Let’s begin by imagining two simple scenarios for coronavirus.
A: each case infects 2 others in a week.
B: each case infects 3 others in five days.
I’m making these up, but bear with me. Assuming each case infects this many and no more, and is infectious only for this long, after 10 weeks, scenario A has clocked up about 2,000 cases. B is heading rapidly to 10 million.
Such is the power of repeated multiplication that A also approaches 10 million about three months later. But it would, at least, give us valuable time—and could imply a lower peak.
An inch and a mile
What’s striking is how differences that might seem marginal at the outset can entail wild divergence. It’s reminiscent of Chaos, and a warning to those who think they know how this will play out: uncertainty needs only an inch to take a mile.
Is either scenario likely? I’ve no idea. No “real” model would be as neat as my formula—that much we can say, as there’s plenty more than an inch of uncertainty about how any scenario will unfold. And that’s the point: every uncertainty could make any of us hopelessly wrong.
Our fallibility isn’t much helped by trying to take control of events. Sure, we can change our behaviour with things like hand-washing and self-isolation measures, and try to disrupt the virus. To some extent, we’ll succeed. Experts who can tell us how to go about it will be invaluable—I’m not for a moment claiming they know nothing at all.
But how the public will react, and with what effect, is hard to say. For instance, we might not do as we’re told. Shut down our sporting events and we might go elsewhere. We’re both unclear what the virus will do and unclear what people will do.
If you’re not feeling uncertain yet, you should be—because it gets worse. Estimates of how many will die range from less than 1 per cent of those infected to over 3 per cent—a difference of well over half a million people if half the UK population falls ill.
The challenge for public health
The uncertainties, in short, are huge. Playing with them can seem grotesque but it helps us see what the public health authorities are up against.
They must devise pragmatic, effective, publicly-acceptable plans—in detail—in the dark. There may be massive uncertainty, but we somehow expect them to be right.
Danger lies in every direction: accusations of destructive panic if the worst doesn’t happen, accusations of fatal complacency if it does.
Worse, we also fill the knowledge void with presumption and beliefs of our own—which we’re all too happy to declare. That’s how some of us can say things like “calm down, flu kills too you know!” while others wear face masks and stockpile loo roll.
All of which is to say that on top of the uncertainty out there, there’s more uncertainty in here, in the variety in our own heads. No number stipulates our reaction. That would be true even if we knew which numbers to use. When they’re this uncertain, it’s an open house for every psychological disposition from terror to indifference.
Justifying those attitudes is easy. Just cherry-pick the data. For example, within the average mortality, there’s big variability for different groups. The young and healthy might face a risk of less than 0.1 per cent. The risk to the old and ill might be 100 times greater.
Or you can choose your level of focus. At one level, there’s the whole population, among whom deaths seem certain, possibly in the tens of thousands or more. There could also be societal effects like a reeling health service and a shrinking economy.
At another level there’s the risk to a representative individual—you, maybe—for whom you could perhaps optimistically assume a 25 per cent risk of infection (because we’re sure to get on top of it before long) combined with a 0.05 per cent mortality rate. This puts your personal risk of death at only 1 in 8,000.
If that’s not reassurance enough, because the ‘1’ in that ‘1 in 8,000’ draws the eye (and as every lottery player knows, it could be you), simply reframe the same number as a chance of survival instead of death: which in this example is a 7,999 chance in 8,000 of being just fine.
Even if you are one of the old or ill—as I guess I might be, by virtue of a rare and weird heart disease—you can always comfort yourself by saying that risk is relative and so, relative to the everyday background risk that you already live with, another 10 per cent is not so bad.
Strange as that may sound it’s close to how I feel. Though if in similar circumstances you feel terrified, you also have a point.
And then there’s the judicious comparison, in which you say that 1 per cent mortality risk, if half are infected, is only a little more than the average lifetime risk of death on a British road: “So,” you might think, “what’s the big deal?”
Equally, you might think: “What? A whole lifetime’s risk on the roads from a fortnight’s illness? That’s terrifying. Have you ever tried to cross the A4251 to school in the morning?”
A temptation when no one knows as much as we want them to is to find answers in our own instincts. Maybe, as the historian Daniel Boorstin once said, that’s because we don’t like to leave our imagination unfurnished; we choose furniture to match our décor.
There is an alternative, though. We could just accept that when we’re deeply uncertain, we really are deeply uncertain. Which means, when it comes to knowing what will happen, that we should aim for humility—and show tolerance to those expected to get it right, because they’re at the mercy of events too.
The road ahead
After that, we have to decide what to do. And here David Spiegelhalter, my colleague at the Winton Centre for Risk and Evidence Communication sums it up: “In the face of deep uncertainty, it’s completely appropriate to try and be resilient to extreme events, and completely inappropriate… to portray these possible extremes as likely to happen.”
I’m not wearing one (yet) but your face mask is fine by me. Just don’t tell me this thing will get me, and don’t tell me I’ll be fine either. I much prefer the uncertainty—which is, at least, reliable.