The wrong path on retirement incomes

There is, as in most of Grattan's work, some very useful supporting evidence, but it takes effort to appreciate the assumptions and caveats involved, and to understand all the implications, limits and uncertainties of the report's headline prescriptions. Other expert groups – such as the Centre of Excellence in Population Ageing Research, which released three documents on retirement income a week before the Grattan report was published, and the Committee for Sustainable Retirement Incomes – have also presented preferred policies but generally with less fanfare and more care to address the objectives of the retirement-income system as a whole.

The Grattan report makes some sensible recommendations but also some highly questionable ones, and has important omissions.

The headline story is the institute's view that there is no need to increase the compulsory superannuation contribution rate beyond the current 9.5 per cent, and to that do so would disadvantage needy working families and cost taxpayers a lot of money.

Indeed, the authors suggest that retired people even today can expect more than adequate incomes compared to their pre-retirement incomes (a significant proportion at both low and high incomes can expect incomes now above their pre-retirement incomes). Not readily apparent from the headlines is that this:

The authors' surprising result is also due to the high average returns on investments in the last two decades.

The report emphasises that, on a similar basis and with the currently legislated 12 per cent compulsory contributions to be phased in over the next few years, all young people would be able to meet the adequacy benchmark of 70 per cent net income replacement, with many on lower incomes having higher incomes in retirement than before. At the current 9.5 per cent contribution rate, Grattan's modelling suggests only a minority of people, all on above-median earnings, would not reach the 70 per cent net income replacement benchmark.

It may be reasonable to view non-super, non-home-ownership savings as available to fund consumption in retirement, but Grattan's calculations are based on average levels of such assets by income decile; no doubt, however, the levels will vary considerably within each decile and will be affected by levels of superannuation and home-ownership savings. The purpose of the retirement-income system, however, is to ensure secure and adequate incomes in retirement for all Australians. It may be that, for some, increasing super contributions is unnecessary because they are already saving in other ways but, if so, they could – and almost certainly would – adjust their other savings (Grattan seems to suggest the other savings would largely be unaffected by increasing the compulsory super contribution).

Grattan dismisses (with a snide reference to self-interest) those who focus only on the age pension, super (mandated and voluntary) and home asset "pillars" of our retirement-income system, and largely ignore other savings. But it is these four pillars that the government must focus on to ensure secure and adequate retirement incomes for all.

Work commissioned by the Committee for Sustainable Retirement Incomes two years ago confirmed that 12 per cent contribution rates would ensure that those up to just below median earnings would meet the 70 per cent net income replacement adequacy benchmark, given access to some age pension. At 9.5 per cent contributions, the replacement rate for couples on median earnings would be under 60 per cent. More recent work by the Centre of Excellence in Population Ageing Research seems to confirm these findings. Those on incomes above the median would generally need to save 15 per cent or more to achieve the adequacy benchmark (i.e. they must voluntarily supplement the mandated 12 per cent contributions). Such contribution levels are particularly important for those who wish – or need – to draw on some of their super before age 67.

These estimates may seem much more credible than the Grattan data to public servants, whose compulsory employer contributions are 15.4 per cent, and who are encouraged to supplement these with their own contributions, recognising they are unlikely to be eligible for much if any age pension. (Those, like me, on the CSS that required effective employer contributions of about 20 per cent, and another 5 per cent in employee contributions, are fortunate to live comfortably in retirement, but the income-replacement rates after tax were generally no higher than the 70 per cent adequacy benchmark, albeit they could access the pension from age 60.)

Grattan is right to draw attention to the fact that the legislated increase in compulsory contributions to 12 per cent over the next few years will add to employees' costs and, over time, will be borne by employees. There is therefore a strong case for carefully phasing in the increases, linking them to real increases in remuneration, just as was done in the 1990s when the first steps were taken to introduce and increase the compulsory contributions. Low real-wage growth at present may justify adjusting the phasing-in arrangements, though these are already quite prolonged. But the case for most people to spread more of their lifetime earnings in order to maintain living standards in and through retirement remains strong, particularly when they can expect more years of retirement living.

One of Grattan's continuing arguments against the current Australian retirement-income system is the claimed cost of superannuation "tax expenditures". The latest report suggests a further cost of $2 billion a year from increasing the compulsory contribution rate. It is important to recognise the basis of these claimed tax-expenditure costs. They assume super contributions should be taxed at employees' marginal tax rates, and that all earnings from the accumulated savings should be taxed the same way, while incomes drawn in retirement should not be taxed (a "comprehensive income tax" approach, or TTE). This is not the orthodox way to tax savings intended to spread lifetime earnings. The orthodox approach is to exempt contributions and fund earnings but fully tax the incomes derived in retirement (a "comprehensive expenditure tax" approach, or EET).

It is simply too hard now for Australia to adopt the EET approach, after more than 20 years of taxing contributions and fund earnings. Work commissioned by the Committee for Sustainable Retirement Incomes, however, examined what "concessional" taxes on contributions and earnings, with no tax on retirement incomes (a "ttE" approach as distinct from TTE), might be broadly equivalent to the orthodox EET approach. The finding was that the progressive tax on contributions introduced by the Turnbull government, and the caps on contributions and accumulated savings for the purposes of our ttE regime, achieves about the same net income in retirement as the ideal TTE approach right across the income spectrum with no significant subsidies to either high or low-income earners. There is no "tax expenditure" involved unless one assumes, as Grattan does, that current taxes on contributions and fund earnings should ideally be doubled.

So the $2 billion a year cost Grattan claims is based on a false assumption. Similarly, its suggestions to further tighten the current tax regime for super are inappropriate. Indeed, while the system is still maturing, there is a good case for relaxing the $25,000 cap on contributions introduced by the Turnbull government, particularly for older female workers with low levels of accumulated super savings. Again, this is likely to be seen as obvious to public servants on higher incomes, whose 15.4 per cent employer contributions already exceed the cap.

Despite these criticisms of the Grattan Institute report, it does include recommendations with which I strongly agree, but these have been overshadowed publicly by its emphasis on not increasing compulsory contributions. These include in particular the importance of increasing rental aid for pensioners who do not own their own home, and of relaxing the recently tightened assets test.

The evidence from Centre of Excellence in Population Ageing Research and the Committee for Sustainable Retirement Incomes, as well as that in the Grattan report, demonstrates overwhelmingly the serious problem of poverty among pensioners in private rental accommodation. A sizeable increase in rental aid is long overdue, not only for age pensioners but also for other social-security recipients.

There is also a strong case for relaxing the assets test, which was unwisely tightened by Scott Morrison when he was social services minister. The test taper was increased from reducing the fortnightly pension by $1.50 for every $1000 of assessable assets above a threshold to $3 a fortnight. This imposed the equivalent of a 7.8 per cent tax, well beyond expected earnings from investments, and beyond the annuity the asset could generate (which includes return of capital). It is difficult to reconcile such a test with the rest of the retirement-income system, which is meant to facilitate careful financial planning to achieve desired levels of retirement incomes and living standards. It can only lead to confusion and distortions in savings behaviour (for example, to invest more in housing than in super) by those in their fifties looking carefully at how best to achieve their desired post-retirement living standards. Grattan suggests lowering the taper to $2.25 reduction in fortnightly pensions for every $1000 of assessable assets; I suspect that is still too tough, noting it would still be tougher than the income test. A merged means test (as existed in the 1960s), or at least a more coherent approach to the separate income and assets tests, could simplify the retirement-income system, making it easier for people to plan how to achieve the retirement income they want.

Tellingly, Grattan omitted any reference to the greatest weakness of the Australia's retirement-income system, and the issue rightly getting the most attention from the Treasury, the Department of the Prime Minister and Cabinet, and external experts. This is the failure to convert the growing levels of accumulated savings into secure retirement-income streams for life. Grattan's calculations – like those of the other researchers mentioned above, as well as the OECD – refer to the income streams that should be possible from accumulated savings, not retirees' actual incomes or consumption levels. As the financial system inquiry revealed, too many retirees are self-insuring for longevity risk by holding onto their savings, unnecessarily limiting their retirement living standards and leaving excessive funds in their estates (this is far more common than people running out of savings).

Longevity risk management requires people to pool savings. The latest government announcement goes some way to address this issue, requiring funds in future to offer retirees with more than $100,000 in super savings a "comprehensive income product for retirement" (or CIPR). (I favour the suggestion from a recent conference participant to change "CIPRs" to "secure lifetime income products for retirement", or "SLIPRs".) Such a product will generally include an indexed lifetime annuity. I strongly suspect that further action will be needed sometime in the future to press people to place more of their super into annuities. Perhaps people may even see merit in the sort of indexed pensions provided under the old CSS and PSS schemes, and start thinking about their super savings not as big sums of money but in terms of the secure, indexed lifetime incomes they may deliver (mostly, with a part age pension). They should be thinking and planning this from about age 50 if not earlier.

Grattan seems to think we have a problem of sustainability in our retirement-income system. That is simply not the case, as the continuing fall in age-pension coverage demonstrates (down already to about 65 per cent of those over 65, with an increasing proportion of pensioners – over 40 per cent now – on part pensions only). It's more important to ensure the system actually delivers secure and adequate retirement incomes for all, including by proceeding with the legislated increase in mandated contributions.

Andrew Podger is an honorary professor of public policy at the Australian National University and was a foundation member of the Committee for Sustainable Retirement Incomes. andrew.podger@anu.edu.au