Final Report: Detailed Analysis
E5. Alcohol taxation
Since Governor Hunter first taxed beer, wine and spirits in the colony of New South Wales, alcohol taxation has provided governments in Australia with an administratively simple tax base. This revenue was once a vital share of government finances. However, government today can raise much larger sums of money from broad-based taxes on income, resources and consumption. These provide a fairer and more efficient way to raise revenue than specific commodity taxes. Narrow-based taxes should be designed primarily to correct particular market failures, not for general revenue-raising (see Section E Enhancing social and market outcomes).
While alcohol excise is administratively simpler than other taxes, its simplicity has eroded over time. Today, commercially produced beer is taxed at eight different rates — depending on alcohol volume, the type of packaging, and whether it is produced for commercial or non-commercial purposes. Brandy is taxed at a lower rate than the domestic spirits rate (which is based on alcohol content), and some imported spirits are subject to an additional duty based on value, on top of the domestic excise rate. Wine is taxed through a separate wholesale tax, based on its value, not its alcohol content.
To enforce the distinction between beer, wine and spirits, a sophisticated system of licensing and bonded warehouses has been devised. This ensures that alcohol produced using one process is not passed off as a lower-tax product. In addition, technological advances mean that alcohol produced using one process can be altered to mimic other beverage types. This has introduced additional complexity — for example, the definition of beer now requires a minimum threshold of 'international bitterness units' and a maximum threshold for sugar. Just to determine the appropriate rate of excise, the Australian Taxation Office (ATO) needs to provide advice on recipes.
While taxes on alcohol should not be used for general revenue-raising, they may have a role in addressing the significant spillover costs on the community associated with alcohol abuse, by changing the price of alcohol faced by consumers. This is a blunt instrument for controlling the spillover costs of abusive consumption, and must be weighed against the wellbeing loss of taxing non-abusive consumers. However, in the absence of more cost effective or better targeted instruments to address abusive alcohol consumption, a tax on the consumption of alcohol can still improve welfare.
Economic studies of alcohol consumption around the world consistently find that higher prices do reduce overall consumption of alcohol products. However, in most cases a 1 per cent price rise in the price of alcohol results in less than a 1 per cent decline in consumption (Fogarty 2008). Moreover, not all individuals reduce consumption to the same degree — evidence suggests that heavier drinkers may be less responsive to the price of alcohol than the general population (Ayyagari 2009).
In theory, if alcohol tax could be targeted at an individual's abusive consumption, it would be imposed on a per-drink basis, at a rate set according to the risk of harm for individual consumers. Drinks more likely to give rise to high spillover costs would be taxed prohibitively, while consumption with no risk of spillover costs would not be taxed. In this ideal world, the price of every glass of alcohol would include the risk of harm associated with its consumption.
In the real world, however, such a tax is technologically and administratively infeasible, and would be unnecessarily intrusive. Accordingly, alcohol taxes are levied equally on all products of a particular class or type. The effect of this is to raise the cost of drinking, but with the cost averaged across all drinkers, not targeted only at those most likely to cause social harm.
This makes excise a blunt instrument for reducing the spillover costs of alcohol use. It means that consumers who enjoy alcohol responsibly face an unnecessarily high price (and pay too much tax). In other words, even though alcohol tax raises revenue for the government, it is not a costless way of addressing alcohol abuse. As with all policy interventions, the benefits of taxation should be weighed against the costs (see Box E5–1).
Box E5–1: Costs and benefits of taxation
Taxes to reduce social harm can be costly when not all units of consumption give rise to the same spillover cost. Chart E5–1 presents a stylised welfare-economics framework, which splits consumers into two groups — one whose consumption has 'no abuse' (left) and one whose consumption has risks of 'abuse' (right). Consumption with no abuse gives rise to no spillover costs, while consumption with abuse gives rise to spillover costs that increase with the quantity of alcohol consumed in a single period.
Chart E5–1: Taxing to control social costs
Source: Adapted from Pogue and Sgontz (1989).
A tax — levied across both groups — reduces the consumption of both groups. An increase in price reduces consumption by the abuse group (from Qa to Qa1). While this may result in a loss of immediate satisfaction (area C) to the individual, on balance society as a whole benefits from a reduction in spillover costs (area A).
However, this gain needs to be balanced from the wellbeing loss of consumers with no abuse. They too reduce their consumption (from Qb to Qb1), resulting in a welfare loss (area B) that is not offset by any reduction in spillover cost.
Taxes on alcohol should not be increased beyond the point where the marginal reduction in spillover cost exceeds the marginal reduction in the wellbeing of responsible drinkers. One corollary of this is that while a tax on alcohol might be used to reduce social harm, it is not an appropriate tool to eliminate it.
Collins and Lapsley (2008) evaluate a number of policy interventions for reducing the spillover costs of alcohol abuse. They find that higher alcohol taxes would reduce overall spillover costs, but also that individual-based interventions (usually by doctors) are an effective way to reduce hazardous alcohol consumption. Stricter enforcement of random breath testing, and reducing the allowed blood alcohol concentration level for drivers would be effective ways of reducing the costs of drink driving. Complete or partial controls on the advertising and marketing of alcohol would also reduce costs.
As with the use of tax to control the spillover costs of alcohol, the costs and benefits of non-tax programs need to be considered. The advantage of these interventions is that they could be closely targeted at actual spillover costs, unlike an excise that is necessarily averaged over all units of production.
The point of production or importation is not the only point at which it is possible to introduce prices that reflect spillover costs. For example, data on police attendance at alcohol-related incidents in New South Wales suggests that more than half of incidents are related to only 10 per cent of licensed premises (Moore 2009).
To improve the targeting of the price signal, State governments might relate the licence fee for an establishment to the number and severity of violent incidents connected with it. In this way, licence fees would act as a more targeted tax than excise. They would provide operators with a stronger incentive to refuse service to high-risk patrons. Establishments that are sustained sources of violence would pass these costs on to their patrons in higher prices — while consumers who drink in a low-risk setting would not pay the additional cost.
Taxes on alcohol should be set to address the spillover costs of consumption, when this delivers a net gain to wellbeing and is more effective than alternative policies.
A World Health Organisation (WHO 2007) expert committee concluded that policies that increase alcohol prices have been shown to 'reduce the proportion of young people who are heavy drinkers, to reduce underage drinking, and to reduce per occasion 'binge' drinking'. In addition, the WHO found that '[h]igher prices also delay intentions among younger teenagers to start drinking and slow progression towards drinking larger amounts'.
To the extent that people want to be inebriated — and that this is associated with social harm — it is the alcohol rather than the form in which it is delivered that drives demand. For this reason, it is the alcohol consumed by an individual in a set period, not its value, packaging or the method or place of production, that is most closely related to social harm. Any tax on alcohol designed to address spillover costs should therefore be levied on a volumetric basis.
Alcohol tax should be levied on a common volumetric basis across all forms of alcohol, regardless of place, method or scale of production.
Spillover costs of alcohol abuse include foetal damage and child abuse, domestic violence, road accidents, crime and violence, increased mortality and a range of diseases and medical conditions.18 These costs can arise directly (for example, in the form of costs on victims of alcohol-related violence) and indirectly (for example, in the form of the cost to the community of additional demand on a publicly-funded health care system).
There are additional characteristics that may be related to social harm in specific cases. Policy intervention to reduce social harm might also consider the alcohol strength, the environment in which the alcohol is consumed, who is drinking (those more prone to violence) and how (binge drinking). Whether these factors can be taken into account in designing a tax on alcohol production depends on the extent to which particular identifiable classes of beverage can be causally associated with greater or less risk of social harm.
For example, many submissions have argued that 'alcopops' have higher spillover costs than other forms of alcohol, at least for certain groups. Similarly, an expensive single malt Scotch whisky may be associated with lower spillover costs than cheap vodka. Relating classes of drink to risks of harm requires detailed information about the relationship between alcohol products and spillover costs. This relationship may also change over time as producers and consumers adjust their behaviour in response to taxation.
In some cases, the welfare gains from taxing products differently may outweigh the complexity costs, provided that a product can be clearly defined and the net gains identified. However, in the absence of specific information, a uniform rate of tax across all beverages is the least complex and most efficient way of imposing an alcohol tax.
Having different tax rates on beer, wine and spirits is a common feature of alcohol taxes around the world. However, these often reflect the pattern of domestic production, rather than spillover costs. For example, major wine producing countries tend to charge little or no tax on alcohol in wine. Over time, this influences consumer preferences and reinforces demand for particular products.
Departure from uniform taxation also encourages producers to innovate simply to avoid tax. Some Australian firms produce grape-based alcohol products (such as Father O'Leary's) that are very similar to spirit-based products (such as Bailey's Irish Cream). Different tax rates also encourage consumers to change from their preferred drink, without necessarily reducing the risks associated with their drinking.
While there should be no discrimination between different types of production, submissions from both public health advocates and producers have generally supported lower rates of tax on lower strength products. This can be justified on the basis that alcohol concentration can be readily measured and is correlated with the level of social harm (see Box E5–2).
Indeed, low-alcohol products can be considered as having a social benefit to the extent that they substitute for higher strength alcohol products that impose greater spillover costs on the community.
Box E5–2: Extremely low-alcohol products are mostly harmless
The body can process about one litre of water an hour (more and life is at risk) and about one standard drink (12.7 mL of alcohol) an hour (more and alcohol starts to accumulate). This implies that consuming a product at 1.27 per cent alcohol content or less should have negligible social harm from alcohol as the water would kill the drinker first.
The rate of tax on alcohol should be based on evidence of marginal social cost.
18 These costs are more limited in scope than those used in the cost of illness methodologies that have been developed in the public health literature (for example, Collins & Lapsley 2008), which also include many of the costs that individuals bear themselves. To estimate spillover costs relevant for setting rates of tax, it is necessary to exclude private intangible costs (such as pain and suffering), and the loss of household production from premature death or sickness. That said, the distinction between private costs and spillover costs is not always clear. For example, if a family utility and decision making model is used, alcohol-related violence against family members and the loss of family disposable income are private costs; but, if an individual utility and decision making model is used, costs borne by other family members are spillovers.
Next Page – E5–2: Current taxes on beer, wine and spirits are incoherent >>
<< Previous Page – E5: Alcohol taxation