Claim that up to 135,000 children have not returned to school needs more context

A Sunday Times article, published on 3 October, claimed new figures show that between 95,000 and 135,000 children did not return to school in the autumn term. The Sunday Times said this figure does not include “normal absences” or children staying at home due to Covid infection—and that in many cases the pupils concerned “cannot be accounted for”.

We’ve since seen these figures replicated in articles published by the Sun, the Mirror, Birmingham Live, and Joe. 

While there is plenty of anecdotal evidence that many pupils have not returned to school since the start of term for a wide variety of reasons, the figures come with important caveats which weren’t included in the media reports. They are based on a simple calculation which uses attendance figures for just one day and can’t provide a reliable picture of how many pupils have been persistently absent since the start of term.  

How were the figures calculated?

The estimate of 95,000 to 135,000 “ghost children”, as the Sunday Times describes them in its headline, is credited to the Commission on Young Lives, a task force headed up by former Children’s Commissioner for England Anne Longfield. 

There’s no detail in the Sunday Times about how the figures were calculated, so we asked the Commission for more information. It sent us its calculations, which were based on government data which isn’t easily comparable and comes with its own set of caveats, as we’ll set out below. 

In order to reach its estimate of the number of children missing from school, the Commission used figures published fortnightly by the Department for Education (DfE), which give a snapshot of attendance on one single day. 

These cover England only, and incorporate both state-funded primary and secondary pupils. The data the Commission used was published on 21 September, and looked specifically at 16 September 2021. 

The DfE figures show that on 16 September, there was an attendance rate of almost 92%, meaning roughly 8% of pupils were absent. Some 1.5% were missing due to some reason connected with Covid-19, leaving an absence rate of about 6.5%. 

In autumn 2019, which the Commission used as a baseline for normal attendance in the autumn term prior to the Covid-19 pandemic, the absence rate was roughly 5%. 

This, the Commission told us, suggests that on 16 September 2021, the rate of absence was around 1.5 percentage points higher than would normally be expected in the autumn term pre-pandemic. The Commission then used this percentage to calculate a rough estimate of the number of missing school children on 16 September, by working out that 1%-1.5% of the approximate pupil population is 90,000 to 135,000. (It’s not clear why this range was reported in the Sunday Times as 95,000 to 135,000—we’ve asked the Sunday Times and will update this story when we hear back).

One day’s data can’t reliably tell us about persistent absence

The Commission says that it believes overall the Sunday Times feature did a good job of exploring the problem of children failing to attend school. But it also told us that there are a number of caveats to its figures which were used, the most important of which is that they are based on calculations from one day’s data, which cannot give us a reliable estimate for how many pupils have been persistently absent since the start of term. A spokesperson for the Commission said: “Many teachers tell us about children who have not returned to school since the lockdowns, however the current available data means establishing the exact extent of this problem is not possible. 

“This analysis attempts to highlight an estimated level of ‘unexplained absence’, and comes with a number of caveats—for example it is just one day’s data, and it does not record or estimate persistent absence.” Neither the Sunday Times article nor the reports that followed provided details of how the figures were calculated, or mentioned that they were based on a single day’s data. There are a number of other more technical caveats to the Commission’s calculation too, as not all the data sets used to come up with the figure covered the same groups. For example, the pupil population figures are from 2020/2021 and cover all pupils including those at private schools, whereas the attendance figures are for the 2021/2022 year and only cover state-funded pupils. 

Also, the autumn 2019 term figures and the snapshot Covid-19 absence figures can’t be exactly compared. While they both try to achieve a similar end—giving us more information about school attendance rates—their methodologies are different and they cover different groups of pupils. 

For example, the autumn 2019 term figures are part of the DfE’s pupil census, which is based on school register data and is processed over a longer period of time. Meanwhile, the Covid-19 pupil absence figures are directly reported to the DfE by schools, and are updated every two weeks.The Sunday Times told us it believes the figures used in its feature are “accurate”.

What other figures are available?

There is no current data which can give us a full understanding of how many pupils have been persistently absent—which is why the Commission came up with its estimate in the first place. And it is possible that the actual number of children who have failed to return to school in the autumn term falls within the Commission’s 90,000 to 135,000 estimate. But given the calculation is from a single day’s data, it’s not possible to say reliably based on that that up to 135,000 children have failed to return this term. 

The snapshot Covid-19 attendance figures are published every two weeks, and have been updated again after the Sunday Times article was published, on 5 October. These latest figures show that on 30 September attendance fell to 89.5%, with the proportion of pupils off due to Covid-19-related reasons rising to 2.5%.