This month's startling statistic comes from The Times, May 18th.
May was a mad month in Britain and there were rich pickings for connoisseurs of silly numbers. The country seems determined to oust California from its pre-eminence as the home of hysteria. There was yet another furore about "Frankenstein foods". The heir to the throne weighed in with an anti-science diatribe that was welcomed in ecological and theological circles. He told us to listen to our hearts. I listened to mine and it said something like "berboom, berboom, berboom……", which did not seem to add a lot to the argument. There was a rather inconsequential "contamination" of oil seed rape, which produced pages of indignation in the press and got the European politicians in a tizzy.
There were some particularly junky-sounding numbers in a story all over the media from Stanford University School of Medicine in California (where else?). As ever, it is not possible to test them as the vital information was left out. The only numbers given were that 496 people newly diagnosed with Parkinson's were compared with 541 people without the disease. They were asked about their exposure to pesticides and weedkillers, and it was deduced that this "may" double the risk of Parkinson's. It relied, of course, on self-reporting. What does "exposure" mean? How many people have not been exposed to pesticides at some time or other? One Dr Nelson said: "This study is the largest yet of newly diagnosed individuals with Parkinson's and it is the first to show a significant association between home pesticide use and the risk of developing the disease." This implies that there were other studies that did not show this effect; so, if the researchers were working to the usual epidemiological probability level of 5%, Dr Nelson was simply the lucky one with his one in twenty chance of the statistical one-armed bandit.
There was a ludicrous manufactured story based on the fact that one of the seventeen highly qualified candidates who failed to get one of the five places for medicine at an Oxford college was from a state school. The Chancellor of the Exchequer, who ought to have better things to do with his time, used this non-event as an excuse to re-ignite the class war, which socialist governments always do when their backs are against the wall. As the Times leader of May 27th pointed out, Mr Brown was wrong in almost every particular of the case. Still, this should not faze a Chancellor who is notorious for the dishonesty of his triumphalist budget speeches – he simply leaves the nasty numbers out and puts them in the following small print. Predecessors like Healey and Howe may have been disasters, but at least they were honest ones. With his notorious stealth taxes on individuals, the dismantling of farming and manufacturing industries and a threat to multinationals that is likely to drive them out of the country, Brown looks like turning his good fortune in inheriting a healthy economy into a rout. No wonder he feels the need for a political feint
All in all, it was not a very good month for British universities. We old cynics were pilloried when we suggested that one of the outcomes of the nationalisation of the British universities would be academics filling in time sheets. This month they were doing just that. The experience drove me to verse. With that combination of guile and unconscious irony unique to the Great British Bureaucrat, they called it the Transparency Exercise. It was ostensibly designed to calculate the real costs of university research by determining how much academic time was spent on research. Only in the bureaucratic mind can such a measurement be made. Anyone who has ever been involved in research knows you are doing it all the time - driving your car, digging the garden or sitting on the loo. There are two real reasons for this waste of manpower; first, to show the insubordinate swine who are their real masters and, second, to allow a bit of prevarication while the bureaucracy tries to sort out what to do with the wreckage it has made of what was once the finest university system in the world. Watch out for some very silly numbers from this source.
Nevertheless, when it comes to really silly numbers, nobody can hold a candle to the anti-smoking freaks, so they take the biscuit this month. The Chairman of Council of the British Medical Association in a letter to The Times (May 18th) wrote:
Tobacco is both a high priority in health and a truly international concern. In 1990, smoking accounted for one in six adult deaths worldwide; in 2020, it will account for one in three. Every nine seconds, someone dies from a disease caused by tobacco - that's 3.5 million preventable deaths a year. In the next 20 years, the toll will rise to one death every three seconds - or ten million deaths annually. Seven out of every ten of those killed will be from poorer countries.
It is difficult to know what to say about this sort of hysterical hyperbole. These ludicrous numbers do not come from any known study, even an epidemiological one. They are simply made up, plucked out of the air. They come from an assertion by the World Bank and propagated by the WHO. They are based on the thesis that every death of a smoker is caused by his habit. You might just as well argue that a half of all deaths are caused by masculinity, or indeed that five sixths of all deaths are caused by non-smoking. There seems to be a bidding process by which someone thinks up a number and then someone else tops it.
When so-called studies produce this sort of number they are easily exposed as lies. One of the most oft repeated numbers from the anti smoking campaign is that it causes 400,000 premature deaths in the USA each year. This number is a total fabrication – see, for example, the article Lies, damned lies and 400,000 smoking-related deaths by Levy and Marimont, Regulation, Vol21 No 4 1988. The calculation, which was made by the Center for Disease Control and Prevention (CDC), involves almost every fiddle in the book. These include unacceptable risk ratios, substantial confounding factors, self-reporting, unrepresentative sample populations and many others. Most startling of all is that 60% of the "premature" deaths occurred at ages over seventy and 17% of them at eighty-five and above. It has been pointed out that the same data can be used to "prove" that tobacco saves 200,000 lives a year.
Now, like many human activities, smoking carries risks (for example it increases the probability of developing lung cancer by a factor between 10 and 20) but what purpose does it serve to tell lies on such a grand scale, other than giving the bureaucrats yet another excuse to interfere in our lives? It is a powerful illustration of the progress made by Political Correctness that the author of such hogwash should be the leader of what was once a respected scientific institution. Does anyone seriously believe that one third of all deaths could be attributed to any single cause?