes.  The investment returns earned in retirement saving schemes are often either not taxed at all or taxed on a favourable basis, compared with investment income directly received.  Finally, the benefits paid at or after retirement often receive favoured treatment compared with other types of income.

As the report notes, the costs of these concessions ("tax expenditures") are hidden in that their quantification depends on a calculation as to what the tax position would have been in the absence of the concessions.  The fact that they are hidden means they tend not to be subjected to the same level of scrutiny as direct expenditure.

The 'loss' calculated in this way is just as much a cost shared by all taxpayers as amounts spent directly on roads, health or welfare.  However, whereas spending on direct welfare payments is very progressive (benefits accrue proportionately in favour of poorer citizens in need), the indirect costs of retirement saving subsidies are steeply regressive (favour richer citizens) because they are uncapped, and because richer citizens are able to save more.  Higher income earners, who usually have higher marginal tax rates, receive higher subsidies than lower earners, who can't save as much anyway.  

The report quotes research showing that the Australian government spends more on the retirement savings of those earning more than $A72,000 a year (about twice average earnings) than the lifetime value of their Tier 1 Age Pension.  Their report doesn't suggest reducing the total amount spent on subsidising retirement saving but thinks there is a better way of 'spending' the value of the concessions.  The current system, according to the report, has two major flaws.  First, minimum wage earners get no help; secondly, "[t]hose in the top income tax bracket receive on average more than $11,000 per annum."

"This paper proposes two alternative models for reform. The first is based on a flat rate similar to the First Home Savers Account.  The second has a flat rate for those earning up to $80,000 per annum, phasing out after $100,000 - similar in design to the recent tax bonus included in the stimulus package.

"These schemes would provide a much more equitable distribution of benefits.
·    Minimum wage earners would be up to $24,000 better off on retirement under the first model and
·    $32,000 under the second; more than a year's additional salary.
·    85 per cent of wage earners would receive higher benefits under the second model.
·    Savings by low-income earners would be encouraged, reducing reliance on the pension.
·    Both models are revenue neutral."

The report suggests that the changed system would improve transparency and may even save the government money in the long term.  That's because lower paid workers will have higher private savings in retirement and so would lose more of their Tier 1 Age Pension through the income and asset tests that apply.

"In reality, Australia's welfare state is already much larger than is generally acknowledged, but its most generous support is reserved for the genuinely well off and is obscured from public debate. By returning the tax system to its primary purpose of collecting revenue, and combining disparate social support schemes into the system of government payments, we could create a more inclusive, more generous and more affordable welfare state."

PensionReforms agrees that countries need to know much more about the "tax expenditures" hidden in their systems of private provision for retirement.  Australia is one of a very small number of countries (the UK and Ireland are two others) that even bother to count their cost.  Others should follow their lead - if you don't know the cost, it's not possible to measure effectiveness.  Both Australia and Ireland now spend about the same amount on retirement saving tax incentives as they do on Tier 1 pensions.

However, PensionReforms suspects that if the direct costs of the report's proposed tax rebates were subjected to the same level of scrutiny as other expenditure, they might not survive unscathed.  The report assumed that Australia should continue to spend about the same amount on "tax expenditures" as now.

The report's suggestions might address regressive inequity, one of the major flaws of current tax concessions, but that is not the only problem with incentives to save for retirement.  Tax concessions are distortionary, favouring one form of saving over another; very expensive; complex; and they impose significant regulatory costs as money in the favoured vehicle must be kept there and is let out only for approved purposes.  Worst of all though is that tax incentives do not appear to raise saving levels.

Australia differs in two regards from other countries in ways that might make the report's proposed rebates more effective.  First, it has an extensive Tier 2 compulsory savings scheme that covers nearly all employees.  So, nearly all employees would benefit directly from the recommended rebates (the the rich would get less).  However, given that Tier 2 saving is compulsory, why might there be a need to subsidise them at all?  The main purpose of tax incentives is to change behaviour, so PensionReforms wonders what it is about the compulsory arrangements that require that type of incentive.  It seems that the rebate system is an unnecessary money-go-round as far as Tier 2 is concerned - perhaps more equitable than the present system but, at least with respect to Tier 2, apparently no more necessary.

Secondly, because of the income/asset tests that apply to the Tier 1 Age Pension, the state will benefit in the long run from more private provision and the consequent reduction in the costs of the state pension.  

PensionReforms thinks that while tax rebates as suggested are much better than the existing tax incentives, distortions and complexity remain.  Forgotten in all this is that tax rebates too have to be paid for by higher taxes elsewhere.  They are unlikely to impact on savings overall.  There is no free lunch. (File size 323 KB; 18 pp)  305