Skip to content
The UK's only independent think tank devoted to higher education.

A Friday thought: Should universities do more to help deliver non-selective schooling?

  • 18 January 2019
  • By Nick Hillman

Over the past week, HEPI has been at the centre of a social media storm about academic selection. Last week, we published a report in our Occasional Paper series which looked at the question of how different school types fare in access to higher education, especially selective higher education, including Oxbridge. Controversially, grammars came out rather well.

Many people have taken strongly against this finding, which is odd as it seems obvious that selective schools are likely to perform well against the entry criteria for selective universities, and – compared to other countries – we have a hyper-selective higher education system. The HEPI paper is based on some wholly new data that needed airing but it doesn’t negate separate pre-existing research showing that grammars may do other things (like offering places to children on free school meals) less well than many other schools. Oddly, we’ve been accused of ignoring the evidence when we’ve actually broadened out a rather narrow, tired and old debate, in part through the use of some previously unpublished data.

The arguments have played out in enormous detail on Twitter, over at Wonkhe and in the specialist press, such as Schools Week. It would be boring to retread them here. There is a little bit of me that hopes I never have to write another word on grammar schools ever again.

But the past few days have got me thinking about one thing in particular. The fuss has shown once more, in case anyone was ever in any doubt, that there are large numbers of academics who fervently believe their research proves academic selection at the age of 11 is morally and educationally wrong. Among the many specific claims are: it’s bad for those in selective areas who don’t pass the test; it’s bad for achievement overall; and it’s bad for social cohesion.

That takes me to the 2016 education green paper, which argued, in contrast, that we should: 1) have more school selection at age 11; and 2) that universities should start sponsoring schools left, right and centre, declaring ‘We believe universities have a greater and more direct role to play in improving school quality and pupil attainment.’

The first half of that has survived to fight another day – despite Theresa May losing her majority at the 2017 general election – via the small Selective Schools Expansion Fund. But the second part never really got off the ground in the first place (notwithstanding rare pre-existing examples, like the two excellent university-sponsored Maths schools, which are discussed in our report alongside grammar schools).

Back when the green paper was published, senior managers at selective universities, most notably perhaps Louise Richardson at Oxford, went out with their guns blazing to oppose the idea that universities have much to offer in the way of school sponsorship.

But isn’t there a disconnect somewhere? If universities are repositories of extensive knowledge about what works before the age of 18, is it completely unreasonable to hope they will put more of that into practice by establishing and running schools? Wouldn’t it enable us to learn even more about what works in pre-university education? There was an inherent contradiction in that old green paper that it could have been illuminating to expose, which was the gap between its support for new selective school provision and its desire to have universities’ education departments help establish new schools despite their common scepticism of selectivity in schooling.

This question is especially live for me, and here I declare a deep personal vested interest, as I live in a selective area with a crying need for more secondary school provision. The clear preference of most of the parents I talk to, as well as my own preference, is to have a new non-selective school in our area.

But when you put together the Government’s Selective Schools Expansion Fund, the bar on new local authority sponsored schools, universities’ general lack of interest in sponsoring new schools and charities’ understandable desire to work in more disadvantaged areas than rural Buckinghamshire, it turns out it is easier for an area like ours to get more selective provision than it is to get more non-selective provision. That doesn’t reflect the desires of parents or educational researchers.

If we in the higher education sector disapprove of this, is it not at least partly in our gift to change it through even deeper.engagement with the secondary school sector? Then, more children would have the benefits of the non-selective education that education specialists believe should be the right of every child while also moving us closer to a non-selective system overall and teaching us more about effective schooling.

If just 10 per cent of the energy spent critiquing our latest paper were now to be spent on creating a new school, it would surely quickly become the best school in the country.


  1. albert wright says:

    Very interesting paper. As a grammar school boy myself in Sunderland who went to Manchester University (first in family to go) and got a 2.1 (in 1970) I am pleased to see such schools continuing to contribute to social mobility.

    Thank you for the strong evidence to support creating more selective Grammar Schools to achieve greater access to University by those living in “disadvantaged areas”.

  2. Mike Ratcliffe says:

    For me, the most alarming part of the attempt to require a university to sponsor a ‘good’ school was the way that was going to be built into regulations. It was immediately clear that this would be a threat, the answer to which would be more hyper-selective sixth forms rather than new schools or the assistance of current struggling schools.

    University sponsorship of schools continues to be a mixed bag – yes we have the maths schools, but there’s the continuing struggle with UTCs and the weakness of some university sponsored MATs. Perhaps more interesting are renewed attempts to bring FECs and Universities together. These underline just how much effort is needed for these projects – at a time when SMTs are lots of other things going on.

  3. John Mountford says:

    At heart, Nick, your commentary here on Iain Mansfield’s report is a cry to respect and respond to the data on the proposed impact of selective schools. Despite mine, and others’ objections to selection, at any point during compulsory education, albeit on other grounds, it is clear that as a data analysis, Mansfield’s report must be judged on the validity of the conclusions drawn from those data. In my humble opinion, it is crucially important that the debate focuses on this aspect, leaving it to others on another day to argue the moral and social consequences of selection. Why is such an approach necessary?

    As stated by Chair of Comprehensive Future, Dr. Nuala Burgess in a commentary on Mansfield’s report for Comprehensive Future, “the use of ‘below median income families’ as a proxy for disadvantage.” is deeply flawed.

    You make the point that, “selective schools are likely to perform well against the entry criteria for selective universities”. The key to understanding this alleged link is selectivity. Put crudely, remove the ‘dead wood’, and access to the gleaming uplands of higher education becomes easier to achieve for the selected FEW. What is at stake is the future of policy over selection based on evidence destined that will ultimately affect the MANY – garbage in, garbage out springs to mind. In our data-led society, ordinary people like me need to be able to feel confident about how robust the facts we are fed are.

    The problem I, and others have, is not the “use of some previously unpublished data.” by Iain Mansfield. It is the concern that those data are too broad and by the governments own admission “unreliable”. Therefore, I disagree with your claim that the use of the “previously unpublished data” has broadened the debate. It’s my view that the DfE, in its advice that, “Caution should be taken in drawing definitive conclusions from these findings (below median income families) until we have completed further work” has been ignored. Such a move surely has consequences for any conclusions reached by those using those data. This is the central concern of those who question the validity of Mansfield’s findings.

    I believe your reference to the 2016 Green Paper is not relevant to the present discussion about selection. You may be right about university’s not taking up the challenge of sponsoring schools. But, that is quite clearly not the issue at stake in regard to selective education. In that debate you are right to believe that bias is a potential problem. Bias may exist for both pro and anti selection supporters. This is the nub of my criticism of Mansfield’s report. It is making the case, albeit perhaps indirectly, for the expansion of selection on the strength of an analysis of the “new data” in question. If the data are unreliable, it is not unreasonable to criticise the conclusions reached. In such a situation it’s not unreasonable to suspect bias.

    As to the situation in Bucks, creating another non-selective school in a selective area, while required on the basis of demographic need, is complex, as Professor Becky Allen points out –

    Her analysis of the situation, points to the need for caution when quantifying the impact of selection. This is why I find it a real challenge to understand why an institution like HEPI, with its impressive list of what I call, ‘big-hitters’, stands by Iain Mansfield’s report which raises more questions than it answers because of its reliance on questionable data.

    Finally, The reason I find it hard to close the debate down is because policy-makers are now at pains to impress joe-public with the fact that their decisions are firmly based on ‘the evidence’ for this, that or the other.

    Surely, in light of the unfolding ill-informed, and even more disastrously managed drama that is Brexit, those who raise their heads above the parapet with claims of certainty about policy choice owe it to the public to satisfy their critics with the sheer weight of hard evidence in support of their claims. If such hard evidence is known to be even harder to establish, lets refrain from making decisive pronouncements about matters that affect the lives of real people. The debate clearly has to go on.

  4. Peter says:

    @John Mountford

    The problems with the methodology are even more glaring than that.

    Mansfield – deliberately and knowingly – excludes more than a third of selective university entrants from non-selective LAs on the grounds that, due to the structure of 16-18 education in their area, they did their A-levels at a 6th form college or FE college rather than their 11-16 school so they can’t be considered a success for the comprehensive system.

    He also completely ignores cross-border movement of children able enough to get into selective universities from non-selective LAs to selective LAs and underplays the impact of a Grammar school on the local private school market on his findings.

Leave a Reply

Your email address will not be published. Required fields are marked *