Taxing Alcohol

Presentation to “Perspectives for Change”, a conference convened by the Alcohol Advisory Council of New Zealand at Ohinemutu Marae, Rotorua, 20-23 February, 1994.

Keywords: Health; Regulation & Taxation; Social Policy

Given the brevity of the time available, this paper simply states a series of propositions about using taxation as a means of regulating alcohol use. The presenter has long been an advocate of an appropriate level of tax on alcohol to limit abuse and to pay the social costs of the abuse. While reaffirming this position, the paper is directed to the thesis that taxation for this purpose is a limited policy instrument, and that other instruments will have to be increasingly used if we want to obtain the socially optimal level of alcohol consumption. This presentation summarizes a series of papers on the economic regulation of licit drugs, which are listed in the appendix. This work draws on the parallel literature on the use and abuse of tobacco, where the similarities and differences are instructive.

(Two issues not discussed in any depth in this paper is the economics of addiction, and the question of different excise duty rates for different sorts of alcohol.)

There is no question that, all other things being equal, the higher the price of alcohol the lower household consumption. However the competitive market system has the effect of keeping the price of alcohol low , as firms compete for market share (of both the market for spending on alcohol and for total consumer spending) .An excise tax on alcohol raises the price of alcohol above the level set in the competitive market, reducing consumption.

1. Raising the excise tax on alcohol increases the price of alcohol, and reduces consumption.

However, alcohol consumption is “price inelastic ” , that is a rise in the price does not markedly reduce overall alcohol consumption. There would appear to be. at least two groups who are more sensitive to price changes than the population as a whole:
– Adolescents: According to (limited) empirical studies the young are more likely to reduce their alcohol consumption than average, if there is a rise in the price of alcohol. The parallel (and better demonstrated) phenomenon for tobacco means that we can discourage the young from tobacco by keeping tobacco prices high. However alcohol policy may have the different objective of developing sensible attitudes to drink among the young.
– Heavy drinkers who are not (yet) addicted: Raising prices cuts an individuals’ real income. The size of this cut is related to the total outlays. Thus heavy drinkers experience larger income cuts than light drinkers, which is likely to discourage their drinking in greater proportion. This is an a priori argument, which I have not seen empirically verified. The policy implication is that higher taxes may discourage heavy drinking, which may be a step to addiction. (The policy parallel is raising the price of tobacco to discourage addiction among adolescents.)

2. Higher prices for alcohol from higher excise duties do not necessarily reduce alcohol consumption markedly. (Demand tends to be price inelastic.) Exceptions may be adolescent and heavy drinking.

While higher excise duties do not substantially discourage alcohol consumption, they do contribute to government revenue. There are equity and efficiency arguments which say that the excise duty should cover the social costs to the nation of alcohol abuse. The social costs are the public costs which drinking generates that would not be incurred by the drinker if they paid the private cost of producing the beverage (i.e. before the duty was levied). I shall not go through here the economic arguments. More relevant is
– we have little idea of the social costs of alcohol abuse (most of the guestimates confuse social and private costs);
– strictly the excise duty should vary from incident to incident, since some drinking has low social costs (i.e. a glass of wine with a meal) while others may have high ones (drinking oneself rotten and then driving home). There is an important distinction between alcohol and tobacco in this respect, because the social costs of smoking tobacco do not vary as much between each incident as do those from drinking alcohol. (It is argued that there may even be social benefits from moderate drinking, although they are often exaggerated.)

3. The revenue from the duty on alcohol can be justified by the existence of the social costs (above private costs and benefits to the drinker) which the drinking generates.

In principle raising excise duty reduces social costs, as individuals reduce abusive drinking. This means there is a fiscal benefit from reduce public expenditures as well as increased revenues. We have seen that any reductions in drinking may not be great from higher taxation, so most of the fiscal benefits are probably on the revenue side.

An additional consideration is that the expenditure reductions may not be immediate. For tobacco -which admittedly has different social costs -the main expenditure gains occur some ten years after the giving up of the smoking, and so the immediate expenditure reductions are small, and the fiscal significance of the reductions is likely to be discounted.

The fiscal advantage, then, of “sin” taxes is they are a means of raising revenue, which has a degree of public support. Thus the anti-alcohol (abuse) lobby and the fiscal lobby tend to have a common policy solution. However there comes a point when the revenue increase from higher excise duties is zero or negative, while the consequential rise in prices remains. (I do not go through the details of this argument.) While excise duty on alcohol may not have yet reached this point (although excise duty on tobacco may well have), it is well to remember that the common interests of the anti-drink lobby and the fiscal prudence lobby diverge as excise duties rise. (In this context the previous paragraph is important, for while it might be argued that the tax should be raised because it will reduce public expenditure, the fiscal lobby is likely to reject this argument because any gains are only in the long run.)

4. There is a common interest at low levels of excise duty between those who want to limit alcohol consumption and those who want to relieve fiscal stress. This alliance tends to break down at high levels of duty, when increases in levels do not (markedly) increase tax revenue

Thus far the analysis has been primarily in terms of trying to reduce overall alcohol consumption -the traditional objective of the temperance movement. In recent years the campaign has enlarged its compass to cover those who recognise that some drinking may be abusive but believe much is not. The policy focus is to move away from reducing alcohol drinking, to eliminate that which is socially damaging while not affecting (or encouraging) that which is social beneficial (or neutral).

Before sketching the fiscal consequences of this argument it is worthwhile contrasting the situation with tobacco usage, where there is no scientific case (as far as I know) that any tobacco usage is a net social benefit. (The argument that tobacco consumption is beneficial to tobacco growers and processors is true but is not a net social benefit one, since the resources deployed there could be transferred to some other usefully beneficial activity, such as alternative horticulture. )

The difference with alcohol is that there is much drinking which is privately beneficial but which is not socially damaging (and which may even be socially beneficial, although that is not crucial for the argument). To see the problem consider the impact of raising excise duty on alcohol and using the increased revenue to reduce income tax:
– the teetotaller is better off, since their income tax burden goes down (and they may suffer less from the consequences of alcohol abuse by others);
-the heavy drinker is worse off, since they are paying more for their liquor, which is not being offset by the reduction in income tax (nor by any reduction in alcohol abuse by others).

Somewhere between these two extremes is the moderate drinker. Those at the more teetotal end will be better off (for reasons similar to -but of lower magnitude than -the teetotaller). Those at the higher end of moderation will be worse off (for reasons similar to -but of lower magnitude than -the heavy drinker).

There is both political and efficiency issues here. The political issue is that as tax rates rise moderate drinkers increasingly join the anti-tax lobby, so the broad coalition of interest groups get narrowed. In efficiency terms, some of the drinking the higher taxation is discouraging is beneficial.

5. Higher taxation on alcohol reduces the welfare ofboth the socially abusive drinker, and for some moderate drinkers. As a consequence the political coalition suppol1ing higher taxation is fragmented, while there can be severe welfare losses among the moderate drinkers, which more than offset any gains from reduced alcohol abuse.

The point is that the tax is a rather blunt policy instrument for alcohol regulation. Like rain it falleth equally on the just and the unjust. This is not to say there is no case for further raising excise duty on alcohol. That is a question of policy judgement, dependent on facts which this short paper cannot canvass.

6. This paper does not prove the case for or against raising excise duty on alcohol in current circumstances. What it says is there is a pragmatic limit to the level to which the duty can be raised.

What is being argued here is another example of the shift in social policy from a focus on pathological behaviour to one concerned with the totality of behaviour, of which pathological behaviour is but one aspect. If our only concern was excessive drinking (or all drinking) then higher taxation would have a much more vigorous role to play than where we recognise that moderate drinking is not a social bad, and that moderate drinkers have rights and interests.

What the analysis tells an economist is that we need policy instruments which discriminate between socially detrimental and non-detrimental drinking. There would be some interest in differential taxation for different drinking circumstances, reflecting the degree of abuse which arises, but it is not evident how such differentials can be effectively implemented. The general conclusion is to move way from too heavy a reliance on taxes to regulate to instruments which discriminate between different drinking situations.

There is not space here to review these alternatives. Instead I want to conclude by contrasting the implications for the differences in policy between alcohol and tobacco regulation (and no doubt the control of illicit drugs which I do not explore here). Tobacco policy is relatively straight forward. How to reduce consumption; how to prevent addiction. We may look around for a variety of policy instruments to do this, but basically they are all targeted on the same ultimate goal of less consumption.

Alcohol policy has a more sophisticated goal of eliminating certain sorts of drinking, while maintaining and even encouraging other sorts of drinking. This requires a variety of policy instruments which may be targeted on different objectives. While tobacco and alcohol are both drugs it would be foolish to attempt to manage them in the same way. Similarly it does not make sense to group alcohol and illicit drugs into the same general policy organization. There may however, be for specific issues which generate efficiencies and synergies. (Treatment of addicts is an obvious area of mutual co-operation.) But one common policy instrument or concern does not create an overall commonality with other drugs with otherwise different characteristics, any more than we should depend on a single policy instrument to attain the complex objectives of alcohol policy.

7. Because only some alcohol consumption is socially damaging, its overall economic regulation is not the same as for other licit (and illicit) drugs. Taxation is a blunt policy instrument for regulating the abuse of alcohol, since it is also likely to impact any privately and socially beneficial consumption.

Easton, B.H. (1967) “Alcoholic Beverages and Tobacco”, Consumption in New Zealand 1954/5 to 1964/5, NZIER Research Paper 10, 1967, p.23-5.
Easton, B.H. (1978) “The Lost Fortnight”, New Zealand Listener, 11 March, 1978, reprinted in B.H. Easton Economicsfor Social Democrats, Mclndoe. 1981, p.131-2.
Easton, B.H. (1979) “Fiscal Drag”, New Zealand Listener, 26 May, 1979, reprinted in B.H. Easton Economicsfor Social Democrats, Mclndoe. 1981, p.95-6.
Easton, B.H. & L.B. Kay (1982) The Impact on Alcohol Consumption, ALAC Research Report No 1.
Kay, L.B. & B.H. Easton (1982) “The Impact of Taxation on Alcohol Consumption: the Problems of Statistical Analysis”, New Zealand Statistician vol 17 (Nov 1982) p.18-23.
Easton, B.H., N. Hunn, & P. Pryke (1985) A Review of the Fiscal Options for the Regulation of Alcohol Consumption, NZIER Working Paper 85/36.
Easton, B.H. (1988) Economic Liberalisation and Social Intervention, opening address to Symposium 88: Whanau, Family and Early Intervention, Waikanae, October 47, 1988.
Easton, B.H. (1991) Offsetting the Effects of Tobacco Sponsorship, a Report for the Coalition Against Tobacco Advertising and Promotion.
Easton, B.H. (1991) Economic Instruments for the Regulation of Licit Drugs, Paper to Perspective for Change conference, November 1991.
Easton, B.H. (1992) “Economy Column”, New Zealand Listener, July 6, p.38.
Easton, B.H. (1992) “The Economic Regulation of Adolescent Use of Licit Drugs”, Proceedings of the 1991 Conference of the New Zealand Association for Adolescent Health and Development, p.119-120.
Easton, B.H. (1992) “How the Economy Impacts on Health and Welfare”, Investing in Health: The Economics of Health Promotion, Health Promotion Forum of New Zealand, Auckland.