Governments losing against Coronavirus: it's no coincidence

“One of the things that makes novels less plausible than history, I find,” says a character in Tom Stoppard’s play Night and Day, “is the way they shrink from coincidence.”

We do try to reject coincidence. “This happened after that other thing happened, so it must have been that other thing that caused this one.” Post hoc ergo propter hoc, the Romans used to say, after that so because of that.

And yet there really are things that are simply coincidence, in the sense that they happen together, or after each other, purely by chance.

London had a population climbing towards seven million over the decade I was a student there (yep: I was once the proud owner of ten University of London student cards). And yet I hadn’t been there many years before I had my first experience of meeting people I knew, without arrangement, on the street. It happened to me several times over the period I lived there.

What? The chances of meeting any particular person out of 6.5m is, naturally, 6.5m to 1. And yet I several times met individuals I knew?

We are not intuitively good at handling statistical reasoning, as the psychologist and economist Daniel Kahneman has pointed out. I had a wide network of acquaintance, which immediately increased the chance of meeting any one of them by chance. And it was not as though I only ever crossed the path of one other person in any one day – on the contrary, commuting from home to work or to college brought me close to hundreds or even thousands of people (hence the difficulty of maintaining social distancing today).

Take that into account, and the probability of meeting someone I knew was significantly higher than I might have imagined. Certainly, high enough for it to be perfectly reasonable that it should happen several times over a ten-year period. Particularly when you consider that most frequently the encounters were in places which act like people-funnels, such as the main railway stations.

In fact, the most surprising such event I witnessed was just last year, as I was catching up with a Turkish friend, Muharrem, who was about to enter Victoria Station ahead of me. I found him in lively conversation, in Turkish, with a woman. After she had left, Muharrem explained to me, “she’s one of the most famous novelists in Turkey, and I met her when I was working at a library where she came to give a talk”.

Despite this kind of experience, we do shrink from coincidence. Indeed, that shrinking becomes a weapon in the armoury of people trying to build certain kinds of argument. They might, for instance, say:

“My friend Sally took that medication. And exactly two years, five months and 17 days later, she had a series of terrible migraines. Coincidence? I think not.”

It’s a great debating trick, because it establishes in advance that only the most naïve and credulous people could possibly respond, “well, actually, yes, I reckon it probably is a coincidence.” As, of course, it is.

After all, Sally probably took other things than that medication before getting her migraines. If anything that happened before them could be regarded as a potential cause, might it not have been the chocolate she had the day before, or the overindulgence in wine, or the fact she drove through a polluted neighbourhood? After that so because of that isn’t an argument, it’s a sloppy and unreliable way of reasoning.

To establish that there’s more than chance at play, we have to show some causal connection between the two incidents.

To take a different example, naturally chosen entirely at random, consider lousy government. Faced with a major and fatal pandemic disease, a lousy government fails to take it seriously. It may attempt to put in place measures to protect its healthcare workers, and even promise to do so, and then take far too long and under-deliver. It may, rightly, identify track-and-trace as the best way to combat the virus, promise to put a system in place and then fail to meet its own deadline.

Or lousy government might just be far too glib in attitude. It may be scornful about simple and effective measures like wearing masks. It may pay lip service to other measures, such as social distancing or quarantining, while flouting them itself and turning a blind eye to infringements by friends or supporters. It may be halfhearted about its lockdown measures and end them too soon.

Now which governments around the world have been most guilty of some or all of these kinds of behaviour?

Why, the governments Trump’s United States, Johnson’s UK and Bolsonaro’s Brazil. Which makes it fascinating to look at the latest figures published by Worldometers. Imagine which are the governments with the highest numbers of deaths due to Coronavirus:

The three nations with the worst Coronavirus death totals
Note that Johnson's UK has the worst death rate per million

A coincidence? I think not.