January 13, 2021

Priti Patel, the Home Secretary, held a press conference on Tuesday, and entirely failed to answer a basic question: if the new variant is more infectious, how come the lockdown rules aren’t as tough as last spring’s?

Realistically, the reason for the weaker lockdown rules is that the Government continues to refuse to take difficult decisions until they are inevitable. But, if Patel were more nimble on her feet, there was a possible, plausible explanation open to her. And that is simply: we know more, now, than we did in March, so we can be more targeted about what we shut down.

For instance: the playgrounds are not closed this time. And that is, I expect, because we know that the virus, both in its old form and its new one, finds it much harder to spread outdoors. Dr Müge Çevik, a virologist at St Andrews University, details here that the risk was around 20 times higher indoors than outdoors. We didn’t have that sort of data in March last year, so we had to be more cautious. (I cannot begin to guess, however, what their reasoning is for allowing estate agents’ offices to remain open.)

Choosing our targets wisely is important. The “wash your hands” message has been drilled into us since the beginning, but the WHO has not found a single clear-cut case of transmissions by surfaces, and research in the Lancet considers it “low risk”. It is important now that we tell people about the risks of being indoors and close to people, rather than about washing our hands until they bleed.

On the same basis, there may be room for being cleverer about schools. British children have not been in their classes since mid-December (except that one mad Monday last week), and before that had only had a few weeks’ in-person teaching since March. The Government plans to revisit its lockdown measures in mid-February, but that seems enormously optimistic: despite the progress with vaccination, the NHS will still be under enormous pressure by then.

Here’s why. By mid-February, the Government hopes to have given almost 14 million people the first dose of a vaccine. That will include all the most vulnerable groups, including all the over-70s. That is, obviously, extremely good news. (If it succeeds.)

Vaccinating those vulnerable groups will hugely reduce deaths. The actuary Stuart McDonald has put the numbers in, and estimates that if we vaccinate those groups on time, Covid deaths recorded by the ONS will drop by about 90% by mid-March. Deaths are hugely concentrated among the elderly.

But there’s a wrinkle. Deaths come down dramatically, and that’s great. But hospital admissions will come down rather less – by about 60%. And intensive care admissions will come down less still – by about 30%. That’s because, as two health economists writing in the Conversation point out, most ITU admissions are younger, with a median age of 62; only about a third are in the over-70 groups who will receive the first vaccines. 

And that means that pressure on the NHS will probably remain very high for a few months. It may start to come down as the usual winter concerns fade away, and as the weather gets warmer Covid might start to fade. But it may not, or may not be enough.

There are about 15 million people in the UK aged 50-70. At two million jabs a week, that’s another seven weeks of vaccinations; plus, because it takes time for the vaccine to be effective and because ITU cases are in hospital about twice as long, it would be weeks later still until the pressure on the health service really returns to normal. Say 10 weeks to have full effect; 10 weeks from mid-February is the end of April.

The current school closures are meant to be reviewed at the February half term, which in my borough is 12 February. But the health economists mentioned above say “The proposed date of February 22 for easing lockdown and opening schools seems optimistic,” and at least one academic I’ve spoken to agrees. The vaccinations just won’t have got that far. Even by the Easter holidays, which start 31 March, we will still be vaccinating the 50-to-70-year-olds, and the ITUs will still be full.

What I haven’t mentioned, of course, is the lockdown. Whether it’s been enough to reduce R below 1 is not clear. There are some early indications, as I write, that the spread has come down – the number of cases each day and the percentage of tests coming back positive appears to have plateaued, at least, and may be falling. 

But since it’s so close to bubbling over, even under this severe lockdown, the fear is schools could be shuttered until Easter or the May half term. I even had a panicked text message from an academic who works on paediatric public health, wondering whether they would have to stay closed until September.

This is where, I think, we can afford to be a bit cleverer. First, let’s talk about whether schools spread Covid. The London School of Hygiene and Tropical Medicine’s modelling of the new variant, which was key to this most recent lockdown, suggested that without closing schools, it would be impossible to get the R value below 1. But as I wrote at the time, the schools themselves seemed to have a relatively small impact on the total number of deaths.

Suggested reading
What's the real cost of sending kids back to school?

By Tom Chivers

So: what’s the evidence of schools’ impact on transmission? It is, I am afraid, hugely uncertain. I spoke to Dr Andrew Lee, a public health doctor at Sheffield University, and he pointed me to a recent preprint with the highly relevant title “Do school closures reduce community transmission of COVID-19? A systematic review of observational studies”. And the answer is, essentially, we don’t know.

The review looked at 10 studies, which all examined whether closing schools reduced the spread of Covid. Some of those studies found it had a big effect; some found a smaller one; some found none at all. “Our results are consistent with school closures being ineffective to very effective,” say the authors, unhelpfully. But there is a bit more information we can glean. 

First, not all those studies are created equal. “We know the quality of the studies are mixed and the findings are mixed,” says Lee. And those two facts are related. “The studies where the risk of bias is least, the more robust studies, tend to find no effect. The ones where they found rates of transmission were affected didn’t control for other measures.” 

Lee himself “is more inclined to believe” the studies finding smaller or zero effect, “because children aren’t effective spreaders of the disease”. He points to reviews of the evidence by the Royal College of Paediatrics and Child Health and McMasters University that suggest low levels of spread among schoolchildren.

Çevik broadly agrees. “It’s a very difficult question,” she told me. “We need to accept that studies looking at the effectiveness of non-pharmaceutical interventions are going to be at risk of confounding.” But, she says, the evidence, as best we can make out, is that children are not major drivers of disease. Transmission in schools seems to be only rarely from child to child, or from child to adult; it’s mainly adult to adult, or adult to child. 

We don’t know exactly why this is. Lee speculates that children, being smaller, breathe out smaller volumes of air, and because their immune system seems to repress the virus more effectively, keeping their viral loads low. Çevik notes that children are less likely to be symptomatic, and asymptomatic cases are less infectious. 

Secondly, and more importantly, so far we have lumped primary schools, secondary schools, colleges and universities together, and when we as a society have discussed school closures, it’s largely been all-or-nothing. (Nurseries are currently open, but otherwise it’s key workers and vulnerable children only.) The LSHTM model I mentioned above had them closed or open, altogether.

But children are not all the same. “A primary-school child is not the same as a teenager or a college student,” says Lee. “The risk is much higher with older age groups.” Çevik agrees: “Susceptibility to infection increases with age. After 15 the transmission dynamics are much more similar to adults.” Her own research backs this up. Smaller children seem to be at much less risk of getting and spreading the disease.

So it seems reasonable to say that the risks of opening primary schools, secondary schools and universities are very different. Lumping them together in one big thing called “education” misses a lot of important detail.

It’s also worth noting that the costs of closing primary schools are greater, on every dimension. Small children require much more parental supervision, making home-schooling far harder for working parents. They are more vulnerable to abuse. And the direct educational impacts are worse. So should we reopen primary schools before the rest of the education system, and if so when? 

A few things to consider. First, it’s worth noting that I have primary-aged children and am hardly impartial about this. And it seems that the new variant does not disproportionately affect children, as was originally feared. But it seems to be better at spreading under all circumstances, by somewhere around 50%. So even if children are still a smaller part of the equation, relatively speaking, than adults, they could still be important. “If a contact has 5% chance of transmission,” says Adam Kucharski, a mathematical epidemiologist at the LSHTM, “and you have multiple contacts, and so do your contacts, then you have yourself an epidemic.” It doesn’t matter if other kinds of contacts would be worse; what matters is the absolute numbers.

We can take steps to reduce the risk for teachers and others when they do go back – enforcing masks, ventilating classrooms, using more municipal buildings and recruiting more staff to minimise class sizes; the National Education Union has some ideas here. And as the year goes on teaching will be easier: in the summer, the virus will probably have abated and the weather will be warm enough to allow outdoor classes. Shortening or moving the summer holidays might be a good idea, although I dare say there are logistical challenges.

With the vaccines on their way, the cost-benefit equation changes: the costs of keeping children out of school indefinitely are obviously much higher than keeping them out for a few weeks. But it may be that the benefits of closing schools, at least in the case of primary schools, are fairly small as well. 

As Patel should have said: our lockdown rules must be as targeted as possible. In the case of schools, that means we shouldn’t treat the whole education system as a single undifferentiated mass. University students and older secondary-school children are essentially adults, in terms of their ability to spread the disease; they’re also more capable of managing without in-person teaching. Primary-aged children seem to be less of a concern from a virological point of view, and keeping them at home has more costs both for them and for wider society.

But in the end, we urgently need a concrete discussion of what tradeoffs we’re willing to make: as cold-blooded as “how many dead people are we willing to accept for a thousand pupil-years of in-person schooling?” It can’t be zero. And if primary schools really do end up being closed until September, that seems too high a price to pay.

Join the discussion


  • January 20, 2021
    Its a metaphor - if you set off a chain of infections there's a grannie at the end of it who dies. And 80% behind restrictions in a democracy kinda does mean they are right. What it certainly doesn't mean is the small minority is right Read more

  • January 19, 2021
    You mean, you can't understand what point I'm making. Read more

  • January 19, 2021
    Can you show point me any grannies killed by bullets? Use emotive arguments and I will use emotive arguments back. That 80 per cent of people are behind the restrictions doesn't mean they are right. Read more

To get involved in the discussion and stay up to date, become a registered user.

It's simple, quick and free.

Sign me up