People living within a 25-mile radius of Michael Gove, MP, are more likely to die of cirrhosis of the liver and alcohol-related tumours than anywhere else in the country. There is a direct correlation between (fairly) close proximity to Michael and very high levels of alcohol consumption. A study carried out by researchers from the Liverpool John Moores University discovered that five of Britain’s most alcohol-saturated areas were congregated around Michael’s constituency in Surrey Heath. The only one that wasn’t nearby was Harrogate, in North Yorkshire — and just a cursory check through the clippings shows that Michael delivered a speech there to the Chartered Institute of Housing, in which he was critical of the government’s house-building policy. So, now that we have uncovered the correlation, we must decide what to do about it. The humane answer would be to relocate Michael to a place where his baleful, drink-inducing aura will be of the least danger to human life — Gruinard Island, in the north-west of Scotland, for example, which has recently been pronounced entirely free of anthrax. I favour the humane approach because Michael’s an old friend whom I like very much indeed and I have to say I was shocked by these latest findings. Who would have thought it?
The Gove Correlation was not the conclusion reached by the researchers from John Moores University — that was my reading of their work having drunk most of a bottle of Sancerre left over from a party on Saturday night. It is, of course, a fantastically stupid conclusion. There is indeed a loose correlation between proximity to Michael Gove and what the researchers call heavy recreational drinking, but only a lunatic or an imbecile or a vindictive drunk would suggest a direct causal relationship. The conclusion the people from the university drew was that heavy drinking is not confined to yer working classes. More than a quarter of people living in Surrey Heath, Runnymede, Mole Valley, Guildford, Mid-Sussex (and Harrogate) drink ‘between 22 and 50 units of alcohol per week’. To translate that into the sort of language used by human beings, that’s a couple of bottles of wine for men and one and a half bottles of wine for women. The newspapers reported these results with great glee and apocalyptic predictions. According to the Independent, the respectable and affluent people of Surrey are ‘sitting astride an alcohol time bomb set to explode some time soon with an epidemic of liver problems, heart disease and cancer’. I have to say that the tone of the report implied that they were quite cheered by this prospect.
We are bombarded on a pretty much daily basis with grim statistics which always seem to support government obsessions and objectives. Now that smoking has been kicked into touch, the focus has shifted to alcohol — just as many of us smokers suggested it would. It has got to the point now that I believe none of them. The study from which I drew my Gove Correlation made the politically correct point that the middle classes were drinking more than the working classes. A closer look at the study showed that this conclusion was almost as fantastically stupid as my own; areas with the real drink problem, where between 7 and 9 per cent of the population drink way in excess of 50 units every week, were the economically and socially impoverished regions of the north-west and north-east of England. The people in ‘leafy’ Surrey (it’s always ‘leafy’ Surrey, isn’t it? As if leaves were an immediate signifier of members of the Conservative party, rather than of, say, trees) drink just a little more than the top quacks and the anti-alcohol pressure groups in the country insist is good for us. In other words, they imbibe maybe two and a half bottles of wine every week rather than two. Even that figure may be exaggerated because most people do not drink a set number of units every day. And are we really about to witness the population of Surrey being wiped out by liver disease, cancer and heart disease? Well, one assumes that in the end they will succumb to some ailment or other, even if they are fortunate enough to live in leafy Surrey — but the increased likelihood of an individual developing liver disease because of an extra few glasses of wine every week is infinitesimally small. Nor is drinking the only thing which can give you liver disease: there is a direct correlation, apparently, between eating pork and suffering cirrhosis of the liver. I know this because I found another in-depth, scientific survey on the internet. On an Islamic website.
There’s a very fine book recently published about statistics and their chronic misuse, The Tiger That Isn’t, by Michael Blastland and Andrew Dilnot. They do not concur with my thesis that all statistics of which the government approves should be immediately ignored. They say instead that: ‘Statistics is far from the dry collection of facts; it is the science of making what subtle sense of the facts we can. No science could be more necessary, and those who do it are often detectives of quiet ingenuity. It is others, snatching at numbers, brash or overconfident, who are more naively out of touch.’ Well yes, sure, but there seems a marked lack of subtlety in the politically loaded statistics with which we’re bombarded every day. A little later in their book, when dealing with the subject of causation, the writers invent a little Gove Correlation of their own: ‘Loud music causes acne. How else to explain the dermatological disaster area witnessed in everyone wired to headphones audible at 50 paces? That is a cheap joke . . .’
A cheap joke maybe, but it is scarcely any more stupid than another study published recently and which, again, seemed to support a government aspiration for the electorate. This one came from the Office of National Statistics, via some bloke at the London School of Economics, and its conclusion — reported without question in all of your morning newspapers — was this: people who are married tend to live longer and suffer fewer long-term illnesses than people who simply cohabit. Further, single mums have the poorest life expectancy and general levels of good health than anyone else. Now, this conclusion is every bit as fatuous as my Gove Correlation and the Blastland-Dilnot joke about loud music causing acne. I rang the Office for National Statistics to find out more about the survey: had it been weighted to take account of economic and social status, I inquired? They weren’t sure, they said. Then they rang back and said they thought it probably hadn’t been weighted for economic status. Ah, there you are. It seems to me people who own homes with a triple garage, badminton court and indoor swimming pool, living close to Michael Gove in leafy Surrey, may live longer and suffer fewer long-term illnesses than those who live under a piece of tarpaulin in the road in Oldham. They are also more likely to be married and far, far less likely to be a single mum. It is general poverty which inflicts long-term chronic illness upon people and also reduces their life expectancy and single mums and cohabitees are, on average, less affluent than people who are married. And yet this report was trumped up to suggest that it was the simple matter of living arrangements which had some sort of causal effect upon the health of the people concerned. Well, perhaps it does, perhaps it doesn’t. This survey told you absolutely nothing either way.
Or take an example from the Blastland-Dilnot book about that current obsession, global warming. The mean temperature rises and there are more cases of malaria in the East African highlands, therefore global warming is causing more malaria in the East African highlands. ‘Convinced by these news stories from respectable broadcasters and newspapers? You shouldn’t be; they are all causation/correlation errors made hard er to spot by plausibility . . . Plausibility is often part of the problem, encouraging us to skip more rigorous proof and allowing the causation instinct to settle too quickly: it sounds plausible so it must be right. Right? Wrong.’
It is not just plausibility — our readiness to believe because it sounds right — but also suggestibility — our readiness to believe because it either confirms an underlying prejudice of ours or confirms an underlying fear. Enemies of Michael Gove, if the chap has any, may even now be repeating my absurd causation/correlation error to anyone who will listen. But equally, we find it agreeable to think of alcohol-induced disease as a great leveller, like tuberculosis once was, affecting rich and poor with similar effect. But, clearly, it is not. By and large, the poorer you are, the more likely you are to drink yourself into an early grave. And simply because the residents of Surrey Heath like an extra glass of Rioja at the end of their meal does not remotely alter this fact.
Offer ends in:
