Chapter 1: Introduction


A tax expenditure arises where the actual tax treatment of an activity or class of taxpayer differs from the benchmark tax treatment.

  • Tax expenditures typically involve tax exemptions, deductions or offsets, concessional tax rates and deferrals of tax liability.
  • A positive tax expenditure reduces tax payable relative to the benchmark. A negative tax expenditure increases tax payable relative to the benchmark.

Benchmarks represent a standard taxation treatment that applies to similar taxpayers or types of activity.

  • Benchmarks may also incorporate structural elements of the tax system; for example, the progressive income tax rate scale for individual taxpayers.
  • The benchmarks used in the 2015 TES are outlined in Appendix B.

Determining benchmarks involves judgment. Consequently, the choice of benchmark may be contentious and benchmarks may vary over time. The choice of benchmark should not be interpreted as indicating a view on how an activity or taxpayer ought to be taxed.

  • To facilitate discussion and understanding of the impact of using different benchmarks, the 2013 TES included an illustrative case study which showed the differences in the estimates for superannuation tax expenditures if an expenditure tax benchmark was used rather than the usual income tax benchmark — see Appendix A, 2013 Tax Expenditures Statement. Although that exercise has not been repeated for this year’s TES, the conceptual points that were discussed in the 2013 TES remain.

The Tax Expenditures Statement

The annual TES provides a description of Australian Government tax expenditures and, where possible, the estimated value or order of magnitude of the tax expenditure.

The TES is intended to facilitate scrutiny of tax expenditures by Parliament and parliamentary committees, the media and the general public. Transparent reporting of tax expenditures also helps inform debate on the efficiency and equity of the tax system.

This TES reflects Australian Government policy up to and including the 2015-16 Mid-Year Economic and Fiscal Outlook.

Estimating tax expenditures

Consistent with most OECD countries, estimates of the size of tax expenditures in the TES reflect the existing utilisation of a tax expenditure, similar to Budget estimates of outlays on demand-driven expenditure programmes. This is known as the ‘revenue forgone’ approach.

  • For example, Budget Paper No.1, Budget Statement 5 — Expenses and Net Capital Investment, reports expenditure on the age pension. Broadly, the amount reported reflects the number of age pension recipients and the amount of pension each receives. Budget Paper No. 1 does not provide any estimate of the hypothetical saving to the Budget should the expenditure cease.
  • Similarly, the estimated size of a tax expenditure reflects the number of taxpayers utilising the tax expenditure and the notional amount of tax expenditure each receives. Revenue forgone estimates do not indicate the hypothetical saving to the Budget should the tax expenditure cease.

In practice, the revenue forgone approach involves estimating the difference in revenue between the existing and benchmark tax treatments, but importantly assuming taxpayer behaviour is the same in each circumstance.

This approach can be illustrated using the GST exemption for water, sewerage and drainage services (H6).

  • By definition, no GST revenue is raised under the existing tax treatment. The benchmark treatment is the imposition of GST on water, sewerage and drainage services. The estimated value of the tax expenditure is therefore the amount of GST revenue that would be raised on water, sewerage and drainage services assuming that consumption of these services remained unchanged under a GST.

An alternative approach involves estimating the impact of abolishing a tax expenditure taking account of the potential changes in taxpayer behaviour, unlike revenue forgone estimates. This is known as the ‘revenue gain’ approach. Because they take account of behavioural responses, revenue gain estimates are often lower than revenue forgone estimates.

  • Introducing a tax expenditure may create incentives for taxpayers to change their behaviour to utilise (or avoid) the new tax provision. Removing the tax expenditure (so that the benchmark tax treatment prevailed) would remove this incentive and may cause a corresponding change in taxpayer behaviour.

In particular, taxpayers may make greater use of other tax expenditures if a particular tax expenditure were to be (hypothetically) abolished.

  • For example, a revenue gain estimate for the concessional treatment of employer superannuation contributions would take account of the potential for voluntary employer contributions to be redirected to other tax-preferred investments.

Revenue gain estimates should be treated with particular caution.

  • They assume that a tax expenditure is abolished, which may be implausible in many cases.
  • In practice, the revenue gain can be difficult to estimate as there is usually little, if any, information on how taxpayers might react to the removal of a tax expenditure. Assumptions about taxpayer behavioural responses therefore need to be made, and these assumptions can be difficult to meaningfully substantiate.
  • Judgments also need to be made about likely policy settings — for example, whether it is realistic to assess the abolition of a single tax expenditure (for example, a particular GST exemption) while keeping other tax expenditures unchanged (for example, other GST exemptions).
  • Revenue gain estimates do not take into account any potential changes in direct expenditure flowing from the removal of a tax expenditure.

Consistent with a recommendation of the Australian National Audit Office in its 2007-08 performance audit of the TES, the TES reports revenue gain estimates for 10 large tax expenditures.1

Interpreting tax expenditure estimates — additional caveats

Tax expenditure estimates in different editions of the TES are generally not comparable.

  • Estimates may change between editions as benchmarks are modified, tax expenditures are modified, revised or new data becomes available, or changes in modelling methodology are made.

Readers should exercise care when comparing tax expenditure estimates with direct expenditure estimates.

  • Tax and direct expenditure estimates may measure different things. For example, the tax expenditure estimate for the Private Health Insurance Rebate (A17) relates to the tax exemption for the rebate, not the rebate itself.
  • Direct expenditure estimates of non-taxable transfer payments effectively include the value of the tax exemption for the payments. Summing the direct and tax expenditure estimates would therefore overstate the cost of the government support to the budget.

It is not appropriate to aggregate revenue forgone estimates. As indicated above, revenue forgone estimates do not take account of potential changes in taxpayer behaviour following the (hypothetical) removal of a tax expenditure. However, in reality such changes in behaviour would be likely to occur — in particular, the removal of one tax expenditure would often affect the utilisation of other tax expenditures. Aggregating revenue forgone estimates therefore risks significantly amplifying the limitations inherent in this method of estimating the size of tax expenditures.

Reliability of tax expenditure estimates

Tax expenditure estimates vary in reliability depending on the quality, detail and frequency of the underlying data and the consequent extent to which calculations mu
st be based on assumptions, and the sensitivity of estimates to those assumptions.

Importantly, the TES reports estimates for future years. In many cases, this unavoidably reduces their reliability because of the inherent uncertainty around forecasts of future economic conditions.2 Estimates with higher reliability tend to be those where future taxpayer behaviour is relatively more predictable because of longstanding stable trends in the historical data, or where only estimates based on historical data are reported.

Unquantifiable estimates

In many cases there is insufficient data to produce a reliable estimate for a tax expenditure item. While Treasury has access to detailed tax data collected by the Australian Taxation Office from tax returns, the scope of this data is limited by the number and nature of questions on the tax return itself. While expanding the tax return could increase the data available for estimating tax expenditures, this would increase compliance costs for taxpayers. Treasury also utilises, for example, Australian Bureau of Statistics data where relevant.

In the 2015 TES, estimates are not available for 2015-16 for around 48 per cent of tax expenditures — that is, 140 out of 290 expenditures.

Large tax expenditures

The 2015 TES reports 290 tax expenditures. 150 of these estimates are able to be quantified. Table 1.1 lists the largest measured tax expenditures for 2015-16.

The table includes revenue gain estimates for several of the largest tax expenditure items. These estimates illustrate the points made above that:

  • significant differences can arise between revenue forgone and revenue gain estimates, particularly because the latter attempts to take account of behavioural change by taxpayers; and
  • conversely, in some cases, revenue gain and revenue forgone estimates are identical or very similar as taxpayer behaviour is assumed to be relatively insensitive to a tax expenditure.

Unquantified tax expenditures have been assigned an order of magnitude rather than an estimate of their value. The largest such tax expenditures are as follows:

  • income tax exemption for prescribed entities (B50);
  • exemption for foreign branch profits from income tax (B10);
  • off-market share buy-backs (B26);
  • statutory effective life caps (B71); and
  • quarantining of capital losses (E28).

Table 1.1: Large measured tax expenditures for 2015-16

Estimate $m
Tax expenditure Revenue
Large positive tax expenditures
E6 Capital gains tax main residence exemption — discount component 30,000 n/a
E5 Capital gains tax main residence exemption 24,500 n/a
C3 Concessional taxation of employer superannuation contributions 16,250 15,600
C6 Concessional taxation of superannuation entity earnings 13,550 12,600
H28 GST — Food 6,800 6,600
E11 Capital gains tax discount for individuals and trusts 6,150 n/a
H16 GST — Education 4,200 3,800
H19 GST — Health — medical and health services 3,700 3,650
H2 GST — Financial supplies — input taxed treatment 3,250 3,250
A35 Exemption of Family Tax Benefit payments 2,250 2,250
C5 Concessional taxation of non-superannuation termination benefits 2,050 2,050
A19 Medicare levy exemption for residents with taxable income below the low-income thresholds 1,880 n/a
B13 Exemption from interest withholding tax on certain securities 1,730 n/a
A17 Exemption of the Private Health Insurance Rebate 1,510 n/a
D10 Exemption for public and not-for-profit hospitals and public ambulance services 1,450 n/a
D14 Exemption for public benevolent institutions (excluding public and not-for-profit hospitals) 1,450 n/a
A24 Exemption of Child Care Assistance payments 1,390 n/a
B3 Local government bodies income tax exemption 1,350 n/a
H5 GST — Child care services 1,270 n/a
A49 Philanthropy — deduction for gifts to deductible gift recipients 1,210 n/a
F7 Concessional rate of excise levied on aviation gasoline and aviation turbine fuel 1,190 n/a
H20 GST — Health — residential care, community care and other care services 1,100 n/a
B73 Capital works expenditure deduction 1,000 n/a
H6 GST — Water, sewerage and drainage 950 n/a
B78 Research and development — non-refundable tax offset 850 n/a
Large negative tax expenditures
F11 Higher rate of excise levied on cigarettes not exceeding 0.8 grams of tobacco -1,995 n/a
F23 Customs duty -1,950 -1,950

Compared to the list of large tax expenditures in the 2014 TES, the biggest change in estimates are to the two tax expenditures relating to the main residence exemption from capital gains tax (E5 and E6). These have increased significantly between the 2014 TES and 2015 TES as a result of the higher than expected increase in housing prices in 2014-15.

1 ANAO Audit Report No. 32, 2007-08, Preparation of the Tax Expenditures Statement, Recommendation 5.

2 For more information on these uncertainties, see 2015-16 Mid-Year Economic and Fiscal Outlook, Part 3, Fiscal Strategy and Outlook, Attachment A, Forecast Uncertainties, Sensitivities and Scenarios.