The Economic Regulation Of Alcohol Consumption in New Zealand

Conference on ‘The Social Cost of Alcohol Abuse’, IRER – University of Neuchatel, Switzerland, 24-25 October, 2003

Keywords: Health;


While New Zealand has some measurement of the social costs of alcohol misuse, which the paper reports, the interest in the country, and this paper, has been the shift to implementing policies whose focus is to minimise harm from misuse.

The paper traverses the policy environment from the initial revenue-raising role of the excise duty in 1840. As the frontier society moved to a settled society, policy from the 1890s moved to restricting the consumption of alcohol, with revenue remaining the main fiscal concern. However, in 1989 a new direction was undertaken in which aimed to minimise restrictions on low and zero harm alcohol consumption, and eliminate as far as was practical harm arising from misuse. Over the next 14 years various measures were taken culminating in the latest tax package of May 2003.

The paper traces through these changes. It argues that the policy transformation is not complete, and also discusses some of the inherent tensions in the new approach. In particular the shift from restriction to a liberal regime which treated liquor as a largely normal consumption good, with targeting on harm minimisation, resulted in easing of prohibitions on advertising of liquor.

But the paper also discusses the limitations of the economic approach, for it is not possible to use the policy instruments to target precisely on harm reduction without also limiting some low and zero harm consumption. This emphasises the need for non-economic policy instruments, most pertinently those which change attitudes to alcohol consumption where there remains a ‘frontier’ spirit.

The New Zealand Background

New Zealand is a high income, democratic country of four million people, located in the south-west Pacific over a thousand kilometres to east of Australia. It was the last substantial land mass to be settled, around 1000 years ago by Polynesians who voyaged from the central Pacific. They evolved in to today’s indigenous people, the Maori. About 15 percent of the population have at least one Maori ancestor, although there has been considerable intermarriage.

The first Europeans explorers arrived around 230 years ago (occasional visitors who either neither landed or left no trace aside) and began to brew beer, introducing alcohol to the Maori who had had none before. Sealers and whalers began arriving 200 years ago, followed by missionaries and settlers, although that latter were not significant until about 150 years ago, following New Zealand becoming a British colony in 1840. The settlers evolved a Pakeha culture, evidently of European origin and connected to the dominant world cultures (as indeed does Maori culture) but with distinctive features.

Initially the European based economy based on extracting the natural resources. It was a ‘colony of exploitation’ or ‘quarry’. By the late nineteenth century, however, sustainable farming became important, particularly pastoral farm shipping refrigerated dairy and meat products and wool to Europe. The transition to a ‘colony of permanence’, or settlement, created many tensions, not least involving the regulation of liquor.

Drinking and drunkeness were a feature of the male dominated, shifting, frontier society. The control of these excesses was a priority to the settlement where, not surprisingly, woman who are central to sustainability in a way they are not in the quarry, played a prominent part in the prohibition movement. In 1918 it almost succeeded, when the domestic voters passed the threshold, but votes from soldiers overseas pushed the proportion just below the required 60 percent. The compromise was to restrict access to alcohol, but to do little to tackle drinking attitudes.

Changes after the Second World War were incremental and slow. Six o’clock closing, when drinking in the public houses stopped for the general public, was not abolished until 1967. ALAC, the Alcohol Advisory Council of New Zealand (Kaunihere Whakatupato Waipiro o Aotearoa) was founded in March 1977, as a result of a recommendation of the 1975 Royal Commission into the Sale of Liquor. It was charged with encouraging and promoting moderation in the use of liquor, discouraging and reducing its misuse, and minimising ‘the personal, social, and economic evils resulting from the misuse of liquor’. In 2000 the statute was amended to replace ‘evils’ with ‘harm’.

The excess social costs are substantial. The most comprehensive estimate, based upon the counterfactual of no alcohol misuse – so that the health benefits of moderate alcohol consumption are not netted out – suggests that alcohol misuse reduces effective GDP by 4.0 percent, may well reduce the effective size of unmeasured (informal) economy by a similar amount, and has also reduced the welfare of New Zealanders via additional mortality and morbidity by 2.0 percent and the population of New Zealand by .8 percent. (Easton 1997)

This estimate is for 1990 and there is no subsequent comprehensive update. It is to be hoped that the proportions are lower, reflecting success of the harm minimisation policies. The study did not include law and order costs, as there were no estimates available at the time.

Social Costs of Alcohol Misuse (1990)

Effect of population mortality $6,000m
Effect of population morbidity $7,200m

Reduced production from mortality $600m
Reduced production from morbidity $1,200m
Additional resources from consumption $900m
Additional resources from not have to treating induced diseases and other consequences $750m

Benefits from consumption -$540m

Total Costs of Abuse $16,110m

Intangible Costs
Total $13,200m
Percentage of total human capital 2.0%

Tangible Costs
Total $2,910m
Percentage of GDP 4.0%

Population Decrease 0.8%

Source: Easton (1997)

For frontier attitudes died as slowly as did the restrictive regulatory environment. Even so, suddenly the old regulatory regime began to be swept away, when the Sale of Liquor Act 1989 stated
“the object of this Act is to establish a reasonable system of control over the sale and supply of liquor to the public with the aim of contributing to the reduction of liquor abuse, so far as that can be achieved by legislative means.” (Italics added)

In future the industry would be regulated by the ordinary commercial laws of the land. The notion that the purpose of the licensing system was solely concerned with reductions in the abuse of liquor (today we are more likely to refer to ‘misuse’ and ‘harm’) and not with restricting access or reducing consumption, was a major break from past legislation.

The change reflected the wider economic reforms of the times, which were focussed on applying market principles in economic and social policy. This paper is primarily about their subsequent development particularly as it applies to taxation, which was affected only later. But first a brief history of taxation of alcohol in New Zealand.

Alcohol Taxation in New Zealand

In 1839, the British Colonial Secretary , Lord Normandy, confidently advised putative Governor Hobson ‘Duties on the import of tobacco, spirits, wines and sugar, will probably supersede the necessity for other taxation …’, (McLintock 1958:90), although beer was not initially taxed. There was a brief six months period from September when the duties were repealed, but since then excise duties on alcohol (and tobacco) have been an ongoing part of New Zealand’s fiscal practice. (Taxation Review Committee 1967: 63)

Normandy’s promise that the revenue source would be sufficient proved unfounded. Admittedly in 1875/6 customs and excise duty made up 91.6 percent of tax revenue, but since then the proportion has been diminishing. In 2002/3 excise duties (including the Good and Services Tax (GST) – a kind of VAT – levied on them) on alcohol and tobacco contributed 3.5 percent of total tax revenue (with a further 1.9 percent coming from customs duty). The decline is not so dramatic relative to total output. Customs and excise duties (including on some other products) was about 4.1 percent of GNP in l875/6, while alcohol and tobacco taxation is expected to be about 1.4 percent of GNP this year. What Normandy underestimated was the revenue demands of future governments.

The decision to tax spirits, tobacco and sugar was initially a matter of administrative convenience. All excise duties generate the possibility of illicit import and production, but these products were reasonably easy to identify, and there were a few key production or transit points at which the duty needed be levied. (Vineyards did not become widespread until after the Second World War.)

Later on, as settler society developed, the taxation of alcohol (and tobacco) was further attractive because its consumption was thought to be indulgent, if not downright evil. Hence the practice of labelling excise duties on them as ‘sin taxes’. In effect alcohol was deemed a ‘demerit good’, whose consumption should be discouraged. The justification for why a good is a demerit (or merit) one is not easily identified by economic analysis. although there is a widely held view that aggregate alcohol misuse is directly correlated aggregate alcohol consumption so a reduction of the first can be necessarily caused by a reduction of the second. However there would also be a reduction of consumption which was not generating and that was only irrelevant if alcohol was not a demerit good.

Ironically alcoholic beverages are also attractive for excise duty because their demand is ‘price inelastic’, that is the quantity purchased is not very sensitive (‘insensitive’) to a change in price. Thus the duties do not reduce consumption, so much as punish by a ‘fine’ those who consumer.

The 1958 Budget and its Aftermath

The notion of taxing less acceptable products was evident in the 1958 budget, when the government, needing to reduce the fiscal deficit, hiked taxes on alcohol and tobacco (and motor spirits). At the time the public saw the ‘Black’ budget as a betrayal of the promises in the 1957 election promise and the punishing of the working man’s pleasure in his ‘beer and baccy’, by a 30 percent rise in the price of alcoholic beverages and a 40 percent rise in the price of tobacco. So dramatic were these hikes that they perceptibly cut consumption despite each product’s price inelasticity of demand (Easton 1967) The public justifications given for these excise hikes were surprisingly thin. The budget speech simply says there is a need for fiscal action and baldly announced the additional impositions. The politically traumatic experience of the 1958 budget meant that subsequent governments were loathe to raise taxes on alcohol.

The 1970s

Tobacco duties were hiked in 1967, and again in 1970, and 1976. The October 1970 Economic Statement said
“Of all the products at present subject to indirect taxation, it is clear that cigarettes and tobacco can be subjected to additional tax without harming in any way the general welfare of the community. In fact it is increasingly argued that discouraging the consumption of these commodities is likely to make a positive contribution to our general health.” (p.12)

This may be the earliest fiscal reference to using taxation to improve health by inhibiting consumption. There was no mention of alcohol and its duty level was left untouched. It is unclear why the two were treated differently. Possibly tobacco excises impacted more greatly on workers who were not the right wing government’s significant supporters. The 1976 announcement contains no justification for the tobacco excise hikes, but this time duties were raised on spirits as well.

As inflation galloped along in the 1970s at double digit rates, the real value of the excise duties fixed in nominal terms was diminishing, which reduced the real value of government revenue, and the real price which made the licit drug use cheaper. A coalition evolved of those who wanted the government to control and reduce its fiscal deficit, and those who wanted the government to control and reduce the use of licit drugs (or perhaps merely punish the drug users). They may have had quite different objectives, but a common policy prescription for both groups was to raise excise duties. In the 1979 Budget there was some linkage of specific taxes to inflation via the imposition of specific sales taxes.

The 1977 budget speech signalled an apparent change in policy. The first paragraph of the section titled ‘Duty on Alcohol and Tobacco’ says
“The Government is concerned at the high level of public expenditure caused directly by the consumption of tobacco and alcohol. The adverse effects on health of smoking and drinking have been well publicised. Alcohol is also a major cause of many road accidents. The cost of providing health care and treatment in public hospitals and elsewhere for those affected is very high.” (Economic Statement 1977:p.41)

Having set the new higher duty rates the section concludes that the additional revenue ‘will help sustain the high level on health, including the extension of community health services’ Thus a new principle has been introduced. While the duties remained primarily for revenue purposes a justification is now in terms of the costs that the consumption impose on the public. Some of the increase in revenue went directly to particular health promotion and treatment programs. However this ‘tagging’ of some tax revenue for particular purposes did not last.

At the end of the 1970s the purposes of special taxes on alcohol (and tobacco) were:
1. They were relativity easy to levy, with relatively low compliance costs compared to many other products.
2. They were levied on products which were price inelastic so a hike in taxation increased revenue.
3. The tax revenue would contribute to the costs of the health system (and, possibly, the spending of other government agencies) incurred as a result of their consumption.
4. Insofar as the higher taxes raised prices and reduced consumption, that would also contribute to population health.

Lurking in some of the public’s mind was the notion that alcohol was a demerit good whose consumption should be discouraged. However the notion is not mentioned in the later policy statements. As time went on it probably plays an decreasing role in the policy debate and in the public mind. But there remained the recognition that the consumption of alcohol could be a ‘bad’ insofar as it caused negative externalities (or harm), the point captured in 3 above (and possibly 4).

The 1989 Budget Reform

The 1980s were a period of major fiscal reform. Among the important changes impacting on alcohols was taxation on wine was switched in 1986 from a sales tax based on value to an excise duty based on alcohol content. A significant effect of this change is that, for the same total revenue, the price of expensive wines would fall and the price of cheap wines rise. (This has considerable significance for econometric studies attempting to estimate price elasticities. Traditionally they have treated the imposed tax as a proportional sales tax, or that the price range was narrow enough so that the excise duty could be broadly treated as a proportional sales tax. This clearly does not apply for wine where there is a great variation in price.)

Following the 1988 Report on the Review on Excise Duties on Alcoholic Beverages and Tobacco Products (Sullivan Report) the 1989 Budget made major changes.

This is another area of unnecessary complexity and controversy. … The basic case for heavier taxation on alcohol and on tobacco lies in the social costs of their consumption. In the case at least of alcohol this is somewhat arbitrary, since everyone faces the higher taxes on alcohol but most drink only in moderation. Nevertheless the Government believes that the social harm that can be caused by excessive consumption of alcohol justifies the collection of a separate tax related to the alcohol content of alcoholic beverages.

As from tonight the excise duty on alcoholic beverages will be related directly to alcohol content, so that similar products with similar strengths of alcohol will be taxed to the same extent. Only two rates of excise duty [exclusive of GST] will apply:
– On beer and wine, $15 per litre of alcohol;
– On spirits, $30 per litre of alcohol. (1989:20).

In practice it has been difficult to assess the aabv in the ‘wine’ range from 6% to 23% aabv. Instead it is divided into three sub-ranges which are taxed on the basis of $21.096 per litre of ethanol at a point somewhere in each range. (Throughout this paper ethanol is used to mean ‘absolute alcohol’ while ‘alcohol’ usually means a beverage containing alcohol.) The effect is that beverages with higher ethanol in each range are taxed at a lower rate per unit of ethanol than beverages in the same range but with lower unit ethanol. For instance ‘light spirits’ at 23 percent aabv alcohol, was taxed as if it is 18.0 percent aabv alcohol, so the excise duty per unit of ethanol was only 76 percent of the set rate. In effect, every fourth drink tax was free in comparison to a purchase at 18.0 percent aabv. Not surprisingly, businesses took legitimate advantage of this anomaly, with spirits based drinks with a low aabv which were taxed at the wine and beer rate. The policy response is described below.

The Treasury View: 1991

The new regime was not without its difficulties In 1991 the Treasury provided a report in a response to the incoming Minister’s request for a review of alcohol and tobacco taxation, the last comprehensive policy framework report available.

It argued that GST (i.e. a VAT) is the most efficient revenue raising indirect tax instrument, but acknowledges two reasons why there may be a case for separate taxation alcohol:
– the private cost of alcohol use is less than the social cost (i.e. taking account of the cost to others) thereby encouraging irresponsible drug use – the ‘user charge’ argument; and
– some individuals need to be protected from making drug use choices that are not in their own best interests – the ‘irrational consumer’ argument.

The Treasury was troubled by the different excise duty rate for spirits from other beverages. It saw it as historic, and is unable to offer any justification for it, other than history and not abruptly disrupting past relativities. The best strategy, it thought, was to hold duty on spirits constant while the excise duty on the remainder rises with inflation, until they attain the same rate, a practice which had been announced in the 1989 budget. .

The report concludes that
– there is a case for specific taxation of alcohol primarily in terms of recovering user costs, but that the best level of these taxes is difficult to judge:
– the specific excise regime, coupled with some form of indexation arrangement, is the most appropriate system for the taxation of alcohol …;
– there should be a goal of uniform alcohol taxation. (1991:6)

There is no prevention element in this strategy. Excise duty is not seen to have a role in discouraging harm, a view which shifted in 2003. Before then we need to look at what happened over the 1990s.

Levels of Consumption

Not surprisingly the liberalisation of licensing laws resulted in a major increase in outlets, transforming particularly inner cities where bars are now common. What is surprising is that despite the increase in outlets available alcohol for consumption (in effect the amount drunk) has fallen. In the March 1989, the year of the reform, it was 10.0 litres per person over 15, rising to 10.2 litres in
1991, after which has fallen to 8.8 litres in 2001 (although they were as low as 8.5 litres in 1997).

The reasons for the fall are complex. Real alcohol prices rose, there has been an persistent campaign to improve attitudes to alcohol, and drink-driving enforcement is tougher. On the other hand, not only has licensing been more liberal, but prohibition on the television advertising of liquor was abandoned in 1991 the legal age of purchase of liquor was lowered to 18 (the voting age) in throughout the 1990s and there has been an increase in the variety of alcoholic beverages, all of which ought to have lifted consumption to some extent. An optimist might argue that the licensing liberalisation has led to more ‘civilised’ drinking, but the most cautious interpretation is that there has been a long term shift away from heavy drinking, which the liberalisation measures did not reverse.

Youth Consumption

There is one major exception. The evidence points to significant increases in alcohol use and misuse among the young.

The annual Auckland surveys (the longest continuous set available) show that the average number of drinks consumed on one occasion among 14 to 19 year olds increased from 3 to 4 in 1990 to 5 to 6 drinks in 1999. The rise was particular strong at the younger end, with the 14 to 17 year olds increasing from 2 to 3 drinks in 1990 to 5 to 6 in 1999. (Casswell & Bhatta 2001)

The National Alcohol Survey found that the average quantity of ethanol consumed by 14 to 17 year olds doubled between 1995 and 2000. It also rose for 18 to 19 year olds, but fell – or was much the same – for older age groups. A similar pattern applies for frequency of drinking (occasions per year). The survey also found that the 14 to 19 year olds were consuming more on each occasion, in contrast to the older age groups whose rate had not changed much over the five years. (Habgood et al).

There are no long term series but currently over a third of 14 to 17 year olds consider themselves as ‘heavy drinkers’. (ALAC Youth Drinking Monitor 2002) The National Survey found that 7 percent of 14 to 15 year old males, 25 percent of 16-17 year old males and 40 percent of 18-19 year old males were involved in a session of at least 6 standard drinks at least once a week. The latter proportion was more than for 20 to 24 year olds, and the 16-17 year old proportion more than occurs for those over 25 years old. The same patterns apply for females, except their levels are slightly lower (7, 22, and 28 percent). (Habgood et al)

These trends are reinforced by the evidence that the age of first drinking seems to be lowering, although the surveys were not well designed to measure this. It seems that the earlier the age of onset of drinking the more likely the young person is to be involved in heavy drinking in the teenage years. (Lynskey & Ferguson 1995; ALAC 2002)

In summary there has been rising drinking among teenagers in the 1990s, older teenagers appear to be at least as prone to heavy drinking as young adults. The reasons for the change are not certain. In part it is an international phenomenon, but it could also be argued that the thrust of the post-1989 policy to make alcohol consumption as ‘normal’ as possible, exposed young people to an environment where they had to learn to drink in a mature way, without much support to help them in the learning task.

The Taxing Harm Report

At the end of 2002 ALAC commissioned me to write a report on alcohol taxation. After a review of the literature and the existing situation, some of which is reported above, I concluded that while much policy toward alcohol was in the spirit of the 1989 licensing reforms, excise duties on alcohol could be better aligned. In particular, while the purpose of the interventions were to reduce harm, this was not the focus of the taxation regime.

Almost as an aside, the review also showed that the tax regime had created a fiscal anomaly. The 1989 tax package had envisaged that there would be a differential between beer and wine on the one hand and spirits on the other. However the implementation was on the basis of whether the absolute alcohol by volume was above or below 23 percent, so diluted spirits were taxed at a wine and beer rate. Now for highly diluted spirits – the flavoured alcoholic beverages with an aabv similar to beer – this may not have mattered, since the cost per unit of ethanol proved similar to existing drinks perhaps because of packaging costs, at the top end it was possible to produce ‘light spirits’ which were astonishingly cheap. A 1.125 litre bottle, containing 23 standard drinks cost as little as $7.95 at a time when the minimum adult wage was $8.00 an hour. A typical drinker could get drunk on the cost of a cup of coffee. Some did. Anecdote said the main drinkers of light spirits were young people, and there are at least two cases where drinking a bottle and a half of light spirits caused death.

The report argued that there were two types of harm. There was ‘consumption’ harm which was mainly related to the quantity of ethanol drunk regardless of the circumstances, and there was ‘drunkenness’ harm which was the result of excessive consumption of alcohol in particular circumstances. I was unable to find any estimates of proportions between these two types – in any case they would depend on the particularities of a drinking environment – but Eric Single cites that 70 percent of alcohol harm occurred within a year of consumption and this might be equated to the share of ‘drunkenness harm’. (Single 2002)

Could taxation be used to reduce the drunkenness harm? The research literature – albeit thinly – and anecdote – richly – suggests that during heavy drinking the quantity of ethanol imbibed is affected by the price of liquor. The higher the price, the less which is consumed in a session. This is much less true for drinking situations which do not lead to drunkenness, say when the liquor accompanies a meal. A rise in price may lead to reductions in the value of the liquor consumed, as for instance people switch to cheaper wines, but not to the quantity of ethanol consumed. Moreover the price of ethanol in the moderate drinking situations is likely to be much higher than the price in the heavy drinking situations. Anecdotally, one purchases liquor suitable for a meal as much as one can afford, for a heavy drinking session one purchases cheap liquor as measured in the cost of ethanol. It seems likely that for drunkenness harm the minimum price of ethanol matters, while for consumption harm the price of alcohol may matter but the price of ethanol probably does not matter much.

The effect of the existing polices was to target consumption harm (insofar as any harm was being targeted), even though drunkenness harm appeared to be the bigger problem. Moreover the policy levied consumption to contribute to the cost of harm, it did not aim to prevent it. An alternative strategy was to focus on the minimum cost of ethanol, since that would have a real effect on reducing drunkenness harm.

This led to the following executive summary, which sets down the analysis of Taxing Harm as well as the recommendations. (Not all of the 46 paragraphs are reported here.)


1. The National Drugs Strategy includes the policy objective of ‘the minimiz[ation] of harm caused by alcohol use to both individuals and the community’ where harm is defined as ‘all adverse effects or outcomes, including harm to health as well as detrimental effects on social and family relationships, loss of actual or potential enjoyment or livelihood, and economic or financial costs.’

2. The report recommends that the specific taxes and levies on alcohol be levied with the primary purpose of reducing alcohol misuse and the consequent harm.

3. A secondary purpose, which assists in setting its appropriate level, is that the taxation should enable the government to recover some or all of the expenditure outlays and revenue losses caused by alcohol harm.

The Effects of Excise Duties: How the Market Can Help Reduce Harm

4. The effect of (an increase in) an excise duty or levy is to increase the price of the alcohol on which it is levied.

5. The effect of an increase rise in the price of alcohol is to reduce alcohol consumption to some extent. The extent varies by type of drinker, by drinking situation, and possibly by the quantity drunk in each situation.

6. The biggest reductions in harm from rising alcohol prices are likely to arise from
– reduced teenage consumption;
– inhibiting moderate and heavy drinkers becoming very heavy drinkers, and
– reduced additional drinking in a session.

7. The strongest economic reasons for extra regulation of the alcohol market (including the imposition of excise duty) are
– externalities and the need to bring market prices in relation to social costs;
– the semi-rational or irrational behaviour of teenagers learning to drink, and of heavy drinkers and addicts (although the latter group may not be particularly susceptible to price signals); and possibly
– inhibiting moderate and heavy drinkers becoming very heavy drinkers and addicts.

8. Excise duties based on ethanol are probably mildly regressive, that is they probably impose proportionally more on the poor than the rich. However the biggest redistributional effect of excise duties is the transferring of spending power from heavy drinkers to moderate drinkers and non-drinkers. This transfer is in the opposite direction to the social costs which heavy drinkers impose on others. Thus the main fiscal effect of excise duties on alcohol is to offset the redistributive impact of social costs of alcohol misuse – that is to reverse (in part) the transfer of social costs which heavy drinkers impose on moderate drinkers and non-drinkers.

9. Higher prices for alcohol impact on different consumers if different ways. The evidence suggests that on average while modest drinkers are not greatly affected, heavy drinkers may cut back their drinking by 1 percent for every 1 percent rise in price, and chronic drinkers are almost completely unaffected. Teenagers reduce their drinking in the face of higher prices, and it seems likely that there is less drinking in extended drinking sessions as the price rises. Higher prices may also inhibit moderate drinkers becoming heavy drinkers and heavy drinkers becoming chronic drinkers.

10. In a modern market economy, market prices coordinate the decisions of producers and consumers. To do this effectively, the prices should relate to social costs.

11. While many people are offended by excessive drinking, which in economic terms may classify it as a ‘demerit’ good, this may not be a compelling reason for rasing alcohol prices.

12. While there is some irrationality (or quasi-irrationality) among some alcohol consumers and on some occasions, it is not evident that price policies are particularly helpful in reducing harm from this course. Insofar as irrationality is a problem, other interventions need to be pursued. The report notes however, that teenagers who are being socialised into moderate drinking need special attention.

13. Alcohol is one of a handful of products, where the costs of production do not roughly reflect the overall costs to society of consumption. The excess social costs are substantial. The most comprehensive estimate suggests that alcohol misuse reduces effective GDP by 4.0 percent, may well reduce the effective size of unmeasured (informal) economy by a similar amount, and has also reduced the welfare of New Zealanders via additional mortality and morbidity by 2.0 percent and the population of New Zealand by .8 percent.

14. The excess social costs may be thought of as the economists’ equivalent of harm, in which case the objective of alcohol policy in economic terms is to reduce social costs. Reducing the gap between the prices which individuals make their alcohol consumption decisions and the social cost to the economy will reduce harm, because individuals are less likely to partake of potentially harmful consumption.

15. It is not practical to require alcohol consumer to pay an appropriate insurance for the social costs they may incur with each drink …

17. While a lift in alcohol excise duty will raise prices for all drinkers, many will be better off, insofar as the resulting tax revenue will be recycled back. Under some assumptions, over 70 percent of adults will be better off as a result of this recycling (and also from reductions in harm). This is because heavy drinking is concentrated in a small part of the population. They will be made worse off as a result. On the other hand heavy drinking is the main source of harm, and so the heavy drinkers would be paying for a greater share of the harm they generate.

New Issues

18. Among the new issues facing strategies to reduce alcohol harm are new beverages, and new legislation which liberalises access and attitudes to alcohol. Excise duties probably can do little to assist with any potential harm these may generate …

19. A major new issue is the rising incidence of teenage drinking over the last decade, and the clear evidence that much of it is potentially harmful. Teenagers are sensitive to higher prices. Higher excise duties would reduce their potentially harmful consumption, while they learn to drink like mature adults. (The report also notes that there is also considerable heavy drinking by adults in their early twenties, which would also be reduced by higher excise duties.)

20. A second new development is the introduction of new beverages, for which the cost of ethanol is exceptionally cheap …

The Proposed Excise Duty Regime

23. An excise duty on alcohol can contribute to reducing harm from alcohol misuse by raising the minimum price of alcohol, thus reducing consumption by teenagers and heavy drinkers, and inhibiting the creep of moderate drinkers into heavy drinkers and to chronic drinkers. However, ideally the excise should not impact moderate drinking which is not potentially harmful. In practice this impact cannot be eliminated but it can be moderated by choice of the duty regime.

24. The most effective excise duty with the purpose to reduce harm is likely to be one which is levied on ethanol (the quantity of absolute alcohol), and is particularly concerned with the minimum price of alcohol. A particular merit of this levy regime is that it impacts relatively less on high value (relative to ethanol content) drinks.

28. Very low alcohol beverages should be exempted from excise duty on the basis that their consumption causes little harm (as well as reducing compliance costs) …

29. As far as is practicable, ethanol should be levied on actual content, and the excise rate should not based on duty bands. Where duty bands have to be used, the within band rate should be set at the top of the band in order to avoid suppliers avoiding actual taxation on ethanol content.

31. Distilled spirits (especially light spirits) can be produced at a considerably lower cost than other forms of alcohol. In order to maintain realistic minimum levels for the price of alcohol, either the base excise duty rate for all alcohol has to be raised, or a differential maintained between spirits needs to be introduced. This report recommends the latter option. …

33. The base level for excise duty is a political judgement to be made through the parliamentary process. That it is a political judgement suggests the need for a wide public debate on the appropriate excise duty rate. An important consideration in setting the base rate is the extent to which the total excise duty should contribute to gross fiscal costs of alcohol harm.


36. The report does not recommend further hypothecation. …

The Consequences

The two forces which allied in the 1970s to raise excise duties, came together again to lead to a quick implementation of some of the Taxing Harm recommendations. The fiscal imperative was aghast at the tax anomaly and, of course, always willing to generate more revenue: those concerned with harm were anxious to raise the minimum price of ethanol. In May 2003 the excise duty on all alcohol from 15 aabv to 23 aabv was raised to the same level as for full spirits. The implementation differed from the Taxing Harm report’s recommendation of only raising it on spirits. A consequence was that the price of fortified wines and liqueurs also rose, much to the chagrin – according to the media – of the elderly who enjoyed a glass of sherry.

The spectacular change was that light spirits, whose price rose sharply and are no longer produced, indicative that its attraction was primarily as a cheap source of ethanol. On the other hand it appears (the data are not yet available) that there has been little reduction in the consumption of fortified wines and liqueurs, which suggests they are consumed for the entirety of the product.

Why did the government not follow the recommendations of Taxing Harm, especially in including non-spirits in the tax hike? From the Treasury official’s papers are available under the Official Information Act we can distill the following reasoning.

Officials were reluctant to target a minimum price of ethanol. ‘ [T]he excise would be based on production costs. We consider this basis of excise taxation to be inconsistent with encouraging efficiency in production and growth’. There is no further analysis. It is not obvious that the statement is correct. Suppose that a producer were to find some efficiency which halved the cost of producing ethanol so the cheapest sources of liquor fell dramatically. Would the excise duty be left at the current rate? Of course not. The government would soon increase the excise to offset the fall. Would this government reaction discourage producers from looking for efficiencies? Not by much. The producer who introduced the efficient gain would still be selling at a lower price that the competitors after the excise duty increase, so would still reap a reward from the efficiency implementation. The light spirits is an illustration of this principle although the lower cost was the result of a tax anomaly as well as a new cheap source of spirits (from cheese whey). (The reference to ‘growth’ in the Treasury statement is not a hope for more liquor consumption, but that the efficiency gains would release resources for other sectors.)

Once the Treasury advice had rejected the relevance of a minimum price of alcohol, the lowering of the high excise duty threshold from 24 percent to 14 percent aabv, for all alcohol followed. It also gave greater revenue gains. However the puzzle – observed in the 1991 Treasury paper – as to why there should be two excise rates remained. (Recall the Taxing Harm paper explains it because the focus on the minimum price of ethanol and otherwise minimizing the impact of taxation, but not trying to maximise excise revenue, means there is a case for a differential rate between spirits and other forms of liquor reflecting the differences in their production costs.)

Thus the official advice reflected the principles it set out in the 1991 Treasury paper of taxing the consumption of ethanol rather than the harm from ethanol consumption. However the papers agreed to by the cabinet state very clearly that the purpose of excise duties on alcohol is harm minimisation. But since the papers also adopt the excise proposals of the officials, based on taxation of consumption ethanol consumption rather than minimising harm, there is an inconsistency between the principles and the practice.

Future Directions

How long can this inconsistency be maintained? The resolution could go one of two ways. The harm principle could be abandoned, although that would go against the direction of post-war policy development. Alternately the excise duty regime could be oriented more towards the harm principle and less to the consumption principle. That seems likely to lead to more attention to the minimum price of ethanol, even if that involves a concern with production costs. To get the head-space shift it may be necessary to transfer responsibility for excise duties to the health desk of the Treasury, or at least to involve them more than the May package seems to have done.

Whether it is possible to further tune the excise duty on ethanol to target harm is not evident given the practical difficulties of differentiating by circumstance. One can imagine an excise regime where the level rose as the imbiber drunk more, but there is no obvious way to administer it. (If it could be shown some forms of alcohol were inherently more or less harmful than others, then differential excises might be practical by product, but there is as yet no really authoritative and quantified consensus to such propositions.) At this point the development of excise duty as a policy instrument may be exhausted. Other less economic instruments – restrictions, penalties for drunkenness, especially while drive, and programs to change attitudes – will remain relevant and will have to be actively pursued.