Public debate on superannuation is currently focused primarily on the level of the guarantee. This is a legitimate debate, but the guarantee is not the most important issue for ensuring Australians have adequate and secure retirement incomes.
Indeed, for some Australians, increasing the guarantee will reduce their retirement incomes unless the other issues are properly addressed – notwithstanding the strong case for proceeding with the next increase.
An international perspective
Australia has long relied upon the means-tested age pension to deliver income protection for most retirees. This has generally protected them from poverty but, unlike most countries’ social security systems, it was not designed to maintain people’s previous living standards. For that, Australians had to have sufficient other savings to draw upon so as to spread their lifetime incomes, and few were doing so or doing so effectively (public sector workers being the main exception through their compulsory superannuation schemes).
Different countries pursue different mixes of retirement income ‘pillars’ (to use the World Bank terminology) to meet the two key objectives of poverty alleviation and maintenance of living standards. These pillars are:
- the base poverty alleviation pillar,
- social insurance (‘pillar 1’),
- mandated contributions (‘pillar 2’),
- voluntary superannuation contributions (‘pillar 3’), and
- non-superannuation savings, especially housing (‘pillar 4’)).
Our emphasis on the base pillar and pillars 2 and 3 with no pillar 1 contrasts with most OECD countries’ emphasis on pillar 1 and/or pillar 2 plus a more modest base pillar. The compulsory contributions rates for pillars 1 and/or 2 are typically around 20% (including both employer and employee contributions).
Interestingly, a shift is occurring in many OECD countries with increasing emphasis on pillars 2 and 3 and some containment of pillar 1. As here, the concerns are to limit the risks for government and better ensure inter-generational equity. Equally, however, there is concern that the shift is unduly increasing the risks individuals bear.
Some critics of our superannuation guarantee see it as excessive government intervention. An international perspective would be more likely to portray it as ‘contracted out’ social security with less government involvement (and risk) than most countries’ schemes, and offering individuals more choice (along with greater risk).
Key challenges for our emerging system
Converting savings into income
The biggest weakness of the Australian system is its failure so far to ensure accumulated savings are converted into secure lifetime income streams. As the Financial Services Inquiry found, the main problem is not that people spend their super too quickly but that they are too cautious and limit their spending for fear of their savings running out and end up leaving more money in their estates than intended. The result is not only lower living standards than their savings should provide but also the waste of the tax concessions involved and increasing wealth inequity amongst future generations.
We need to change the language of superannuation in Australia – instead of focusing on the amount accumulated we need to focus on the incomes those savings should generate. Funds should be required to advise members regularly of the percentage of current real income members’ savings plus current employer contributions are likely to generate by the time of retirement.
Despite measures taken since the FSI to encourage funds to offer products that provide secure retirement incomes with adequate insurance against longevity risk, the number of such products being taken up remains low. If this doesn’t change, serious consideration should be given to the Commonwealth itself selling indexed annuities.
Fixing the means test
The second serious weakness is the complex way superannuation interacts with the age pension. This weakness was exacerbated by the change to the assets test in 2018 which now reduces age pension entitlement by $3 a fortnight for every $1,000 of assessable assets above specified thresholds. Currently, $1,000 will buy a CPI-indexed annuity for someone of age pension age of around $50. So, an extra $1,000 of savings would cause a net reduction in income of around $28 a year ($50 less $3 times 26 fortnights). What would a rational person do? Avoid saving beyond the asset test thresholds (unless the person is confident their assessable savings will be above the test cut-out points ($876,500 for a home-owning couple)); or spend the money as quickly as possible for example on a big holiday to reduce savings to the threshold; or put money into non-assessable assets such as the home.
The penalty involved in this assets test is particularly unjust for those compelled to contribute more to superannuation when their savings are already projected to be around the test threshold: saving more will take them backwards. For some, increasing the guarantee will precisely do this.
David Knox (from Mercer) and I suggested last year that we return to a merged means test replacing the current separate income and assets test, with a taper that ensures extra savings do indeed lead to increased retirement incomes, while still containing the cost of age pensions.
Excessive fees and poor fund performance
A third issue concerns the poor performance of some funds and the common experience of high fees and multiple accounts. Fortunately, the Government has announced in the Budget its intention to act on the 2018 Productivity Commission report which presented sensible recommendations that would ensure the default funds most people rely upon do in fact perform well and that people do not inadvertently end up with multiple accounts with multiple fees as they change jobs. The Government’s planned measures do not go quite as far as the PC recommended, but they represent a very considerable reform.
The SG Level
The level of the superannuation guarantee is fundamental to ensuring adequacy. Broadly, ‘adequacy’ in terms of preserving pre-retirement living standards requires a net income replacement rate of about 70%, at least at median to average earnings. In our submission to Mike Callaghan’s Retirement Income Review, Mike Keating and I suggested the Review examine carefully the modelling by various expert groups to identify the SG level most likely to achieve this benchmark for those on incomes up to the median, leaving those on higher incomes to use additional voluntary contributions to achieve the benchmark suitable for them. The modelling we contributed to a few years ago indicates that 12% is appropriate but some others, including the Grattan Institute, claim that the current 9.5% is sufficient. Careful study of the assumptions used is essential (such as how best to measure pre-retirement income and living standards, how accumulated savings should be converted into income streams, the appropriate indexation factor and the way the age pension interacts with superannuation at retirement and over time).
Unfortunately the Government has not yet released Callaghan’s report. I would be surprised if his modelling reveals that 9.5% is sufficient but it is possible that 12% is more than required if those on median incomes have access to some age pension under the assumed means test. Whatever figure emerges, those unlikely to be eligible for any pension will need to contribute more than the SG to maintain living standards.
Notwithstanding the claims of some in the union movement, increasing the SG will increase labour costs, will involve a wages trade-off and will defer consumption. That was recognised when the SG was first introduced and first increased in the 1990s.
Whether this fact justifies further postponement of the next increase in the SG is not clear. It will be seven years since the last increase and already it will not be until 2054 that retirees will have 40 years of contributions at the current 9.5%. Moreover, there are obvious and more effective alternative ways to promote consumption to stimulate economic recovery in the short term.
The current 15.3% employer contribution rate is not excessive as few retired Commonwealth employees will be heavily reliant on the age pension. But, once the SG debate is settled, there is a strong case to introduce new flexibility: to allow low-income employees and those under financial pressure to reduce their employer contributions to the SG level (the money then added to their wage).
While the shift away from defined benefits was justified, the removal of access to indexed pensions was not. In line with the recommendation of the 2007 review of military superannuation (which I led), career Commonwealth employees should be allowed to buy indexed annuities at a price set by the Government Actuary (and at no extra cost to the Commonwealth). Arguably, the public should also be allowed to do so.