Image by Barney Moss CC 2.0 via Flickr

Broadly speaking, the Government has come up with a clever package which addresses some equity concerns. The proposals unwind some of the generosity of the 2006 changes that were introduced by former Treasurer Peter Costello. But overall, they leave the basic structure of our over-complicated superannuation system intact. In most cases the impacts are confined to between 1 and 3% of contributors. Since the basic structure has some serious structural flaws, it is likely that this will not be the last word in superannuation tax reform.

In broad terms the savings measures contribute $6 billion to the budget over the forward estimates period from 2016-17 to 2019-20. The spending measures cost $3 billion, for a net gain over the forward estimates of $3 billion. This is a relatively insignificant amount in a super tax concession system which costs $30 billion per annum and rising, as measured in the Tax Expenditure Statement.

The Government has endorsed an objective for the superannuation system, to be enshrined in legislation: “to provide income in retirement to substitute or supplement the Age Pension”. It does not refer to retirement income adequacy – an objective which has been advocated by the Committee on Sustainable Retirement Incomes among others.

The main changes are as follows.

Tax on fund earnings

The amount that high earners can transfer into retirement phase accounts is capped at $1.6 million. This will have an impact on current balances (called by some, retrospective), which can be reduced to $1.6 million or tax can be paid on the excess (now to be in accumulation phase) at the 15% rate.

The tax exempt status of earnings for transition to retirement income streams will also be removed, for a revenue gain of $640 million over the forward estimates. This is an overdue change as many people are using the transition to retirement rules to reduce tax to zero or 15% while continuing to work. Some new retirement income products will be supported, with the tax exemption for earnings in the retirement phase extended to products such as deferred lifetime annuities and group self-annuitisation products. This is a sensible change.

It seems likely that many people will continue to keep excess balances in superannuation as this is still a concessional environment. The zero rate was always unsustainable and unfair. The impact is not stated but it probably affects fewer than 3% of superannuation retirement accounts. The estimated revenue gain is $2 billion over the forward estimates.


The contributions tax threshold will be lowered so that people with earnings and superannuation contributions in excess of $250,000 per annum will pay a surcharge, taking the contribution tax rate to 30% for high income earners (the previous threshold was $300,000). The estimated saving over the forward estimates is $550 million.

As our top marginal rate is 45-49%, this measure, with the reinstated Low Income Superannuation Tax Offset (see below) moves the whole retirement system closer to a uniform rebate of approximately15-20% on contributions at all income levels. The Henry Review recommended a uniform 20% rebate, but the Government’s approach, while less consistent, is administratively easier.

The cap on concessional contributions will be lowered to $25,000 per year, from $30,000 under age 50 and $35,000 for 50 and over. This concession is a major contributor to the net cost of superannuation tax concessions, so the cap is sensible although the Grattan Institute recommended a much lower cap of $11,000.

In a new measure, ‘catch up’ contributions are allowed over rolling 5-year periods if the superannuation balance is less than $500,000. The stated goal is to benefit those who take time out of the workforce, for example some women, although they would need to have the financial ability to make such contributions. The revenue gain is $2.4 billion over the forward estimates, less $350m for the ‘catch up’, and affecting only about 3% of contributors. There will be a similar tax impact for unfunded pensions.

The $500,000 lifetime cap on non-concessional contributions applies to all such contributions after 1 July 2007 and compares with current annual caps of $180,000 or $540,000 every 3 years. Revenue gain over the forward estimates is $550 million. The Grattan Institute recommended a lifetime cap of $250,000. It is appropriate that this loophole be tightened as these non-concessional contributions are in addition to concessional contributions of up to $25,000 per annum, and allow savings from outside of super to be placed in a very concessionally taxed environment.

Contributions will be allowed up to age 75 without the current work test, at a slight net cost. This is an administrative simplification. There will be a concessional deduction for the self-employed and wage earners not utilising the full cap, which should reduce the attraction of salary sacrifice arrangements.

Finally, the Government has conceded ground on the low income superannuation tax offset (LISTO) which will replace the Low Income Superannuation Contribution that the Abbott/Hockey government had abolished effective I July 2017. This effectively refunds contributions tax up to $500, up to an income level of $37,000. It is a good measure which helps to address the regressivity of the 15% contributions tax for those who would otherwise pay little tax on their income – but it falls well short of the value of the 15-20% rebate available to higher income earners.

Some women with a higher income spouse may benefit from the spouse contributions where the receiving spouse income limit is to be raised from $10,800 to $37,000, with an offset worth up to $540 at a slight net cost to revenue.

Modest and sensible changes: But we still have a ramshackle system

The changes proposed are modest and sensible. They are clearly designed with one eye on the politics, and in most instances affect only a tiny minority of contributors. However they really only amount to tinkering with a basically unsatisfactory system. What the superannuation tax system really needs is a root and branch reform, as it is hugely expensive and severely regressive in its impact overall.

We suggested one such reform in Ingles and Stewart, Superannuation tax concessions and the age pension: a principled approach to savings taxation (TTPI WP 7/2015). Our preferred approach is an expenditure tax regime for superannuation, an approach favoured by many tax experts. However the normal approach used overseas, which is EET (Exempt contributions, Exempt earnings, Tax payouts at marginal rates) is no longer an option in the Australian system because of our existing reliance on contributions and earnings taxes.

An alternative approach is feasible, which is a pre-paid expenditure tax, or TEE. Contributions would be fully taxed at marginal rates, but earnings and payouts would be wholly exempt. This is conceptually a defensible and simple system.

A TEE system would raise another $12 billion a year compared to current revenue, based on Treasury estimates. However, we argue that such a new exemption for fund earnings should not apply to pre-existing accumulations, which in our proposal would be segregated and taxed at 15%. This approach therefore has the potential to raise up to $20 billion per annum, after adding fund earnings tax back into the tax expenditure calculation.

Improving the age pension

Some of this extra revenue could be used to reform the age pension means test, which is to become much tighter in July 2017. The impact of the new rules is to impose a 7.8% annual wealth tax on pensioner’s assets, which corresponds with a 15.6% deeming rate and is well above the returns available to them in normal investments. This creates an effective marginal tax rate on asset incomes of up to 200 %.

By contrast we propose a simple deeming rate of 6% per annum on all assets, and modifications to the income test so that effective marginal tax rates on pensioner incomes are in the range of 35 to 40%– half of those now prevailing. We expect that these changes, costing $5-10 billion, would have a big impact on work participation. By encouraging savings, it would offset any adverse impacts of tighter superannuation tax. Earnings replacement rates would be broadly similar but improved in the middle income ranges. This proposal reduces distortions on both the tax and the age pension and has the potential to reduce appreciably the efficiency losses in both these systems.

This article has 3 comments

  1. The superannuation system is an example of how not to go about tax design. We have converted what is supposed to be part of the personal tax system into entity taxation. Entity taxation does not lend itself to equivalence with progressive personal tax rates.
    If the objective is to tax a member’s share of the taxable income from superannuation funds at 15%, then it could be done in the same way as Division 6 trusts with (now) a bipartisan policy of an exemption of $75-$80k after retirement. (Why do we insiste on deprecating pension and accumulation accounts?) Essentially if we want super to be equated to the personal direct tax system we should adopt a schedular system of personal tax of 15% with thresholds or offsets. As for concessional contributions, they could be included as schedular income with rebates.
    All contributions to and income from superfunds would be aggregated in the personal schedular system.
    How much does the community pay in fees (of the fund and advisers) to run the current and proposed systems. Taking the tax impact back to personal taxation makes the tax impact and costs of the system transparent. And hopefully, the system could be changed so that assistance with compliance could be given by tax agents (rather than “registered tax (financial) advisers”) at a low cost.

  2. Pingback: Budget Forum 2016: Does the Budget Address Women’s Economic Security In Retirement ? - Austaxpolicy: The Tax and Transfer Policy Blog

  3. Pingback: Election 2016: Tax Policies of the Major Parties Part I - Austaxpolicy: The Tax and Transfer Policy Blog

Leave a comment

Your email address will not be published. Required fields are marked *