We will just do them together, then. I will say a bit about the Helsinki visit. Personally, we found the price of alcohol to be a sensitive issue. We learned very quickly how expensive it was, even given the rate of the pound against the euro, to purchase alcohol in Alko shops, which are the national retail outlets for alcohol, or in a restaurant. The price is so extortionate that it limits purchases. The problem for Finland is external, with Estonia being so close. Because the prices for alcohol in Finland are so high—of course, the tax revenues go to the state—people simply make a short trip to Estonia to load up. There are also special boat trips to Estonia. Cheap alcohol can even be ordered on the internet. For all I know, it is possible for people to get it delivered to their door, like Tesco and Sainsbury’s deliveries.
Again, it is a cultural matter. Finland had a temperance movement, and it is still pretty active. It has one month when it tries to get people to abstain from drinking alcohol but, being clever Finns, they picked February because it has only 28 days. I thought that that was rather charming. I thank you for the observations. They were an extremely useful backdrop to our next item of business.
By way of background to our Scottish work, I say that we were initially commissioned by the Department of Health in England to do a systematic review of the literature on pricing and consumption, pricing and harm, promotion and advertising, how consumption and harm are related, and what we could expect if consumption drops by a certain amount. That was the first report to come out of the project. The second report was a modelling study that was not dissimilar to the current one but which also covered advertising, general price rises and taxes targeted at lower-price alcohol. Slightly more policy options were considered than in the Scottish work, which focused on the discount ban and minimum pricing.
We should not lose sight of that third driver when we get into the debate on the pricing issue.
No. The evidence statement did not refer to our modelling, which we had not done at the time. There is one paper on the subject, which we would not call relevant to the UK study; nevertheless, it has been peer reviewed and published in a decent journal. However, there are other studies, such as the Paul Gruenewald study, which looks specifically at what happens if the price is increased for low-priced products only, which is what happens with minimum pricing. Although it is not a minimum pricing study, its findings are applicable. That is why we said that there is some evidence on the effectiveness of minimum pricing. However, nobody has done any research on minimum pricing in a similar context to us, in a western culture, because nobody has tried it outside alcohol monopolies except in Canada, where the situation is again not comparable to the situation here.
My next point was to be about the Canadian experience of social reference pricing. We have corresponded with Tim Stockwell in Canada—I believe that he has made a big grant application and I hope that he succeeds, because the policy needs to be studied. There is no published evidence on social—
I do not quite agree that those people will continue to drink in exactly the same way and will experience the relevant harms. When people who are now 30, 40 and 50 were 18 to 24-year-olds, the pattern was the same—they drank more in the on-trade. People subsequently switch to the off-trade, bringing with them their high drinking patterns but tending to stay at home to drink.
For background, it is important that we get the definitions right. I hope that you agree that although we use the elements of the abstainer, the moderate drinker, the hazardous drinker and the harmful drinker, there is a continuum. However, those are the generally accepted groups and we have definitions of them.
No, the groups are mutually exclusive as far as our modelling is concerned. The harmful drinkers are harmful drinkers and the hazardous drinkers are those who drink below the harmful level.
I do not know exactly how the Scottish household survey defines that when it reports its results. I can say only that we used the Scottish survey, but we did not use its definitions. We split up the population into those who drink up to 21 or 14 units a week; those who drink over that but below 50 or 35 units a week; and then the people who drink above the harmful levels.
The literature suggests that harmful binge drinkers, as one group, tend to be less price elastic than moderate drinkers. However, the only review of that is the Wagenaar analysis. Basically, he included studies on young binge drinkers, so they were not harmful or dependent drinkers per se.
The figures are 16 per cent for moderate drinkers, 23 per cent for hazardous drinkers and 35 per cent for harmful drinkers, in terms of their average spend.
Yes, that is the spend—that is not a consumption reduction, I hasten to add.
I want to put on record that page 29 of the Sheffield report states:
I am not in any way qualified to comment on Scotland’s cultural relationship with alcohol. It is not my area of expertise.
So you think that Scotland can just be lumped in with any other country.
That is not what I am saying; I am saying that I cannot comment on that.
Could we have one question at a time?
The point is that the cross-price elasticity of demand is the key active ingredient for—
Okay. I am not sure what you mean by “the substitute effect”, as I understand that to be the cross-price elasticity, which we did estimate. As you saw in the report, there is a mixture of Scottish and English data at the moment, but the next iteration will be based on Scottish data.
Did you consider the cross-price elasticity of demand between one drink and another? As I have mentioned before in the committee, the point has been raised with me that many young people in the Highlands may find illegal drugs more attractive should there be a minimum price for alcohol. It is therefore a matter of the cross-price elasticity of demand for two goods, or within one good.
We did not consider the effects with drugs. There is no good evidence on which we could base such a model. It is acknowledged in the discussion section of the report that that would have to be monitored carefully.
We are considering the overall consumption of alcohol. In Finland, when high prices were applied to drinks, it was discovered that people actually consumed more, because of cross-border trading with Estonia. All that happened was that the duties were lost. The issue is the overall consumption of alcohol.
Just a minute—there is supposed to be a convener here. I have a light touch, but I am not invisible. We will move on, as Ross Finnie and Michael Matheson are waiting. I will then call Helen Eadie and Rhoda Grant. You can come back in after that.
I think so. We thought long and hard about how best to approach things. We had to make assumptions, and I think that we were open about where assumptions were made. The work went through various rounds of peer review and people generally agreed that our assumptions were the best that could have been made. On occasions, people suggested alternatives, which we used as sensitivity analyses. For example, when people were not happy with the econometrics that we used, we used alternative evidence from Chisholm and from Huang, who did a previous UK study—I think that it was for HM Revenue and Customs. We checked how sensitive our model was to alternative assumptions, and although there were some variations we generally found that it was not far out in terms of the scale of effect.
Do the figures not relate to particular sub-groups?
I will come back to you. I want to let members ask more questions.
So this type of modelling is not a new approach to dealing with such matters—it is a well established, academically recognised approach.
You mentioned that you are doing some further modelling work and referred to a few areas in which you will seek to use additional Scottish data. Can you say more about the extent of that new modelling? I am conscious of the fact that your findings may be substantially different once that work has been done. As a committee member, I would like to question you again when your new report has been published. What impact do you think the work may have? Will you use a similar model?
I advise the committee that we have scope during stage 1 to consider and ask questions about the supplementary report, if we wish.
The same methodology will be used for the updates. The issue is that new data have become available. There is interest in using the most recent Scottish health survey—for 2008—which was not available when we started the work, to update the consumption data. We also have the price data from Nielsen, the market research company that has detailed Scottish prices against which we can validate the expenditure and food survey data that we use at the moment.
So you assert that everything that you say is based on evidence.
It is not an evaluation—it is based on real data and evidence.
Of course in Scotland. I am just not sure that I have those figures to hand. Can you make a note of that and I will supply the information in the next few days?
Certainly.
Dr Meier has already addressed that point.
Recently, I have attended many European meetings, and you are right to say that a holistic approach—in other words, an approach using multipronged policies—has been recommended. However, at every meeting, it has almost been taken as read that price is the most effective lever that Governments have at their disposal and that something must be done about it. It has also been made clear that something has to be done about availability.
Have you published any papers on or carried out any review of the alcohol policies in the 29 member states?
On page 2 of your submission, you set out the percentages for the overall reduction in consumption for different minimum prices coupled with a discount ban. I take those to be the average figures. Have you worked out any figures for different income groups? Is the impact different for, say, low income, average income and high income drinkers?
It depends. In general, the impact on moderate drinkers would be fairly minor, especially if the minimum price were in the lower range—up to 50p per unit, say. If I remember rightly, the estimated increase for moderate drinkers was £11 per year. Even for a low-income drinker, that would not be a dramatic amount. There might well be income effects for harmful drinkers, but we do not know. On average, harmful drinkers spend about £2,000 per year on alcohol, so an increase of 5 or 10 per cent would be a substantial amount. The difference that it made would depend on how much money the person had available to buy alcohol. As I said, we have not looked at the issue in detail.
I think that it includes people who drink more than one unit a week, if I remember right. It includes some very low-level drinkers and some people who drink a little more.
We have established that it is an average.
Yes.
I will let Rhoda Grant work that out. I am getting a headache at the thought of working out the increase in cost for someone who drinks 21 units a week. What would that amount to?
That is fundamental.
My next question was going to be whether you had done any modelling by income group of the effect on hazardous drinkers and dangerous drinkers. Can I assume that you have not, given that you did not do so for moderate drinkers?
I found your paper very interesting. Earlier, Dr Simpson used the quotation:
I think that it has strengthened it.
I want to ask one more question about the report to the Department of Health. It says:
Not at all. There are the same kinds of responses on advertising restrictions, for example. Health people tend to be in favour of such things and cite the evidence, but the industry will be more cautious. Similarly, every time the issue of tax is raised, the health lobby will say, “Yes, that’s a good idea” and the industry will say, “No, that would be an absolute catastrophe for the country.” So the responses do not surprise me very much, although I am surprised that there has been so much of a response from the health lobby. When such policies are proposed, the health lobby is usually slower off the mark to comment than the industry. The divide between respondents and how they responded was entirely predictable.
I encourage the committee or researchers in general to look into whether it is true that moderate drinkers buy the very cheap stuff that is targeted. You should consider whether that is just an anecdotal example of what one couple buys, and how common that is. You would want to know those things, and to know how people on low incomes would be affected, and then you will balance that somehow against the health benefits that you assume. At present we have not looked into that, so I would hesitate to say that that is the general pattern of consumption of low-income drinkers. My gut feeling is that they are probably more likely to drink in the on-trade, where minimum pricing will not have much effect, but we need to establish the facts before we can draw those conclusions.
There is no doubt that every single member around the table is concerned about those whose drinking is hazardous and harmful, whom we are trying to target the most. I was interested to read the papers that are before us, one of which points out the fundamental contradiction in the Sheffield report, namely that the modelling assumes that heavier drinkers are the most price sensitive, even though the systematic review cites studies that suggest the opposite. It states—
It is in the Centre for Economics and Business Research paper that was placed on the table this morning. It states that the University of Sheffield modelling systematically shows
By whom? It was not placed by the clerks.
We do not have it.
My researcher brought it to my attention.
Okay. Can I possibly have the report passed to Dr Meier? She may not want to answer the question.
Manning et al stated, in 1995, that the 5 per cent of heaviest drinkers have
The literature review by Ludbrook, in 2004, outlined that although there is “unconvincing evidence” that price affects consumption in heavy drinkers, there is more convincing “indirect evidence” that it does, which comes from studies that have shown a decrease in alcohol-related problems following increases in taxation. Would you comment on that?
My question follows on from Rhoda Grant and Richard Simpson’s point on income elasticity and so forth. “The Scottish Health Survey 2008” says:
I am not sure how that differs from the earlier point on the health—
I am sorry—I may have misunderstood you.
You do not have the figures on how that splits between men and women.
We can probably assume that there are more hazardous and dangerous drinkers who are men, so the percentage will be high.
No problem.
I want to ask about what was taken into account in the modelling. As I mentioned earlier—I think that Dr Meier was in the room at the time—in France, wine was previously untaxed yet wine drinking fell steeply while consumption of vodka, which was highly taxed, actually increased. However, the increase in consumption of spirits was masked by the fall in the consumption of wine. For me, that contradicts the whole minimum pricing model. I know that such things are difficult to capture in a model, which is not based on reality, but did your model take into account trends and fashions? You mentioned advertising, earlier.
No—it is notoriously difficult to predict what will happen with such things. As soon as one tries to make a prediction, someone will have a different attitude about what will go down and what will go up. I am not sure what the pricing levels of wine and spirits are in France. Spirits are taxed, but are they still quite cheap, or are they very expensive?
The French were taxing vodka particularly heavily, because they viewed it as the target drink. It was made very expensive even in comparison with its price here, although alcohol is quite cheap in France. Wine is very cheap because, politically, the French cannot tax it, so they taxed vodka very highly because they saw it as the problem, and consumption has still been increasing. They did not find that it was to do with advertising, on which they have strict rules. It has not been researched, but it has been suggested that there is a global culture in drinking, with which the French were out of synch because they drank a lot more wine and very little by way of spirits. A convergence was said to be going on, and we are perhaps operating on a different level here, with the volume of wine drinking going up and that of spirits drinking going down. The French tried to use very high taxation to stall consumption, but it did not succeed.
Richard Simpson asked a question earlier that was along similar lines to what was going to be my final question, so I would like to follow up his question. He spoke about the marginal effect—almost expressing it as a tax—of an increase in price on low-income drinkers, and he put a question to you. I understood your answer in general terms, about the difference between price and tax. I followed that. However, I was not clear—whether it was in Richard Simpson’s question or your answer—about whether a 10 per cent increase in tax, as opposed to the imposition of a minimum price, would have the same effect in health terms as is suggested in your report. I am sorry if I am mangling Richard Simpson’s question.
No—you are not.
I understand that. We have a 10 per cent price increase and the imposition of a minimum unit price at 40p. Would they have the same effect on public health?
No. The reference is not to the Scotland report, but to the Department of Health report, which I do not have with me. If we consider the total valuation of the changes to harm, we find that the 40p minimum price comes out stronger than the 10 per cent across-the-board price increase because, in the balance of effect, the minimum price had a greater effect on the harmful drinkers than it did on the moderate drinkers. Although the populationwide consumption reduction was very similar, the effects on harm were slightly increased by the minimum price compared with the 10 per cent across-the-board price increase.
That is helpful. That leaves us only the unanswered point about the proportionality within those elements. You were not asked to answer this point, but the issue of the proportions among the moderate, harmful and hazardous groups—and lower, middle and higher incomes—is left outstanding, and only once we know about them can we tell who would be affected by a marginal increase in price of that sort.
Thank you—that was an interesting question, Ross.
On the cultural aspect—I hope that you agree with this—we were told that wine drinking was part of rural culture at one time, with a lot of cheap wine being taken without water. Consumption is dropping now, but people are drinking more good-quality wine. I think that we all accepted that it would be impossible politically to increase tax on wine in France, because just about all French politicians have a small vineyard in their area, and a tax increase on wine could mean jobs being lost and politicians having to look over their shoulder. Very few of them are prepared to go down that road. Politics, both internal and external, played a substantial role in what we found in Paris. Does somebody else want to comment on the French experience, before I go on to the Helsinki one?
I just want to add to what Mary Scanlon said. One of the things that was told to us almost as an aside, and which seemed to fascinate us all, was that young people under the age of 24 had decreased their drinking substantially. No research had been carried out into why that was happening; the only thing that they could point to was the adverts that Mary Scanlon mentioned. They influenced parents, who then did not drink so much in front of their children, so the children did not pick up those habits. I am interested to know why there was such a substantial decrease in the amount of drinking among young people.
Obviously, we were in two countries with very different approaches to alcohol. Finland is still regarded by the World Health Organization as a low alcohol consumer, while the WHO still considers France to be a high alcohol consumer. I suppose that our experience was slightly predicated on the fact that, essentially, we met health officials. I do not disagree with what the convener said regarding what we found, but the debate was portrayed in health terms. It was interesting that, in both places, health officials were concerned about an increase in binge drinking, but their real, long-term concern was the increasing trend of persons resorting to alcohol in quantities that, in the opinion of the health officials, were not good for their long-term health. There was an absolute divide between health departments and departments of the economy, trade and industry. As the convener has observed, the parliamentarians in France were simply not prepared to consider the health issue. Indeed, although they tax spirits, it is interesting to note that their levels of taxation across the board are still considerably lower than the levels of taxation that prevail in the United Kingdom.
That is helpful. You paper refers to Gallet and Wagenaar as the major meta-analyses on the subject.
That was not very relevant.
That is excellent.
We will collaborate on that work, so we will know more in due course.
I do not have the figures in front of me—I have only those that include a discount ban—but that sounds right.
The modelling study refers to a group that concerns us—18 to 24-year-olds, who are drinking volumes more and are binge drinking. That group will produce the harm for the future and its behaviour reflects the rise in consumption. Your study shows that MUP would reduce consumption by only 0.7 per cent among those people. What is the reason for the difference in that group?
That is not quite true. I will clarify how—
Let Dr Meier answer.
Dr Meier is here specifically to answer to the Sheffield report.
Yes.
No, I just—
Dr Meier, can you answer all those questions if they are put to you in a list like that? I want to give you the opportunity.
I am trying to write notes.
Hold on, Mary. Your questions are not a problem—it is just a matter of hearing them one at a time, so that Dr Meier can answer them. Your first one was?
Yes—and the substitute effect. The issue is how people overcome a minimum price.
That is important.
Good morning. I will ask a question that is different from the one that I originally intended to ask. I followed Richard Simpson’s line of questioning, and I want to give Dr Meier the opportunity to put it into context. We have to evaluate what your report actually demonstrates, and, as with any report, there will be pluses and minuses.
I will ask a brief supplementary on that point before I move on to my second question. It is clear that, on that basis, you have made assumptions; the question is whether they are reasonable or unreasonable.
I have not often looked at models, and some of the formulae left me slightly askew. Because you started with reference points in relation to indicators of harm, I found it difficult to go back—I am thinking about the table that Richard Simpson mentioned. The other problem is that of course the report would go on for a mile if we had all the data, so you picked an illustrative minimum price of 40p in the five main appendices—there is nothing wrong with doing that; it means that you get a series of figures. There is the overall 2.7 per cent reduction in consumption if the minimum price is 40p, and there is the 5.4 per cent reduction if there is both a minimum price and a ban on off-trade discounting. That seems perfectly all right. For harmful drinkers—the ones whom we are interested in—the reductions are 4.7 per cent and 8.7 per cent.
Ours is a population-based model, so it is not useful to consider the individual drinker and say, “Joe Bloggs, who drinks 48 units at the moment, will reduce his consumption by exactly 5 per cent and thus reduce his personal risk.” We looked at population risk. We can work out how a 5 per cent across-the-board decrease in consumption in a certain group translates into reduced numbers of deaths and hospital admissions—the information on harmful drinkers is in the table on page 88—and we can put a price tag on that, using standard health economics valuations. However, because the model does not consider the individual, a narrative around the amount by which an individual’s drinking would be reduced would probably not be helpful in this context.
They relate to the sub-groups that we specified. We made separate estimates for moderate drinkers, but we are talking about all moderate drinkers, rather than predicting what an individual moderate drinker would do.
Can I ask about an issue that Richard Simpson and Ross Finnie raised?
It is common now to model the effect of policies, especially where there is uncertainty because they have not been introduced in exactly the same way before in the same country. The National Institute for Health and Clinical Excellence recommends that approach and uses it in all cases. We have just done work for NICE on possible alcohol policy options around screening, brief intervention and so on. Cost-effectiveness modelling is a standard part of such work.
Yes.
I am not entirely sure what the publication schedule is, but our draft report is due by March. There will be a short period of back and forth, but it should be available in the not-too-distant future. We have some preliminary results that I am not happy to share in detail, but it does not appear that there will be substantial changes as regards the overall effectiveness of the different policies.
I must let Dr Meier respond to that.
I do not see how modelling can be placed on one side and evidence on the other. We used a recognised approach of making predictions based on actual data, not fictional information that we just made up.
It is based on data, but not on evidence of policies that have been implemented elsewhere.
My only experience of modelling concerns the transport policy of the City of Edinburgh Council—we know of the chaos to which that led.
Okay.
In Scotland.
What modelling studies have you done or what specialist marketing opinion has been published on the likely market response to minimum unit pricing?
As you will probably acknowledge, the impact will be different for different income groups. An average drinker in a low-income group, who might buy a value brand bottle of spirits and a cheap bottle of wine a week, will pay substantially more if there is, say, a 40p minimum price. For example, their weekly spend will rise from £6.95 to £10.50 for the spirits and from £3 to £3.65 for the wine, which is about a 41 per cent increase. However, someone on a moderate income who buys a bottle of malt and an expensive bottle of wine will feel no impact at all. Can we assume that there will be more of an impact on lower income groups and that the average reductions in consumption that you set out in your submission apply more to those groups rather than to moderate or higher income groups?
What do you base that level of consumption on?
On what the average moderate drinker buys.
So it includes people who do not drink on a weekly or a monthly basis.
But that does not tell us the cost of the proposal to a moderate drinker who drinks up to the number of units that it is considered safe to drink.
You mean a moderate drinker who drinks exactly 21 units per week.
No—I am talking about what the increase in cost would be. That is like the story about the bath and the buckets of water.
Have you done any modelling to find out what people in lower income groups drink? They will obviously not buy malt. Have you looked at what they consume?
As I said, no separate modelling has been done by income group. Income is accounted for in the econometric model in terms of how people respond to price, but we were not asked to produce separate tables on low-income groups. That is something that could be done with the data.
Yes, but that information is not in the report. We have found out that that work has not been done.
Yes.
Yes.
Has the extra research that you thought should be done strengthened or weakened the case for suggesting that
You are clearly involved in the issue of how we tackle alcohol misuse and your response to Helen Eadie’s questions on the European studies that have been undertaken showed the different approaches that could potentially be used to deal with it. It is clear that there is a healthy level of scepticism among committee members about the modelling and the assumptions that have been made, despite the fact that we are talking about a well-established modelling process that is used for assessing policy decisions.
Yes. It was among our papers that were put on our table this morning.
Perhaps it was my researcher who handed it in.
As I said, we conducted the Wagenaar sensitivity analysis. The Manning study is done quite differently and it is quite old data, so we have not included it as another analysis. It is not particularly relevant. The Wagenaar study is relevant, but, as I said before, the people in the Wagenaar meta-analysis are mainly binge drinkers rather than heavy, dependent drinkers, so they are different from our hazardous and harmful drinkers. Our analyses are based on the most recent data that was available in the UK. Of course, you can find alternative evidence abroad and so on, and we are happy to show how different our results would have been if we had based our work on that international, older literature, but we maintain that our model is based on local purchasing data.
We will move on. Mary Scanlon has some questions.
It would not be appropriate at this committee on this particular bill.
I will let in Dr Meier. She looks as if she is about to explode.
Could you please tell me where you are?
Page 58.
I did not have the figures to hand in putting the question earlier, convener. I apologise for that.
I have quoted your predictions on what will happen in the first year of implementation. All the figures that I quoted from pages 58 and 59 of your paper relate to the first year of implementation.
I will lower the excitement a bit, I suppose.
Yes—that is true for hazardous and harmful drinkers, but it is not so true for moderate drinkers.
I would be grateful for those figures, in order that we can place the debate in context.
Yes—in tax, not price. There is a difference.
Item 2 involves an oral report on the fact-finding trip to Finland and France that was undertaken by members in January as part of the committee’s stage 1 consideration of the Alcohol etc (Scotland) Bill.
It was an excellent visit. It showed us two different ways of dealing with the issue. I was keen to find out why the drinking levels in France had been falling, and it was interesting to see that that decline in drinking levels was masking a rise in binge drinking and the drinking of spirits rather than wine. We also saw that the low level of taxation on wine did not encourage people to drink wine and that the high level of taxation on spirits did not discourage people from drinking spirits. That struck me as strange. People told us that the issue was to do with a merging of cultures across the globe and that we in northern Europe were starting to drink more wine while people in France were drinking less wine and increasing their consumption of spirits to match that in northern Europe. The visit taught me more about culture than pricing.
I want to talk about both.
I point out that, in 10 years of the Parliament, I have had a trip round the care homes in the Western Isles and an overnight trip to London with Helen Eadie to look at commercial health care providers but the visit was my first out-of-the-country trip. I appreciate that resources are scarce, but the trip was enormously helpful for us. I put that point on record because I know that there is a lot of apprehension about asking MSPs to make such trips, but I have come back enormously better informed.
I do not completely agree with that last point.
First, I welcome Dr Petra Meier and thank her for coming here. I know how busy university life can be these days. Her team’s literature review has been enormously helpful in allowing me to formulate my views on the subject. The review was done for the Department of Health, which wanted the team to look at the particular issue. I have some questions by way of an introduction to the subject.
We sometimes hear about the three big As—affordability, availability and advertising—as the three major things that may affect culture. If alcohol is freely available, is quite cheap and is advertised widely it is imaginable and plausible that that shapes cultural attitudes in a major way.
That was the Cook study on an aboriginal community in Australia.
We have just heard that that grant application succeeded.
One of our concerns is that we have heard that the consumption of spirits is rising in Canada. That is an interesting phenomenon for the approach that is the closest to our minimum unit pricing.
That is because minimum pricing is targeted at the off-trade—at cheap supermarket prices—whereas a major share of what 18 to 24-year-olds drink is drunk in the on-trade, so they would be less affected by minimum pricing than some other groups would be .
So that figure is correct.
As ex-students, we probably realise that that is the case. The trend continues to pensioners, among whom consumption is even lower. That is the trend over time, but the volume and extent of drinking by 18 to 24-year-olds concern us.
I have a supplementary question to Richard Simpson’s point about binge drinking—I apologise for interrupting.
I read from your report.
I will explain. Econometrics is part of the model. Whether econometric estimates are produced for the total population or for a sub-group is determined by how many data are available. All that that does is tell us whether moderate drinkers and harmful drinkers, for example, respond slightly differently to price increases. We found that to a degree, but not in a major way. Although we do not have separate econometric price elasticities for binge drinkers in particular, we know how different age and gender groups and moderate and harmful drinkers respond to prices and we can model that. If we know how people respond to a price increase in on-trade beer, for example, and we know that binge drinkers tend to buy a certain amount of a particular beer, we can estimate the effect on consumption.
Right, but in relation to the definitions in the household survey of 8,500 people in Scotland, are another 7 or 8 per cent of men harmful drinkers, on top of the 30 per cent of men who are hazardous drinkers, or does that 30 per cent include both groups—let us call them excessive drinkers?
I accept the categorisation. Perhaps someone else can help us with the issue. I just want to find out the total number of drinkers.
Yes, it must be, because you have had a good slice and I have a queue of members waiting to ask questions.
I know—I am sorry.
Most 18 to 24-year-olds are hazardous drinkers and there are very few harmful drinkers.
Exactly. The argument that we sometimes hear about how dependent drinkers would respond is not covered by that meta-analysis because there were no separate estimates that could be used.
So if the purpose of introducing a minimum unit price is to deal with the most harmful drinkers—the ones who go to health professionals for treatment and about whom there is concern—it will not be as effective as straight price increases would be.
That is not true according to what we found in the modelling in our econometric analysis. We found that harmful drinkers might actually respond more. We found that minimum pricing is targeted more at harmful drinkers because they select cheaper alcohol. The issue is not so much how they respond to price changes; it is that they consume more of the products that are targeted by minimum prices. Our modelling shows that the decreases in consumption among harmful drinkers would be far more than the average that you mentioned earlier.
But what I am asking is important. I have read the Sheffield report. We do have a unique relationship with alcohol. I picked out 16 examples of what you did with information on elasticities, market research, off-trade discounts and morbidity. You assumed that health conditions are the same in Scotland as in England and Wales. I could go on. You state:
Of course we would have wished to have Scottish data on everything, but we have made substantial efforts to take into account all the Scottish health data, crime data, expenditure on food data, survey data on purchasing and Scottish consumption data. Accusing us of not making the best effort to take Scottish data into consideration is not quite fair. However, we are currently updating the model. We have been commissioned by the Scottish Government to take into account even more Scottish data, as they have become available, such as the 2008 Scottish consumption data. The expenditure and food survey is going to be validated according to Scottish Nielsen data—purchasing data from Nielsen that are just on Scotland—to see whether that information holds up. We have new data on crime and health conditions—a newer period of data. We are just modelling that at the moment.
I am not accusing you of anything; I am stating a fact. I will read whatever you produce in the future, but all I have at the moment is the data that are in front of me and in front of every member of the committee—the Sheffield study.
There is a difference. For the chronic harms, we would expect a time-lagged effect—one hopes that it will occur. With crime, we would expect a fairly rapid effect if the consumption reduction occurred among the group of people who committed the most crime, that is young males. I have not looked at the figures in detail, although I can do so if you want me to.
The effects on chronic health harms tend to lag 10 years behind—the effect is long term—so we would not expect to see the effects in the shorter term. However, we might expect to see an effect on the number of accident and emergency and other hospital admissions.
And we have not seen that.
They will be answered, but they are all—
They are all on the same topic.
They are all related—the marginal propensity to consume, the cross-price elasticity of demand and the substitute effect.
Okay. I thought that you were talking about price elasticity, not cross-border trade.
Just to—
Someone like me, who is unfamiliar with public health modelling of this kind, must constantly remind themselves that, although we get down to precise figures, they relate to populations and not to individuals, which leads us to a different conclusion.
I do not know. We used a model because the policy has not been introduced, and we had to project what would happen. It is like the weather forecast; you do not evaluate it afterwards. It is a model. If the policy is introduced, we will obviously want a very strong evaluation to be carried out—the whole world, not just Scotland, will be interested in that. Scotland is currently the focus of the international community with regard to minimum pricing precisely because such a policy has not been attempted before, and people want to see what happens if it goes ahead.
Like Ross Finnie, I am not familiar with the modelling approaches that are used in public health. I am interested in whether Dr Meier’s approach is commonly used when people are trying to understand the impact of public health policies on the population. That will help me to understand whether the report stands out in its approach or whether a common approach has been used.
Before you respond, can you advise us when the supplementary report will be available?
Those comments are helpful.
So the committee is at the point of being able to say clearly that your approach has been based solely on modelling and not on evidence. My closest parallel—
It is based on evidence—it is not an evaluation.
It is not based on evaluation of any evidence of any other practice elsewhere in the world.
It is not correct to say that it is not based on evidence of any other practice. It is not based on evaluation of minimum pricing elsewhere.
That was specifically excluded from what we were tasked to do.
Can you comment on European work on this issue? For example, a report on alcohol that was produced last October by the European Economic and Social Committee and a report by Peter Anderson, who was to be a witness this morning, took a public health perspective on alcohol in Europe and both recommended a much more holistic approach to alcohol issues. What are your views on that?
The European Economic and Social Committee’s verdict is that pricing is important. However, it is talking about pricing in general, as distinct from minimum unit pricing, with the implication that it is the responsibility of the state to recover any duties. The majority of that committee, which has 129 members—by coincidence, the same number of members as the Scottish Parliament—concluded that pricing, not minimum unit pricing, was the issue, with only five members taking a minority view.
I do not see that the issues are different. The committee highlighted pricing to allow member states to decide the pricing mechanisms that work best for them. For example, we have quite a complicated tax system that treats different products differently. Other countries are starting to think about introducing a tax based on alcohol strength, and there is the option of a more targeted approach, such as minimum pricing, which affects only the cheapest part of the market.
No. Peter Anderson, who would have been here, is the person for European comparisons. We are not particularly interested in that issue.
Members who wish to ask Peter Anderson anything should tell the clerks, who will put the questions into a letter and seek his response.
We would love to do that work, but we were not commissioned to look at different income groups for this report.
My assertion is that to someone who is on a low income, an increase of £4 or £5 a week—
The estimated increase is £11 per year, which is less than £1 per month, or 25p a week.
Does the moderate drinker category include people who do not drink regularly? My back-of-an-envelope calculations were based on the assumption that a moderate drinker would drink within the recognised levels.
We are talking about the average moderate drinker. That does not include people who do not drink, but does include a range of people who drink, from those who do not drink regularly to those who drink right up to the limit.
You could work that out, but we have not looked at that. For us, a moderate drinker is anyone who drinks below the threshold. The large majority of moderate drinkers do not all drink 21 or 14 units a week; there is a spread.
A bottle of vodka and a bottle of wine, or something like that.
It would depend on what the person drank. If their 21 units were made up of whisky, the minimum price would not change anything. If their 21 units were made up of cheap cider, the effect would be more noticeable. The extent of the increase would depend very much on what we call the basket of goods that makes up a person’s average consumption.
At the time, there were various opinion polls on minimum pricing. There was, for example, the north-west drink debate, or the big drink debate, as it was called. High levels of support for minimum pricing were found. Support for minimum pricing was higher than that for taxation, for example.
I want to pursue the income group issue, because that is fundamental to one of my objections to minimum unit pricing. Seventy per cent of people are moderate drinkers or abstainers. Roughly 10 per cent are abstainers, so minimum unit pricing or taxation would have an effect on 60 per cent of the population. However, let us consider a tax being put on safe, moderate drinkers of modest means who buy less expensive alcohol because of the constraints on their means. Let us consider a couple who buy cheaper alcohol—my colleague Rhoda Grant gave the example of spirits and wine—and spend £10 on it every 10 days. A pensioner couple has come to me to complain about the potential policy. With minimum unit pricing, the price of their alcohol would be £14.15—the own-brand vodka that they currently buy for £6.95 would go up to £10.50 and the wine that they buy for £3.05 would go up to £3.65. They spend roughly £10 on alcohol every 10 days, which puts them in the upper bracket of moderate, safe drinkers, but nevertheless within the confines that the health lobby tells us is appropriate. Under the minimum pricing mechanism, they would be taxed at 41.5 per cent. If VAT went up by 10 per cent, which would produce as great a response as minimum unit pricing, according to your studies, those people would pay an extra £1. In my view, we will all have to pay a price to deal with the alcohol problem, but I am radically opposed to any system that attacks people who are moderate drinkers and of modest means. Such people make up a substantial proportion of the population. If the proposal had a massive effect on the hazardous drinking group, we might still have to consider it, but that group will be affected in a similar way, in that those on low incomes will be affected but those who have higher incomes will not. Although the problem is skewed in relation to deprivation and the skew has got worse, it is not sufficient to lead us to implement a new policy that has no evidence base, apart from the matters that we have discussed, and which is based almost exclusively on a modelling study. It will have a serious effect on moderate drinkers with low incomes.
You have made your point, Richard.
Just a minute, Helen. Can you tell us where that is, and in which paper?
I am a bit lost. Members have not seen that paper. Is it the document “Minimum alcohol pricing: A targeted measure”?
No, I can answer it. I am very familiar with the CEBR report on our work.
The clerk will pass the report to you—that is only fair. I am sure that you are perfectly able to comment, but I want to ensure that you have got it, because none of the rest of us know what this document is.
CEBR took two out of several hundred elasticities from our report and said, “Oh, look. This shows that hazardous and harmful drinkers are less price sensitive compared to moderate drinkers.” Those were overall aggregate elasticities, not the actual ones that we present later, which the CEBR report ignores, and which include cross-prices: how people shift between different products if faced with a price change. We know that harmful drinkers are more likely to shift when confronted with price changes, so that makes a difference to the elasticities.
That study is from 2004 and the Wagenaar study is from 2008. It is the same kind of study, but Wagenaar is slightly more up to date. I have seen studies that suggest that dependent drinkers are price sensitive because they tend not to have much money, but I have seen other studies that suggest that binge drinkers are not very price sensitive. We can only work from the data we have about how it works in the UK.
Do you accept that Ludbrook is a valued study and that it has integrity equal to your own studies?
Yes.
I was going to say how inappropriate that would be.
Do not look at me when you are saying that, Mary.
That confirms the earlier point on elasticities. Price changes are not so severe for people who have high incomes.
Not quite.
So, are you saying that minimum pricing would affect a managerial or professional woman in the highest income quintile who drinks more than the weekly average in the same way that it would affect someone in a deprived area who is less well off? Are you honestly saying that their response to minimum pricing will be the same?
As I said, we have not modelled it, but we know that managerial and professional women drink more.
Earlier, when I mentioned the figure of 9.3 per cent for the reduction in consumption between 2003 and 2008, you said that you could not predict health outcomes. In fact, you do predict the effect of minimum price—
It feels as if the jackets are off, Mary. I ask for courtesy please, ladies.
The point is that—
Please do not talk over each other. I have a little headache coming on.
The point is, what happened in 2005 to 2008 is very different to your predictions. Our real-life experience of a reduction in consumption and the effect on our population is the opposite to your predictions.
I am not sure that that is true. The evaluation stands as it is. You have a figure for 2003 to 2005, but you cannot simply apportion it to year-on-year consumption. For example, we know that the consumption curve went up first and then down again.
It is interesting that cancer has been mentioned. In a discussion during our visit to Paris, I learned that excessive consumption of alcohol is the second most common cause of many types of cancer. Not many people are aware of that, but it ties in with the point about the long-term effects.
Surely not. You are not known for that.
I have a simple question, because I am getting confused by the various figures. How many Scots adult men are hazardous and dangerous drinkers? What is the percentage when those two figures are added together? What proportion of Scottish women are in that category? I am confused about whether the 30 per cent figures includes all those drinkers or only some of them. Can you spell that out for me?
I do not have the male versus female split with me, but I can give the total numbers of the population who come into those categories. There are 2.4 million moderate drinkers. There are about 950,000 hazardous drinkers and about 273,000 harmful drinkers. Those are our baseline figures before modelling the effects of any policy changes.
I do not have those figures to hand, but I can provide them.
That would be useful.
I am not familiar with the figures, the results or the details of who drinks vodka and how they are affected by its price. I cannot really comment further.
We cannot expect you to.
It might be worth putting that point to Peter Anderson.
Perhaps it would be better if Dr Simpson were to repeat the question. He posited a 10 per cent increase in tax.