Matt Ridley: The Rise Of Fake Science

  • Date: 28/02/19
  • Matt Ridley, The Spectator

Pseudoscience is on the rise – and the media is completely hooked

‘The whole aim of practical politics,’ wrote H.L. Mencken, ‘is to keep the populace alarmed (and hence clamorous to be led to safety) by menacing it with an endless series of hobgoblins, all of them imaginary.’ Newspapers, politicians and pressure groups have been moving smoothly for decades from one forecast apocalypse to another (nuclear power, acid rain, the ozone layer, mad cow disease, nanotechnology, genetically modified crops, the millennium bug…) without waiting to be proved right or wrong.

Increasingly, in a crowded market for alarm, it becomes necessary to make the scares up. More and more headlines about medical or environmental panics are based on published scientific papers, but ones that are little more than lies laundered into respectability with a little statistical legerdemain. Sometimes, even the exposure of the laundered lies fails to stop the scare. Dr Andrew Wakefield was struck off in 2010 after the General Medical Council found his 1998 study in the Lancet claiming a link between the MMR vaccine and autism to be fraudulent. Yet Wakefield is now a celebrity anti-vaccine activist in the United States and has left his long-suffering wife for the supermodel Elle Macpherson. Anti-vax campaigning is a lucrative business.

Meanwhile, the notion that chemicals such as bisphenol A, found in plastics, are acting as ‘endocrine disruptors’, interfering with human hormones even at very low doses, started with an outright fraudulent study that has since been retracted. Many low-quality studies on BPA have pushed this theory, but they have been torpedoed by high-quality analyses including a recent US government study called Clarity. Yet this is of course being largely ignored by the media and the activists.

So the habit of laundering lies is catching on. Three times in the past month, pseudo-science flew around the world before the scientific truth had got its boots on (as Mark Twain did not say, but Jonathan Swift almost did): in stories about insect extinction, weedkiller causing cancer, and increased flooding. The shamelessness of the apocaholics is increasingly blatant. They know that even if a story of impending doom is thoroughly debunked, the correction comes too late. The gullible media will have relayed the headline without checking, so the activists have made their fake-news hit, perhaps even raised funds on the back of it, and won.

Take the story on 10 February that ‘insects could vanish within a century’, as the Guardian’s Damian Carrington put it, echoed by the BBC. The claim is, as even several science journalists and conservationists have now reported, bunk.

The authors of the study, Francisco Sánchez-Bayo and Kris Wyckhuys, claimed to have reviewed 73 different studies to reach their conclusion that precisely 41 per cent of insect species are declining and ‘unless we change our way of producing food, insects as a whole will go down the path of extinction in a few decades’. In fact the pair had started by putting the words ‘insect’ and ‘decline’ into a database, thereby ignoring any papers finding increases in insects, or no change in numbers.

They did not check that their findings were representative enough to draw numerical conclusions from. They even misinterpreted source papers to blame declines on pesticides, when the original paper was non-committal or found contradictory results. ‘Several multivariate and correlative statistical analyses confirm that the impact of pesticides on biodiversity is larger than that of other intensive agriculture practices,’ they wrote, specifically citing a paper that actually found the opposite: that insect abundance was lower on farms where pesticide use was less.

They also relied heavily on two now famous recent papers claiming to have found fewer insects today than in the past, one in Germany and one in Puerto Rico. The first did not even compare the same locations in different years, so its conclusions are hardly reliable. The second compared samples taken in the same place in 1976 and 2012, finding fewer insects on the second occasion and blaming this on rapid warming in the region, rather than any other possible explanation, such as timing of rainfall in the two seasons. Yet it turned out that there had been no warming: the jump in temperature recorded by the local weather station was entirely caused by the thermometer having been moved to a different location in 1992. Whoops.

Of course, human activities do affect insects, but ecologists I have consulted say local populations of some species are often undergoing huge changes, and that some species regularly die out in one location and are then regenerated by migrants. This is not to be confused with species extinction. The real evidence suggests that insect species are dying out at a similar rate to mammals and birds — which means about 1 to 5 per cent per century. A problem, but not Armageddon.

Curiously, 41 per cent cropped up in another misleading story the same day, 10 February. This is the claim that exposure to glyphosate, the active ingredient in Roundup weedkiller, increases the incidence of a particular, very rare cancer, non-Hodgkin lymphoma (NHL). ‘Exposure to weed-killing products increases risk of cancer by 41 per cent,’ said the Guardian’s headline.

Once again, this paper is not a new study, but a desktop survey of other studies and its claim collapses under proper scrutiny. According to the epidemiologist Geoffrey Kabat, the paper combined one high-quality study with five poor-quality studies and chose the highest of five risk estimates reported in one of the latter to ensure it would reach statistical significance. The authors highlighted the dubious 41 per cent result, ‘which they almost certainly realised would grab headlines and inspire fear’.

Full post (£)

 

Recent Popular Articles


We use cookies to help give you the best experience on our website. By continuing without changing your cookie settings, we assume you agree to this. Please read our privacy policy to find out more.