Professor Tim Stockwell:

That is an important point to clarify. Moving on to Saskatchewan, the importance of this particular case study is that, of all the provinces, Saskatchewan introduced minimum pricing closest to what has been proposed in Ireland and Scotland, which is minimum unit pricing. Overnight, on 1 April 2010, the province developed a comprehensive set of minimum prices for all alcoholic beverages. It increased the rate of that and applied it to the strength of each drink. Spirits had three strength categories, while beer had four and the higher the strength the higher the minimum price. It is not precise but it is as close as we have got in Canada.
I will now show committee members some of the results. The graph shows per capitasales of six different types of alcohol and total ethanol sales before and after the change. We did statistical analysis control for a lot of variation, but if one eyeballs this seasonally adjusted data, there is a sudden change in the pattern of results. Overnight, therefore, consumption patterns tended to go downwards and there was some switching between products. In fact, there was switching from high strength to low strength products.
I have summarised some results that were published in the American Journal of Public Healthshowing that overall consumption went down 8.4% for a 10% increase in minimum prices. That is describing the overall relationship. It was greater for off-licences than for bars, which one would expect because the minimum prices affect sales in off-licences more. I have already mentioned the shift from high to low strength beers and wines. Increased government revenues and increased profits for the alcohol wholesalers were also evident in that case study.
What about harm? For Saskatchewan so far, I can cite only anecdotal reports from the police. A local police inspector reported that police calls per month related to intoxication or disturbances had halved since the introduction of minimum pricing. It seems to be obvious. These are good witnesses to public order problems on Friday and Saturday nights.
We have more objective data from British Columbia on crime outcomes. I will also show the committee some acute hospital admission data.

These data are relevant to crime because the majority relate to injuries, including a disproportionate number of violent injuries, arising from motor vehicle accidents and so forth. Our objective was to identify patterns of change over an eight-year period, specifically the relationship between changes in the minimum alcohol price and the rates of acute hospital admissions, including those involving violent injuries. The overall relationship was significant, with the data showing that a 10% increase in price was, on average, associated with a 9% reduction in these types of hospital admissions. The data are a very good indicator of crime outcomes, because when people are injured they go and seek help. In general, crime data tend to be influenced by enforcement strategies, so these particular data are a good objective indicator.

We have another paper which is not yet published but the findings of which my co-authors have permitted me to share with the committee. This study was based on nine years of crime data across 89 local areas of British Columbia and used the same type of modelling. Taking account of economic variables, seasonal factors, demographic profiles and so on, we are finding large and significant impacts on correlated traffic offences, with an almost 20% reduction for a 10% increase in minimum alcohol price. The study involves large confidence intervals and we cannot say for sure that the reduction would be exactly 19.5%. It might be somewhere between 4% or 60%, but our statistical model is pointing to a relationship of that magnitude. The other issue to be clear about in considering these findings is that the minimum price did not actually increase by 10%. In fact, it was going up and down all the time; our concern was to describe the relationship between the price and this particular outcome. We found no significant impact on non-alcohol-related traffic accidents. This was nice because it was in accordance with how the data should behave if there is indeed an alcohol-specific effect. We also found significant impacts on property crimes and crimes of violence. As I said, those data remain to be published.