Hansard

There has been minimum pricing in Canada in around eight of our 10 provinces for many years in most instances. It is part of a suite of pricing strategies that can target the ethanol content of alcoholic drinks. The Alcohol (Minimum Pricing) (Scotland) Bill would do precisely that by specifying a unit of alcohol as having a minimum price. That is a very good example. The rationale for that, of course, is that the risk of alcohol-related harms depends on the dose. That applies to whether the harms are acute and short term, such as injuries or poisoning, or chronic harms that arise from various diseases resulting from alcohol consumption.I have dose response curve illustrations. The slide shows one for breast cancer, which is a summary of all the studies that have ever been published on the risk of breast cancer. You can see the level of alcohol consumption going up to 20g, 40g, 60g and so on along the bottom. Each of the lines would be a bit more than two units. The point is that there is an exponential increase in risk with rising consumption.The next slide relates to cancer of the oesophagus, which is another common cancer. There is a similar exponential curve.The effect also plays out at the population level, of course. The next slide shows data on hospitalisations in the United Kingdom caused by alcohol. Members are probably very familiar with those data, which are from a period of rising per capita consumption. I understand that consumption levelled off shortly after that, although perhaps not in Scotland.The principle of raising alcohol prices to reduce consumption is well established. There are two high-quality reviews that effectively consider every single published study on the subject. One of those reviews was done by somebody from the business world—a marketing professor in the United States—who identified 132 studies published after 1945. Across the board, his estimate was that a 10 per cent price increase leads to a 5 per cent decrease in consumption on average.An even more careful study by Alex Wagenaar from the University of Florida had a higher bar for quality. It considered 112 studies that had been published since 1823 with 1,003 estimates. It came up with a similar conclusion but also concluded that prices affected the heavy drinkers as well. I will return to that topic.As you are aware, there is also evidence that price and taxation have an impact on harm. We do not necessarily need to worry about effects on consumption, but there are studies that show directly that increases in price and tax reduce rates of alcohol dependence, liver cirrhosis, road trauma, assaults, sexually transmitted diseases and, in fact, all forms of alcohol-related mortality and morbidity.Those studies concern across-the-board price rises, so why introduce minimum pricing? The theoretical reason is that we know that the heaviest drinkers gravitate towards the cheapest alcohol. One US study found that the top 10 per cent of drinkers with the highest risk paid on average 79c, compared with the bottom 50 per cent—the light drinkers—who paid nearly $5 per standard drink. We also know that young people and high-risk drinkers are especially responsive to minimum prices.There are some good theoretical reasons for minimum unit pricing, but there have been hardly any studies. In fact, until we published our work, we could not find a single empirical study into whether a minimum price would have any effect.I note that our national alcohol strategy in Canada recommends—I believe that there are 43 recommendations—that every one of the Canadian jurisdictions should be encouraged to adopt minimum prices, which the strategy calls “social-reference prices”, that they should be indexed to the consumer prices index and that they should be reviewed annually to report compliance.I also note that a number of other provinces have minimum prices in bars, restaurants and hotels. The social-reference prices mostly apply in liquor stores. The Government in British Columbia owns about 200 liquor stores in the province and there are about 800 privately owned stores. It has a monopoly on distributing the alcohol, so it can set the starting price. In its own stores, it will hardly ever sell below the minimum prices—there is one exception, which I will explain.Some jurisdictions also index the minimum prices to inflation, but British Columbia has not done that.One of the loopholes is that, if the products are not selling well in its liquor stores, the Government can delist them and sell them for a lower price, which surprised me. However, those delisted, cheap products account for less than 1 per cent of sales. It is a tiny proportion.Slide 13 gives you an idea of the kinds of minimum prices that are charged for a Canadian standard drink, which is nearly double a British unit; it is almost 14g, compared with your 8g unit of alcohol. However, my observation is that Scottish and British people drink alcohol in pretty similarly sized servings to the Canadians—actually, I think that they are larger servings.The slide shows how the minimum prices vary by beverage type and across different provinces. British Columbia has some of the lowest minimum prices; Ontario is intermediate; and Saskatchewan is among the highest. Those three examples show the range in Canadian provinces.One of the loopholes concerns 75 per cent strength rum, which is at the bottom of the slide. Because the minimum price is not set to a standard amount of ethanol, it can still be cheap per standard drink, compared with a 40 per cent spirit, such as a tequila.The next slide is a map of Canada. British Columbia is on the left and Saskatchewan is a bit to the left of the middle. I will give you some of our analysis of the impacts of minimum pricing in those two provinces.The first case study concerns BC, where the Government alcohol monopoly has set minimum prices for more than two decades. That is the study period in our paper. In fact, only spirit prices were regularly updated during that time. Beers, wines and alcoholic sodas were hardly ever adjusted—only two or three times—but that makes for a nice natural experiment, because there were long periods with no change followed by a sudden, substantial change and then nothing. That enables us to compare what happened with the different beverages and over time.One of the unfortunate consequences of that system from a policy point of view is that every time the Government dares to raise the minimum price it makes the front page of the newspapers and there is a great deal of alarm and criticism of the Government getting into our pockets, robbing us and so forth. Such moves are extremely unpopular, whereas in Australia, where the excise taxes are adjusted on a quarterly basis, no one notices. There, it is routine for the excise taxes to be adjusted in line with the cost of living.I mentioned that the Government monopoly is partially privatised as regards where the alcohol is sold. Someone who was visiting British Columbia who went into a store would find it hard to notice whether it was a Government store or a private store; they would not know the difference. Given that the prices that the Government sells at in its stores are the prices at which it sells to restaurants and bars, which then add on their own profits, the minimum prices that it sets affect in some way, directly or indirectly, the prices of all alcoholic products that are sold, regardless of the type of outlet in which they are sold.The graph on slide 15 shows the four main beverage types: spirits, beer, wine and coolers. We have quarterly data over the 20-year period—in fact, we have 84 quarters’ worth of data on prices and sales. Each little adjustment represents an adjustment for CPI or an adjustment in the minimum price rate that has been applied to a litre of spirits, a litre of beer or a litre of wine. The top line, which shows the price of spirits, is the only one that has kept pace with inflation. The figures are in standardised, CPI-adjusted dollars. The price of the other three beverages, particularly coolers or alcoholic sodas, has tracked downwards and has not kept up with inflation.In our time-series analyses over 20 years, we have controlled for the effects of time of year, overall trends in the data, trends in average alcohol prices and trends in household income, and we have used inflation-adjusted minimum prices to predict the volume of alcohol sold for each main beverage. Our results are that, overall, for any one alcoholic beverage, our model suggested that a 10 per cent increase in its price would reduce its consumption per capita by 16.1 per cent in comparison with consumption of all the other beverages.There are some substitution effects. If we look just at individual beverages and forget what is happening with the other types of beverage, we find that there is a relatively small but significant effect on beer—a reduction in consumption of 1.5 per cent—whereas there are larger effects for the other beverage types: spirits, wine and alcoholic sodas.Overall, because of the substitution effect, we find that a 10 per cent increase in minimum price produces what appears to be a fairly modest 3.4 per cent reduction in overall consumption but, even though that looks small, we should bear in mind that it is the impact on total alcohol sales, whether they are made in restaurants and bars or private or Government liquor stores. Given that the minimum prices affect only a minority of the products that are sold, it is quite surprising that they should affect such a large proportion of total alcohol consumption.Since then, we have looked in more detail at possible impacts on harm. We have used quarterly data on alcohol-related hospitalisations across 89 areas of the province over eight years. We have not submitted this work for publication, but I will share with you that our estimate is that a 10 per cent increase in the minimum price across the board resulted in a 4 per cent reduction in acute alcohol-related hospitalisations over the period, controlling for all the other factors. We are talking about injuries and poisonings. There was a less pronounced effect for chronic illnesses, which is pretty much what we would expect, given that we were looking at short-term impacts. It takes a number of years for chronic illnesses to develop, so that is the kind of result that we would have predicted.18:15 Before I finish, I turn quickly to Saskatchewan. We have another paper in preparation on the situation there. There is interest in Saskatchewan because it decided to adopt a fairly radical approach to increasing minimum price. Late in 2009, Saskatchewan had a problem with a shortfall in revenue—I believe that it was to do with provincial potash sales, which are apparently important for the economy there. As luck would have it, there was a public health-inspired proposal to adjust all minimum prices in accordance with the strength of alcohol, pretty much like the Scottish Government proposes to do, although the Government in Saskatchewan did not put it in terms of the standard drink and just set different rates for different strengths. The proposal was adopted as a means of raising more revenue to make up the shortfall. There were four strength categories for beer. One category was for beers below 6.5 per cent and the highest rate was for beers above 8.5 per cent. The proposal resulted in a range of minimum prices, with higher ones for stronger alcohol. There were two strength categories for wine, two for alcopops and five for spirits.I hope that the graph on slide 20 shows what happened, without giving you any statistics. We seasonally adjusted the data, because there were big seasonal variations in the sales of each beverage type. When we smoothed those out, we got detailed data for two years before the intervention and one year afterwards. It is clear that, before the line in the graph, which indicates the introduction of minimum pricing, there is a trend of increasing consumption and, after the line, there is a trend of steady or decreasing consumption for all the beverages. We used the same approach as we did in British Columbia and our statistical assessments suggest that there was a similar overall effect on consumption. In this case, there were stronger effects for beer, particularly high-strength beer, than there were for other beverages. There was a massive reduction in the sales of beers above 8.5 per cent in strength.There was a stronger effect on sales off-premise rather than on-premise, which is what you would expect, because a minimum price has a more direct effect on prices in off-premises—off-licences if you like. There was also a stronger effect for higher alcohol strength varieties of beer, wine and spirits. The graph in slide 23 shows an example of the shift that occurred—the red part is the stronger beer and the green part is less strong beer.In Canada, minimum pricing has been implemented fairly inconsistently and has differing impacts on consumption. Loopholes permit sales of very cheap alcohol and most of the minimum prices have failed to keep pace with inflation. However, periodic price increases that have occurred have triggered reductions in consumption and, it appears, some harms. It appears that reduction in consumption is more likely when the minimum price is significantly higher. That is fairly obvious—if there is a low minimum price, it will not have such a big effect. The reduction in ethanol consumption is greater when there are across-the-board increases for all beverage types instead of increases that target a particular beverage, which allows a lot of substitution.Many people object to pricing strategies, but minimum pricing is perhaps one of the least objectionable strategies. I am sure that that is why you are considering it in Scotland. All pricing strategies have the most impact on heavy drinkers, but minimum pricing especially targets heavier and younger drinkers, because they mostly prefer cheaper drinks. Minimum prices can be adjusted so that they are higher for more hazardous products, which the Scottish Government proposes to do.A high proportion of all alcohol consumption, whether it is in the United Kingdom or in Canada, is done in a way that is not consistent with low-risk drinking guidelines. About two thirds of all alcohol that is consumed is potentially putting the drinker at risk of some kind of harm. That should be taken into accountWe are doing further studies, which are needed to confirm whether there is a disproportionate impact on alcohol-related harms from adjusting minimum prices. I am happy to answer any questions.