Health, Social Care and Sport Committee

Meeting date: Tuesday, November 7, 2023

Official Report 640KB pdf

---

Alcohol (Minimum Pricing) (Scotland) Act 2012 (Post-legislative Scrutiny)

Agenda item 3 is an evidence session with Public Health Scotland as part of our post-legislative scrutiny of the Alcohol (Minimum Pricing) (Scotland) Act 2012. I welcome Clare Beeston, public health intelligence principal; George Dodds, chief officer; Lucie Giles, public health intelligence principal; and Tara Shivaji, consultant in public health, all from Public Health Scotland.

Before I open up the meeting for questions, I invite the panel to give an overview of Public Health Scotland’s evaluation of the minimum unit pricing of alcohol.

Thank you for the opportunity to attend the committee. The focus of Public Health Scotland’s work has been to reduce the preventable harm that is caused by alcohol consumption in Scotland. We know that people in our poorest communities are five times more likely to die from alcohol-related disease than those in the wealthiest communities. If alcohol consumption trends continue, we expect life expectancy to fall and the cost of providing additional health and care services to increase by an estimated 3 per cent over the next couple of decades.

One approach that is encouraged by the World Health Organization is to reduce demand for alcohol by using pricing mechanisms. That is what Scotland has done, and the World Health Organization continues to advocate that as part of a range of measures to address the harm that is caused by alcohol. My colleague Dr Shivaji will briefly explain the convention that we have used in public health evaluation, and we will then hear from Lucie Giles about the key findings in our report.

I will briefly touch on the rationale for the methods used in our evaluation. Public Health Scotland follows the WHO’s recommended best ways to prevent alcohol-related harm. We recognise that policies should be of a multi-component nature and that those components are interdependent and act synergistically. Our evaluation took an approach that was quite cutting edge for public health research. We asked a number of questions not only about whether the policy works, but about the context. We also looked for any unintended consequences and tried to set out the strengths and limitations of the policy in the context of a multi-component approach.

MUP was implemented in Scotland on 1 May 2018 and it set a minimum price of 50p per unit of alcohol. In order for MUP to have the desired impact on alcohol-related health and social harms, a plausible chain of events—which we would call the theory of change—needed to be realised. The theory of change is, in itself, based on the evidence that was available prior to the evaluation and it is endorsed through consultation with a range of stakeholders and experts. That theory-based approach is recommended by the Medical Research Council for use in evaluating the effectiveness of complex policy interventions such as MUP. PHS believes that the evidence shows that, through that plausible chain of events, MUP has had a positive impact on alcohol-related harms, as I will now outline.

The evaluation showed that MUP was well complied with by retailers. There were some infrequent, isolated instances of non-compliance, but they were not typical and were generally associated with what we might call teething problems. Broadly, retailers have found the legislation easy to follow and apply, hence it being so well complied with.

Sales of alcohol priced at below 50p per unit virtually disappeared, compared with 40 to 50 per cent being sold at below 50p per unit prior to the implementation of the policy. The average price of alcohol went through an immediate and sustained increase of around 5p per unit, with products such as strong ciders and own-brand spirits, which were more likely to have been sold at below 50p per unit prior to the policy being implemented, increasing in price more. Sales of alcohol per adult decreased by an estimated 3 per cent. That reduction was entirely driven by sales through the off-trade, and particularly by reductions in the same products in which we saw the greatest increases in price. We found no impact on sales per adult through the on-trade.

Alternative data sources show that household purchasing of alcohol reduced, with the biggest reductions being estimated for those households that bought the most. One study showed that households in the top 5 per cent for volume purchased reduced their alcohol purchasing by nearly 15 per cent, while those in the lowest 70 per cent for volume purchased did not change their purchasing at all. Self-report survey data presents a bit of a mixed picture on the impact of consumption among different groups, but in general there is a consistent reduction among the heaviest drinkers.

MUP has been estimated to reduce alcohol-specific death rates by 13 per cent—about 150 deaths a year—compared with what we would have expected to happen had MUP not been implemented. A smaller estimated reduction in alcohol-attributable hospital admissions of around 4 per cent, equating to around 400 admissions per year, was also found. The largest reductions were for chronic conditions such as alcoholic liver disease among males and those living in the most deprived areas of Scotland. Conditions such as alcoholic liver disease are experienced only by people who drink at levels that are sufficient to do themselves harm.

I hope that you will see that the findings follow a logical sequence of events, with one preceding the other. Many of the findings were consistent across a number of different studies and data sources and, importantly, they were specific to the timing of MUP. PHS is therefore confident that the evaluation provides robust evidence that, overall, MUP has had a positive impact on population-level health outcomes and alcohol-related health inequalities.

The findings are all from large quantitative studies that use routinely collected data and statistical methods to analyse the data in order to understand and isolate the impact of MUP at population level. However, as Tara Shivaji said, it was also important to understand people’s experiences of MUP and the strategies that they might adopt to account for it. We did that by including in the evaluation a range of different qualitative studies whereby individuals were asked about their experiences. From those qualitative studies, we found some evidence of individuals, as a consequence of MUP, possibly engaging in potentially harmful strategies such as reducing their spend on food and, for those who already used drugs, increasing drug use. Although those findings are obviously important and it is really important to minimise harm for individuals, it is also important not to lose sight of the main findings that there was no evidence of widespread social harm and that we have seen population-level improvements in relation to consumption levels and alcohol-related harms and inequalities.

I will draw our comments on our findings to a close. What we have illustrated is that it is a complex and nuanced picture. There is a really strong need for a range of interventions, particularly to support people with established dependence and those who are affected by alcohol dependence, as well as other measures to address youth usage. We estimate that, in the absence of the implementation of minimum unit pricing as a policy, the number of people who would have died as a direct result of alcohol would have been higher and the inequalities in health due to alcohol would have been wider. On that basis, we recommend the policy as an effective component in addressing the levels of harms that we see in Scotland.

Thank you for that overview. We move to questions from committee members, starting with Ivan McKee.

Thank you for coming along and talking us through the approach to the data. There are a couple of things that I want to dig into a wee bit deeper, if that is okay. First, just to get one thing out of the way, when you talk about a 3 per cent reduction in sales, is that by value or volume?

It is by volume of pure alcohol.

Got it—thanks. Some modelling work was done in advance of the implementation. I think that it was in the University of Sheffield study of 2016. It would be useful if you could refer back to that, say what was fundamentally different there and say what you believe the outcomes were that related to that. I am also interested in digging into the theory of change. There was a 13 per cent reduction in deaths caused directly by alcohol, which you compared, in effect, against the counterfactual—what you think would have happened otherwise. Is that trend going upwards or downwards? What has been the actual difference in alcohol-related deaths during the period since MUP was implemented? Can you give us a number and say whether it is higher or lower than previously?

I do not know the exact numbers. Alcohol-specific deaths went up during 2020, 2021 and most recently 2022, but that was following quite a large dip in 2019 after MUP was implemented.

What is the difference between the latest year and the year prior to MUP being implemented?

I do not know what the actual, absolute difference is, but the number is higher than it was prior to MUP being implemented.

The number of alcohol-related deaths is higher now than it was prior to the implementation of MUP?

It is.

Just to be clear, the theory of change is telling you that the number was increasing anyway and that it would have increased by more had MUP not been introduced. Is that right?

That is essentially what we are saying. If we had not seen that dip in 2019, we could assume that we would be at a higher number now, had MUP not been implemented.

Okay. The other thing that I want to explore a wee bit more is where MUP has had an impact. I would certainly expect that there would be some impact on heavy drinkers and those who are, unfortunately, at the stage where they are quite likely to succumb to alcohol-related deaths. However, I assume that a big part of the policy objective is to address people who are starting to drink by reducing their access to cheap alcohol—young people, perhaps. Is there any evidence about that? The sad reality is that price will be an issue for somebody who is going to drink excessively but, frankly, it is an issue that they will deal with if they need to get a drink, whereas a big part of the impact will be on early-stage drinkers. Is there any data to support any behaviour change at that stage?

I ask Clare Beeston to comment on that.

We did a qualitative study with young people under the age of 18 who are already drinkers—it was not a quantitative study on general impact—and we heard their stories about how MUP had impacted on them. It is fair to say that there was not much evidence that price was a big factor in what they choose to drink. That was driven more by peers, trends and so on. There was no evidence of MUP having a substantial impact in changing what young people who are under 18 are drinking. They are people who are starting out on their early drinking careers, but we do not have evidence on how their habits have changed. We would need to look at that over the longer term, so it is not a question that we can answer at this stage.

Do you have any comments on how the data that you arrived at in the evaluation compares with the data that was on the table prior to implementation of MUP?

Do you mean the data from the modelling?

Yes.

I would caution against comparing what we have produced with the data from the modelling, as they are different things. The modelling was there to illustrate how different types of policies or levels would alter outcomes such as consumption and harm. That said, however, I note that the modelling and what we have produced on consumption are quite similar. The 2016 modelling suggested a reduction of 3.5 per cent, which is fairly comparable with what we found.

It is also a wee bit difficult to compare the findings in relation to harms because they were typically looking 20 years ahead. If I recall, they estimated some 2,000 deaths over 20 years. I do not want to assume that the impact of MUP will continue at the current level over time, but you can see that the findings are broadly comparable in terms of that annual increase.

09:30  

Okay. Thanks very much.

I declare an interest as a practicing national health service general practitioner, and I suppose that it is important to say that I met three of the four panellists and we had a discussion about MUP last week.

Lucie, I was deeply disappointed not to hear you say in your statement that the 4 per cent reduction in hospitalisations is not statistically significant. That is quite an important statement that you left out. You went on to mention other studies, but what studies back up what you said, which was that deaths have reduced by about 150 annually and there has been a 4 per cent reduction in hospitalisations?

I did not specifically say that there were other studies looking at admissions and harm. On deaths, the study that we in Public Health Scotland performed was the only study that looked specifically at that outcome. There were lots of different studies looking at things such as purchasing data, price and consumption, and they generally showed a very consistent picture.

You are absolutely right that the overall admissions figure was not statistically significant, based on the p-value. I would now prefer to look towards confidence intervals, which give a much better picture of how likely it is that there will have been a change in one direction or the other. The confidence interval was very largely to the right of zero, indicating that it is much more likely to have been a reduction than anything else.

Did it cross zero?

It did just cross zero, yes. If we look at chronic conditions, which is where we have seen the biggest impact, that reduction was statistically significant, as was the reduction in admissions in males. They were all statistically significant, and the figures on three of the four most deprived areas of Scotland were also statistically significant, if that is the most important thing.

Accuracy is.

We will stay with you, Dr Gulhane, for your further questions.

Thank you, convener. Tara, what is your definition of a dependent drinker?

A dependent drinker would be somebody who has a physiological and psychological dependence on alcohol. It has less to do with the volumes that people are consuming. You would expect a dependent drinker to be consuming a higher volume of alcohol, but when that stops, they will experience withdrawal symptoms as a result.

Looking at the bill and the modelling, what would you have expected and what did you actually see when it came to the spend of dependent drinkers?

The patterns of spend in people who are the heaviest drinkers tend to involve the lowest and strongest alcohols, and their purchases tend to be from off-sales. We would therefore have expected minimum unit pricing to have had a particular impact on purchasing decisions in that group. However, we see in the broader context that dependence is quite a complex phenomenon, and we saw reports of individuals prioritising their spend on alcohol over other commodities where household budgets were finite.

They spent less money on feeding their kids and things like that. What mitigations were put in place to help dependent drinkers, who we knew would be spending more money on alcohol?

I guess that that is outwith the scope of the policy evaluation, in some ways. A wider range of interventions are necessary, including support for dependent drinkers and recognition of the impact on their families and those around them. Those are key interventions that we would want to see.

There is a concept called the prevention paradox, whereby those who are particularly affected by a policy may not benefit from it the most, while we expect those in the middle to benefit substantially. There is therefore a need to have targeted measures that are focused on those at the very highest risk, those people being dependent drinkers and those who are affected by alcohol dependence.

Given that you are a consultant in public health, and given your specialisation, can you say whether there has been a decrease in alcohol brief interventions when it comes to referral rates for people seeking help? Surely something should have been put in place to help those people if a policy was being put in place that, as you have just said, would possibly affect those who most need help. Perhaps some extra money should have been put in to help those who needed it the most.

Public Health Scotland produces indicators and statistics on the numbers of alcohol brief interventions that are delivered. They are short conversations that are designed to change people’s consumption patterns. They are not actually aimed at people with dependence. We also produce statistics on referrals to treatment.

Looking back over the past 10 years and thinking particularly about the brief interventions programme, there was a standing start of zero, after which there was a substantial increase in the numbers of recorded and reported brief interventions that have been delivered across the healthcare sector, although that has started to tail off over the past five years. Similarly, there has been a decline in the number of referrals to treatment. We do not yet have explanations for that, or an understanding of what is driving that decline in referrals to specialist treatment. We are undertaking a piece of work on that at the moment. Given the work that has been done in England, where there was a similar problem, we would expect the explanation to be multifactorial in nature.

From what you have said, it seems to me that we have sort of abandoned our dependent drinkers, but thank you for your answers.

Good morning to panel members.

In relation to the consumption of alcohol, the different types of drinkers and the Sheffield model, can you expand on how minimum unit pricing has affected the different types of drinkers and their consumption?

I ask Clare Beeston to pick that one up. A few studies have examined different types of groups and have found a bit of a mixed picture. We should bear it in mind that that comes from self-report survey data, which has limitations. Do you recall the harmful drinker study and specifically what it said, Clare?

The harmful drinker study had three components. One involved drinkers recruited through treatment services, who were screened to have probable or likely alcohol dependence. There was no consistent evidence of a reduction in their consumption. There was a mix: some people said that they had reduced consumption, and some said that they had not. There is no consistent evidence on people with probable alcohol dependence recruited through services.

Regarding people recruited through the community, there was, again, no consistent evidence one way or the other. Some people said that it was a qualitative matter, which was not generalisable, some people said they had reduced consumption, and others said that they had not.

There was a study using a market research company called Kantar, which does what is called an alcovision survey. That surveys lots of people who drink and collects detailed evidence. It found no change in consumption. Again, that is self-reported. The survey found that the proportion of people who were drinking at a harmful level did not change, but there was a significant reduction in the proportion of people drinking at hazardous levels.

We have to be a bit careful about self-report data. It is subject to recall bias—how much people remember what they drank—particularly when you are looking retrospectively. Therefore, the sales data is a better measure of population change, but sales do not tell us who has changed what. That was the harmful drinking study.

As Lucie Giles referred to earlier, the purchasing data found that the households that purchased the most reduced their purchasing the most after MUP was introduced. Therefore, the heavier-drinking households reduced their purchasing most.

What about moderate drinkers and the overall consumption of alcohol?

One of the studies of household purchasing data to which Clare Beeston referred found that the bottom 70 per cent—the people who you would call the moderate drinkers—did not change their purchasing habits.

Good morning, everybody. I am interested in how we compare with other countries that have introduced minimum unit pricing. I know that Canada, Wales and Ireland have done so. I have in front of me a World Health Organization report that talks about how we are reducing alcohol deaths by introducing minimum unit pricing.

What work has Public Health Scotland done to look at other countries? Canada introduced MUP in 2014. Is there something that we can learn from other people?

Scotland was the first country to implement MUP that applies to all alcohol. In Canada, it has been introduced in different areas in different ways and applies to only certain types of alcohol, so it can be a wee bit difficult to draw comparisons. As yet, we do not have a lot coming out of Wales, because the Welsh were behind us, but we see a broadly similar picture there, where MUP was implemented in a similar way.

Thank you very much for the evidence. I am interested in the various income groups. Early on, there were concerns about MUP disproportionately affecting low-income groups and, on the other side, whether it would have an impact on people in more affluent areas.

I am interested to get clarity on the current pricing. For MUP to work, do we need to increase the price? Will it continue to have the same broad effects on those groups or do we have any concerns about it disproportionately affecting lower-income groups because of the other crises in income that people face?

The evidence from the evaluation shows that the increased expenditure or the price that people pay for alcohol is much more closely linked to the volume that they purchase than to income. From the data that we have, there is no systematic patterning that shows that lower-income households increased their expenditure more. The pattern is much more closely linked to how much people buy.

It is difficult to answer your question about the impact of the current cost of living crisis, Ms Mochan, but there are considerations that are worth sharing. In a situation where there is wider economic difficulty, alcohol sales, and alcohol consumption, as measured by sales, fall as alcohol becomes less affordable across the population.

09:45  

In the context of the cost of living crisis, there are important considerations about how inflation affects different commodities in different ways. Therefore, it is important to consider what that means for alcohol and how to keep the value of minimum unit pricing such that it continues to have the effects that we have seen. However, it is somewhat outwith the scope of our evaluation to touch on that.

The final thing that is worth reflecting on is the impact that the cost of living and economic crises have on people’s health and mental health. Other countries have experienced an increase in mental health difficulties, problematic substance use and problematic alcohol use. The drivers of that are loss of income, unemployment and the distress around that.

In that context, we need to think about how we protect people who are most vulnerable to those effects, including people with established dependants and young people who may be more touched by an economic crisis than others. Thinking about the issue in that broader context, there is a need for a general preventative policy that allows us to address the harms that are associated with alcohol and a need for targeted interventions that support people who are at the highest risk of harm.

Is it helpful to think and talk about minimum unit pricing as part of a package of public health measures that aim to change the direction in this country away from alcohol harm?

That is exactly the approach that we are taking.

Thank you very much.

I want to raise concerns about poly substance use—for example, using benzodiazepines and alcohol. Have you observed a substitution effect in people who have problematic substance use generally where there is a price consideration? Do they substitute with other products that are potentially more harmful?

I will start and then hand over to Tara Shivaji. We found no evidence that people who did not use illicit substances started to use illicit substances. There was evidence that some people who already used drugs substituted substances. It is probably worth saying that there were a number of that type of unintended harmful consequence, and there were exacerbations of existing tendencies. People who already took drugs perhaps took more drugs, and people who had to make decisions about their food spending had to make more decisions about that. There was evidence that people who already took drugs occasionally made different decisions in relation to taking drugs instead of alcohol.

I hand over to Tara on what that means and the wider impact.

My remit also covers drugs, and benzodiazepines are one of the key substances that contribute to the higher level of drug harms and drug deaths in Scotland. They are a very common finding as part of poly substance use among people who die and among people who experience overdose.

It is important to bear in mind that the drugs market is a global market, and there have been shifts in that market in the past five years. It has moved from diazepam and temazepam, which featured commonly in the early 2000s, through to what we call street benzodiazepines. I refer to substances such as etizolam, which is currently being replaced by a new substance called bromazolam. That shift has a lot to do with wider market forces. Globally, it has a lot to do with regulation. As substances are banned and regulated by the United Nations, synthetic substances are manufactured.

However, alcohol and benzodiazepines are depressants and, where benzodiazepines are used, they are commonly used together. Therefore, as Clare Beeston says, it is less about a substitution effect. We need to think much more holistically about our approach to people who use substances, and the support that we provide needs to take into account much more the range of substances that people use, including alcohol and benzodiazepines, because they do not tend to use one or the other.

If the policy is influencing behaviour and causing substitution in any way, can you suggest any mitigations that could assist in reducing the harms that might be present? I know that the interdependency that you described is complex, but are there any specific measures that you might consider?

We did not see that substitution effect, other than in people who already had established dependence. That brings us back to the need for support and treatment services and for outreach services—proactive care that meets people at the point that they are at and deals with the issues. Those issues often occur in the context of other complexities, such as homelessness, so there is a need for targeted interventions that sit alongside MUP.

Obviously, 100 per cent of the additional revenue that is generated by minimum unit pricing flows to the private sector, not the public sector. Do you have a view of how much revenue has been raised as a result of the policy?

There was one study using our data. We did not do the study; it used our sales data. It examined the additional revenue in Scotland and compared it to England. The study assumed that the revenue trends would have been the same in Scotland if MUP had not been introduced and it estimated that there was £270 million of additional revenue over four years, which equates to £67.5 million each year.

Could any adjustments be made to the scheme that would allow for the public sector to capture a share of that? Is it possible?

That is a good and challenging question. It was not possible to say from our study on the economic impact on the alcohol industry how that extra revenue had translated into additional profits and how much had landed with producers versus retailers and large retailers versus small retailers. It is not that everybody benefited uniformly, and not every retailer benefited uniformly. Overall, there was a net increase in revenue, but we were unable to say where that landed. Therefore, we cannot answer the question with a policy solution.

I appreciate your time.

Ivan McKee has a question. Is it on this theme?

It is directly on it.

On the revenue raised by retailers and the private sector as a consequence of the policy, I think that Clare Beeston is saying that the tax levers are currently reserved. If those tax levers on alcohol duty were devolved, the Scottish Government would be in a position to benefit from and bring some of that revenue into the public coffers. Is that correct?

I think that that is raising—

In theory.

Convener, might I help a little here? We are trying to describe for the committee’s benefit the approach to an impartial study. I totally respect that question, but, with respect to the committee, it takes us into a space around levels of taxation and who can tax who, and, for an independent organisation, that is probably outwith the scope. It would be great if members could respect that response.

Thank you.

Your point is noted, Mr Dodds.

I will follow on from the previous two questions. Were any comparative studies undertaken or commissioned to compare how the money raised in other countries by minimum unit pricing is used, or is that a gap that you feel should be looked into?

It was not covered specifically in the evaluation.

I am not aware of any studies looking at that in other countries.

The report says that the theory of change hypothesised that the alcoholic drinks industry might make changes to the size of products. To what extent has that happened as a result of minimum unit pricing?

There is evidence of changes to product size. For example, large 3-litre bottles of very strong cider have largely disappeared from shelves in Scotland, and there has been a move towards smaller 1.5-litre bottles. It is important to remember that the effect was limited because Scotland is a relatively small part of a UK-wide industry, and it is difficult to disentangle the effect of MUP from other things that might drive producers to make different sizes and different strengths of alcohol products.

We were not able to say that strength had changed, but there was some evidence that some products had got smaller. There was evidence that single containers had got smaller and that there were fewer bottles or cans of beer in a pack.

The report also said that there was no discernible positive or negative impact on the drinks industry as a whole. Can you give us more insight into the evidence that brought Public Health Scotland to the view that there was no discernible impact one way or another?

One study undertook quantitative analysis of five metrics of economic performance—number of business units, employment, turnover, gross value added and output value—and was unable to determine any impact from MUP on those measures. In terms of qualitative analysis—that is, from speaking to industry and people on the evaluation advisory group, survey responders and participants in interviews—the general message was that MUP is now business as usual and that, “We’ve dealt with it. It’s what we do now.”

It is fair to say that there was not clear evidence that any increased revenue that retailers accrued from MUP was being passed on to producers. There were discussions about how that was shared, but it was not clear that it was being shared. The impact on individual businesses depended on what they sold or made in the first place.

A retailer who only ever sold alcohol that was a lot above 50p per unit did not see much difference, because their products were not affected. Some of the retailers who sold a lot of products that were affected said that they had seen a negative impact on their revenue.

10:00  

Ivan McKee has some questions.

I have covered everything that I have to cover.

Including on theme 8, on policies and modelling?

Yes, I have covered modelling.

Good morning.

Alcohol-specific deaths are at their highest level since 2008. How does that fact correspond with Public Health Scotland’s report, which shows that MUP reduced deaths by 13 per cent?

It is about the question that we asked. On the impact of MUP, the question that we asked was not whether the number of deaths that occurred after implementation of MUP was lower than before. The figure is a comparison with what might have happened had MUP not been implemented in the first place. We talked earlier about the dip in 2019 and the increases since then. Had that dip in 2019 not occurred, we would potentially be at a higher level of deaths now.

So, is it an estimate?

It is an estimate compared with a counterfactual situation.

Okay. Do you recognise that alcohol deaths are at their highest since 2008?

Yes.

Figures for alcohol-specific deaths registered in 2022 show that the number of female deaths tragically rose by 31 to 440 while, as you mentioned earlier, the number of male deaths remained unchanged. Why is that?

That is outwith the bounds of evaluation of MUP. I do not want to speculate about what those changes are about. It is obvious from the evaluation work that we have done that the reduction in deaths has been greater in males, so there might be something else going on there. I do not want to speculate.

I do not know whether Tara Shivaji wants to add anything to that.

Just give us a view if you can, Tara.

Women make up about 40 per cent of people who access treatment services, so the rise in alcohol-specific deaths among women has to get us thinking about what might be the particular risk factors that affect them. The stigma of alcohol use is one. Especially in the context of women being parents, that can be a real barrier to accessing support and care, and to engaging with treatment services. There are a number of gender-specific barriers that we need to focus on and address. Stigma is one; others include experience of domestic violence and co-existing mental health problems. For a lot of people, dependence follows a series of previous life traumas, and it is common for substantial alcohol use—alcohol dependence—to co-exist with serious mental health conditions. That is another area where we need to see strengthening and improvement.

There is a range of factors. In particular, with minimum unit pricing the question is about the products, consumption and where women purchase alcohol. However, I return to the point that we need multipronged intervention. MUP is one of a range of measures that we need.

So, we need more data.

A Public Health Scotland report from June 2022 found no clear evidence that MUP led to reduced alcohol consumption or reduced levels of alcohol dependence among people who were drinking at harmful levels. Will you explain how that finding corresponds to the June 2023 report?

The finding that you cited is from the harmful drinking study, which looked at a particular group of people who were drinking at harmful levels—predominantly, people with alcohol dependence. Although there is a relationship between drinking at harmful levels and alcohol dependence, they are not the same thing. People with alcohol dependence are a subset of the people who drink at harmful levels, but far more people drink at a harmful level. The reduction in deaths and hospitalisations illustrates that there has been a reduction in harmful drinking, because a death is the ultimate harm that is caused by drinking. They tie up in that way.

Is that an estimate?

The study that you cited was an estimate of the impact that MUP had had on reducing drinking by people with alcohol dependence, who were recruited through treatment services.

So, one result is an estimate and one is a fact, which makes it difficult to draw comparisons.

I am not sure that I answered that question very well, because that is not what I meant to say.

Everything that we are presenting today is based on evidence and data that have been collected in one way or another. We use the term “estimate” to convey the idea that there is still some uncertainty around some of that, but all research has assumptions and uncertainty associated with it, so use of the word “estimate” does not undermine the results and findings that we have presented in the evaluation. I am slightly concerned by the suggestion that it would.

I have a quick supplementary question. We went into lockdown on 23 March 2020, just two years after the policy was introduced. What effect did the pandemic have on your research and on alcohol consumption? Tara Shivaji mentioned women in response to Tess White’s question. I am interested in that area, too, but we have not talked about the pandemic. Did that have an impact on your research on alcohol consumption?

I will answer about consumption first. The pandemic obviously had an impact on consumption because on-trade services ceased to operate for a number of months and we saw off-trade sales going up as a result of that. Overall, at population level sales were lower in Scotland and in England and Wales. I will explain why I am talking about England and Wales in a minute. We estimated that, during the first three months of lockdown, sales were 6 per cent lower in Scotland and in England and Wales.

The pandemic also impacted on our research. We dealt with that in a number of ways. We used a control area for a lot of our studies, which was already planned and was something that we would have done anyway. By “control”, I mean an area where the policy was not implemented, which is why I am talking about England and Wales. Most of the time, we used England, or England and Wales, as our control area, which allowed us to account for external factors that we might not have expected. If those happened in both areas, that essentially levelled the playing field. That is one way that we accounted for things.

Some of our studies were of only the first year of MUP and so were not impacted at all, but in other studies we added into our modelling data to account for the restrictions that were introduced during the pandemic. The pandemic absolutely had an impact on consumption. Some of the data from lower-than-population level shows that different people changed their habits in different ways. People who were lower or more moderate drinkers prior to the pandemic tended to stay the same or to reduce their consumption, but those who were drinking at the higher end were more likely to increase their consumption. There was definitely an impact and we have done our best to account for it within the studies that we have conducted.

Good morning. Tara, in your opening remarks, you said that unintended consequences of the policy would be one of the things that would be looked at. Can you expand on that?

That relates to nuances, in that when the policy was implemented, not everybody in the population was impacted in the same way and to the same extent. My colleagues—either Lucie Giles or Clare Beeston—can describe that in more detail. Within the population, different subgroups have different purchasing and consumption patterns. For example, as a population, people who drink within the chief medical officer’s low-risk drinking guidance of no more than 14 units have a particular purchasing and consumption pattern, and were affected differently to people who drink above the guidelines. As we have discussed, people with quite severe dependence were impacted in various ways. Therefore, what I was referring to was that we set out to try to identify the impacts on the key groups.

With hindsight, I say that it is really difficult to identify all the groups that you want to learn from. However, some of the findings about young people that Clare alluded to demonstrate that we might have wanted them to be a key consideration at the start, to see how the policy would affect them. The subgroups of interest were set out in our protocol for investigation but, with hindsight and given the impact of the pandemic, we would probably want to broaden that and to think much more about equalities and equalities groups within that.

Thank you. Are there gaps between impacts in different areas, such as between rural and urban areas or island and mainland areas?

We did not specifically look at the impact in urban and rural areas. Clare—do you want to add anything?

The study of people drinking at harmful levels did case studies of rural and urban areas, and there was not really any difference between them. The main difference was seen in areas that are close to the border with England because people travel across the border to shop, depending on where the nearest supermarket is, and that continued. I think that one study found that there was less evidence of an impact on purchasing for people who live within 12 miles of the border. However, the issue of cross-border purchasing was very limited—to people who lived near the border. It was not widespread.

Lucie Giles said that 70 per cent of people did not change their purchasing habits. Why do you think that was?

That was the finding of one study in particular, which looked at the household panel purchasing data. That 70 per cent were the lower-purchasing households—the more moderate drinkers. This is speculation to some degree, but based on that pathway—that plausible chain of events—I can only assume that those households were not impacted by the change in price because the products that they were purchasing prior to implementation of the MUP were already above the 50p premium. However, that is speculation. I do not know the answer on the basis of that specific study.

That is speculation, but it makes sense, does it not?

It does.

10:15  

It makes sense to say that most people—including everyone in this room—are not affected by MUP with the type of alcohol purchases that are made. Why 50p and not £1, £10, £20 or £50?

That is not something that we have included in the evaluation. We have evaluated the impact at 50p, which is the level that was set by Parliament. I guess that that decision sits with you.

When MUP was introduced, Buckfast sales surged by 40 per cent. There were also increases in sales of Mad Dog 20/20 and Dragon Soop. Those drinks are all associated with heavier drinking and antisocial behaviour. Why do you think there were increases in the sales of those types of product?

I do not think that it is necessarily to do with the specific type of product; I think it is to do with the level at which they were priced prior to MUP, and whether they were impacted by the policy or not.

So, people moved into purchasing those drinks, then.

The sales work that we did cannot 100 per cent say that individuals switched from one product to another, because we do not have that data at an individual level, but we did see sales reductions in cider, perry and spirits, a smaller reduction in beer and an increase in fortified wine. It makes sense to think that some people, potentially, were switching from one product to another.

Tara said earlier that MUP needs to be part of a package or range of measures. What are the other measures that have come in with MUP?

There are the other measures that are set out in the Scottish Government’s “Alcohol Framework 2018: Preventing Harm”. The measures that I am alluding to would be the World Health Organization’s “best buys” and those that are set out by the WHO European framework. They relate to restrictions on availability, perhaps through licensing, but also through structural separation of alcohol. There are examples from Ireland on that.

There are also restrictions in marketing. Public Health Scotland’s focus on restrictions in marketing would apply particularly to marketing that targets children and young people. As has already been discussed, there is a need to strengthen early access to treatment and the quality of treatment. Those are the broad measures that have been taken.

Of those, what has actually been introduced with MUP?

Alongside—[Interruption.]

I am asking the panel as a whole.

Alongside MUP, there is the alcohol licensing legislation, and we have a programme of brief interventions and treatment. Those measures, out of what has been recommended, are the things that are currently available.

You said to me earlier that the alcohol brief interventions have plateaued and fallen.

I do not know who to direct this question to. What has been the impact of the push towards 0 per cent alcohol drinks that we have seen in the past couple of years?

From a public health point of view, the answer is that we do not know yet. That is currently a subject of research. There has been growth in what we would call the lower end—drinks with no alcohol, or 0 per cent—and in drinks with lower alcohol by volume—lower ABVs. We are asking whether that presents an opportunity for public health, through improving health by reducing consumption. The thing that we do not really know is whether people are switching from full-strength products to low-strength products.

Does that present particular opportunities among key groups—pregnant women, for example? The advice is that women who are planning to conceive, are pregnant or are breastfeeding should abstain from alcohol. The question is whether that shift offers particular solutions, but we do not have the answer yet.

I have one final question. Am I right in saying that MUP is not a panacea or magic bullet to reduce health harms with alcohol, and that your argument is that it should be introduced with a suite or package of measures?

Yes. Our approach would be that it would be part of a package of well-calibrated measures to respond to the high levels of alcohol-related harms that we see in Scotland.

Before we move on, I remind committee members that it is me who is convening the meeting and that you speak through the chair, not across the tables.

The latest Public Health Scotland report states:

“We therefore cannot completely exclude alcohol treatment as an alternative explanation for the observed impact on alcohol-attributable deaths and admissions.”

Does Public Health Scotland plan on doing any more work on alcohol treatment services and the effect that they have on alcohol-related hospitalisations and deaths?

We are currently conducting a review or investigation into what led to the decline in referrals to treatment services. That should help to inform what we need to do to improve access to care.

There is also wider work across the UK, and there is four-nations guidance on alcohol treatment. It is the first time that we have had guidance on the quality of alcohol treatment services. Work will need to be done to implement that guidance and to explore the extent to which it is improving care and people’s experience, and their move into recovery. We expect to have a role in that, but at this stage we do not know what that will be.

I have one follow-up question for Dr. Shivaji. Is MUP, in your view, being billed as the magic bullet, to the detriment of other support and solutions for people with alcohol dependence? You have highlighted that further work will be done. I suppose that my concern is that surely addiction to alcohol should be addressed holistically rather than using just one lever.

I would answer that question by referring to the fact that the harms of alcohol are quite broad. Dependence is a particularly serious harm associated with alcohol, but alcohol is also related to cancers and hypertension, which we project will increase substantially in the next 20 years. That is why we are concerned about the 23 per cent of our population who are drinking above the low-risk threshold.

Of course, those who are drinking in the most harmful and risky way—people with dependence—are at the highest risk of experiencing harm. What we are saying is that a mix of measures is needed. The primary prevention measures—as we would call them—including minimum unit pricing, are often very useful for targeting the harm that is associated with alcohol in the wider population, who are not at extremely high risk. That is why a mix is needed. That is how we would frame the problem. It is very important to have both.

Thank you.

I know that there is a lot of work going on regarding sales, marketing and advertising. I am interested in following what is being done in Ireland and the evidence for segregating sales.

I want to pick up on what Clare said about cross-border purchasing, because there needs to be some myth busting and debunking of the idea that folk are driving fae Ecclefechan tae Carlisle to pick up whatever alcohol they want. If they did that, they would have to buy 33 bottles of vodka to save the five quid on petrol that it costs to go the 20 miles fae Ecclefechan tae Carlisle. Also, my understanding is that the price of alcohol is the same in Hawick and Berwick, so if you live in Coldstream you would be crossing the border to go for your shopping anyway. There are not the booze cruises that keep being touted.

I would be interested to hear about the research that is debunking the myths about cross-border purchases. Can you tell us about that?

There are a number of aspects to what we did on that. For example, we looked at the number of licences around the border. If lots of booze cruises were happening, you might expect a boost in licences at the border to service those booze groups. We did not see that.

Another important element that we looked at, which might be the work that you are referring to, was the costs that are associated with driving across the border, in terms of petrol and time, and how much alcohol one would have to buy and what it would cost to make the savings. On the whole, it is not economically beneficial. We did the research in 2020, I think, so we used fuel prices for 2020: fuel has gone up a lot since then. Therefore, the economic argument for travelling has got less over time. It is largely not beneficial to cross the border just to buy alcohol, because the savings do not offset the cost.

It does happen, as you said, when people already cross the border because that is where they do their shopping—because that is where the biggest supermarket is, or they work over the border and are just going to the shops on the way home, or whatever. That is largely where it is happening.

Conversely, if I want to get my shopping delivered to Ecclefechan fae Asda in Carlisle, there would be a price for delivery of the groceries, as well.

Yes.

So, it doesnae make economic sense to shop across the border, especially since, as I have just said, the price of alcohol in Hawick is the same as the price in Berwick. I guess that debunking that booze-cruising myth is something that we should be doing.

Yes. With a lot of those things, we can say that they are not happening on a large scale. You will be able to find an individual who says, “I do that”, but I agree with you about the larger scale; we found no evidence that shopping across the border was happening on a large scale.

Okay. Thank you.

I thank the panel for their evidence. The committee has certainly learned a lot this morning, and I am sure that it will help us in our post-legislative scrutiny of MUP. We will take a short break to allow panels to change. Thank you.

10:28 Meeting suspended.  

10:38 On resuming—