High-profile academy saw year seven to nine pupil numbers drop by 15 per cent after chain’s takeover –but no comment from inspectors in Ofsted report

Pupil year groups in an academy shrank by nearly 10 per cent in a single year after it was taken over by a major chain –but the dramatic change went uncommented-on by Ofsted inspectors.

The Harris Federation, England’s second-largest academy trust, is celebrating after one of its more recently-taken-over schools, Harris Academy St John’s Wood in Westminster, central London, registered a “good” rating from the inspectorate.

This was a big improvement on the “inadequate” judgment handed to its predecessor school.

But, Education Uncovered can reveal, specific year groups within the school reduced by a total of 81 pupils between the last year of the predecessor school and the first year after Harris arrived.

This was nearly 10 per cent. And pupil numbers within individual cohorts continued to drop in the following year, meaning that there was a 15 per cent fall in the size of the three pre-GCSE year groups which remained in the school throughout the years 2017 to 2019.

Ofsted, which on occasion seems recently to have been seriously investigating schools with shrinking pupil rolls, does not comment on individual reports. But it told me in response that pupil cohort data would not be mentioned if inspectors had “no concerns” about it.

Yet the figures as revealed here are bound to raise fresh questions, not just about the rigour with which it is investigating pupil number changes, but also in terms of the consistency of judgments, since this school’s level of pupil cohort changes is actually higher than that of another Harris secondary, which was criticised for it in an inspection report published only four months ago.

The detail

Harris Academy St John’s Wood (HASJW) replaced Quintin Kynaston, which at the time was the largest state school in Westminster, in September 2017, five months after an Ofsted report placing the latter in special measures had been published.

I looked at the most recent three years of pupil cohort data for this school, based on official Department for Education censuses in January of 2017, 2018 and 2019.

This makes it possible to track the size of individual year groups in schools, as they progress in age. So I compared the size of the year seven group at Quintin Kynaston (QK) in January 2017 before it closed, to that of year eight in January 2018, after the school had joined Harris.

The size of year 8 in 2017 could also be compared to year nine in 2018, and so on.

The data show that, while there were 214 pupils in year seven in QK in January 2017, this had fallen to 192 in year eight in HASJW by January 2018: a drop of 10 per cent.

Meanwhile, the 220 pupils in year eight in QK in 2017 had fallen to 195 in year nine in HASJW a year later (an 11 per cent fall). There were 208 pupils in year nine in 2017, but only 190 in year 10 a year later (a nine per cent drop). And there were 203 pupils in year 10 in 2017, but only 187 in 2018 in year 11 (an eight per cent fall).

Overall, the year groups eight to 11 at the Harris school in January 2018 were 81 pupils smaller – or 9.6 per cent – than they had been when these students were in years seven to 10 at QK the year before.

Year groups continued to shrink at the school the following year, with numbers dropping off between 2018 and 2019, though not by as much as that initial fall. There was a drop of 39 pupils, or five per cent, between the 767 who were in years seven to 10 at HASJW in 2018 and the 728 who were in years eight to 11 the following year.

Overall, the successive reduction in cohort numbers in 2018 and 2019 meant that the number of pupils in years seven to nine at QK in 2017 – 642 – had shrunk by 15 per cent to 547 by the time the same year groups reached years seven to nine in 2019. That is one in seven pupils having left the school in these year groups, net, over the two years.

Two thirds of this drop came from the number of boys, rather than girls, shrinking. There were 332 boys at QK in years seven to nine in 2017; two years later, there were only 267 boys in years nine to 11 at HASJW. This is a staggering 19.6 per cent fall, meaning that the number of boys in these year groups had reduced by a figure of one boy in five.

Such large changes in cohort size raise questions not only as to where these pupils went, but as to the potential impact on key results indicators against which schools are judged, not least, still, in Ofsted inspections themselves.

What Ofsted said in its report on the school

Yet none of this was commented on in Ofsted’s just-published report on Harris Academy St John’s Wood.

The school got an excellent write-up from the five-person inspection team, with the only area for improvement being that inspectors criticised HASJW’s two-year key stage three, a finding which featured in media discussion on that issue in recent days.  

Inspectors found that “whatever their background, pupils go on to achieve very well in GCSE examinations and in a wide range of courses in the sixth form” and that “disadvantaged pupils achieve very well in the GCSE subjects they take, including English and mathematics”.

The shrinkage in individual cohort sizes between the two versions of the school helps explain why, as of the latest official census in January 2019, HASJW overall was 12 per cent smaller, at 1,235 pupils, than the 1,401-pupil school it replaced.

But there was no mention at all of pupil number changes in the Ofsted report. The only reference to pupil movements appeared to be a statement that “the school uses [two named] alternative providers to educate a few of its pupils full-time”.

What was the explanation for these changes? Comparison with other local schools

Asked for a response, Harris itself, which generally does not respond when I email them, again had yet to do so in time for my deadline for this piece.

But in the past, it has sometimes put shrinking pupil rolls at its other schools down to what it said was “turbulence” in London in terms of its student population, with parents often moving their children in and out of particular secondaries.

But a comparison with schools in this local authority shows that the scale of cohort number reduction at HASJW in the 12 months leading up to January 2018 was specific to that comprehensive.

The school actually went from being the largest comprehensive in the borough in 2017, with 1,018 pupils in years seven to 11, to being the second-largest in 2018, with only 954.

Of the 11 secondaries in the borough, the shrinkage of QK/HASJW’s year groups between 2017 and 2018, as described above, was double the rate of any other school: two other Westminster secondaries shrank by five and four per cent respectively, with two per cent the highest rate among the other schools.

The fact that all but one of the schools across this borough saw pupil rolls shrinking in individual year groups between 2017 and 2018 (see chart below) will be of interest to readers, and does raise questions as to why this might be. But, as pointed out above, QK/HASJW’s figures were particularly striking.

How did the school’s pupil characteristics change?

It is also possible to look at pupil characteristics data to check whether the large reduction in student numbers changed the balance of the school. For example, did it reduce the number of children eligible for free school meals, as would be expected if such pupils were over-represented among those leaving the academy.

Again, I used official DfE data, drawn from successive “school performance” statistics, to make this comparison.

These data do indeed show changes, with the percentage of HASJW pupils who were boys; who spoke English as an additional language; who were eligible for free school meals; or who ever had been eligible for free school meals, all dropping compared to the figures for Quintin Kynaston.

Indeed,  HASJW had the highest drop-off, between 2017 and 2018, of any school in the borough for the percentage of its pupils who had ever been eligible for free school meals, as well as far and away the largest drop in the number of its pupils overall.

That said, when I looked at all schools in Westminster across a number of indicators, HASJW was not always the swiftest-falling in terms of indicators which are often associated with deprivation. For example, Pimlico Academy, run by Future Academies, saw its percentage of pupils eligible for free school meals drop by 8.4 per cent over the two years. This was more than double the rate of the second-fastest changing school on that measure: HASJW.

Ofsted has been mentioning pupil roll changes in other reports

Ofsted has been talking of a crackdown on pupils leaving schools for the past couple of years now, and on occasion has seemed to act in individual reports on “off-rolling”: cases where a child is alleged to leave a school at the institution’s instigation, but without the school recording it as an exclusion.

To be clear, data analysis such as the above is not in its evidence of “off-rolling”, which implies a deliberate effort on the part of the school to remove children from its books. But it does invite questions as to why cohort numbers have changed so swiftly, and of the potential impact on a school’s published results.

What an Ofsted report on another Harris secondary said about this
 

As Education Uncovered revealed in September, Harris Academy Orpington, in Bromley, south London, became the first in the chain to be rated “requires improvement” after inspectors raised a number of concerns, including pupil departures.

That report stated: “Although the number of pupils who leave the school part-way through secondary school is reducing, it remains too high and is much higher than typically seen.”

Remarkably, however, Harris Orpington’s numbers, in terms of the totals leaving school “part-way through secondary school” actually appear better than those of QK/HASJW.

Pupils “leaving part-way through secondary school” must mean students who depart before year 11. Falls in pupil numbers between years seven and eight, eight and nine and so on over successive years, of course, will all be such pupils.

But on the statistical comparisons used throughout this piece, again Harris Orpington actually fares better than QK/HASJW.

Comparing the size of the year seven to 10 cohorts in one year with year eight to 11 the following year, Harris Orpington saw a drop in size of 3.7 per cent between 2017 and 2018, and 4.2 per cent between 2018 and 2019. Over the two-year period of 2017 to 2019, the 2017 year seven to nine cohorts had shrunk by six per cent by the time they reached years nine to 11.

But, for Harris Academy St John’s Wood/QK, the comparable figures were 9.6 per cent, 5.1 per cent and 14.8 per cent.

So, the latter school saw larger falls on every measure. But this received no mention in the Ofsted report.

Responses

I showed my statistics on the shrinking pupil rolls at HASJW to Ofsted.

It offered only the following in response: “Inspectors do consider pupil cohort data, including the number of students on roll and any significant fluctuations. If there are no concerns, then inspectors may not mention it in their report.”

The Harris Federation generally does not respond to requests for comment from me. It has yet to do so in this case.

 

Snap analysis

It seems remarkable that such large changes in pupil cohort data have not been thought sufficiently worthy of mentioning in an Ofsted report.

A 15 per cent fall in the number of pupils who were in years seven to nine at Quintin Kynaston in 2017, compared to those still in the school by the time they were approaching their GCSEs in years nine to 11 two years later is staggering.

As stated above, this is more than one in every seven pupils having left the school in these year groups over this period, although in terms of the actual numbers of children departing, that total may be even higher, since these are only net statistics, with any student arriving reducing the net leaver figure.

The obvious, and most important, question arising from this statistical investigation is what happened to these pupils: what were their destinations? The Ofsted report offers no insights on that, given that its only mention of any children possibly leaving was that reference to “a few” – not 81, then – using alternative provision.

Harris has said in the past that “turbulence” in pupil populations in London can explain rolls shrinking. But that would surely affect many secondaries and other local schools did not see this phenomenon that year.

It could also be argued that parents might have reacted to the former school’s special measures verdict from Ofsted by moving their children. But this seems unlikely, given that parents will not move a child mid-school lightly and that Harris, bringing with it a very successful Ofsted record, was to arrive so quickly to take the school over.

Did parents decide they did not want the Harris approach? This is possible, though again it sits oddly with the chain’s overall inspection record. So the shrinking cohorts here remain something of a mystery.

Whatever happened to these pupils, the inspectorate should surely also have been alive to the potential impact that big changes in cohort data could have on the effectiveness of the school, as measured by the inspection system.

For there are other potential implications for the success of a school – and thus how it might be judged by an inspectorate – from a large change in the number of pupils such as this.

Ofsted’s report on Harris Academy St John’s Wood includes lavish praise for the school on behaviour management. But did this get easier with the school having seen such a large number of students depart?

And did the leaving students, whose departures, as I show above, came in the context of the academy’s data on a number of indicators becoming less disadvantaged, help it to improve its results? After all, those results also appeared the subject of praise in the inspection report.

The above again suggests Ofsted has many questions to answer about its investigations into changing pupil rolls. More broadly, how deep is the inspectorate investigating, and is it resourced, or inclined, to do the kind of investigative work which is required given what now rests on the outcome of its work, for individual schools?  And how consistent are the judgments on which schools’ futures rest?