Item 2 is the first of two additional oral evidence sessions that the committee agreed to undertake in advance of consideration of amendments to the Alcohol etc (Scotland) Bill at stage 2. Today’s session has been arranged in the light of written evidence received by the committee from Scottish health action on alcohol problems on the purchasing patterns for low-price off-sales alcohol in Scotland. Tomorrow morning we will continue with a further two panels of witnesses, who will address issues including the sale of caffeinated alcoholic products in Scotland.
I thank Professor Ludbrook for coming. I read her paper with considerable interest. Was it commissioned by SHAAP?
Yes.
Is it correct that the paper’s main conclusion is that low-cost alcohol is purchased by every group and therefore minimum unit pricing is likely to affect every group?
Yes, that is correct.
Do you agree that the proportion of cheap alcohol, at least at 40p a unit or less—a new minimum price of 45p has been announced, but let us stick with 40p, because your paper uses the range of 40p to 50p and does not have a 45p break point—in the basket of alcohol purchases of people with lower incomes is greater than it is for those with higher incomes? I am not talking about the amounts in each, which appear to be broadly similar, although there is some differentiation. I am talking about the fact that the proportion of alcohol in that price range that the lowest three deciles purchase appears to be greater as part of the total consumption.
Of that decile.
Of the lowest three deciles.
Part of the problem is that if you look only at purchases, the proportion that are made at a lower price for those deciles is higher, but if you look at the whole group—those who do and those who do not buy alcohol—there is a much lower impact on those three deciles, because many of them do not buy any alcohol at all. It is a question of whether you want to look at the effect on the whole group or on only those who purchase alcohol.
At the moment, I am concerned only with people who drink. Those who choose not to drink choose not to drink and that is absolutely fine. We are not about abstinence or being teetotal; we are about trying to affect the consumption of drink. I calculate that a moderate-drinking couple who consume between them—this is like one of those maths questions—one 70cl bottle of own-brand vodka, which is 26 units, per week, will now have to pay, with a minimum price of 45p, £200 extra a year for the same moderate consumption of alcohol. I cannot see any flaw in that mathematical/statistical statement. If that is the case, what is more regressive: an increase in duty on all alcohol or a minimum unit price, which appears to me to represent, for a poor pensioner couple who drink moderately, an additional tax of £200, which is a gross imposition?
I do not have the benefit of having done the calculation but, if you were considering a tax increase that would have the same effect as minimum pricing—one that raised the price to the same sort of level—you would be talking about similar impacts on the moderate couple who buy a bottle of vodka.
There is an argument for an increase in price. It is clear from all the graphs that we have been shown that the overall relative price of alcohol against income dropped over the years until the past three or four years, when it stabilised. However, to place a tax—a duty—on all alcohol so that the price goes up for everybody, no matter what their income is and what sort of alcohol they choose to buy, seems to me to be a lot less regressive.
I am not sure about that last point. I think that I saw some figures that suggested that the proportion of harmful drinkers was highest in the lowest-income group.
No, it was deaths that were highest.
I think that the figures showed that the lowest-income group had the most non-drinkers and the most hazardous and harmful drinkers, but I would have to check.
The fundamental question is which is more regressive: a minimum unit price or a general increase in taxation?
I have not considered those calculations, because no such tax increase is before the Parliament. It depends on which aspects you think will be affected.
My question is directly related to Richard Simpson’s. I ask him to repeat his alcoholic drink and its price.
The current price of a bottle of standard own-brand vodka, which is 26 units, is £7.95 but, at a minimum unit price of 45p, it would be £11.80, therefore the increase would be £3.80 per week, which is almost £200 a year.
Right. My question, Ms Ludbrook, is this—you will wish that you had brought a calculator with you.
Or a blackboard.
If we take the bottle of vodka that has gone up from £7.95 to £11.80 following the imposition of a minimum unit price of 45p, the £3.80 increase is borne by everybody. It does not increase if the vodka is purchased by Richard Simpson, but undoubtedly the lower income group has to bear that £3.80 increase.
The main difference that would arise is that the increase in tax would be paid on every product. For example, you would pay an increase in tax on a bottle of standard vodka that currently sells at a much higher price: it might be selling at £13, and you would pay £16.80 for it.
I want to expand on something that Richard Simpson said. I will quote from page 2 of your submission, Professor Ludbrook, which states:
I do not think that we have the full calculations to support such an analogy. My paper sought to demonstrate that all income groups will be affected by the imposition of minimum pricing, which did not appear to have been considered in evidence up to that point. Undoubtedly, the more income you have, the more you will be able to afford and absorb any price increase, but it is not clear whether the measure will have a disproportionate effect. There will be a price increase and therefore a price effect for people on higher incomes—unless, of course, the rules of economics do not apply to them—and the measure will reduce the amount of alcohol consumed across the market.
You say that a full analysis has not been carried out, but this histogram sets out the figures. The number of people in income decile 3 purchasing alcohol at less than 30p per unit is more than twice the number in income deciles 9 and 10. According to your figures, there is no doubt that there will be a change.
I would not interpret the analysis in that way; instead, I think that it shows that there is some effect on all income groups.
We now know what the minimum price will be. Your figures in figure 5 set out the percentage in each income decile likely to purchase alcohol at less than 30p per unit. When that unit price becomes 45p, the people on income decile 3 will at the very least be faced with a 50 per cent price increase. On the other hand, in income deciles 9 and 10, far less than half the number in income decile 3 will purchase minimal amounts of drink at 30p per unit. Surely, given that those people purchase very small amounts of drink at less than 30p per unit and that the majority of their purchases are in the yellow band—that is, between 40p and 50p per unit—we can conclude that when the price is raised to 45p per unit there will be minimal, if any, impact on them.
The impact will not be minimal. You are referring to the percentages in figure 5, but I point out that the people in the higher income deciles actually purchase quite a lot, because they are more likely to purchase in total. You have to look at the total amounts that are purchased as well as those percentages.
By far the majority of their purchases are already in the range of 40p to 50p per unit.
On these particular figures, yes, I agree, but that does not mean that they are not bearing part of the burden.
I said that the impact would be minimal, compared with the enormous effect on low-income earners.
I direct you to figure 1, which shows the purchasing patterns in terms of the total number of units for each decile. In decile 10—the very highest income group—we are seeing much less purchasing of units at 30p and 40p. If you take the 30p and 40p bands together, most of the middle and higher-income groups buy more units at those prices than do the lower income deciles, so they will pay those price increases on more units.
Yes, and the majority of their purchases are at more than 40p.
The majority of them, yes.
They buy more alcohol.
They buy more alcohol. They have more money, so they buy more alcohol.
Across the range?
Across the range.
To get away from deciles, your point is that they buy more alcohol than do lower income groups, across the whole range. Ian, do you have a supplementary on this, too?
It is all on the same sort of point really, is it not? On figure 1, please correct me if I am wrong, but it seems to me that if you offset the rise in alcohol price by the number of people in each decile who drink, they almost come down to about the same level. Is that right?
A lot of the explanation is to do with the fact that there are households who do not drink at all. Whether that compensates completely—
I would have thought that it compensates almost completely—I am doing a quick measurement with my thumb.
I have seen only the figures in other papers that are before the committee. I think that something like 9 per cent were very heavy drinkers, who drink more than the equivalent categories in other income deciles.
Much more, yes. There is therefore more ill health in that group. So, if you rule out the 67 per cent who do not drink at all and you rule out the percentage who drink extraordinarily heavily, the number of people who actually fall into Dr Simpson’s category of people who drink a bottle of vodka between two of them a week is probably quite small. Is that right?
Yes, it is.
I have a couple of quick supplementaries. Are the tables that have been provided based on the ordinary price of alcohol or do they include alcohol that was discounted or special-purchase alcohol?
They would have included alcohol that was sold in any form. The data are from the expenditure and food survey, which collects information on what households spend and what households purchase. Therefore, the only calculation that we can do is simply to divide one by the other to get the price. Some of the alcohol will be products that are sold at a low price all the time and some of it will have been discounted, but we cannot tell which is which.
Which groups are more likely to be price inelastic? I assume that, if the price was raised, the higher income groups would either pay the extra money because they can afford to or move on to a different kind of alcohol. Is that your assumption?
I have not seen price inelasticities calculated separately for the different income groups so I cannot comment on whether there would be a range. However, we also have an income elasticity, which tends to show that people buy more as their income increases. Those are the two ways in which these things are usually looked at. In a sense, the income effect is felt through the higher level of purchasing, but if the relative price changes for every group, we would expect that to change people’s consumption patterns.
Okay, but it could be that they buy something different.
It could be a smaller effect.
I have one last supplementary, just for information. If people in higher income groups tend to drink more hazardously, has any research been done to determine how many of them subsequently fall into the lower income groups that Ian McKee talked about? If they develop a drink problem while they are higher earners and can afford to drink more, they might fall into the lower groups later because of their drink problem.
I am not aware of any such research. That is not in my area. However, it is certainly a plausible explanation of why there are hazardous drinkers in the low income categories.
I have a supplementary question. When we discussed figure 1, you agreed that, if we factor in the number of people in each decile who drink—it is 70 per cent in the upper group and 33 per cent in the lower group—that flattens out the curve a lot. Does the average person in the upper decile drink more than the average person in the lower decile, if there are more of them? Do you follow?
There is still some gradient when we examine income in relation to alcohol consumption. It is not in those figures, but the general household survey collects data on income and alcohol consumption, and it shows that there is still a gradient when we take into account the whole population.
If we take the average drinker as opposed to people who do not drink at all—
The figures are something like 15 units for men in the lowest income group and maybe 19 or 20 for men in the highest income group. It is actually a steeper gradient for women.
Thank you.
Do not look at me. [Laughter.]
That is terrible.
Absolutely—there will always be more than price involved, and people’s tastes and preferences come into it. These things change. The popularity of a drink might change in a way that is not just a reflection of price factors. We might consider the phenomenal growth in wine drinking in this country, in which a number of factors are at play, including people sampling things when on they are on holiday abroad. Whenever we consider a price effect, we can take it only with all other things held equal. The problem with the real world is that all other things are not held equal. Inevitably, whatever happens on price, other things that are changing in society at the same time can have a reinforcing effect or an opposite effect.
That brings into question what happens in England, where the pricing structures are the same and there is a similar culture to ours. People there seem to consume quite a lot less alcohol.
Absolutely, but there must be something different between the culture of England and that of Scotland, such that the levels of consumption are different.
For the sake of continuity, I will continue talking about England. I read the two papers that are before us today, and I did not see anything on cross-border sales or internet sales. My understanding is that internet sales comprise the fastest-rising area of alcohol sales. Did you consider that at any point in your research? If a minimum price is imposed, will people be more likely to buy from England, Northern Ireland or the internet? The same thing happened in Finland, where people simply went on the booze cruise to Estonia.
I have not been able to look into that. The data do not include those identified sources of purchases.
Did you consider the substitute effect for younger people should alcohol become more expensive? The point has been raised with me by people in some parts of the Highlands—if alcohol becomes more expensive and difficult to buy, there might be a substitute effect through illegal drugs. Did that come into your research?
I considered that point recently, because I was asked to comment on a paper on the subject, which was published recently in Alcohol and Alcoholism. Considering the evidence in that paper, it seems that illegal substances are more likely to be a complement to alcohol consumption than a substitute.
Retailers, and indeed producers, are likely to increase their profits significantly as a result of the minimum price. It has been mentioned—or alleged—that, given those increased profits, retailers might be able to afford to reduce the price of alcoholic drinks that are over 45p per unit. Although the minimum price might come up to 45p a unit, producers could afford, thanks to the increased profit, to reduce the prices of some of the more expensive alcoholic drinks, and that might make alcohol even more attractive to higher-income earners and could therefore increase consumption. We know that a decrease in price increases consumption, although there is no evidence about an increase in price.
Yes, there is great uncertainty over what would happen to the increased revenue that would go into the supply chain, and how much of that would stay with the retailers and how much would go to producers. Equally, I have heard arguments that, if the price of lower-priced products goes up, producers might wish to raise the price of other products to maintain differentials. So, it could go in either direction. Given that there is an element of below-cost selling, supermarkets might transfer that marketing, as we might call it, to other products by reducing prices elsewhere in their basket of products.
There are no predictions, forecasting or modelling on that issue.
It depends entirely on the commercial decisions of the actors in the supply chain.
Mary Scanlon asked about the risk of displacement from alcohol to drugs if minimum unit pricing came in. Would there be a risk of displacement from more expensive alcohol to cheaper alcohol if there was a large increase in tax, because that would put up the prices of all goods? Therefore, if the Famous Grouse went up by £3, people might move down to supermarket alcohol and not decrease their intake. Is that a possible risk?
That certainly comes through in the modelling work that was done by the University of Sheffield and in a Swedish study. It was found that if prices are increased across the board, people trade down to cheaper products and therefore the effect of an across-the-board price increase is rather less than one that is targeted at cheaper products.
I will continue the vein of questioning that Mary Scanlon started. You have rightly pointed out the disadvantage of minimum unit pricing as a windfall to the industry, mainly retailers. That is extremely unlikely to be offset to any great extent by a levy—certainly not the full £140 million. You have speculated as to how that largesse would be returned to the wider community. What will happen to consumption if the bulk of the new money goes on advertising and reducing the standard price of alcohol to just above 45p?
I have not done that calculation, but it would be a concern and would tend to undermine the intended effect of reducing alcohol consumption. That could only be addressed over time—I cannot predict which way it would go. The House of Commons Health Committee, in considering the topic, suggested that, if minimum pricing were accompanied by increased taxation, the revenue effect for the retail and producer sector would be clawed back to an extent.
So you suggest that Her Majesty’s Government should work in tandem with the bill to claw back the profits from the retail industry.
That is one way of addressing concerns about that element.
Potentially, it is a real flaw in the bill, as we have no control over that. We anticipate that a massive profit of £140 million might be made.
On the other hand, there is a concern that a large change in alcohol sales would have an impact on producers and the alcohol industry. They would have to adjust to that. An element of the increased revenue from minimum pricing might offset the reduced sales and therefore an impact on employment would be less likely. So, there are benefits to having increased revenue going to the industry during the adjustment process to a lower alcohol consumption society.
But we know from the questioning that Rhoda Grant led that there will not necessarily be a reduction in consumption. That has not happened in France. I suppose that we will just have to let that one stick to the wall.
Before we move on, I have a question. I have never been able to square the circle around why the supermarkets are so opposed to minimum unit pricing if they are going to get a bonus out of it and make bigger profits. If I was getting more profit, I would support the measure.
Some of the supermarkets, such as Tesco, do not oppose it.
If retailers will make more money, I do not understand why they do not support minimum unit pricing.
As I said, some retailers have indicated support for it. It is also important to say that the model shows that they would increase their revenue but not necessarily their profits. The supermarkets are highly competitive and—this is only supposition—it is far more likely that increased revenue from minimum pricing for alcohol would be used to market other products through loss leaders. We might hope that they would be on staples such as baked beans and bread, but they might not be.
Convener, may I ask a supplementary question?
I was going to let Helen Eadie in.
Carry on; I am happy to be patient.
I will let you chair the meeting.
I want to ask something about one of Professor Ludbrook’s answers.
The committee was talking about the supply chain as a whole. We do not know whether the revenue would lie with the retailers or the producers, but if duty is raised, the proportion of any revenue that goes to the Treasury as opposed to the supply sector would shift towards the Treasury.
Convener, I apologise if I tried to take over convening the meeting. I was just trying to be helpful.
The most recent figures mirror the stabilisation or drop in consumption that has been seen as a result, it would seem, of the economic conditions. For once, alcohol has become less affordable simply because of the impact of the recession on people’s incomes.
Would not the data have been collected in 2007-08 rather than in 2009? Of course, they were only reported in 2009, which was really before the recession hit us so badly.
Certainly, the drop in consumption is most noticeable in 2009, but there was a levelling off before that.
I am sorry, but I want to go back. I have been mulling over Richard Simpson’s bottle of alcohol.
I think that the answer would be yes. The products that would be most affected by such a reform would be drinks such as cider, which has a particularly low price and tax base at the moment. It would also increase the duty on beer and lager. It would also increase the duty on wine, which might be less regressive, but the lowest prices and cheapest products would be most affected by such a reform in duty.
That is helpful. We can construct a fairly simple arithmetic table that demonstrates the regressive nature of VAT. It is not desperately affected by an individual’s purchasing pattern, largely because certain goods that would have a disproportionate effect on low-income groups are exempt from it. However, in the case of alcohol, the nature of the regression is accentuated, I think, by virtue of low-income individuals’ preference for drinks at a lower price—below 30p per unit—and a higher alcohol content. Will you comment on that?
I would have to go away and do the calculation. It is much more complicated because of the amount and distribution of purchasing. If we examine the VAT exemption on food, we find that the patterns of expenditure across different income groups are perhaps not so different.
We have a little time in hand. Do you want to comment on any issues that we have not asked about but which we should have asked about?
Gosh—that is an open invitation. I am committed to the use of evidence so, having followed the committee’s discussions, my only slight concern is that the value of the modelling work that has been done is sometimes misunderstood. When Professor Beath gave evidence, he sought to allay the committee’s concerns about that work.
Thank you for your evidence. That concludes today’s business.
Previous
Subordinate Legislation