This paper was prepared in October 2009 for the Law Commission when it was reviewing alcohol policy. It is released under the Official Information Act.
Keywords: Health; Regulation & Taxation
Terms of Reference
The Law Commission has asked for a report describing how a minimum price regime for alcoholic beverages might be implemented, and what would be its benefit (especially in relation to a reduction of harm) and its costs.
1. At the lowest price demand for absolute alcohol is probably more price elastic (more sensitive to price) than other beverages. These ‘cheap drinks’ are therefore of particular interest to pricing policies for alcohol control purposes.
2 . It appears that cheap drinks contributes more than proportionally to harm from alcohol misuse.
3. A Sheffield University study has quantified the effects of Scotland raising the minimum price of absolute alcohol. It found reductions in mortality and morbidity, and in public spending on health care and crime and from employment related harms. The magnitude of the reductions depends on the minimum price level.
4. This suggests that alcohol control policies should seek to increase the minimum price of absolute alcohol.
5. The report explores the possibility of regulating the minimum price of alcohol by a Retail Price Maintenance (or a price control regime which sets a minimum price)
6. The benefits of regulating the minimum price of alcohol include
– reduce their consumption;
– reduce the harm from cheap drinks;
– increasing on-licensing drinking;
– with little impact on moderate drinkers.
7. However, regulating the minimum price of alcohol would also involve
– administrative difficulties;
– a regime which conflicted with the economic policy of the last twenty-five years (although the different regime might be justified by having a particular social purpose);
– changing the balance of economic power in the supply chain, probably to the detriment to the consumer;
– increasing the revenue and profits of suppliers without their providing any additional services to consumers;
– providing the incentive and the funds (from the increased profit) for suppliers to market cheap drinks more vigorously, thereby increasing sales of the cheap drinks, and increasing harm.
8. The report also considers using excise tax to set a minimum price for absolute alcohol.
9. The benefits of using a taxation regime to set a minimum price of alcohol include
– reduced consumption;
– reduced harm from cheap drinks;
– simpler to administer (although there would need to be an additional effort to obtain minimum price information);
– the revenue from the additional outlays would go to the exchequer to be used for reducing other taxes and/or increasing public spending and/or reducing the deficit.
10. However using a taxation regime to set a minimum price of alcohol would also involve
– a smaller increase in on-licence drinking than from regulating the price of alcohol to the same minimum level;
– a greater impact on moderate drinkers, although this could be offset by the recycling of the additional excise revenue.
11. This report sets out possibilities and feasibilities. It makes no recommendations.
In recent years there has been an increasingly realisation that alcoholic beverages, like all consumer purchases, involve a bundle of characteristics. Confining the concern – as for most of this report – to off-licence purchases the characteristics can be reduced to the quantity of absolute alcohol (which is easily measured) and quality. Quality is rather poorly defined, but explains why individuals pay more than the minimum price for an alcoholic beverage, since they are also purchasing desire this intangible quality.
Formally then (but put simply) the beverage purchaser is making a decision involving three characteristics – the quantity of absolute alcohol, the beverage quality and expenditures on other products .
The analysis is a little tricky since the absolute alcohol and the beverage quality cannot be purchased independently of one another. This report eschews the mathematical formulations, but their essence is that an increase in excise duty which increases the overall price of alcohol beverages may not necessarily reduce consumption of absolute alcohol (or not reduce it by much). Rather, purchasers may reduce their consumption of the beverage quality while maintaining the consumption of absolute alcohol by shifting to lower beverage quality products. Of course some purchasers will reduce absolute alcohol consumption in order to maintain a higher beverage quality, but the insight from the analysis suggests one of the reasons why the consumption of absolute alcohol is relatively price inelastic (that is total consumption does not change much when prices change).
The one group of consumers who cannot trade quality down are those are already consuming beverages for which there are no lower beverage quality. Such beverages are identified by their having a low price of absolute alcohol since the product does not contain anything but the minimum beverage quality. The beverages are typically sold off-licence, since an on-licenced premises sales are more expensive because of the cost of supply of drinking facilities (which are a part of beverage quality).
As a consequence, it seems likely that the consumption of those who drink the cheapest alcoholic beverages is more price elastic (sensitive to a change in price) than those who consume more beverage quality.
There has to be a lot of conjectures in such analyses. The comprehensive empirical evidence does not exist, mainly because it is very hard to get relevant data bases. While there is a vast literature on aggregate demand for alcohol beverages , there are only a handful of studies which assess the impact of price changes on particular fractions of drinking, and each is subject to some flaws.
Even so, on the basis of what is known and the general theory of consumer behaviour it seems likely that those alcoholic beverages which are purchased primarily for their absolute alcohol content (i.e. having a minimum of beverage quality) are likely to be more price elastic compared to other alcoholic beverages.
It follows that pricing strategies for alcohol control purposes are likely to have their maximum impact on the cheapest beverages measured by absolute alcohol content.
Cheap Drinks As A Source of Harm From Alcohol Misuse
The Law Commission’s Alcohol in Our Lives describes the harm that alcohol use and misuse generates. Probably all consumption causes some harm although it may be an offset by benefits.
At one stage alcohol control policy tended to focus on reducing the aggregate consumption of absolute alcohol. Subsequently, there has been greater attention to targeting certain types of drinking or situations which involved drinking, where it is believed that the greatest harm occurs. (The Law Commission report is an example of this approach.)
Under the new approach it is thought that the targeted interventions will be insufficient to reduce harm to acceptable levels. without forgoing the benefits from consuming alcohol. The approach therefore still requires an excise duty on alcohol. In New Zealand the level of the duty is largely related to the level of absolute alcohol, although there are various ‘anomalies’. The justification for the regime is that absolute alcohol is the best indicator of harm that can be readily taxed. (Note that for a given level of harm, the success of the targeted interventions reduces the necessary excise duty.)
The proposal for a minimum price for absolute alcohol arises from the belief that cheap drinks contributes more than proportionally to harm from alcohol misuse. In particular cheap drinks are associated with binge drinking, heavy drinking and alcoholism. The belief is that if the cheap end of alcoholic beverages were more expensive, there would be less drinking of them, and less harm.
These effects are explored in Sheffield Alcohol Policy Model of the University of Sheffield, prepared for the Scottish Government which is proposing to introduce a minimum price regime. 
Here are the simulations based on a minimum price of 40p per unit. ( A unit is 10 mls of absolute alcohol – say a ‘standard drink’). This is substantially higher than current minimum prices, which are below 25p per unit.
The estimates are particular to Scotland which has different drinking patterns from New Zealand. However in order to get a sense of the magnitudes involved, it may be useful to note the Scottish population is slightly larger than New Zealand’s (around 5.2m against 4.3m) while the British pounds could be multiplied by 2-and-a-bit to convert into New Zealand dollars. The Scots consume absolute alcohol more per head than New Zealanders.
Under the minimum unit price of 40p, overall consumption falls 2.7%, and is estimated to be reduced by on average 22 units per person per year. Consumption reductions are greatest for harmful drinkers (3.7 units per week), who are 6 percent of the population over 11 years old. Hazardous drinkers, who are 21 percent of the population, drink 0.5 units less. Moderate drinkers, who are 64 percent of the population, are affected in a small way (approximately 0.1 units less per week). 
While off-licence consumption falls in volume terms, it rises slightly in value terms (i.e sales) because of the rise in price. On-licence drinking increases in volume and value terms. It could be argued this is a desired effect since, probably, supervised on-licensed drinking is less harmful than unsupervised off-licenced drinking among some drinking groups.
The effects on health are estimated to be substantial with deaths estimated to be reduce by approximately 40 within the first year of implementation and a full effect after ten years of around 210 a year. Deaths are differentially distributed across the groups, with approximately 10 amongst moderate drinkers, 60 amongst hazardous drinkers and 140 amongst harmful. drinkers. Illness also decreases with an estimated reduction of 1,500 chronic and 500 acute illnesses at full effect. Hospital admissions are estimated to reduce by around 800 in year 1, and a full effect after 10 years of 3,600 a year.
The societal value of these harm reductions were estimated as £540m in total over the 10 year period modelled. In the first year, the estimated societal value of the harm reductions were as follows: NHS cost reductions, crime costs saved (£1.0m), and employment related harms avoided (£20.5m). (Note the substantial gains from employment related effects including higher employment.) Additionally there are improvements in the quality of life estimated as £8.6m for health and £0.7m from reduced crime. The societal value of harm reductions is distributed differentially across the groups, with harmful drinkers accounting for £360m of the total value, hazardous drinkers accounting for £110m and moderate drinkers £70m.
However, overall revenue to the British Treasury (from excise duty and VAT receipts) falls by approximately £4m, although there are offsetting expenditure gains from the lower demand for public services.
These figures are indicative and subject to wide margins of error. But they illustrate how, and in some of the ways, a minimum price of absolute alcohol reduces harmful drinking with a relatively low impact on moderate drinkers (and a positive impact on non-drinkers). 
The analysis also implicitly draws attention to three major problems with the strategy.
First, harmful drinkers are drinking over 12 times as much as moderate drinkers (6.0 units a week vs 5.1 units). The difference remains almost as large after the minimum price is imposed (61.3 vs 5.0).
The implication that a minimum price for absolute alcohol will reduce harm and harmful drinking, but it will not eliminate it.
Second, the policy is not tested for effectiveness against any other policy – such as a general increase in excises.
Third, total sales increase £87m or 3.4 percent (on a fall in the volume of 2.7 percent). Since revenue to the Treasury falls, the supply industry is receiving a substantial boost in its revenue (and hence its profits, for there is no increase in its costs). There is little discussion in the report as to what happens to the additional industry profit, although it becomes a critical issue when the policy option is evaluated (below).
A Minimum Price Strategy – The Resale Price Maintenance Option
What has been thus far demonstrated is that higher prices for cheap alcoholic beverages has some beneficial outcomes in terms of reducing harm. This section looks at how this objective might be pursued by a regulated minimum price, and the consequential problems.
Most of the advocates of a minimum regulated price do not discus in detail how the policy could be implemented. They seem to have in mind a kind of price control. New Zealand abandoned price controls for products sold in shops over twenty years ago, and there is a diminishing capacity on the part of the bureaucracy to implement such a regime. In any case, the proposal involves setting a minimum, rather than maximum, price.
Alternatively, the minimum price could be regulated via a system of resale price maintenance. This is probably more administratively practical; all its weaknesses apply also to the price control option.
Resale Price Maintenance (RPM) occurs when a supplier of goods specifies the minimum price at which a reseller can sell those goods. The practice is prohibited under section 37 of the Commerce Act 1986 (and so if it was to be used it would, as in the case of the price control option, require separate legislation). The primary policy reason for the prohibition is that RPM restricts the reseller’s ability to compete with others on price and therefore is an anti-competitive conduct.
The application of RPM to maintain a minimum price for absolute alcohol would involve the wholesale suppliers of alcoholic beverages requiring their retailers to sell the products at a set minimum price which would be regulated by an authority to have the effect of setting a minimum price for absolute alcohol. Presumably the system would only apply to those products which are at the cheap end of the market, although it may not be easy to define the line between that for which the RPM applies and that for which it does not.
There is only a handful of large suppliers of alcoholic beverages and sales are dominated by a handful of supermarket and off-licence chains. Any application of RPM to some or all alcoholic beverages sold off-license would involve a change in the balance of market power between wholesalers and retailers. There is no systematic study of the existing situation, so it is not possible to assess the implications of such a change, although it seems likely that consumers would be worse off.
An even more serious problem may be that, as the Sheffield study draws attention to, the RPM (or a price control) raises the revenue from sales, and hence the profits of those who supply the alcoholic beverages.
What the industry would do with these additional profits is uncertain, but there would an incentive for it to vigorously market the now more profitable beverages which have a set minimum price. Insofar as the marketing was successful, some or all of the gains from reduced harm from cheap drinking could be wiped out. Presumably there could be restrictions on marketing – realistically on all marketing, since again the boundary between cheap and moderate liquors would be difficult to define.
Were that successful the public might still object to a policy regime which gave the alcohol supply industry gratis increase in profit at their expense. It is possible – although not obviously so – that an ad hoc taxation regime could be implemented to pull back theses profits from the producers. It is likely to be fraught with difficulties, including how to treat offshore producers.
In summary while a RPM regime would be feasible – and is administratively more practical than a minimum price control regime – it is subject to major weakness including changing the balance of market power within the supply chain and increasing industry profits. While it is not possible to assess the first effect, measures to reduce the second would involve increasing interventions and may not be successful.
Either the price control or RPM regime would be quite out of character with overall economic policy of the last two decades. This, does not in itself, preclude the use of such mechanisms, if the social gains were substantial enough. (That tradeoff is a political judgement.) However, as the next section shows, there may be a more practical alternative.
This said, the Scottish government is further advanced in its planning of a minimum price scheme, although details are not yet available. It will be facing the same challenges and developments there should be monitored.
A Minimum Price Strategy – Targeting Excise Duty Option
The following section is illustrated by the effects of the policies on a 330ml bottle of 4% aabv beer costing $1.33.  The illustration is indicative. Nor should the illustration of a minimum price of alcohol at $1.20 per standard drink (10 mls of absolute alcohol) be considered a proposal. The purpose of this report is to set out the principles behind a minimum price strategy. If it is decided to proceed with one, the choice of levels is for a later stage in the policy development.
An alternative to a regulated minimum price in order raise the price of cheap alcoholic beverages is to target excise duty to attain this objective.
One of the current purposes of the excise duty regime is to raise the price of absolute alcohol, in order to discourage consumption and thereby reduce harm. Because it is a largely uniform rate on absolute alcohol, it raises the price of cheap drinks more than those which are higher priced (which are also charging for their beverage quality).
The level of the excise duty has a substantial impact on the price of cheap drinks. For instance including the GST on the excise duty, over a quarter of the price of a $1.33 a 330ml bottle of 4% aabv beer can be attributed to excise duty. 
The level of excise duty rate can only really be explained in historical terms. In the far past it was set by politicians making a nice judgement between what they thought was politically feasible and the excise revenue that they needed for fiscal purposes. Subsequently the rate was updated for inflation – as is now done annually – plus any further change as the balance between political feasibility and the desperation for revenue changed.
There has been some consideration as to whether the aggregate revenue from excise duty covers the fiscal costs of harm from alcohol but that calculation has not been decisive in setting the duty rate. (In any case, the economic logic is that tax revenue should also cover some of the private costs of alcohol harm.)
Thus there remains a question as to what is the proper excise duty rate. Insufficient is known in order to identify some optimally appropriate rate (e.g. one which gives the maximum social benefit from the consumption as offset by consequential harm). But the logic of the earlier part of the report is that perhaps the minimum price of absolute alcohol should be taken into consideration.
Suppose it was decided that the minimum price for 10ml of absolute alcohol was to be $1.20. (This price is chosen entirely for illustrative purposes and is not a recommendation.) The calculations for beer are below, but for comparison purposes it would raise the minimum price for a 750ml bottle of 12.5% aabv wine to $11.25.
Suppose, too, the previously discussed 330 ml bottle of beer with an aabv of 4 percent was the cheapest source of absolute alcohol. It contains 13.2 mls of absolute alcohol. That would set its price in the minium price regime at $1.584 (1.32 X $1.20), or 25.4 cents higher than its current resale price. Instead of using a regulated price in which the additional 25.4 cents (less GST) would go to the suppliers, the excise duty could be increased by 67 percent to 63.1 cents (including GST) on the bottle and the increment in revenue would accrue to the government.
Accordingly the excise duty on beer would be raised from $25.4756 per litre of absolute alcohol, to $42.5443 per litre (with GST additional). The broad effect form this example would be to increase the price of every standard (10 ml of absolute alcohol) drink by 19 cents(including GST), plus whatever margin the suppliers paced on them, irrespective of the beverage quality.
This change involves only an existing administrative system (and a simple amendment to the Customs and Excise Act 1996) , so it could be readily and effectively introduced.
(At the time of the new legislation it would also make sense to remove some of the anomalies– especially the ones which favour wine – and, arguably, to introduce a lower (or zero) rate for low alcohol beer. The spirits excise rate – i.e. the rate for alcoholic beverages over 14% aabv – would be aligned so that the cheapest available source of absolute alcohol would be the same as for beer and wine.)
The current practice for increasing the duty in line with consumer inflation could be continued.
However there is no reason to expect the cost of beverage production and distribution would increase at the same rate as inflation while it is possible that overseas sources of alcoholic beverages could undercut existing prices.
Means of observing the lowest price source of absolute alcohol include.
1. inspection; or
2. requiring retailers to report their lowest price beverage; or
3. using point-of-sale electronic records.
In the event of a the lowest price falling below the target minimum price (after adjusting for consumer inflation) it would be necessary to raise further the excise duty.
The major weakness of this regimen compared to the regulated minimum price is that according to the Sheffield University analysis the regulated price regime would have little impact on moderate drinkers rasing their prices only a little. There are two responses to this.
The first is that the revenue accrues to the government who will recycle it via tax reductions and expenditure increases. Moderate drinkers and non-drinkers could be net beneficiaries from these changes (as well as from a general reduction in harm).
The second is that it might be possible to construct a drawback arrangement whereby , as the price of the absolute alcohol rises, the excise duty gets phased down. It may not be practical since it probably would have to be applied at the wholesale level (and hence would not discriminate in favour of on-licence drinking) and there may be opportunities for gaming the system especially as the excise is levied at the wholesale stage but the relevant price is the retail one. Note also moderate drinkers on modest incomes would be paying more excise duty per unit of absolute alcohol than those who could afford dearer drinks.
It should also be noted that there will not be as strong an incentive towards drinking on on-licenced premises as for the regulates prices option.
Undoubtedly alcohol control policies need to pay attention to the minimum price of absolute alcohol, for it appears that cheap drinks contribute a more than proportional share of the harm from alcohol misuse. than the more expensive ones.
However a regime which raises a the minimum price of alcohol involves practical difficulties.
A regulated minimum price regime is likely to affect the balance of market power between producers and retailers, while increasing the profits of the supply industry; some of the additional profits may be used for increased marketing. The regime has two major advantages, in addition to the reduction of harm which comes from a higher minimum price. It favours supervised on-licence drinking and it has a low impact on moderate drinkers.
An alternative would be to use excise duties to raise the price of the cheapest drink to some desired level. This is far more administratively straight forward than a regulated minimum price, although it involves additional effort collecting price information. In contrast to the regulated price option, it does not favour on-licensed drinking as much, and moderate drinkers are also affected by the higher excises. Even so most are likely to be better off when the additional revenue is recycled (via lower taxes and/or higher public expenditure) and by the reduction in harm that consuming cheap drinks causes.
This report sets out possibilities and feasibilities. It makes no recommendations.
 Another reason is container size. The mathematical analysis assumes that the beverage is available along a continuous spectrum. In practice it is purchased and drunk by the can or bottle or glass.
 There is a differential rate between spirits and other alcohol based beverages. While the rate seems to have arisen from historical circumstance, it could be justified by observing that were there to be a uniform rate, spirits would be relatively cheaper, and that (depending on the assumptions about the rate) would lead to a higher rate of absolute alcohol consumption or higher average prices.
 But see Model-Based Appraisal of Alcohol Minimum Pricing and Off-Licensed Trade Discount Bans in Scotland: A Scottish adaptation of the Sheffield Alcohol Policy Model version 2 ScHARR, University of Sheffield, September 2009. http://www.scotland.gov.uk/Resource/Doc/285795/0087053.pdf
 The simulations include the effects of of other minimum prices, and also of prohibiting discounting.
 The remaining 9 percent of the population over 11 are non-drinkers.
 Note however that moderate drinker outlays rise by 2.2 percent.
 The example was chosen because it is an example in another Law Commission paper.
 The actual excise duty is currently 33.6 cents plus 4.2 cents GST a total of 37.8 cents on the bottle – the proportion is therefore 28 percent.