The almost sycophantic enthusiasm for the advice of so-called ‘experts’ displayed in some quarters of the press – notably The Guardian – has always seemed wholly misplaced to me. Do they mean the economists who failed to predict the 2008 recession? Or the parole-board psychiatrists who can’t tell a psychopath from a cycle path and regularly let out criminals to murder again? For the complex socio-economic and medico-economic events that are so important in politics, very few experts are reliable prophets.
The coronavirus pandemic has brought this problem into sharp focus. The expert upon whom the government mainly relies is Professor Neil Ferguson, head of the Department of Infectious Disease Epidemiology at Imperial College London. He is an advisor to the World Heath organisation, the European Union and the governments of the UK and the USA.
But there is a difference between eminence and competence: Professor Ferguson’s track record is distinctly chequered. On several occasions in the past he has made apocalyptic predictions which turned out to be entirely false. In 2001 it was his predictions about foot and mouth disease that led to the unnecessary – many claim – slaughter of millions of cattle. He also predicted that up to 150,000 people could die of bovine spongiform encephalopathy (BSE or mad cow disease) and its equivalent in sheep when fewer than 200 deaths from the human form of BSE have so far been recorded. And when researching the swine flu epidemic of 2009 he predicted that as many as 69,000 could die, when in fact only a few hundred did so.
At first Professor Ferguson did not challenge the government’s laid-back approach which prevailed early in March. But then, on March 16th he published a paper which predicted firstly that more than half a million people would die if nothing at all was done; secondly that this would fall to 250,000 people with the government’s then moderate restrictions, but thirdly, that the death toll could be cut to 20,000 by a stringent lockdown.
These conclusions were challenged by other epidemiologists, notably Professor Suneptra Gupta of an Oxford University team who published a rival paper a week later suggesting that a high proportion of Britain’s population might already be infected, conferring a herd immunity which would greatly limit the impact of the coronavirus. It later emerged that there is a long-standing rivalry between the Oxford and Imperial teams, springing from a bitter personal dispute of 20 years before, though one in which Professor Ferguson was not involved. This does not invalidate the integrity of either of the two studies, but it does emphasise that scientists have their quarrels and feuds like everyone else – another reason for mistrusting ‘experts’.
Since the total death toll rose to more than 10,000 by April 12th, with daily totals running at 700-800, it would seem that Professor Ferguson was right, but one is left with the feeling that this is more by luck than judgement and that, with the evidence of numerous recent pandemics to draw on, epidemiology should by now be a more exact science.
Behind all this, though, lies a deeper problem: an excessive reliance on mathematical modelling. Back in the sixties, Edward Lorenz, a mathematician and meteorologist at MIT found that very small changes in the input data fed into his computer-generated models could lead to very large changes in the predicted outcomes. This led to the publication in 1972 of a paper entitled : Predictability: Does The Flap of a Butterfly’s Wings In Brazil Set Off A tornado in Texas?
Computers have become enormously more powerful since then, but these epidemiological experiences strongly suggest that the same principle still obtains. When all this is over, perhaps we should look more closely at the reliability of climate change forecasts and challenge some of the more extreme findings.
Grumpy Greybeard 