We don’t know whether teachers are more likely to catch Covid

“Coronavirus rates among teachers are 333% above the average.”

Claims that Covid-19 infections among teachers are up to 333% above the average have been going viral on Twitter.

The claim comes from an article published in TES (formerly the Times Educational Supplement) and is not reliable as it is based on comparing two uncomparable datasets. 

It is possible that teachers may be at higher risk of infection than the general population, especially when schools are open and other workplaces are not. But other sources suggest that teachers may face an average, or lower-than-average, risk of infection. 

Where did 333% come from?

The 333% figure quoted in the TES comes from comparing data on infections among secondary school staff with infections among the general population collected by Leeds City Council for a six-week period up to 20 November last year. The comparison was made by the NASUWT teaching union.

The council itself noted that the prevalence rates among teachers and among the general population could not be compared accurately because they were compiled differently. The TES article included this caveat, saying: 

“[Leeds City Council] said this was because the local authority rates were based on confirmed PCR test results, while the [teacher prevalence was based] on self-reported cases.”

That means that a teacher who suspected they had Covid-19, but who had not received a positive test result may have been recorded in the data for teachers. But a plumber who similarly suspected they had been infected, but had not been tested, would not have been included in the dataset for community prevalence. 

Also, Leeds City Council noted that teachers may have been counted more than once, because their illness may have lasted through more than one week, while cases in the community may not have been.

These two differences combined could have meant that prevalence of Covid-19 among teachers appeared higher than if the figures were based solely on positive test results.

Another problem with using this figure to talk about all teachers is that it only came from one area, which may not be representative of the rest of the country, even if you ignore the comparability issues.

The TES article also included data from Birmingham City Council and Greenwich Council. We asked these councils whether their figures had issues similar to those indicated by Leeds. Birmingham City Council, said that the information appears to have been misunderstood or misinterpreted in the TES article. We’ll update this article if they or Greenwich Council provide us with any further information.

Are teachers more at risk?

NASUWT told us that, while it acknowledged the data on teachers wasn’t comparable with the whole local authority infection rate, the pattern of higher infection rates among teachers was reflected in other data it was seeing.

It pointed us towards calculations it had done for infections in Scotland, based on data published by Public Health Scotland in November on cases among people employed in the education and childcare sector. 

However, Public Health Scotland subsequently published a report looking at infections among staff following the return of pupils to schools in August 2020 and up to 6 December 2020. 

This found that, while the risk of infection confirmed by a test was around 47% higher for teachers than the general population during this period, this may have been affected by higher testing rates among teachers.  The report also found “no difference” in teachers’ risk of hospitalisation, compared with the general population.The Office for National Statistics (ONS) also published an analysis covering the period from 2 September to 16 October, showing that teachers were not significantly more likely to have tested positive than other workers. Although there was a high degree of uncertainty in the findings. 

New variants of the Covid-19 virus complicate matters further. It’s possible that children may play more of a role in transmitting the new variant, potentially putting teachers at higher risk than they would have been earlier in the pandemic. 

However, this may just appear to be the case, because the new variant emerged during the second English lockdown, when schools stayed open, allowing it to spread more easily among children in particular. 

In summary, the evidence on teachers’ risk from Covid is very uncertain. The results reported by TES are not reliable and there’s reason to believe the risk faced by teachers in October and November may have been different to the risk they would face in other situations.