In the evidence he gave to the UK Parliament on 26 May 2021, Boris Johnson’s former chief advisor Dominic Cummings added to the impression informed commentators had already reached that the UK policy-making process in early 2020 that resulted in the first Covid-19 lockdown at the end of March was conducted in an atmosphere of panic and chaos.
It is true that it was in a matter of mere weeks, and in some aspects days, that policy was changed from one of ‘mitigation’ of the outbreak of the SARS-CoV-2 virus to one of its ‘suppression,’ even though suppression had to involve by far the most draconian government intervention in the life of the entire society in peacetime history. Mr Cummings’ criticism is that, but for that panic and chaos, the UK would have been able to respond to the outbreak with more extensive and therefore effective lockdown measures than it did.
We think it fair to say that Mr Cummings’ criticism, no matter how vividly expressed, is something of a damp squib. The main reason is that it has long been acknowledged by all involved that the epidemiological advice given to Government in early 2020 was based on highly imperfect information. The Government’s response has been justified by the pressures of operating in a perceived extreme emergency; it was predicted that 510,000 lives would be lost in Great Britain.
Mr Cummings’ criticism follows from his own acceptance of this prediction. But, putting aside those who maintain the arguable position that lockdown is not a justifiable response even to such an emergency, those who are capable of independent thought have from the outset been unable to dispel from their minds a lingering anxiety about the scale and nature of a suppression policy based on such imperfect information.
However, imperfection of information, no matter how marked, is by no means the main reason that the Government’s policy must be regarded, not as insufficiently drastic, as Mr Cummings would have it, but as a disastrous overreaction. The problem was not at all that the 510,000 figure was only accurate to within margins of error acceptable in the circumstances. It was that this figure was a fantasy scare number that had no possible reference point in the empirical world.
The most important document produced during the process of policy formulation that led to lockdown was a 16 March 2020 Report by the Imperial College London COVID-19 Response Team entitled Impact of Non-pharmaceutical Interventions (NPIs) to Reduce Covid-19 Mortality and Healthcare Demand. Formed in January 2020 to provide advice about the possible pandemic spread of an almost certainly newly emergent respiratory virus, SARS-CoV-2, the Response Team has had enormous global influence on policy. The Report led to a most rapid and extensive shift from a previously restrained, even relaxed policy because, having reviewed what was then known of the infectiousness and severity of the virus and its suspected presence in the UK, the Report predicted that the resultant respiratory disease, Covid-19, would cause 510,000 deaths.
A virus’ infectiousness is highly dependent upon its ability to pass from infected to susceptible individuals, and in the inevitable absence of a vaccine for (or other pharmaceutical interventions against) SARS-CoV-2, this is a virological matter of innate biology. But the rate of infection is also, of course, determined by the rate of contact between infected and susceptible individuals. In the case of human contagion, the rate of contact is a matter of human social interaction and government policy. The Report was considering what could be done to reduce the rate of infection by ‘non-pharmaceutical interventions’ which would reduce contact. The Report’s most important advice was to extensively limit human contact to suppress the virus. That advice was given in the full knowledge that it would require unprecedented draconian interventions into the economic and social life of everyone.
With these considerations in mind, let us look more carefully at how the Report presented the 510,000 figure. The Response Team began its statement of its results by saying that: ‘In the (unlikely) absence of any control measures or spontaneous changes in individual behaviour [a spread of infection would take place so that] we would predict approximately 510,000 deaths in [Great Britain] and 2.2 million in the [United States].’ It is essential to recognise that it was, to put it generously, extremely misleading to describe this scenario as ‘unlikely.’ Its description elsewhere in the Report as being the result of ‘do[ing] nothing’ was even more misleading. For there was absolutely no possibility that there would be no spontaneous changes in behaviour of the sort that would have taken place given an outbreak of, say, influenza or the common cold.
Once it was recognised that Covid-19 was a significant respiratory disease, extensive, society-wide, spontaneous mitigation, certainly including what the Report identified as ‘Social distancing of those over 70 years of age,’ would inevitably have taken place. Nor was there any possibility of the Government not taking some control measures, including steps to support such social distancing by, for example, enforcing conditions of entry to care homes. The scenario which the Report predicted of ‘an uncontrolled’ or ‘unmitigated epidemic’ resulting in 510,000 deaths was not merely unlikely; it was a scenario which could never possibly obtain.
The incredible point remains, however, that in producing the 510,000 figure the Report actually did model a set of empirical circumstances that have never existed and could never exist. We search for the correct word to describe just how troubling it is that this figure was presented as in some way connected with the empirical world, and indeed as an empirical claim of the highest importance. The error involved here has nothing to do with imperfection of information. It is instead a profound logical mistake.
The 510,000 figure, which turned the world on its head, was a fantasy number churned out by absurd modeling of a zero-probability event. The main way in which Mr Cummings, and in this respect he is a significantly representative figure, is going dreadfully wrong is that, when judging government policy as a response to emergency, he does not see that the reason why the SARS-CoV-2 outbreak escalated into the public policy catastrophe that it became was the fundamentally flawed Imperial College London modeling on which the government based its response.
[An earlier version of this article was first published as ‘The fantasy prediction of 510,000 deaths’ in The Spectator Australia on June 1st 2021.]Disregard-of-the-Empirical