Australia's Future Tax System

Final Report: Detailed Analysis

Chapter E: Enhancing social and market outcomes

E5. Alcohol taxation

E5–3 A common alcohol tax would better address social harm

If alcohol taxes are to be effective in reducing social harm, the taxation of beer, wine and spirits needs to be reformed. The ideal tax structure would be a volumetric tax on all alcoholic beverages, set to balance the reduction in spillover costs20 of alcohol abuse with the cost of taxation on non-abusive consumers, and recognise social benefits of lower-strength products (see Box E5–4).

Policies that are unrelated to social harm, including industry assistance, regional development, and the promotion of small business, undermine the capacity of alcohol tax to target social harm, and should not be delivered through alcohol taxes. To the extent that these programs are desirable, direct government funding or concessions should be delivered in a manner unrelated to the method or quantity of production.

Urgent structural reforms are needed to remove specific exemptions or concessions for certain forms of alcohol most open to severe abuse, including cheap wine. However, convergence to a common volumetric rate of alcohol tax might occur over a longer period, to ensure that the gains from reform are not overwhelmed by immediate shocks to producers and consumers.

Box E5–4: Social benefits from taxing wine on a volumetric basis

Even a low rate of tax on alcohol can significantly reduce spillover costs.

Experience with the Northern Territory Living With Alcohol Program and its associated levy provides evidence of the link between alcohol price and social harm. Beginning in 1992, the levy raised the cost of a standard drink by around five cents. It was wound down following a High Court case that clarified the limits of state power to impose excise duties.

An evaluation of the program found many benefits, including a reduction in alcohol-caused road deaths and Northern Territory government savings in excess of $124 million over the first four years due to the reductions in alcohol-attributable deaths, hospitalisations and road injures (d'Abbs 2001).

As part of a more wide ranging alcohol management plan implemented in Alice Springs in 2007, restrictions on the sale of cask wine were introduced. This resulted in a dramatic fall in the amount of alcohol sold in wine casks, accompanied by an increase in sales of full-strength beer, with a fall in the level of overall alcohol sales. Police said they believed that the restrictions had led to a reduction in violent assaults. Although people had switched from wine to beer, they were not as drunk as before (Senior et al. 2009).

A common alcohol tax would better satisfy consumer preferences

Even before taking into account the spillover costs of alcohol, moving to a single volumetric would minimise the biases introduced by the tax system as to the form in which people choose to consume alcohol.

Removing tax distinctions between different production processes and beverage types would allow for better satisfaction of consumer preferences at the same time as targeting social harm. The range of products available would be less influenced by disparities in tax rates. Moreover, tax treatment based only on alcohol content would reduce complexity, and improve the long-term sustainability of the tax system in the face of technological innovation.

A common alcohol tax base would introduce a 'floor price'

A uniform rate of tax on the alcohol content of all beverages — whether produced as beer, wine or spirits — would relate the alcohol tax base more closely to social harm. It would improve the price signal faced by consumers, currently distorted by a range of tax rates applying to different types of drink. A common alcohol tax would provide a floor price for alcohol (although alcohol could still sometimes be sold below cost or given away).

The maximum amount of alcohol that could be purchased with a fixed amount of money would fall, because forms of alcohol that are currently the cheapest would become more heavily taxed. Products that currently face high effective rates of tax per unit of alcohol, such as high-value wine and spirits-based products, would be taxed less heavily under a common alcohol tax.

For all beverages, the first 1.15 per cent alcohol by volume in any beverage would be exempt from taxation. This recognises that consumption of very low-alcohol products is unlikely to lead to social harm. It also provides a further incentive for producers to reduce alcohol concentration in all products, as less concentrated alcohol products would be taxed proportionately less. For example, Chart E5–3 shows that spirits currently face effective tax rates that are 2.2 times that of full-strength packaged beer. Under a common alcohol tax, full-strength spirits would be taxed around 1.3 times that of full-strength beer.

Chart E5–3: Relative taxation of alcohol under a common alcohol tax(a)

By beverage type (alcohol by volume)

By beverage type (alcohol by volume)

  1. The tax per unit of alcohol is measured relative to full strength packaged beer.

Note: The 1.15 per cent low-alcohol threshold reduces the effective tax payable on beer below the statutory rate. For example, the effective excise payable on full-strength packaged beer is 1–(1.15/4.9) = 76% of the statutory excise rate. This treatment would be extended to all beverages under a common alcohol tax. The effective WET liability is calculated based on a 750ml bottle of high-value wine retailing at $40 and a 4 litre wine cask retailing at $12.99.

Source: Treasury estimates, ABS (2009c).

Transition to such a system would need to be managed carefully, to avoid potentially harmful shocks to consumers, and to reduce the impact on existing industry arrangements. For this reason, transitional arrangements might be designed to stabilise the nominal price of some products, and to limit the rate of price change (both upward and downward) for others, while monitoring the effect of these changes on spillover costs.

Setting the rate of alcohol tax

The current wide variation in rates of tax on different types of alcohol makes it difficult to calculate an efficient alcohol tax rate. This is because the distribution of abusive and non-abusive consumption is partly determined by these tax arrangements. In order to estimate the weighted average marginal spillover cost of alcohol, it would be necessary to first reform the structure of alcohol taxation, by removing the current biases to consume alcohol in one form rather than another.

Until reliable estimates are available, the average spillover cost per litre might be used as a proxy. As the current full-strength packaged beer excise rate is closest to the estimates of the average spillover cost of alcohol in Australia, it would be appropriate for rates of tax on other products to converge at this rate over time.

During the transition, data on spillover costs and consumer behaviour should be systematically collected to inform the process of setting the optimal tax rate. The final rate of tax would be intended solely to optimise price signals facing consumers. It should be set without regard to the government's fiscal position, and irrespective of any specific spending commitments related to alcohol abuse. The rate of tax should be indexed to ensure that the real rate of the tax only falls in the event of an adjustment to social cost estimates.

To ensure it is credible and sustainable, the process of determining the tax rate should be based on the best available information and an agreed and transparent methodology (see Section G5 Monitoring and reporting on the system). Consistent with Action 4.1 of the National Preventative Health Taskforce (2009), this process should incorporate independent modelling, in consultation with the departments of Health and Ageing and Treasury and an industry panel.

Estimates and models should be periodically re-assessed to take into account the effectiveness of alternative policies for reducing social harm, as well as changes in culture that affect consumption. If more targeted non-tax measures to reduce social harm are successful, the alcohol tax rate could be lowered.

Taxing alcohol consumed in Australia

Alcohol produced for export should continue to be exempt from tax, as alcohol consumption should be regulated in the jurisdiction in which the spillover costs are borne (that is, in the destination country). Imported alcohol products (including by travellers) should be taxed at the same rate as domestically produced products, as the spillover costs of alcohol in Australia relate to alcohol consumed here, regardless of where it is produced.

Alcohol used for industrial, manufacturing, scientific, medical, veterinary and other purposes that poses no risk of social harm from human consumption would continue not to be taxed.

In principle, the home production of alcohol (for example, home brew or wine-making) would be subject to tax. In practice, this is unlikely to be feasible.

Reducing compliance and administration costs

The excise system has evolved over the past century. In 2008–09, almost three hundred entities in the alcohol industry paid excise duty. This system is administered by the ATO. Excise-equivalent customs duty on imported goods is currently collected through a separate system by the Australian Customs and Border Protection Service.

Much of the existing apparatus for licensing the manufacturing, importing, storing or dealing in excisable products is due to differential (and sometimes high) rates of tax on many different classes of alcohol. Further, the additional value-based tariff on some imported spirits increases compliance costs associated with tracking goods in warehouses.

Treating all alcohol products on the same basis, regardless of how or where they are produced, would make much of this additional regulation redundant. The taxation of alcohol could be brought into a single regime, to ensure consistent compliance obligations between industries and to remove the administrative costs of running multiple systems for taxing alcohol. Such an approach would need to be developed over time in consultation with industry, and could also be extended to include other excisable goods. Businesses that produce a range of alcohol products would not need to deal with multiple tax systems.

The current excise system is particularly costly for smaller producers. However, wine producers, 80 per cent of whom have a turnover of less than $2 million per year, are not currently in the excise system. There are also concessions for microbreweries and 'brew-on-premise' beer.

These small entities need not be immediately included in the current excise system, but should ultimately be brought under a common alcohol tax. The Australian government should explore streamlined arrangements for small taxpayers based on their volume of production. Small alcohol producers might be subject to less stringent licensing conditions, relaxed rules around calibration and testing of equipment, and be allowed to report their liability through the business activity statement. This would help ensure that compliance and audit resources are targeted on a cost-effective basis.


A common alcohol tax that does not discriminate between beverage types would remove production and consumption biases from the alcohol taxation system, reduce compliance and administration costs, and better target the spillover costs of alcohol consumption.

Recommendation 71:

All alcoholic beverages should be taxed on a volumetric basis, which, over time, should converge to a single rate, with a low-alcohol threshold introduced for all products. The rate of alcohol tax should be based on evidence of the net marginal spillover cost of alcohol.

Recommendation 72:

The introduction of a common alcohol tax should be accompanied by a review of the administration of alcohol tax, to ensure that alcohol taxpayers do not face redundant compliance obligations.

20 The weighted average marginal cost, not average total cost.