Models cannot prove, they only predict

JUST when the tide appeared to be turning in favour of the Scottish Government's minimum pricing proposals for alcohol, an academic has bowled a scorcher at the middle wicket. Dr Petra Meier of Sheffield University told the parliament's health committee yesterday that there is no firm evidence to show the policy would work.

The projected effects of the minimum pricing plan, she said, was "like the weather forecast" because her work was just "a model" of what might happen. It is a Sheffield University study that has formed the central argument behind the Scottish government's Alcohol Bill and which ministers have claimed "proves" the case for minimum pricing.

As with some of the research offered as validation by the "global warming" school, a healthy scepticism needs to be applied to models which purport to establish a robust, unchallengeable link to a real-world outcome.

Dr Meier's remarks do not invalidate the argument of a link between the price of alcohol and its consumption. But they invite caution when asserting specific particular outcomes at specific particular levels of price. Forecasts – as investment bankers and central bank governors are now painfully aware – are no infallible guide to actual performance. The challenge to this bill is not to its laudable aims but whether it will do what it says it will do on the tin.