Published:

33 comments

Will Dry is a student from Chichester.

Conservatives think so little about conservatism that they have failed to notice that a certain breed of it has performed remarkably well in the 21st Century, and will be increasingly required in the remainder of it.

The breed can be summarised as two broad beliefs. First, that in the face of the immense epistemic limitations that both individuals and societies face, a key aim of governments should be to try and at least prevent catastrophically bad outcomes.

Second, thathe best way for them to do that is to trust and use things that have historically proven to be effective.

The antitheses of this view are either a misplaced arrogance, or paralysis, generally arising from the mistaken belief that what an individual politician or programme sees and understands is all there is to see and understand – confusing absence of evidence with evidence of absence.

Policy makers and financial elites often flatter themselves that they understand complex systems so well that their interventions and investments are bound to yield the results they desire, and such hubris often ends in catastrophe. Think of the Communists and the Great Leap Forward, the Neocons and Iraq, or Wall St banks and the sub-prime crisis.

A question that hasn’t been given enough attention is: why did our chief scientists and medical advisers fail to give good enough advice earlier this year? If we don’t answer this question, we risk making the exact same mistakes the next time a Covid-sized challenge drops down from the sky.

I think the fundamental answer to this question is that they failed to comprehend and deal with the uncertainty surrounding the disease.

Look at how SAGE reacted. At its first meeting on January 22, it concluded that there were “no practical preventative actions that [the government] might undertake.” It then proceeded to dismiss a whole host of interventions that we would later adopt because, at the time, they didn’t have any evidence these interventions would work.

They dismissed the idea of screening new arrivals from China. On February 3, they dismissed the idea of banning direct flights from China, saying the extra five days it would buy the NHS to prepare would be of “limited value.” Just two days later, they’d found some evidence to the contrary – “a delay in the arrival and spread [of Covid] … would be beneficial for improving the NHS readiness.” Around this time they were also agreeing that “shutting down public transport and suspending public gatherings would probably be relatively ineffective”, and that “there is limited to no evidence of the benefits of the general public wearing facemasks.”

The common thread behind all of these errors is a conflation of absence of evidence with evidence of absence. SAGE seemed to think only the interventions that it knew worked were worth doing. The trouble with this approach is that peer review studies of Covid, and the effect of masks, buses et al, were a bit sparse in January and February!

Nassim Taleb has wrote that in times of uncertainty it’s far better to focus on payoffs rather than probabilities. People naturally do this. Everyone on SAGE is far ‘smarter’ than my Dad, a Yorkshire lad who won’t mind me saying his O Levels resemble an alphabet soup stripped of any As, Bs, and Cs. But in January, he, like lots of people, was buying masks, hand sanitisers, and a bit of surplus soup. He wasn’t focused on the probability of the shops not having any soup in a few weeks, but the negative payoff from not having any soup if we were going hungry. (I confess I shamefully mocked him after his first trip. More fool me).

The exact same principle should have been applied at the national level. Instead of confidently asserting that “there is limited to no evidence of the benefits” of masks all the way to April, SAGE should have said: “We haven’t seen any evidence yet that indicates masks are beneficial. The worst that can happen if we recommend them is that people get clammy chins. The best is that they reduce transmissions and deaths. Given these payoffs, we should ask folks to wear them.”

If this catastrophe-averse conservatism has a leading voice, it’s the aforementioned Taleb. On the January 25, when the total number of global cases was fewer than 2,000, he co-wrote a paper warning governments that the combination of the virus and our increasingly inter-connected world represented a complex system that could deliver an extreme outcome. He said the best solution was to restrict mobility, between and within countries – to panic early so we didn’t have to panic later.

He didn’t need a complex multivariate model dependent on various fragile assumptions to come to this conclusion because, in his words, that if there’s signs of an avalanche at the top of the mountain, you don’t need a model to know it’s best to get off ASAP. Reminder: he was writing this was when SAGE was concluding there were “no practical preventative actions that [the government] might undertake.”

The purpose of this piece isn’t to play a blame game. I’m certain everyone on SAGE was working extremely long days throughout the early part of this year, and I’m grateful they were. It’s solely important because Covid will certainly not be the last challenge we face where there is considerable complexity and uncertainty, and the outcomes range from ‘okay’ to ‘very bad’ to ‘catastrophic’.

For example, the global debt-GDP ratio is at a record high. That’s fine when interests rates are low – but what happens if they rise? Professor Graham Allison of Harvard continues to place the odds of terrorists exploding a nuclear bomb in a major Western city at over 50 per cent in the next decade (Warren Buffett shares this fear – banning any of his insurance companies from offering insurance against such an event).

Some risks are even existential, as Toby Ord’s brilliant book The Precipice makes clear. He sums the collective risk of nuclear war, climate change, engineered pandemics, unaligned artificial intelligence, along with other threats, and places the probability of human extinction within the next 100 years at 16.7 per cent!

Given where we sit on the conventional (largely defunct) right-left spectrum, we should aim to institutionalise this catastrophe-averse conservatism in our government, civil service, and politics. A start might be a comprehensive review (with as many outside experts invited as possible) of our capabilities against all known potentially catastrophic/existential outcomes (no matter how low probability we assign them) with the aim of determining some low-cost actions we can take now to prevent them from occurring or to help manage them if they do.

Next, reform of the workings of government and the civil service is critical – I don’t think its hyperbolic to describe it a matter of life and death. The next time one of these challenges emerges, whatever it is, the experts’ strategic advice can’t be as abominable as SAGE’s was earlier this year.

The Government won’t get a medal from the media if it does this. But it should remember that neither would the FBI had they introduced security scanners at airports on September 10, 2001. They’ll probably get the vote of my Dad, though.

33 comments for: Will Dry: Small-c conservatism is the key to better disaster planning

Leave a Reply

You must be logged in to post a comment.