Australia's Future Tax System

Final Report: Detailed Analysis

Chapter E: Enhancing social and market outcomes

Key points

Not all taxes should be imposed primarily to raise revenue. Sometimes governments intervene in markets to achieve more efficient or equitable outcomes. In the case of goods or services provided by government, direct user charging rather than taxpayer funding can lead to more efficient outcomes. Taxes can also be used to influence the prices faced by the consumers or producers of particular goods or services.

Markets generally allocate resources to their highest value use. But markets do not always arrive at the prices that are best for society. Individuals might make decisions based on their own private interests — including the prices they pay — but fail to take into account costs that spill over onto others. An example is traffic congestion — where motorists will often choose the quickest route for themselves, but not consider the impact of their decisions on overall levels of congestion.

Tax can sometimes be used as a tool to align private incentives with social incentives. If a socially costly activity becomes more expensive because of the tax, the level of that activity generally falls. But not all market failures or spillover costs can be effectively addressed using a tax. Sometimes other approaches, like better defined or enforced property rights, would have better outcomes. In other cases, the lack of information needed to ensure the tax rate is efficient, or the cost of getting it, may limit the usefulness of the tax.

Improving the efficiency of markets

The most efficient way of financing some government-provided goods or services is not to tax at all, but rather to charge users a price that reflects the avoidable cost of providing the good or service. The practicality of user charging partly depends on the degree to which those who are not willing to pay can be excluded from receiving the benefit. This may change over time as new property rights are clarified or new technology becomes available to cheaply restrict access or monitor use. The circumstances in which user charging is appropriate are discussed in Section E1 User charging.

Where direct user charging is not possible, tax can sometimes be used as a tool to improve efficiency, if it can more closely align the price of an activity to the individual with the cost of the activity to society (that is, the social cost). Private costs may deviate from social costs as a result of unpriced spillovers (sometimes called externalities). A spillover occurs when a person's actions impose involuntary costs (or benefits) on others. That is, in addition to the private costs and benefits that accrue to the decision-maker, some costs and benefits can 'spill over' on to others. By imposing taxes that reflect the spillover costs, it is sometimes possible to bring the private cost of the activity (which now includes a tax) closer to the social cost (which counts costs and benefits to everybody). Such taxes give people incentives to make better decisions — about when and where to drive, how much alcohol to consume or whether or not to smoke tobacco (see Box E–1: Where tax avoidance is good for society).

Box E–1: Where tax avoidance is good for society

Many taxpayers are happy to contribute their fair share to society, but almost everyone faces incentives to pay less tax. Tax administrators are constantly identifying schemes, loopholes and arrangements that people use to reduce their tax liability. Governments are constantly responding to this by amending the tax law — often making it more complicated and costly for compliant taxpayers.

But there are some taxes that the government wants you to avoid — for example, you can minimise alcohol tax or tobacco excise by drinking less or quitting smoking. Sometimes the government even helps you to do this. For example, government subsidies to help people quit smoking help people avoid tax. Here, the health of Australians is more important than the tax they pay.

A congestion tax works in the same way. People would reduce or avoid it by sharing rides to work, cutting down on unnecessary trips, and using public transport where possible. The government might also help by building additional rail lines or putting on extra buses for otherwise car-dependent communities.

Some trips at busy times will still depend on cars — for these drivers, the tax would be unavoidable. But they should benefit from faster, more reliable trips, as some of those who do have the chance to avoid the congestion tax are no longer on the road. For example, a tradesperson would spend less time in a van, and more time on paying jobs.

While it is easy to identify many activities with spillover costs, a tax will only be an efficient instrument to deal with them if the marginal spillover cost of the activity can be estimated. A tax that recovers the total spillover costs of an activity from those who cause the harm might be seen as equitable (such as a tax imposed under the polluter-pays principle), but it is only efficient if it provides incentives to improve behaviour. If the costs are high but uneven, other instruments such as regulations may be more appropriate (see Box E–2: A fat tax?)

For example, a tax on drinkers to cover the full spillover costs of alcohol abuse may be seen as equitable, but to be efficient it also needs to encourage drinkers to cut down on levels of drinking that do more harm than good to society.

Box E–2: A fat tax?

The National Preventative Health Taskforce (2009, p. 104) has called on the Australian Government to conduct research into policies and tax incentives that would promote the production, access to and consumption of healthier foods. However, this would not be a simple case of imposing a per unit tax on fat, sugar or salt in food.

While obesity does involve significant health and productivity costs, the relationship between these costs and the consumption of particular products is complex. The risk of obesity is affected by lifestyle, such as diet and physical activity, as well as inherited and social influences.

This makes it very difficult to estimate spillover costs, if any, of identifiable foods or food types. In addition, any quantifiable health benefits of imposing the tax would need to be weighed against the loss to those people who are at low risk.

For reasons of practicality, a tax is usually applied only on inputs to the behaviour (an easily measurable commodity) rather than the amount of the spillover cost itself. The less closely related the taxed input is to the spillover cost, the less efficient the tax will be in creating incentives to reduce the costs. For example, taxing petrol to reduce urban congestion is not particularly effective, even though fuel is used in cars that contribute to congestion. New technology might allow better measurement or targeting of spillover costs.

Market-correcting taxes do not provide a 'free lunch' or costless revenue for the government. Taxes that cannot directly target spillovers will also impact on and reduce activity that does not give rise to spillovers. For example, a tax on alcohol reduces the spillover costs from abusive consumption, but it also reduces the level of consumption and satisfaction of non-abusive consumers. In determining the net benefits of the tax, governments should consider these broader effects.

Unpriced spillover costs are prevalent in the areas of the environment (see Section E2 Taxes to improve the environment), road transport (see Section E3 Road transport taxes), alcohol (see Section E5 Alcohol taxation) and tobacco (see Section E6 Tobacco taxation).

Improving individual welfare

As well as addressing spillover costs on other people, governments sometimes seek to influence people's choices for the purpose of improving their individual long-term wellbeing. This is particularly the case for people with reduced capacity, possibly through reason of age or addiction, to make and execute decisions consistent with their long-term wellbeing.

Addiction is a complex phenomenon, which involves, among other things, highly compulsive use, use despite harmful effects, relapse following abstinence, and cravings (Collins & Lapsley 2008b, p. 10). These effects mean that particularly addictive goods can impair the capacity of individuals to weigh their immediate choices against future consequences. Some goods and services, including alcohol, tobacco and gambling, can be highly addictive for some consumers (see Box E–3: Comparing three potentially harmful and addictive commodities).

Governments use many tools to influence choices, including public education, direct interventions in behaviour, product regulation, product bans and taxation. These interventions have different effects both on the consumption opportunities of individuals, and on their chances of living lives that they value.

Taxation has the potential to improve people's long-term wellbeing if it provides a price signal that reflects the potential for future harm and possible addiction. People may consume a product excessively if they lack information to assess its harmful properties or if they underestimate the difficulty of giving up an addictive good. A tax can bring these costs forward to the present and potentially prevent a pattern of addiction. But, for a tax to be appropriate, the costs and addictiveness of a particular commodity, and the tendency for people to over-consume it, would have to be both well-understood and be broadly uniform across different individuals.

Non-tax regulations to address these problems can also restrict competition, and therefore create economic rent for suppliers of these goods. Taxation might be capable of capturing this rent.

Box E–3: Comparing three potentially harmful and addictive commodities

Alcohol, tobacco and gambling have different consumption patterns and different effects on their users. The proportion of the population that uses these commodities, uses them at risky levels, and suffers significant damage as the result of using them, varies widely from commodity to commodity. In particular:

  • Around 83 per cent of Australians aged over 14 consume some alcohol; around 20 per cent sometimes place themselves at risky or high-risk levels of short-term harm, and around 10 per cent place themselves at risky or high-risk levels of long-term harm (AIHW 2007, p. 37). Surveys from the United States suggest that the top 2.5 per cent of drinkers account for around a quarter of alcohol consumed (Greenfield & Rogers 1999).
  • Around 19 per cent of Australians aged over 14 smoke tobacco (at least occasionally); of whom around 86 per cent smoke daily (AIHW 2007, p. 25). A long-term study of smokers in the United Kingdom found that 'about half of persistent cigarette smokers would eventually be killed by their habit' (Doll et al. 2004).
  • Around 15 per cent of adults gamble regularly, excluding lotteries and scratchies. Of these, roughly one in ten are problem gamblers, who are responsible for a large share of player loss (Productivity Commission 2009a, p. 4.1).

The taxation of such commodities is a complex area, both theoretically and practically. While prices and taxes undoubtedly influence aggregate patterns of consumption — potentially improving the wellbeing of some — they are not well suited to deal with extraordinary forms of human behaviour, and in complex situations can give rise to adverse, unintended consequences. The impact may often fall on the most vulnerable members of the community for whom direct and personal intervention, rather than population-level treatment, is most appropriate.

The appropriateness and efficacy of taxes on alcohol, tobacco and gambling are considered in Section E5, Section E6, and Section E7 respectively.

Equity and compensation for specific taxes

Taxes that are designed to achieve specific social or market outcomes depend on creating incentives for people to adjust their behaviour. They are typically designed without reference to distributive effects, as attempts to shield one group from the effect of the tax means that the spillover costs continue to be borne by another group.

The distributional consequences of market-correcting taxes must, of course, be taken into account. In many cases, it may be judged desirable to compensate affected groups for the loss of purchasing power. However, this should be done using the most appropriate tool. The personal tax and transfer system is generally the most effective way of redistributing income through society, and therefore will often be an appropriate way to compensate taxpayers and transfer recipients for increased taxation on specific goods and services to reduce the costs of market failures.

In other cases, specific compensation to particular groups might be provided by way of specific outlays — for example, additional investment in public transport infrastructure in specific locations may provide a geographically targeted form of compensation for congestion charges.

Another equity-based argument for taxing specific commodities is that the social costs of an activity that harms some people should be borne by those who engage in the activity rather than those who do not. This is essentially an argument based on transferring some of the responsibility for harm to all users of a commodity.

For example, society permits smoking, but arguably requires that smokers collectively must meet the social costs (such as public health costs) that arise, even though not every smoker will ultimately impose those costs. There is also an equity argument that non-smokers should bear some of their own costs. Such arguments operate independently of efficiency considerations, and the scope of their application relies on social and political judgments.

While the main purpose of the taxes and charges discussed in this section should be to align social and private incentives — fundamentally an efficiency purpose — equity arguments are influential in the design of some taxes, particularly those colloquially referred to as 'sin' taxes.

Ensuring sustainability

  • rates of taxes designed to improve market allocations should be set by reference to marginal spillover costs, not a revenue target. When a tax is used as an instrument of regulation, its goal is to change relative prices faced by producers and consumers, not to fund specific spending programs.
  • because the tax necessarily raises revenue, there may also be a perceived conflict between the revenue-raising potential of taxation, and the desire to improve market outcomes. This can make it hard for governments to introduce such taxes, or to keep the tax rate close to an efficient level (in the case of congestion, for example). For this reason, institutional arrangements to monitor spillover costs and systematically monitor taxes in this area are essential (see Section G5 Monitoring and reporting on the system).

While this Report concentrates on taxation, it is only one policy instrument available to governments. The appropriateness of taxation should be assessed in the context of the costs and benefits of other, potentially more targeted forms of both supply-side and demand-side regulation, public information and specific spending programs.


Taxes on specific goods and services can be used to influence relative prices faced by consumers, in order to address spillover costs, or self-control problems related to highly harmful and addictive goods and services.