This Appendix provides additional technical detail on some matters that are frequently raised with regard to the superannuation tax expenditures.
A.1 Choice of benchmark
The Tax Expenditure Statement uses a comprehensive income tax benchmark to estimate the value of tax expenditures on savings, including superannuation. Some argue that an expenditure tax benchmark would be more appropriate.
There are three conceptual taxing points for superannuation: when contributions are made into a superannuation fund; when investments in superannuation funds earn income; and when superannuation benefits are paid (see page 134).
The treatment of superannuation under a comprehensive income tax benchmark is for: superannuation contributions to be funded from after-tax income, earnings to be taxed at marginal rates, and benefits to be untaxed. Under an expenditure tax benchmark, on the other hand, contributions and earnings are exempt from tax while benefits are taxed at marginal tax rates. This is in fact equivalent to an approach where contributions are taxed at marginal rates while earnings and benefits are exempt from tax. When expressed in this way, it becomes clear that the point of difference between the two benchmarks is the taxation of superannuation earnings. Estimates of superannuation tax expenditures using the two benchmarks will be quite different.
The use of a comprehensive income tax benchmark for superannuation is consistent with the general principle adopted in the Tax Expenditures Statement that benchmarks should represent the standard taxation treatment that applies to similar taxpayers or types of activity (see page 134). Superannuation is a form of saving and the earnings on all other forms of saving (with the exception of owner-occupied housing) are taxed as income. This includes interest on bank accounts, term deposits and debentures, investment returns on shares and managed funds, rental income and net capital gains.5
The exception of saving through owner-occupied housing is not comparable to superannuation, since the former is not simply a savings vehicle but also provides current consumption (i.e. a place to live). Instead, superannuation more closely resembles other financial investments which are all taxed under some form of income tax benchmark. For example, savings in a bank account are made from after-tax income, interest on the account is taxed at marginal rates, and withdrawals from the account are untaxed.
While the Tax Expenditure Statement applies the comprehensive income tax benchmark to superannuation for these reasons, the 2013 Tax Expenditure Statement (Appendix A, page 193-4) included experimental estimates of superannuation tax expenditures using an expenditure tax benchmark. This was done in response to interest around the choice of benchmark and to facilitate discussion and understanding of the impact of utilising different benchmarks.
A.2 Choice of alternative investments
The use of the comprehensive income tax benchmark and the need to measure the income tax that would otherwise have been paid on investment income requires some assumption of where people would otherwise have invested their savings. Questions are sometimes raised as to the treatment of investment income under this counterfactual scenario, particularly with regard to people’s ability to minimise their marginal tax by investing in alternative tax-effective investments.
The revenue forgone estimates are based on the assumption that the assets held outside of the superannuation system would have the same portfolio allocation as those held inside the superannuation system, and earn the same before-tax rates of return. This includes assets with deferred capital gains such as shares and property. Imputation credits as well as the costs of generating earnings such as insurance, rates and maintenance for property are also taken into account in the estimates.
The revenue gain estimates, which allow for a behavioural response, make further allowances for the fact that people can minimise their marginal tax by making use of trusts or company tax structures or by investing in negatively-geared assets, lower-taxed foreign jurisdictions or owner-occupied housing.
A.3 Accounting for effects on the Age Pension
One effect of superannuation is to reduce outlays on Age Pension. Some commentary argues that these expenditure savings should be recognised in the estimates of superannuation tax expenditures.
Tax expenditures seek to measure the deviation in the actual tax treatment of an activity or class of taxpayer from the benchmark tax treatment (see page 134). They are therefore a more limited construct than a budget costing and, by their nature, do not seek to measure the full budgetary impact on related current or future government expenditure since these are not taxes.
Other forms of analysis from time to time cover these broader considerations. For example, a 2013 analysis estimated the fiscal impact of a phased increase in the Superannuation Guarantee rate from 9 per cent to 12 per cent, including the balance between tax losses and Age Pension savings (see Charter Group, A Super Charter: Fewer Changes, Better Outcomes: A report to the Treasurer and Minister Assisting for Financial Services and Superannuation, 2013, page 11).
A.4 Reliability of estimates
Why do some of the estimates of superannuation tax expenditures exhibit considerable volatility, both over time and with the annual revision of estimates?
Some drivers of the superannuation tax expenditure estimates, such as Superannuation Guarantee contributions, are reasonably stable and predictable. However, a number of the estimates rely on either forecasts of equity returns or forecasts of individual behaviour based on those equity returns (for example, discretionary superannuation contributions). These elements are difficult to estimate and are typically volatile. For example, estimates of the tax expenditure on superannuation earnings were particularly volatile around the time of the global financial crisis. This uncertainty is why the reliability of the estimate for this tax expenditure is clearly identified as ‘low’ in the Tax Expenditures Statement.
5 The 50 per cent capital gains tax discount for individuals holding a CGT asset for more than 12 months is itself a tax expenditure.