Pupil premium grammar schools


…Or ‘An exercise in policy design’.


This post considers proposals emerging for new selective schools that would select on the basis of ability or attainment and socio-economic disadvantage.

It covers the following ground:

  • The context provided by the selection green paper and the Opportunity Areas policy.
  • Recent Advocacy for ‘pupil premium grammar schools’ and similar entities.
  • Analysis of the model proposed.
  • An alternative, far more efficient model for the system-wide support of disadvantaged high attainers.


The selection green paper

Schools that work for everyone’ (September 2016) proposed increased between-school selection, including new fully or partially selective schools, but subject to certain conditions.

The introduction says:

‘We want more good schools, including selective schools, but we want selective schools to make sure they help children from all backgrounds.’

This immediately suggests some potential resistance to new selective schools reserved solely for learners from disadvantaged backgrounds.

The chapter on selective schools argues that:

‘The vast majority of the existing 163 grammar schools have demonstrated that they provide an excellent education for the pupils, of all income groups, who attend them. They provide a stretching education for our most able pupils, regardless of their background, which recognises, and enables them to meet, their potential.’

It claims that:

‘Some studies have found that selective schools can be particularly beneficial for pupils on lower incomes who attend them.’

But it can cite only one study, whose findings are not exactly as they are reported. Evidence to the contrary is ignored and, since the green paper was published, that evidence has been expanded and updated.

This evidence base includes my own analysis of performance table data which suggests that some grammar schools are far better than others in this respect.

The green paper does acknowledge the low proportion of pupils receiving FSM attending grammar schools, while arguing that:

‘…there is no clear understanding of the number of children of ordinary, working families in selective education or the relative incomes of parents. We believe there is a case for looking at the wider impact of selective education of those on low incomes or who just about manage.’

The menu of options suggests that some or all new or expanding selective schools will be expected to ‘take a proportion of pupils from lower income households’ and one of the consultation questions is:

‘What is the right proportion of children from lower income households for new selective schools to admit?’

There is also a proposal to revise the school admissions code to support fairer admissions practice:

‘All schools have been able to start the process of prioritising the admission of pupils entitled to the Pupil Premium since the current School Admissions Code came into force in December 2014. We need to increase the pace at which selective schools are ensuring fair access. We therefore propose to require all selective schools to have in place strategies to ensure fair access. Legislation would require selective schools to prioritise the admission of, or set aside a number of specific places for, pupils of lower household income in their oversubscription criteria.’

But nothing in the green paper suggests the introduction of selective schools exclusively for learners from disadvantaged backgrounds.

We expect a response to the green paper, and presumably a white paper too, at some point in the next three months.


Opportunity areas

The schools white paper ‘Educational Excellence Everywhere’ (March 2016) announced that the schools improvement effort would include:

‘New intensive focus…on areas of chronic, persistent underperformance, where not enough children have a high quality school place and where there is insufficient capacity to drive improvement.’

Multiple interventions would be targeted at a set of ‘Achieving Excellence Areas’ to be piloted from September 2016 and rolled out from September 2017. A methodology for defining these Areas was published alongside the white paper.

Then Greening replaced Morgan as Education Secretary and announced the first six areas, now rebranded as ‘Opportunity Areas’, in her October 2016 speech to the Conservative Party Conference.

These six – Blackpool, Derby, Norwich, Oldham, Scarborough and West Somerset – would receive £10m funding apiece. There was an associated DfE press release.

A second press release in January 2017 announced another six Areas – Bradford, Doncaster, Fenland and East Cambridgeshire, Hastings, Ipswich and Stoke-on-Trent – with funding increased to £72m. So the share for each Area has fallen substantively, from £10m to £6m apiece.

The speech containing this announcement places Opportunity Areas in the wider context of a drive to improve social mobility, creating:

‘…a country where everyone has a fair chance to go as far as their talent and their hard work can take them, where it is that talent and hard work that matter, not where you were born, who your parents are or what your accent sounds like.’

There have also been press releases marking visits to Blackpool (December 2016) and Derby (February 2017), the latest saying that Area plans are being drawn up, but giving no deadline for their publication.

Comparatively little is known about the support package that Areas will access:

  • Support will be tailored to each area’s needs. They will have ‘prioritised access’ to other support programmes including ‘a teaching and leadership innovation fund worth £75 million over 3 years focused on supporting teachers and school leaders in challenging areas to develop’.
  • DfE will ‘target its programmes to ensure children get the best start in the early years, to build teaching and leadership capacity in schools, to increase access to university, to strengthen technical pathways for young people, and work with employers to improve young people’s access to the right advice and experiences’.
  • Local partnerships will be formed with ‘early years providers, schools, colleges, universities, businesses, charities and local authorities’. There will be a mixture of in-school and out-of-school support. Successful teachers and schools will partner with those requiring additional support. The EEF will establish a research school in each area, with a budget of £3.5m. They will act as ‘local excellence hubs for evidence-based practice’. Teach First will increase its activity in some Areas.
  • There will be support from DCMS and from business, including organisations such as the CBI, Local Enterprise Partnerships, Young Enterprise and City Year. The Careers and Enterprise Company will improve careers advice. The National Citizens Service will be targeted on these Areas. Every secondary school in an Area will have an enterprise adviser to give access to work experience. Support has been promised by PWC and KPMG – more companies are needed to become Area partners.


Opportunity areas and grammar schools

It is noteworthy that no explicit connection has been made between the Opportunity areas and new grammar schools. The Education Secretary’s social mobility speech made no reference to increased selection, let alone prioritising Opportunity Areas in this regard.

However, the Spring 2017 edition of the Grammar School Heads’ Association Newsletter summarised the points arising from its recent meetings with the government, one of which was:

‘The new opportunity areas and other cold spots are clearly priorities for new selective schools, but no areas are excluded where there is parental demand.’

The GSHA’s note also revealed ‘a determination to address the needs of ‘JAMs.’’ (just about managing families)’ and noted that new schools were unlikely to open before September 2020 (so after the next General Election).

A Centre for Social Justice Report ‘Selective education and social mobility’ (December 2016) made a preliminary connection between new selective schools and Opportunity Areas.

In ‘Grammar schools and social mobility – Further analysis of policy option’ (also December 2016) the Education Policy Institute did not make the explicit link, but did set out a series of ‘location principles’ for new selective schools, concluding that parts of only six local authorities fit the bill, all six having below average levels of disadvantage.

In its response to the green paper United Learning proposed the introduction of ‘non-selective grammar schools’ [sic] into Opportunity Areas’. They would:

  • ‘Teach a highly academic curriculum, including three separate sciences and several ancient and modern foreign languages;
  • Seek out highly academically qualified staff;
  • Offer an extensive programme of team sport, musical ensembles, dramatic productions and other extra-curricular activities (including outward bound activities and/or uniformed organisations such as the CCF);
  • Introduce subject-based societies, speakers and activities to develop and deepen scholarship;
  • Develop a strong programme of university entrance;
  • When over-subscribed, admit children by lottery, giving all applicants equal chance of admission.’

It is not yet clear whether the Northern Powerhouse Strategy might also draw on new selective schools, although they were not mentioned in the independent review of the schools dimension (November 2016).


Recent advocacy

In November 2016 Anthony Seldon published a Telegraph article calling for the establishment of 100 ‘May schools’ by 2020.

They would:

  • Be free schools, new-built or occupying existing buildings where there are surplus places.
  • Be located so each serves a local population of 500,000 and would offer a weekly boarding option. This would impact only slightly on the intake of existing schools because there would be one for every 20 secondaries.
  • Select in the ‘bottom 25%’ by economic and social background and ‘those who would benefit from a highly academic education’ by means of testing at ages 11, 13 and 16 (to cater for late developers).
  • Provide an academic curriculum, complemented by ‘a rich offering of arts, sports and character education’, each school having an independent school partner and an affiliated university.

A second article published in March 2017 adds that they would:

  • Use ability tests to select four classes of 25 learners at 11, a further 25 at age 13 and operate a sixth form with 150 places. So in steady state 15,000 students would leave each year.
  • Assuming 100 new-builds the total capital cost is estimated at £3bn.

Also in March the TES reported that two MATs are actively exploring the creation of pupil premium grammar schools, one unnamed, the other the Inspiration Trust. This story was picked up by the Mail and Telegraph. Collectively they suggest that the Inspiration Trust had ‘floated the idea’ of locating such a school in Norwich, one of the Opportunity Areas, but no further details were forthcoming.

It is not clear whether this idea has any purchase with the embryonic local partnership behind the Norwich Opportunity Area. The Inspiration Trust might have floated the idea in the face of determined opposition from other partners.


A critique of pupil premium grammar schools



The fundamental problem with any support strategy for disadvantaged high attainers based on new grammar schools is that a structural solution is being imposed on a standards-based issue.

New grammar schools are a ‘bricks and mortar’ response that creates a ‘postcode lottery’, directly benefiting only those who live close enough to attend them (and with limited wider reach through outreach), or else the huge cost necessary to provide access regardless of location.

Seldon’s call for one-off capital investment of £3bn is pie in the sky, given the state of the public finances, not to mention the huge opportunity costs associated with such a sum.

If we accept his capital cost estimate of £30m per school, it would require £360m even to locate one institution in each opportunity area. That is funding which could be invested directly in support for disadvantaged high attainers.

And he has not even begun to factor in recurrent costs.

SFR63/2016 supplies a 2015-16 net current expenditure figure of £5,846 per secondary pupil in England, but classes of 25 and extensive arts, sports and character education don’t come cheap, let alone boarding costs.

Even if we assume a conservative per capita sum of £8,000, a 1,000 pupil school would cost £8m per year; a 15,000 sixth form student cohort would cost £120m a year and 100 1,000-pupil schools would cost £800m a year.

One might offset against the cost the additional school places provided, but those could be provided far more efficiently. This is not good value for money.

A pupil premium grammar school located in Norwich would struggle to attract sufficient pupils.

In 2016-17 there were 10,559 pupils in the whole of Norfolk eligible for deprivation pupil premium. There were 1,539 in the Norwich North Constituency and 1,218 in the Norwich South Constituency, so 2,757 in total.

Those totals include learners in Years 7-11 inclusive, so the number in Year 7 might be closer to 600 for the two Norwich constituencies and 2,200 for the whole of Norfolk.

Only a proportion of these pupil premium students would be equipped for a grammar school education. In 2016, 35% of Norfolk’s disadvantaged pupils (‘ever 6’ FSM and in care) reached the expected standard at KS2. Nationally only 2% of all disadvantaged students reached the higher standard.

If the entry standard for the Norwich school was pitched at that higher standard – and if Norfolk’s performance at the higher standard equals the national figure – then 44 students from across Norfolk would qualify per year – 12 of them from the two Norwich constituencies.

Even if were to assume, rather heroically, that 20% of all pupil premium students would potentially meet the cut, that would give a total Norfolk-wide pool of just 440 potential students per year group. The equivalent pool from the two Norwich constituencies would be around 120. All of them would need to be recruited to fill the places in a 4FE school.

The only other option is to impose an artificially low pass mark for the test of selection by attainment/ability. But the GSHA note reports the government to be keen on a selection rate of 10%, higher than most existing grammar schools.

And at what point does academic selection become pointless? Wouldn’t a ‘non-selective grammar school’ a la United Learning be preferable to a grammar school with a 50% selection rate?

If eligibility on the basis of disadvantage was also extended to learners from ‘JAM’ families, the potential cohort might increase by about half as much again, but it would still be a struggle to fill all the places.

So, in a nutshell, Seldon’s plan costs too much and the Inspiration Trust’s plan is not viable.



Other issues

Even if the logistical problems can be overcome, there are other problems with a solution built around pupil premium grammar schools:

  • Any provision that reaches beyond the learners attending the school depends on outreach activity, which needs to be staffed and funded alongside the core business. Even if an additional income stream can be identified, the potential reach is strictly limited.
  • Many students will be faced with a choice between weekly boarding and travelling excessive distances to get to school. Boarding might suit some students; but it will not suit others. Unless travel costs are met, adding another item to the already substantial central budget, many disadvantaged families will find the option prohibitively expensive.
  • While Seldon might argue that ‘creaming-off’ high-attaining disadvantaged students by his ‘May schools’ might not be too damaging, the effect will be far more pronounced for those schools unfortunate enough to be located in close proximity. The effect on its neighbours of a pupil premium grammar school in Norwich would be far more profound. Some of the Inspiration Trust’s own schools would be affected.
  • Even with provision for entry in Year 9 and Year 12, the introduction of separate grammar schools creates a two-tier solution. It will be much harder for ‘late developers’ to move between institutions than it would be to move within the same institution.
  • There will be an incentive to tutor for the selection test which will favour those from families that can afford the cost, resulting in unofficial selection on income, so potentially distorting the intake. The tutoring industry will benefit at the expense of gullible parents.
  • Finally, it is likely that some stigma will be attached to the students of these latter-day ‘poor schools’, who will be singled out as disadvantaged. Some will experience mockery and bullying and this will be harder to combat when the cohort is segregated in a different institution.

For all these reasons, the idea of selective schools exclusively for disadvantaged students is misguided and unhelpful.


An alternative approach

I have been proposing for some time an alternative model that would support all disadvantaged high attainers in their existing secondary schools and colleges from the beginning of KS3 to HE entry – and potentially beyond.

Let’s call it a National Entitlement (NE). It might if necessary be piloted in Opportunity Areas before being rolled out to the rest of the country.

The Entitlement would be designed to provide a tailored learning and support programme for all disadvantaged high attainers, from Year 7 through to the completion of Year 13.

Provision during KS3 would be open access, made available to all pupil premium students, or it could be reserved for those achieving at least the expected standard at the end of KS2. It would be designed to familiarise learners with the Entitlement, to raise awareness and build aspirations for the future.

Continuation into KS4 and KS5 would be conditional on attainment and progress thresholds, pitched at a level consistent with the achievement of AAB+ A level grades at the end of KS5.

The core problem is to overcome fragmentation on both the demand and supply sides of the market to secure a coherent long-term programme for each participant that builds their strengths, addresses any weaknesses and culminates in the achievement of a place on a selective higher education course (or equivalent alternative).

This can be achieved by:

  • Placing school and college-based learning at the core of the programme and providing the support necessary to ensure a much more consistent quality of core provision system-wide, so all disadvantaged high attainers are given the in-house challenge and support they need to attain highly and eradicate top-end underachievement (see my 10-point plan).
  • Publishing a framework of provision that covers all aspects of learning and support that participating students need to achieve the desired outcome – attainment in their chosen subjects, aspirations, cultural capital, family and peer support and so on.
  • Using the framework to inform a process of needs assessment, commissioning of out-of-school opportunities and annual review in schools, with arrangements in place to ensure continuity when students move between institutions, especially to post-16 institutions at the end of Year 11. Records would be maintained in each student’s personal portfolio.
  • Using the framework to catalogue all the out-of-school learning and support opportunities offered by the supply side, whether they are universities, third sector providers, commercial organisations, independent schools, public sector organisations or businesses. This online catalogue might include Amazon-style reviews by previous participants.
  • Schools and colleges would use the catalogue to select suitable opportunities tailored to their participants’ needs. Suppliers would use it to identify gaps in the market, to respond to the emerging pattern of demand. Increasingly suppliers might be expected to collaborate to deliver support ‘packages’.
  • Only tutoring within the programme would be admissible – any student found to be accessing tutoring outwith the programme would forfeit participation.
  • Selective universities would be encouraged to make contextualised offers to students who had successfully completed the full programme. 


The participant’s perspective 

A typical student who completed the full seven years of the NE might benefit in Years 7-9 from university taster sessions, advice and support with choosing GCSE options, aspiration-affirming sessions, joint planning and review sessions with parents or carers.

In Years 10-11 he might opt for enrichment programmes in space science, receive intensive tutoring in English or maths and undertake an online course in another of his GCSE subjects.

In Years 12-13 he might attend a university summer school, receive independent partnership support to consolidate attainment in one of his A level subjects, attend a workshop on preparing UCAS applications and benefit from mentoring support.

He might combine support from, say, Oxford University, Eton College, the Sutton Trust, Brilliant Club and Villiers Park.

One of his teachers would act as facilitator, ensuring that his choices were consistent with his needs, that external support was complemented by his in-school learning experience and that his overall programme was coherent and effective, avoiding unnecessary duplication and maximising value-added.

His progress and needs would be reviewed annually, his personal targets adjusted and his success celebrated.

If he moved institutions at 16, his portfolio would accompany him and transition would be managed between his old facilitator and his new one.

When applying to university his portfolio would accompany his UCAS application.



Each participant in the NE would have access to a personal budget held by the school or college. At KS3/KS4 much of the cost would be met from an annual £50m topslice from the pupil premium budget.

At KS5 they would be met from an equivalent pro-rata topslice from the funding available to support disadvantaged 16-19 year-olds.

Universities would also be expected to meet part of the cost, in recognition of their role in attainment-raising to support fair access. They might contribute to the core budget, as well as subsidising their own outreach opportunities so no cost is passed on. Other providers might do the same.

Ideally all selective universities would be required to contribute a variable minimum sum to the central budget from their access agreement spending, determined according to how far they undershoot new pupil premium access targets (aka performance indicators) imposed by OfS.

According to the Russell Group:

‘In 2016-17, the 20 Russell Group universities in England alone will be investing £243 million in scholarships, fee waivers, bursaries and outreach activities aimed at the most disadvantaged.’

A similar approach could be taken with independent schools that are not pulling their weight on partnership with state schools.

The independent sector is coy about the total current value of partnership activity, but asserts that in 2016 87% of ISC schools were engaged in some form of partnership.

We do not know how many pupils meeting the pupil premium eligibility criteria currently attend independent schools, perhaps benefiting from bursaries and scholarships provided, but it might be reasonable to seek an annual contribution of, say, £20m from this source.

Arrangements would be set in place to ensure that the HE and independent sectors did not succumb to the perverse incentive to pass on their costs through charges attached to their support provided within the framework.

The cost of additional workload in schools and colleges would be met through the deprivation weighting in the appropriate funding formula and, in the case of schools, through the normal pupil premium allocation.

All this funding comes from the reallocation of existing budgets. The burden on those could be reduced substantively by the creation of an endowment fund to which institutions, philanthropists and alumni might contribute.

The central administrative cost would be capped and kept low by tendering for an independent entity, operating at arm’s-length from the government, which would also be barred from offering services within the framework. A university might undertake the role if it erected ‘chinese walls’ between the administering body and its own widening participation/fair access activities.

The pilot would also be independently evaluated.


Numbers and costings

In steady state there would be seven participating cohorts in Years 7-13 inclusive. There are different options for design and roll-out, depending on the size of the budget available.

My preference would be to begin immediately with KS3 provision for a substantial cohort of disadvantaged secondary school students, building the essential foundations and helping to tackle the underachievement identified by Ofsted and others.

To speed up implementation I would also launch the programme with a pioneering Y10 cohort that would work through the programme over the next four years. So the roll-out would look like this:

  • 2018/19: Years 7-10
  • 2019/20: Years 7-11
  • 2020/21: Years 7-12
  • 2021/22: Years 7-13

Assuming that provision in Y7-Y9 is available to all disadvantaged learners then, on the basis of 2016 figures, there would be around 186,000 pupils in each cohort. Of course this figure will vary depending on whether the incidence of deprivation rises or falls.

For the purposes of this illustrative costing, let us assume that KS3 provision is limited to those achieving the expected standard at KS2. On the basis of 2016 outcomes that would give an annual cohort of 39% of disadvantaged pupils – so about 73,000.

Allowing for some improvement in the success rate, and assuming unchanged incidence of pupil premium eligibility, we might allow for 80,000 participants per year (so 240,000 in total).

It is hard to estimate what proportion of these students might be expected to progress into KS4 and then KS5, partly because we have negligible information about the proportion of pupil premium students likely to achieve AAB+ A level outcomes, or a commensurate Attainment 8 outcome at the end of KS4.

For the purposes of these costings I have assumed that 25% of KS3 participants continue into KS4 – so a cohort of 20,000 students, slightly more than 10% of the entire national disadvantaged cohort. I have also assumed that 75% of these progress into KS5, giving a cohort of 15,000 – the same size as that generated by Seldon’s ‘May schools’.

Each applicant would be allocated an annual personal budget of £250 in Years 7-9, £1,500 in Years 10-11 and £2,000 in Years 12-13.

For these purposes I have assumed only three contributing funding streams – the pupil premium topslice, the corresponding 16-19 contribution and the HE ‘fair access’ contribution.

Personal budgets set at these levels would demand a total budget of:

  • 2018/19 – (240,000 x £250) + (20,000 x £1,500) = £90m, of which £50m is from pupil premium topslice and £40m from HE access agreement budgets.
  • 2019/20 – (240,000 x £250) + (40,000 x £1,500) = £120m, of which £50m is from pupil premium topslice and £70m from HE access agreement budgets.
  • 2020/21 – (240,000 x £250) + (40,000 x £1,500) + (15,000 x £2,000) = £150m, of which £50m is from pupil premium topslice, £30m from 16-19 contribution and £70m from HE access agreement budgets.
  • 2021/22 – (240,000 x £250) + (40,000 x £1,500) + (30,000 x £2,000) = £180m, of which £50m is from pupil premium budgets, £60m from 16-19 contribution and £70m from HE access agreement budgets.

Administrative and evaluation costs should be no more than £3m annually, but let’s say £5m in the worst case scenario.

That gives a steady state cost of £185m, compared with a conservative estimate of £800m per year for the recurrent costs of 100 1000-pupil ‘May schools’. All of the funding is recycled from existing budgets – no new money is required.

The initial capital outlay of £3bn is saved, though a fraction of that sum would provide an endowment fund that could eventually meet the entire cost of a National Entitlement.

The funding saved could be dedicated to increasing the number of good and outstanding school places, so untangling these two objectives which have been so unhelpfully welded together in the green paper.

This approach is also fully scalable, so additional funding would permit the inclusion of additional learners from other JAM families.



March 2017


3 thoughts on “Pupil premium grammar schools

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s