The UK government’s plan to begin gradual school reopening from 1 June has provoked huge controversy. Some teaching unions have opposed it and have been joined, in a significant intervention, by the British Medical Association. Local government leaders in parts of Manchester, Liverpool, Hartlepool and Tyneside have indicated that they do not intend to comply, and none of the devolved nations appears willing to follow England’s lead either.
To counter the growing rebellion, prominent cabinet ministers have been deployed in the media, asserting their confidence in both pupil and teacher safety while simultaneously invoking a quasi-military “duty” on the part of teachers to return to the classroom. As ever, we’re told the government is following the science (while appearing reluctant to submit it to wider scrutiny).
The debate about school reopening provides a backdrop against which to examine these claims to scientific legitimacy. It’s important to draw a distinction between two types of science going on here. One is theoretical, meaning it is based on models that attempt to predict how things will behave. The other is empirical, meaning it is based on observation and experience of how things have actually behaved.
To begin with, facing a novel disease like Covid-19, there was no empirical data. All that was known was: it is a coronavirus that can cause a potentially fatal respiratory distress syndrome. Countries such as Taiwan, Singapore and South Korea based their strategies on the nearest empirical “fit” – their experience of dealing with past, serious coronavirus diseases such as Sars. Testing, tracing and isolating (TTI) cases and contacts was at the heart of their approach from the outset, as were border controls to limit importation of new cases. Countries such as Britain relied instead on theoretical modelling to inform their response, but these models were based on influenza viruses (for decades, the assumption had been that the next pandemic would be a flu virus).
As time has gone on, the UK modelling has been adjusted to try to incorporate the empirical data that has become available about Covid-19. A seismic shift occurred in mid-March, when real-world data, including from Italy, was fed into the model the government was using to inform policymaking. The resultant, dramatically upscaled predictions for deaths and hospitalisations led to the sudden, panicked introduction of lockdown and the breakneck drive to expand ICU capacity, establish Nightingale facilities and empty hospital beds (which had the effect of discharging coronavirus-spreading patients into care homes).
Britain has suffered an appalling level of mortality, but the shocking predictions of late March, showing NHS capacity being exceeded many times over, never came to pass. The modelling must either have seriously overestimated the strength of the surge, underestimated the transmission-dampening effect of lockdown, or failed to anticipate how many fatal cases would occur in nursing homes and the community rather than in hospital – or most likely a combination of all three. The conclusion that ought to be drawn from this is quite how poorly the theoretical science is performing, yet the government continues to follow it seemingly without question – something that prompted the former chief scientific adviser, Sir David King, to establish a scientific advisory group in order to give voice to alternative perspectives and approaches.
How does this affect the question of schools reopening? At time of writing, there are just two things we can say with certainty about Covid-19 in children. First, out of every 100 symptomatic patients, only two will be children. We don’t know whether this is because children are less likely to contract the virus, or whether they acquire it at the same rate as adults but relatively few of them develop symptoms. We also don’t know whether, or for how long, children are able to transmit the infection. The second certainty is that, notwithstanding a very few tragic cases, the infection rarely causes serious disease in younger age groups. Insofar as anything in life can be said to be safe, sending children back to school will be safe for them. But whether it will be safe for teachers and other adults in the school environment, and indeed for the communities into which pupils will return at the end of the day, is currently unknown.
We know from government pronouncements in early March that its modelling predicted that school closures would not make a significant difference to transmission rates. But a group led by the Norwich Medical School has evaluated the outcomes of different lockdown measures imposed in 30 European countries. While girded with caveats, their conclusion was that school closures were the most effective measure in suppressing transmission.
The widespread resistance to the early reopening of schools has its roots in our demonstrable overconfidence in theoretical science to predict how Covid-19 will behave in the real world. Ministers cite the example of Denmark, which has successfully reopened primary schools in socially distanced format, to justify their policy. This entirely ignores glaring differences between our two countries. Denmark suppressed its outbreak through early, widespread TTI and border controls. Danish schools are reopening in a completely different and safer context.
There is now ample evidence globally as to how best to counter Covid-19. We should acknowledge that our theoretical science has failed us, and continue to further suppress the epidemic while emulating the systems that experience abroad has shown to be effective, including greater freedom for specific regions to modulate their response according to their local conditions. That would truly be following the science.
This article appears in the 20 May 2020 issue of the New Statesman, The Great Moving Left Show