Among the cohort of secondary students who took their GCSEs in 2017, there were 55,000 “unexplained” school exits between Year 7 and Year 11 that are likely to have been instigated by schools.

That is the striking finding of research published by the Education Policy Institute today, badged by the thinktank as the “most comprehensive analysis to date of unexplained pupil exits”.

So does that really mean that a shocking 8 per cent of pupils in England’s secondaries have been off-rolled? Well, it’s not quite as straightforward as that.

Read: 55,000 ‘unexplained moves’ raises ‘off-rolling’ concerns

Revealed: The pupils most at risk of ‘unexplained move’

Need to know: What is off-rolling?

Let’s go back to basics: off-rolling is the name for the illegal practice whereby schools unofficially try to remove pupils from their rolls for nefarious reasons, whether that’s to boost their GCSE results or manage pressures on their budget.

But as you might expect, it’s not always easy to identify cases of off-rolling or to quantify how often it happens.

Unlike formal exclusions, there is no requirement to record the reason why a pupil has left a school’s roll.

This means it’s hard to establish whether such removals are happening due to legitimate, parent-led decisions (such as a house move, or parents deciding to send their child to a school which they think will cater better for their needs), or whether it’s being instigated by schools.

To try to flush out the latter (which could include cases of off-rolling), EPI has come up with a methodology which tries to distinguish between school moves that have been driven by parental choice and those which haven’t.

When is a school exit ‘unexplained’?

Using DfE data, they have discounted official permanent exclusions, as well as school moves relating to:

• Pupils with parents in military service
• Pupils with Gypsy, Roma or Traveller ethnicity and pupils with any absences due to their family travelling for occupational purposes
• Pupils who move from any type of school into a special school (EPI assumes that these moves are likely to be decided with parental consent and in the interest of the pupil)
• Pupils who move to a school with a higher Ofsted grade
• Pupils who move to a different ‘lower super output area’ (i.e. a change of address which is likely to require a change of school)
• Looked-after pupils who are adopted
• Looked-after pupils who experience a change in their legal care status
• Pupils who join the school system at any point after Reception (EPI assumes these pupils are more likely to exit the school system, perhaps to return to the independent school sector)
• Pupils who speak English as an additional language who arrive after Reception (more likely to exit the system for reasons including migration out of England)
• Pupils who live on the Welsh or Scottish border in the term of their move (EPI assumes these parents could plausibly be taking their child out of school because of a preference for the education system in Wales or Scotland)

After removing these pupils, EPI is left with pupil exits that it says cannot be easily explained by the available data. For the 2017 cohort, there were 55,000 such “unexplained” moves.

Clearly, that number can only be a rough proxy - it’s likely to encompass many pupil moves that are entirely appropriate.

And you don’t have to be a data expert to find quibbles with the list of pupils which EPI discounted. In the case of Gypsy, Roma or Traveller pupils, for example, while these families are highly mobile, they are also a group which could be especially vulnerable to off-rolling (something which EPI acknowledges).

Significantly, as Leora Cruddas from the Confederation of School Trusts points out, moves to alternative provision are included in the “unexplained exits” where they are not driven by a permanent exclusion. Ms Cruddas says that a move to high quality AP is “often in a young person’s interests”.

But with EPI’s research flagging up that vulnerable pupil groups with poor education outcomes are at heightened risk of unexplained exits, it feels like it’s on to something. 

The analysis also identifies 330 schools with particularly high rates of unexplained pupil exits. That is a similar number to the group of 300 schools that lost significant numbers of pupils between Years 10 and 11 over a two year period identified by Ofsted (in separate research using a different methodology).

Boxing clever

Off-rolling - and whether a move is genuinely for the good of a pupil - is hugely contentious. The reason it is able to happen at all is because it is so hard to prove.

An added complicating factor for this report is that it’s funded by the NEU teaching union - an organisation which some people will suggest is not entirely impartial when it comes to the debate which swirls around off-rolling.

However, EPI has been skilful in its approach - in tandem with the release of its findings, it is consulting on the methodology it used.

The thinktank argues its findings raise questions about the scale of off-rolling - but it says it is just starting a conversation on the topic. 

Crucially, it has held back the really controversial stuff (the areas and multi-academy trusts with the most unexplained exits) for a second report to be published in the summer, after it has received feedback on its methodology.

The fact that so many education bodies have cautiously welcomed the research - while making clear that not all unexplained moves represent off-rolling - shows that EPI has boxed clever in this instance.