Their worthless model made some very specific predictions about Scotland's first year under MUP:
Alas, consumption has gone up by four per cent (year on year) since the policy was implemented in May - this, against a backdrop of falling consumption over the long-term.
Sensing that the wheels could be falling off already, the Sheffield team are trying to move the goalposts...
The emerging narrative is that the rise in consumption would have been even higher if minimum pricing had not been introduced. This theory has the convenient benefit of being impossible to prove either way and it allows for even more meaningless modelling in the future.
In an interview with the Herald, Sheffield's Colin Angus said:
"A more relevant comparison is to look at alcohol consumption over the same period in England, where MUP is not currently in place.
"Alcohol sales in England rose by substantially more over the same period, suggesting that in the absence of MUP we might have seen a bigger rise in drinking in Scotland.
As tempting as it is to look for immediate impacts of the policy, this volatility means that the real test of MUP will be in how drinking in Scotland changes over the next few years."
Suddenly those predictions for the first year, which were such a big part of the campaign for this regressive policy, are being shoved down the memory hole. "Did we say alcohol consumption would fall? No, no, no! We meant to say it would rise, just a bit more slowly than in England." It's hard to believe than even the SNP would have proceeded with minimum pricing if that was the prospectus.
It's still too early to say what will happen to alcohol consumption, let alone alcohol-related deaths, in the first year. A big decline in the second half of the period is still possible. I have always said that trying to predict the effect of the policy is a fool's errand, other than to say that a shift to spirits is highly likely. But the response from SARG shows that it doesn't matter how far off the mark their predictions turn out to be, they will always find a way to claim minimum pricing 'worked'.
And yes, it was an unusually hot summer (even in Scotland) and there was a World Cup on (which Scotland wasn't in). Both may have been factors in alcohol consumption rising, although there is an international football tournament every two years so that really should have been accounted for in the model. But we were told that minimum pricing is 'the most effective of a range of policy options' (Scottish Government) and 'a highly effective tool' (every self-proclaimed alcohol expert). It was seen as such a powerful weapon against excessive drinking that the Scottish Government went to court to defend it - and commissioned research from (you guessed it) SARG which concluded that minimum pricing was the single most effective anti-alcohol policy available.
You'd think that a game-changing policy of such potent efficacy would be able to overcome the effects of a few weeks of sunshine and a bit of football on the telly, wouldn't you?
But apparently not. In the space of six months, the rhetoric around minimum pricing has gone from it being 'the most potent and evidence-based approach to reducing the population’s consumption of alcohol' to 'er, well, it could be worse.'
If minimum pricing continues to have so little effect on alcohol consumption, a rectally sourced regression model will be the only way to spare the campaigners' blushes. We have seen it all before with the Mexican sugar tax, plain packaging and various smoking bans. In each case, the targeted outcome either failed to decline or declined at the same rate as it usually did after the 'game-changing' intervention. And yet, thanks to opaque modelling that can never be tested or replicated, campaigners claimed success - take a look at this study by Jill Pell to see the lengths to which activist-academics go to conjure a positive finding out of nothing.
This is one of many ways in which 'public health' differs from public health. Real public health interventions deliver observable improvements in health. If rates of disease stay the same or rise after a vaccine is introduced, for example, we would rightly conclude that the vaccine doesn't work. But when the intervention works, the results are obvious and immediate. They do not require statistical models to (supposedly) filter out the other factors - though other factors undoubtedly exist. We can see the effects with our own eyes because they do what they are supposed to do. They actually work.
No comments:
Post a Comment