Thursday 16 November 2017

Looking forward to minimum pricing

Now that the SNP are free to introduce minimum pricing, it's worth looking at what we're supposed to expect.

Back in 2009, when minimum pricing became a live issue, the Sheffield modellers predicted that a minimum price (of 40p in those days) would result in a drop in alcohol consumption of 2.7 per cent and a decline in alcohol-related deaths of 40 in the first year, rising to 210 per annum after ten years.

Given the 'public health' lobby's absolute obsession with this policy in the years since, we must assume that they regarded these as game-changing numbers. Imagine if alcohol consumption fell by 2.7 per cent! What a victory for health that would be.

We don't need to imagine because consumption fell by much more than that after 2009 without any notable policy change. In 2007, Scots were drinking 11.8 litres of alcohol a year. By 2016, this had fallen to 10.5 litres. This is a drop of 11 per cent - four times as great as the decline Sheffield said would occur if minimum pricing was introduced.

You probably haven't heard much about this, but if minimum pricing had been introduced in 2009 you would have never heard the end of it. Not only did alcohol consumption fall by 11 per cent, but the alcohol-related mortality rate fell from 34.6 per 100,000 to 30.0 per 100,000 for men and from 16.7 per 100,000 to 9.0 per 100,000 for women. This is a drop of 13% and 46% respectively.

It is interesting to see the decline in both drinking and alcohol-related deaths in Scotland in recent years and yet I do not see much interest in it from the denizens of 'public health', presumably because they can't take credit for it. 

Here are the alcohol-related deaths for men and women. Scotland is the top (light blue) line.

It's worth noting that the UK as a whole has seen a decline alcohol consumption of around 18 per cent since 2004 and yet Scotland is the only part of it to have seen a significant fall in alcohol-related deaths. This implies that the fall in alcohol consumption in Scotland has been driven by heavy drinkers consuming less whereas the fall in England, Wales and Northern Ireland has been driven by moderate drinkers consuming less and more people becoming teetotal.

Looking at the actual number of deaths below, you can see that mortality increased sharply between 1993 and 2003 before falling by about 20 per cent. It has not followed drinking trends perfectly, however. Note that there was a relatively large number of deaths in 2016 despite per capita consumption being at a twenty year low.

The latest Sheffield predictions for Scotland predict that a 50p unit price will reduce consumption by 3.5% and will reduce the number of deaths by 58 in the first year and by 102 per annum after ten years. This is what the SNP and its allies have been fighting for all this time. This is the promised land.

But despite all the wild celebrations from the neo-temperance lobby yesterday, these projected outcomes are so trivial that they would get lost in the noise of the data. If there is a 3 or 4 per cent downturn in per capita consumption in the first year of minimum pricing, you could plausibly attribute it to minimum pricing, but (a) it could just as easily be part of the longer term decline, and (b) so what?

The aim of the policy is to reduce alcohol-related deaths, but if minimum pricing did 'save' 58 lives, it would be impossible to tell from looking at the numbers because they fluctuate by more than that on a regular basis. Between 2015 and 2016, for example, they rose by 115 for no obvious reason. Between 2011 and 2012 they fell by 167.

Regardless of whether the figures rise, fall or stay the same over the next few years, it is inevitable that a regression model will be published - probably by the monopoly providers at Sheffield University - claiming that there were fewer deaths than there would have been in the absence of the policy. Such a regression model will be politically driven rubbish, but even if someone made a serious attempt to create a regression model, it would be impossible because they would not be able to project future trends. Why? Because they don't know the reason for the recent trend.

The stark reality is that the projected impact of minimum pricing, exaggerated though it almost certainly is, amounts to a rounding error too small to be seen with the naked eye. Even if it does everything its advocates claim it will, the impact of this supposedly world-leading policy will be too small to measure. The policy of doing nothing and selling alcohol at 'pocket money prices' in the last decade seems to have been vastly more successful than the most optimistic projections of Sheffield's activist-academics. 

No comments: