The world according to SIR Ferguson
Pessimism in Public Health
Hello,
Just a quick remark. It seems to me that a re-occurring theme in the UK government’s modelling of the coronavirus’ spread appears to have been Pessimism.
In Imperial College’s original doomsday report (published March 16th 2020), Professor Sir Neil Ferguson and his team considered the impact of various Non-Pharmaceutical Interventions (NPIs) on the virus’ spread. And they did so burdened by gloom. For starters, the authors note that their modelling uses “plausible and conservative (i.e. pessimistic) assumptions” (p.5) about the impact of each NPI it studies. Then, their much publicised forecast of half a million deaths (p.7) within three months was made with the assumptions (which they recognise as unlikely) of zero government interventions and zero voluntary behaviour changes from the public (p.6). Though they shed the former in their modelling of NPIs, the latter appears to have been preserved1.
All of which more-or-less guaranteed that government would be presented with a scenario urgently demanding intervention.
Similarly, during an extraordinary exchange with the Spectator’s Fraser Nelson in December 2021, the chairman of SAGE’s modelling committee admitted that during the Omicron wave, his team had only presented government with worst-case scenarios that would ‘require’ (I feel the urge to quibble, but i’ll contain it) restrictions to inform their decisions. This, as Nelson’s twitter exchange highlights, included assuming that Omicron was as virulent as the Delta variant, contrary to evidence coming from South Africa.
The reason for this assumption, Medley said, was that:
We generally model what we are asked to model. There is a dialogue in which policy teams discuss with the modellers what they need to inform their policy.
[…]
Decision-makers are generally on only interested in situations where decisions have to be made.
In other words, Medley and his team only presented government with bleak scenarios that ‘demanded’ restrictions and so, like with Ferguson’s modelling, more-or-less guaranteed that the government would be biased towards heavy-handed intervention. Weirdly, in that exchange, Medley doesn’t seem to understand why that’s a problem - he doesn’t see the issue with only presenting government with the scenarios that would ‘require’ intervention, irrespective of the likelihood with the scenarios that ‘would not’.
As Nelson points out near the end of this article, this calls into question the quality of the advice that government has been presented with throughout the pandemic. Both of these men and their teams were influential in pushing the government to impose lockdowns, school-closures, and all the other science-following violations that we were subjected to. As a result, any serious COVID-19 inquiry needs to examine - what we can call - government science’s ‘methodological pessimism’. It needs to establish why this was deemed an appropriate attitude to adopt, and whether or not it should continue to be into the future.
Now, obviously I don’t have good answers to this, but I do have one or two preliminary thoughts.
The first is that a charitable explanation (i.e. one that assumes both good intentions and internal rationality) of Ferguson and Medley’s methodological pessimism follows from a particular version of the precautionary principle – one that we could call the ‘Better Safe Than Sorry’ version.
Say that you’re faced with a threat of uncertain severity - e.g. a novel variant of the coronavirus. You don’t know much about its virulence and if it will prove to be as bad as delta or no more vicious than a flu. On ‘Better Safe Than Sorry’, it is better to act as if the worst-case-scenario is true and to intervene heavily (or to model so as to encourage heavy intervention) such that you’ll be ready if it does come to pass. And if - as it turned out - the new variant is actually relatively benign, then your COVID-19 case count won’t suffer too much anyway and at least you took no chances of it doing so2. Better safe than sorry, right?
Note how ‘Better Safe Than Sorry’ is always biased towards intervention. As the Medley-Nelson exchange seemingly illustrates, relatively little weight is given to the respective probabilities of the ‘best-’, ‘okay-’ and ‘worst-case scenarios’, with the latter always being given the most weight in the decision-making process. This, in effect, means that, the decision-maker is de facto given two options: intervene and be sure to have low case-numbers or don’t intervene and risk high case-numbers.
This, I think, makes two assumptions, both of which are pretty contentious:
That, at any given point, a political decision-maker’s focus should be reduced to single things like coronavirus ‘case-numbers’ or ‘deaths’, and
That the costs of these heavy interventions are negligible.
Assuming that political decision-makers aim at fostering their subjects’ well-being, narrowly focussing on a single concern assumes that there are times when the components of that well-being (needs, desires, vulnerabilities, etc.) can all be placed along a single scale of importance, relative to a single concern. In the case of COVID-19, this was framed in terms of ‘essential’ versus ‘inessential’ activities. Depending on the decision-makers’ assessment of the threat’s size, you might find seeing other people’s faces, seeing elderly relatives, going for a run, buying a puppy, or having sex with someone you don’t live with deemed inessential relative to the goal of slowing the spread.
Unfortunately for Ferguson and Medley (and, in consequence, for us), human well-being simply does not work this way. There is no single scale. There are, of course, the basic necessities of subsistence such as water, food, and oxygen, but they are not sufficient for well-being. Beyond them, however, human variability becomes so marked that there can be no single scale, even within the same society. Activities that may seem perfectly trivial to one might form the basis of another’s well-being. To illustrate, consider the role of horse-riding in the following account of the father’s experience of the lockdown:
[Sara, an eight-year-old with cerebral palsy, epilepsy, and complex needs] did not see a physiotherapist, a paediatric consultant, an epilepsy consultant, an occupational therapist, an orthopaedic surgeon, an optician, or a GP for the best part of two years. She was deprived of the health benefits and pure joy normally afforded by her weekly sessions of hydrotherapy and riding for the disabled. Her mobility declined. Her mental health suffered terribly. Her seizures got to the point where they lasted so long that she was turning blue and choking. We have almost lost her several times.3
Though pony-riding may evoke frivolity, rich girls, hairbands, and jodhpurs to some, the reality is that for kids like Sarah, it sits at the heart of being able to live well and be healthy.
Even if COVID-19 was indiscriminate and hurt all demographics alike, and even if its infection fatality rate (IFR) was 25 or even 50 percent like for some estimates of Ebola, there would be no single scale that could do justice to the multiplicity of people’s needs and vulnerabilities4.
As such, what central decision-makers advised by Ferguson and Medley deem ‘safe’, relative to ‘COVID-19 case-numbers’ or ‘deaths’, may in practice be life-ruining and even deadly to many. Using the uncertainty of a particular threat to justify stripping people of their ability to manage their own needs and vulnerabilities as ‘Better Safe Than Sorry’ does is a bastardisation of the precaution5, and has unsurprisingly resulted in widespread and immeasurable harm6.
Which brings me onto the second assumption: that harms of the intervention are minimal or relatively negligible. This is the only charitable way that I can find of explaining Medley’s failure to present the ‘better-case’ scenarios that doesn’t collapse into accusing him of callousness. And it is, by now increasingly, demonstrably false – I won’t rehearse why here.
Instead, the point I’m trying to make goes back to an observation made by Friedrich Hayek in The Road to Serfdom7: when decision-makers impose a series of widespread, granular interventions according to their own scale of value, they are usually blind or insensitive to the costs that do not affect them. It is very easy to present increased restrictions as the ‘safe’ option when you can work from the office of a three-bedroom house in the ‘burbs, overlooking a tastefully chaotic garden. Just as it is easy to call for more masking when you do not work in a shop, or depend on therapy, or aren’t a victim of grievous sexual assault. What the decision-makers and their advisors consider to be a guarantee of safety according to their measure, may in fact translate as a compounding of misery and suffering for others.
As such, if it does focus on methodological pessimism, the inquiry also needs to look at the process of decision-making and advice in government8 and public health more broadly. It needs to look at why ‘Better Safe Than Sorry’ came to dominate and why only one or two concerns were seen as relevant to the question of well-being. And to my mind, it needs to propose far-reaching legal and infrastructural reforms that prevent this from happening again.
Frankly, I’m not optimistic.
In any case – until next time,
Max
Raising an uncomfortable question about the contemptuous suspicion that - often leftwing - academics have of the masses’ ability to self-govern…
Assuming that lockdowns work to suppress case numbers, which, to be clear, I think they probably do.
p.93, Cole, L. & Kingsley, M. (2022) The Children’s Inquiry. London: Pinter & Martin.
This is, of course, not a full argument but rather an illustration. I hope to provide fuller arguments of why needs cannot be arranged on a scale in a future post/essay.
Not least because few things seem more lacking in precaution than radically messing with a complex and somewhat self-regulating system like society, especially when faced with a complex problem.
There is an obvious counter-example to this: the war economy. When a nation is faced with attack, most – even Hayek? – think that it is appropriate for the government to intervene and re-direct efforts and activities into the war-effort. Why shouldn’t the same be true of a disease event? As it stands, I don’t have any good answers, but I do have some suspicions about the analogy with war and the war-economy. For starters, war doesn’t deal with an ‘enemy within’ like disease that risks poisoning our everyday interactions with suspicion and fear. And secondly, the war-effort doesn’t involve the mass atomisation of people like the lockdowns did. That said, it does require a rational reorganisation of our activities so it is a case that I will have to consider.
pp.68, Hayek, F. von (2001) The Road to Serfdom. Abingdon, Oxon: Routledge.
For those interested, Roger Koppl’s 2021 Public Health and Expert Failure is an excellent, critical primer.