March 12, 2020 - 11:55am

President Donald Trump has announced (sort of) a travel ban (sort of) from Europe (or parts of it). It’s attracted a lot of commentary about his motives: he’s covering his own back, he’s panicking, he’s applying his generic nativist rubber-stamp “build a wall” policy in all situations. What I haven’t seen much of is people asking: will it work?

In my last piece I quoted Dr Cassidy Nelson of the Future of Humanity Institute in Oxford, speaking at length on the 80,000 Hours podcast, which I strongly recommend. I spoke to her quickly this morning.

She says that immensely draconian travel bans in ideal situations — ending more than 90% of travel from the affected country — could delay the outbreak peak by about three weeks. That sounds good — we’re all keen on flattening the curve — but, she says, in this case, it’s highly unlikely to work.

First, that implies that there’s only one affected country. The new travel ban affects Europe. One of the worst affected countries, Italy, is in Europe, but the virus is all over the world. And 90% is an incredibly high level with vast economic impacts so it is very unlikely to happen. A 40% reduction would apparently slow it by about three days or less.

Secondly, it creates perverse incentives. If travel is banned from mainland Europe to the US, but not from mainland Europe to the UK and from UK to the US, then people wanting to flee to (or get home to!) the US are incentivised to fly to Heathrow first then home.

“I imagine you’d get people routed through Heathrow,” Dr Nelson says. “The US government is trying to get airlines to find out where people were in the last 48 hours but at the moment there’s no system in place. If you came from Italy to Heathrow and got a ticket to DC, there’s no way the airline will know.”

Thirdly and finally, it’s shutting the stable door after the horse has bolted. The US already has significant person-to-person transmission and is probably hugely undercounting its cases. It may be that as little as 1% of the new daily cases are coming from overseas anyway; if that’s right a ban can make almost no difference.

The World Health Organisation does not recommend travel bans. A 2014 systematic review found that they had “a limited impact” on flu epidemics; Covid-19 is not flu, but that’s OK, because a 2020 systematic review found little effect on other epidemics as well.

This isn’t to say that travel bans should never be implemented. The WHO study finds differences in effectiveness under different circumstances. But what that means is that there should be a cautious evidence-based approach. Perhaps in well-isolated small nations, or when potentially devastating diseases are still geographically confined, they might be useful.

But Trump’s travel ban will likely do little to curb the disease, and a lot to curb the economy. As Dr Cassidy says, a travel ban is big and visible: “it’s hard to have a [PR] win by saying we’re doing more tests or being more efficient; a travel ban looks like we’re being tough”. But it probably won’t help.


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

TomChivers