28 October 2020

Supermodel or superforecaster? Why Caprice should be on SAGE


In a narcissistic world obsessed with self, the hardest thing to find is the outside view. A friend who will tell you hard truths is a treasure beyond price. And what is true of individuals is also true of nations.

That’s why the UK should put Caprice Bourret – businesswoman, model, and long-time reality TV personality – on SAGE, and let her have substantial influence over the pandemic response. Caprice proved her coronavirus forecasting chops on the Jeremy Vine show back on 16 March, where she advocated for a short severe lockdown and a big push for East Asian-style mask use by the general public. For this she got roundly scolded by the show’s medical “expert”, Dr Sarah Jarvis (a GP, not an epidemiologist), and near-universally mocked on Twitter by the usual list of overpaid media personalities. Months later, with lockdowns now a common tool for controlling the spread of the virus, and masks mandatory in shops and on public transport, Caprice emerges entirely vindicated.

The scientific literature on geopolitical forecasting offers some clues as to why the outside view is valuable, and why Caprice was able to offer it. From the US intelligence community’s forecasting tournaments, we know that successful forecasters tend to start with the base rate. When confronted with, for example, the US presidential election of 2016 – successful forecasters tend to start with the question that a newly arrived Martian might ask: “how often do Republicans win elections?” The answer is about half the time, and so the base rate probability for a Trump win is around 50%. This is then refined on the basis of polling, the state of the economy, who most recently held the presidency, and the qualities of the candidates. And so a reasonable probability on election night gives Trump a good chance, perhaps around 30%. But the key is to start with the alien’s-eye, outsider view.

Likewise, when trying to find the best way to limit deaths from Covid-19, pandemic control methods tried and tested across the centuries – like quarantines – are a good place to start. As Caprice rightly realised, another very reasonable place to start is to ask what worked in nations that, encountering Covid-19 a few weeks ahead of the UK, managed to control the virus and keep fatalities in the single figures (Taiwan), hundreds (Hong Kong, South Korea) or low thousands (China, Japan). Although some avoided national lockdowns (Taiwan, Japan), all were extremely vigorous in using travel restrictions and quarantines to stop household transmission and prevent outbreaks being seeded by incoming travellers.

By contrast, the UK’s initial pandemic response was derived from its own pre-existing influenza pandemic plan, mechanically followed with little to no reference to events elsewhere, until soaring cases and growing political pressure forced the Government into national lockdown. Just 12 days away from lockdown, the Deputy Chief Medical Officer, Dr Jenny Harries, justified the Government’s refusal to ban mass gatherings in remarkably British exceptionalist terms, referring to her faith in the UK’s “excellent modellers”.

A Martian might perhaps capriciously wonder whether it was really likely that Britain’s epidemiologists were so much more proficient at modelling than those of Taiwan or Japan. In the same interview, Dr Harries rubbished the idea that the general population should wear masks, bizarrely claiming that this would do more harm than good. Given the observed trajectory of the virus in Asian countries, a Martian would likely have concurred with Caprice that masks were clearly not doing any harm, and were perhaps extremely useful.

Unusually successful geopolitical forecasters are not clairvoyants: in large-scale tournaments, they normally do about 10-30% better than the crowd. Caprice clearly underestimated the prevalence of Covid-19 in Britain as of 16 March. If she had realised just how widespread the virus already was, she would have known that a two-week lockdown would never suffice to eradicate it. Had she been able to assess the risks entirely accurately, she likely would never have been in the TV studio at all!

Everyone is wrong some of the time to some degree: the key is to be consistently less wrong than the rest. The cost of not locking down when Caprice said we should may well have been at least 10,000 lives. Relatively small improvements in accuracy can yield huge benefits in a crisis.

But the meltdown of Britain’s testing infrastructure meant that everyone was flying blind and relying on guesswork: the Chief Scientific Adviser’s public estimates of how far the UK was behind Italy were consistently over-optimistic, out by a good two weeks. In the presence of radical uncertainty, estimation of key metrics like the number of infections and the rate of reproduction should be based on a very wide range of guesses from diverse sources. The higher the uncertainty, the less we can or should rely solely on expert consensus. Everyone – including former supermodels – should be encouraged to have their say.

Caprice’s interesting career offers a few intriguing reasons to think that she might be one of those rare people who are consistently 10% ahead of the curve. After her big break in modelling, she was one of the first celebrities to realise the commercial power of her own brand, entering into a licensing agreement with Debenhams for a lingerie line all the way back in 2000, before buying out the license to launch her own business in 2005. Today, By Caprice largely sells luxury bedding across the world. Celebrity licensing deals are common these days, but were not at the time: ex-models running their own successful businesses are still very rare. Staying relevant and in the public eye for 25 years is perhaps a clue that she’s good at cultivating her personal brand – frequent updating being another predictor of good forecasting ability.

I would guess that Caprice travels frequently for work and has friends around the world, and so was perhaps in a better position to know how far-flung nations responded to the pandemic than the UK’s domestic policy establishment. Niche policy subcultures like public health are probably especially vulnerable to insular groupthink, and would be most likely to benefit from challenging outsiders. Dominic Cummings’ attempts to hire “mavericks and weirdoes” for the benefit of the system have not always gone smoothly to date, but the idea is surely the right one. The outside view always needs to be represented, and we need people with good forecasting track records, who can express their views clearly, but at the appropriate level of confidence.

So put Caprice on SAGE, or give her a Cabinet Office Fellowship. The second wave still confronts us with more radical uncertainty, even though far more is known about the virus now than in March, and we will surely confront further crises marked by an absence of high-quality data in the years to come. At the very least, if you really need to communicate urgently to the general public, Caprice would surely guarantee that everyone tuned in to watch.

Click here to subscribe to our daily briefing – the best pieces from CapX and across the web.

CapX depends on the generosity of its readers. If you value what we do, please consider making a donation.

Jonathon Kitson is an independent researcher and forecaster.

Columns are the author's own opinion and do not necessarily reflect the views of CapX.