The Prime Minister has announced that England will follow Scotland’s lead and introduce a minimum price per unit of alcohol. That would, it is claimed, protect responsible drinking in pubs but deter the binge-drinkers and the addicted. Unfortunately he has got a bit ahead of himself.

Originally the Scottish government pressured the industry to agree the minimum price, and enforce it, amongst themselves. Despite this being the land of Adam Smith, it took a while for the penny to drop that such price fixing would be illegal.  Then it transpired that the Scottish government setting the minimum price would, or might, transgress EU legislation.  A visit to Brussels by the Scottish Health Minister last month did not resolve the matter but we should hear from Brussels shortly.

Stepping back a bit, minimum unit pricing, i.e. the price per centilitre of pure alcohol, works as follows.  If it was introduced at 40p, then the minimum price of a bottle of wine (9 units) would be £3.60 and a bottle of 70o spirits (units) would be about £10.  The products that would be priced up, and possibly out, would be the high alcohol beers and ciders.  And, it is claimed, they are the drinks of choice of the binge-drinkers and the addicted, i.e. those who drink for the effect rather than more socially and moderately.

Sheffield University has some sophisticated models which indeed show minimum unit pricing would have relatively more impact on the target “harm” groups, but would also cut total alcohol consumption, and Treasury revenue, possibly due to a ripple upwards effect on other pricing.  Of course the anti-alcohol groups applaud.  They see reducing total alcohol consumption as the way to reduce alcohol harms.  According to them this is statistically proven. Actually the theory goes back to Sully Ledermann 56 years ago and it does not stand up at all.  The UK is in the middle of the EU pack in terms of per capita alcohol consumption. It has been falling steadily for some time and faster than the EU as a whole.  Yet the UK has more alcohol problems, alcohol related A&E admissions for example, than elsewhere in the EU.  If Ledermann was right, how come problems are increasing when consumption is falling?

In Ireland half the populace does not drink at all, leading to a low per capita total, despite the best efforts of their compatriots to make up for it.  Per capita averages are misleading.

Another problem is that the number of units does not appear on the packaging.  It would be very difficult for retailers to implement minimum pricing if they do not know how many units are in the package.  The drinks industry agreed with government, in March 2011, to have the number of units appear on the packaging of 80% (by volume) of alcoholic drinks sold in off-licences by 2013.  This, sensibly, lets the very small (by volume) sellers such as Chateau Lafite off the hook.

Apart from one Canadian province, and that is a very different situation, minimum unit pricing has not been tried anywhere in the world. It may turn out to hurt moderate more than "problem" drinkers. What works in Sheffield’s models and what happens in practice are two different things.