I have this rule, which is that if there’s a shocking statistic in the news, it’s probably worth assuming it’s false. You’ll be wrong quite often, but I think probably less often than if you took the opposite rule of automatically believing it.

With that in mind, do we really think that 70% of Leave voters and 58% of Remain voters (in England; slightly smaller majorities in Scotland and Wales) “believe violence against MPs ‘is price worth paying’ over Brexit”?

That is the headline finding of a poll by academics at Cardiff University. And it is correct that those are the numbers involved.

But have a look at the actual answers people could give. The question was “please tell us whether you think [violence directed against MPs] would be a price worth paying or not worth paying for Brexit?”

Among English Leave voters, 66% said “I see it as a risk but it’s worth it to take back control”. Not “we should attack MPs to get what we want”, but “it’s possible that this may happen, but I think the risk is worth the potential reward”. We don’t know from that question whether they think the risk is large or not; in a separate question, though, we learn that just 35% of them think violence against MPs “likely” if we leave the EU.

By analogy, I cross the road every day; I think that is worth the risk of being hit by a car, not because I don’t think being hit by a car is bad, but because I think the risk is low. I don’t think “being hit by a car is a price worth paying for crossing the road”. That may well be how respondents understood the question.

The Norwegian writer and fact-checker Doremus Schafer uses a different analogy on Twitter: “If I put if to you that ‘accepting refugees carries the risk of increased violent crime’ (whether by a minority of refugees or hate crimes against them/politicians) — would you answer ‘we should close our borders then’, or ‘we should still do the right thing and accept them’?” If you would say the latter, then you would be saying that the increased risk of violent crime is a price worth paying for accepting refugees.

(Schafer also points out that it’s not that Leave voters are saying they’ll riot: they’re saying that, if Leave wins, *Remain* voters might riot, while Remain voters are saying the opposite.)

There’s an ongoing theme here, which is that the media is constantly saying things about how Brexit is causing a spike in hate crime or whatever, which usually turn out to be overstated. I worry that by constantly telling us that everyone’s on the edge of violence and that the country is a tinderbox inside a gunpowder keg balanced on a landmine etc and so on, these reports might actually make it true. First and foremost, though, stick to the rule: if a statistic is shocking and in the news, start by assuming it’s not true.

---

Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.

TomChivers

---