Australia's Future Tax System

Architecture of Australia's tax and transfer system

3.1 The economic impacts of the tax‑transfer system

Many taxes but only three tax bases

Individual tax bases are often described in terms of the nature of the tax and the things that are taxed. For example, payroll tax applies to the payrolls of business (subject to a threshold), land tax applies to land used for particular purposes, the GST applies to the consumption of particular goods and services and the queen bee levy applies to the production and export of queen bees.

All taxes, whether payroll tax, land tax, GST or the queen bee levy, are ultimately borne by individuals on the earnings from only three factors of production: labour, capital and land (including natural resources). Taxes are levied either on the income derived from these factors or on the use of that income to consume goods and services. Individuals end up paying taxes in a range of ways, including as consumers through higher prices, as employees through lower wages, or as shareholders or investors through lower profits. Chart 3.1 provides a schematic overview of how some of the main Australian taxes relate to income derived from these factors of production.

Chart 3.1: Relationship between economic bases and taxes

Chart 3.1: Relationship between economic bases and taxes

It does not matter whether the taxes are levied on activities, entities, goods or transactions, all taxes are ultimately paid out of the incomes of individuals. See Box 3.1 for a discussion of the relationship between income derived from the three factors of production and the use of that income.

Box 3.1: The relationship between sources and uses of income

The earnings from the three factors of production — labour, land (including natural resources) and produced capital — comprise the income of individuals. This includes income received by individuals in the form of transfers from the government, as such payments are derived from taxes levied on the earnings from the factors of production.

The aggregate of each individual's income from labour, land and produced capital represents the income of the nation. This income can be either consumed or saved to finance consumption in a later period (see Chart 3.2).

Chart 3.2: Sources and uses of income

Chart 3.2: Sources and uses of income

The equivalence of income to consumption and saving, means that in the long run taxes on consumption (such as the GST and excise) can be broadly thought of as taxes on labour (see Appendix B). For example, a payroll tax that reduces the after‑tax income of a worker means the worker has less money to buy goods and services. The payroll tax can therefore be viewed as a tax on labour income or, equivalently, as a tax on the consumption of goods and services. In theory, a tax on goods and services and a payroll tax should therefore have similar effects on the incentive to work, since both reduce the goods and services that can be purchased through working. Indeed, all taxes — from small transaction based taxes to broad based income taxes — tax people's jobs (by taxing their wage income), their savings (by taxing their return to saving) or the returns they derive from natural resource endowments. In practice, however, the impacts of direct taxes on labour income and taxes on goods and services may differ. This reflects imperfect information about the ultimate burden of different taxes and the tendency for individuals to make decisions about complex matters such as the burden of different taxes using relatively simple procedures.

The aggregate of each individual's income from these sources represents the income of the nation (see Box 3.2 for a discussion of alternative measures of national income). The exact breakdown of national income between the three factors is difficult to determine for a number of reasons. In particular, many small businesses only report overall income and it is difficult to determine how much of that reflects a return to invested capital and how much is a return to labour. Returns to land and resources are also difficult to distinguish from returns to capital. However, it is clear that more than half of national income comes from labour.

Box 3.2: Measuring national income

The Australian Bureau of Statistics produces a number of indicators of national income. Different ways of measuring national income are suitable for different purposes. For example, there are important differences between income generated in Australia and the income actually accruing to Australians. Some of the key measures of national income are discussed below.

Nominal gross domestic product (GDP) is the value of all goods and services produced in Australia. Nominal GDP is the major driver of revenue collections. However, it is not a good measure of wellbeing because it includes price changes, which can lift nominal GDP but not raise Australians' actual spending power.

Real GDP shows the volume of goods and services produced in the economy by removing price impacts from nominal GDP. Real GDP is a measure of the productivity and participation outcomes in the economy. However, by removing all price impacts, it ignores the income changes that arise from changes in the terms of trade (as Australia has experienced during the recent boom in export prices).

Real net national disposable income (RNNDI) shows the amount of income accruing to Australians that can be consumed without diminishing the economy's productive capital. It is derived through several adjustments to real GDP. Adjusting real GDP for the income effect of the terms of trade results in real gross domestic income (GDI), which has grown significantly faster than real GDP during the terms of trade boom. Net income accruing to foreigners is then subtracted (resulting in real gross national income (GNI)), reflecting the fact that this income is not available to raise Australian wellbeing. Net income accruing to foreigners has grown strongly during the recent terms of trade boom. Finally, depreciation of the capital stock is deducted to achieve RNNDI. As RNNDI is a real measure, it captures the capacity of Australians to consume actual goods and services.

Chart 3.3: Alternative measures of Australian income and output

Chart 3.3: Alternative measures of Australian income and output

Source: ABS cat. no. 5206.0, National Income, Expenditure and Product, March 2008.

Tax and transfer incidence

Who bears the burden of a tax bears the 'incidence' of a tax. 'Legal incidence' is borne by the person who is required to pay the tax to the administrative authority. 'Economic incidence' is borne by the person who ultimately bears the cost burden of the tax. The economic incidence of a tax is less apparent than its legal incidence, but it is the economic incidence that is important when considering the efficiency and equity implications of the tax‑transfer system.

Taxes can be shifted from one person to another through changes in the prices of inputs to the production process, through changes in the price of goods produced, or through changes in the distribution of the returns to economic activity.

There is no authoritative view as to the economic incidence of our tax‑transfer system. Determining precisely the economic incidence of a tax is extremely difficult. Further, the interrelationships between the influences that determine incidence need not be static, due to changing economic conditions and tax‑transfer settings, so that the economic incidence of a tax may vary over time.

While determining the exact incidence of the tax system or a specific tax is problematic, economic relationships provide insights into the likely economic incidence of at least some taxes (Box 3.3). Some generally accepted outcomes are that the economic incidence will fall to a greater extent on:

  • a good or factor, the demand or supply for which is unresponsive to a change in its price;
  • a good with no ready substitutes; and
  • a factor of production that is relatively immobile.

Where a tax is levied on a good or factor which does not have these characteristics, or on a business entity, the burden of the tax will tend to be shifted to goods or factors of which the demand or supply is relatively less price responsive, less substitutable or less mobile.

For example, foreign investment is generally considered to be more mobile than the resident labour force. It is generally accepted that for existing investment, the incidence of a change in company tax falls on shareholders in the short‑run. However, in the longer run, it is more likely the incidence falls on land and labour, particularly where domestic consumers can substitute foreign goods for goods produced by the resident company. Under these conditions, a high rate of tax on capital income may discourage some new capital investment resulting in a smaller stock of productive capital. With reduced capital investment, average labour productivity could be expected to fall, and with lower labour productivity wages would be likely to be lower. It is through lower wages that, in the longer run, labour may bear the economic incidence of the company tax.

Similarly, it can be difficult to determine who ultimately benefits from some forms of transfer payment. In most cases the incidence of untied cash transfers to individual recipients is likely to fall on those individuals, with the possible exception of transfers to the elderly, which may also benefit future generations by allowing more savings to be passed on as bequests. Where cash payments are tied to the purchase of particular goods or services, as with rent assistance and public transport, part of the benefit of the concession may be captured by the provider of the goods or services in the form of higher prices to which the concession is applied, depending on the demand response.

Box 3.3 Understanding the economic incidence of taxes

Consider a market for a factor (labour, capital or land, including natural resources) or for goods and services (intermediate or final). A tax, whether paid by the buyer or the seller, places a wedge between the buyer price (Pb) and the seller price (Ps).

Chart 3.4: The effects of taxation

Chart 3.4: The effects of taxation

This results in:

  • a fall in quantity produced and consumed from Qo to Qt (unless one of demand or supply is perfectly inelastic — that is, unresponsive to a change in price)
  • a rise in the price paid by the buyer to Pb (unless supply is perfectly inelastic or demand is perfectly elastic)
  • a fall in the price received by the seller to Ps (unless demand is perfectly inelastic or supply is perfectly elastic)
  • a transfer of revenue to the government (of area A + B)
  • losses of economic value to buyers (of area A + C) and sellers (of area B + D)
  • a net loss of economic value (C + D — which is equal to the loss of producer and consumer surplus less the revenue transfer to the government), referred to as the 'efficiency cost' of taxation.

The share of the final economic incidence of the tax borne by buyers and sellers depends on the relative price responsiveness of the demand and supply curves, with the less elastic side bearing most of the final burden. To minimise the loss of economic surplus, taxes need to be higher on relatively inelastic goods.

Note: Strictly speaking the demand curve represented in Chart 3.4 should be the compensated demand curve.

Incentive and income effects

All taxes — whether on labour, capital or land and natural resource income — have two types of effects.

  • First, taxes affect individuals' incentives by encouraging them to shift from taxed to untaxed activities and goods or from heavily taxed to lightly taxed activities and goods (that is, the substitution effect). For example, taxes on wages may discourage people from working additional hours, or encourage them to find alternative forms of remuneration that are not taxed. Taxes on capital income encourage people to save less or shift their savings into vehicles which are taxed less. The loss in economic value due to these incentive effects is what economists call the efficiency costs of taxation (see Box 3.4).
  • Second, taxes reduce individuals' incomes, which can also affect their behaviour. For example, taxing wages might encourage an individual to work longer hours if they desire a given level of income to meet their spending requirements. Similarly, taxing the return to saving might encourage individuals to increase their level of savings to achieve a desired level of income in retirement.

The incentive and income effects of taxes can interact in complicated ways. For instance, a higher tax on labour income would have an incentive effect of discouraging people from working, while the income effect may encourage more work, to ensure that the individual or family has enough after‑tax income to meet their expenditure needs.

Government spending, particularly on transfers to individuals, can also have important incentive and income effects. Transfers act like reverse taxes. Rather than raising money in ways that affect individuals' incentives and reduce their after‑tax incomes, transfers give money in ways that affect individuals' incentives and increase their after‑transfer incomes. For example, transfers can reduce the incentive to work, particularly means tested payments, and to save.

Government spending, particularly on transfers to individuals, can also have important incentive and income effects. Transfers act like reverse taxes. Rather than raising money in ways that affect individuals' incentives and reduce their after‑tax incomes, transfers give money in ways that affect individuals' incentives and increase their after‑transfer incomes. For example, transfers can reduce the incentive to work, particularly means tested payments, and to save.

Box 3.4: Efficiency costs of taxes and transfers

Tax revenue is used by government to fund goods and services, including transfers — past, present and future. The revenue raised by government is not a cost to society as a whole. Revenue collections are transferred from one set of Australians to another through the tax‑transfer system and the broader functions of government. The impact of this transfer on wellbeing depends upon the value assigned by individuals to the goods and services provided by government.

In contrast, efficiency costs represent losses to the Australian community. The vast majority of taxes and transfers affect the choices that individuals and businesses make by altering incentives to work, save, invest or consume things that are of value to them. Individuals and businesses generally respond to taxes by choosing more of lower taxed items and less of higher taxed items than they otherwise would. (They may respond to transfers in ways that increase the payment they receive.) These changes in behaviour can ultimately leave the economy and society as a whole worse off than if the revenue had been raised (or distributed) without affecting their behaviour. It is this consequential loss of value that is referred to by economists as efficiency costs.

The size of these efficiency costs varies across different taxes and transfers. For example, taxes that alter production decisions will tend to have higher efficiency costs than taxes that alter consumption decisions. This is because such taxes alter both the mix of business inputs used to produce a good or service, as well as the final consumption choices of individuals. At a system level, the efficiency costs of taxation in Australia have been estimated to be around 6 per cent of GDP (Freebairn 1998). These estimates are broadly consistent with 'rules of thumb' developed from studies of the efficiency costs of taxation in the United States (Government Accounting Office 2005). Additional efficiency costs could be expected to result from the effects of the transfers system on individuals' choices. However, it is the marginal efficiency cost of raising the last dollar of revenue from a particular tax that is relevant for comparing the efficiency of different taxes or additional spending. Some taxes, such as taxes on economic rent, are likely to have negligible marginal efficiency costs if well designed, whereas the marginal cost of other taxes can be high.

There are also additional costs of raising revenue. These are the costs of administering the tax‑transfer system and the costs to taxpayers and transfer recipients of complying with the requirements of the system. Administration and compliance costs are considered further in Section 11.

The existence of these efficiency, administration and compliance costs does not automatically imply that reducing taxes will result in increased GDP or social wellbeing. Provided that the goods and services supplied by government are of sufficient value to society to offset these costs, the overall wellbeing of society is enhanced. It may, however, be possible to reduce efficiency costs by altering how some taxes are used to raise revenue.

Minimising efficiency costs through tax design

In the absence of administration and compliance costs, and with perfect information, an optimal design for a tax system would be where the rate of tax is inversely related to the responsiveness of economic choices to the imposition of the tax. That is, minimising the efficiency costs of taxation would require higher tax rates on factors of production, the supply of which is relatively unresponsive to the rate of tax, and on goods and services for which demand is relatively unresponsive to the rate of tax.

Such a tax system is impractical, due to deficiencies in our understanding of the efficiency impacts of individual taxes and because the considerable costs of administration and compliance associated with such an approach would likely outweigh much of the gain in economic efficiency. The principles underlying this approach are, however, applied in many aspects of tax design. For example, high rates of excise are applied to goods for which demand is relatively unresponsive to the rate of tax and for which there is a degree of control over the supply of untaxed production. Similarly there is a widely held view in academic circles that capital should be taxed at lower rates than labour or consumption, particularly for a small, open and geographically isolated country like Australia, because of the higher international mobility of capital relative to labour.

It is also this general principle that has underpinned an OECD wide trend towards financing lower rates of tax by broadening the tax base and applying a uniform rate of taxation across the tax base. One rationale for this approach is that if all goods and services are taxed at the same rate, relative prices will be unaffected and therefore there will be less impact on the decisions of individuals and firms. A supporting rationale is that efficiency costs tend to increase more than proportionately with the rate of tax and, hence, for a given revenue target efficiency costs will be lower where revenue is raised across a broad base.

In applying this approach, however, a key question is how widely the base to which a uniform rate of tax is applied should be defined. For example, is it more efficient to apply a uniform rate of income tax across all forms of income, or to have differentiated rates across different types of income, such as that accruing to labour, to the extraction of natural resources and to capital?

If there were no limitations to our understanding of the efficiency implications of different tax settings and if economic efficiency were our only policy objective, it would be appropriate to differentiate the tax base so long as the efficiency benefits exceeded any additional administration and compliance costs. In reality, information constraints and other policy objectives will be important determinants of the extent of differentiation in the tax base.

Inflation may also affect the efficiency of the tax system. Some of its impacts are outlined in Box 3.5.

Box 3.5: Inflation and the tax system

A major issue in the Asprey Report (1975) was the adverse interaction between a nominal tax system and inflation on incentives, especially for savings and investment. This was at a time when inflation was over 10 per cent. The 1985 draft White Paper also had a significant focus on the impacts of inflation, including reasons why tax on income such as capital gains should exclude the inflationary component. There have been major developments in monetary policy over the last 20 years, including independence of the Reserve Bank in setting interest rates to target an inflation range of 2 to 3 per cent. This has helped reduce the negative interactions between tax and inflation, especially unanticipated changes in inflation. However, even with lower and more stable rates, inflation can have significant impacts on the incentives created in the tax system. Its impacts can be very different for labour and capital income.

Australia does not have a formal system for annual indexation of income tax thresholds like many OECD countries but the frequent adjustments to thresholds over the last decade have more than compensated for inflation over the past two decades. Australia generally indexes its pensions and allowances to at least compensate for inflation.

Inflation can represent a very significant part of the return to capital. For instance, with an inflation rate of 2.5 per cent and nominal interest rates of 6 per cent, inflation accounts for some 40 per cent of the amount assessed as income for tax purposes. This proportion has changed over time, varying from up to 90 per cent in the 1970s, through to around 30 per cent in recent years. The inflation component of the return to capital does not increase the purchasing power of the investor. Taxing the inflation component under a nominal income tax results in an effective tax rate on the change in purchasing power that is higher than the nominal income tax rate.