How should England close its excellence gaps?


First edition: January 2018


This is an important turning point in the evolution of national education policy.

Departing Secretary of State Greening has only just published her social mobility plan Unlocking Talent: Fulfilling Potential (December 2017)

It was preceded by the wholesale resignation of the Social Mobility Commission’s Board, protesting at lack of political leadership and insufficient support.

New Secretary of State Hinds has been appointed amid rumoured plans to refresh and prioritise Government policy on improving education standards. Hinds himself is also keenly interested in social mobility.

Some sixteen months on, the Government still has not published a full official response to consultation on its pro-selection green paper Schools that work for everyone (September 2016) – which prompted almost universal criticism.

Greening is understood to have delayed publication, though a document has several times been promised ‘in due course’.

And we still await the outcome of DfE’s consultation on ‘ordinary working families’ Analysing family circumstances and education (April 2017), apparently a major issue for the Prime Minister when it was launched.

This will now become bound up with the more recent consultation on Eligibility for free school meals and the early years pupil premium under Universal Credit (November 2017) which will likely change the basis on which socio-economic disadvantage is defined, so problematising trend analysis.

Despite frequent scaremongering to the contrary, the opening of new 11+ grammar schools seems firmly off the agenda, at least for the time being. It would require legislation which, so far at least, has not been forthcoming and, last we heard, is not planned.



Any bill would inevitably generate much contentious debate in Parliament and, given the Government’s precarious majority, every possibility of ignominious defeat. Greening herself would be odds-on to lead the Tory opposition.

The justification for new grammar schools advanced in the green paper was anyway weak, resting predominantly on a single piece of contested evidence that they benefit disadvantaged high attainers:

‘Some studies have found that selective schools can be particularly beneficial for pupils on lower incomes who attend them. For example, one study reported that the educational gain from attending a grammar school is around twice as high for pupils eligible for free schools meals, compared to the overall impact across all pupils.’

But there is still considerable scope to expand the existing tranche of 163 secondary grammar schools, to reform their admissions so they take in more disadvantaged learners, and to promote a variety of other approaches to between-school and within-school selection.

Thus far we know only that the Grammar Schools Heads Association (GSHA) will enter into a formal agreement with DfE to ‘improve admission rates for disadvantaged pupils’ and that DfE will also work with other partners to increase the flow of low income learners to grammar schools.

Turning to other ideas for extending between-school selection, it seems that:

As for within-school selection:

  • Grammar streams are still favoured in some quarters, including by some who vigorously oppose between-school selection, despite the strong evidence against streaming per se, arguably much stronger than the evidence against setting.

Greening’s social mobility plan steered clear of selection entirely, opting instead for:

‘…a new £23 million Future Talent Programme to trial approaches and present clear recommendations on ‘what works’ to support the most able disadvantaged children, particularly during key stage 3 when they so often fall behind’.

This is helpful as far as it goes – but it does not go very far.

There is an outstanding commitment from January 2017 to announce in due course plans setting out:

‘…the best ways to support the most academically able pupils across the full range of state schools, particularly those from the most disadvantaged backgrounds.’

The Future Talent Programme does not fully satisfy that commitment, which raises the question, ‘what else – including some of the approaches to selection listed above – might the government undertake’?

This post is dedicated to addressing that question in some depth. It does so by developing, updating and synthesising the line of argument I have advanced in several previous posts, including those linked to above:

  • Reintroducing excellence gaps, explaining what they are and why they are significant – and exploring how disadvantaged high attainers are typically sidelined in much contemporary educational discourse
  • Presenting recent statistical evidence of the problem (relying predominantly on the Government’s own data and preferred indicators, wherever these exist)
  • Tracing the evolution of recent Government policy in support of disadvantaged high attainers
  • Assessing the likely impact on excellence gaps of the Future Talent Programme and the wider social mobility plan
  • Specifying what further steps would be necessary to close excellence gaps systematically, within an affordable budget, including the contribution that might be made by selection in its various forms.

I have made a few contentious statements in this piece, all of them based on the evidence I have at my disposal. This is deliberately to provoke others to bring more information into the public domain.

If there is substantive evidence to the contrary I am fully prepared to adjust the offending statements but, if no evidence is forthcoming, they must be assumed to stand.


Reintroducing excellence gaps

The social mobility plan justifies the Future Talent Programme on the grounds that some schools are not yet using pupil premium funding effectively and:

‘This can be especially true for their more able disadvantaged pupils, who fall back compared to their more affluent peers, even after a promising start.’

This is a reference to the oft-cited but rather limited Sutton Trust research brief Missing Talent (June 2015).

But the fundamental justification for targeting support at disadvantaged high attainers is not their faltering progress over time, relative to advantaged learners and particularly at KS3, significant though that problem is.

It is the persistent and substantial excellence gaps which exist between high-attaining advantaged and disadvantaged learners at every high attainment threshold throughout the key stages.

This excellence gaps concept is already well-established in the United States. Over the last seven years I have made repeated efforts to embed it in domestic analysis of the limitations of support for disadvantaged high attainers.

Three years ago I published an extensive and well-received analysis of excellence gaps in England and the associated research – Part One and Part Two (September 2014) – which set out a working definition:

‘The difference between the percentage of disadvantaged learners who reach a specified age- or stage-related threshold of high achievement – or who secure the requisite progress between two such thresholds – and the percentage of all other eligible learners that do so.’ 

More recently, I investigated the limited impact of pupil premium on excellence gaps (February 2016), deciding to retain this definition. Reviewing the wording again now, in the light of recent substantive changes to the assessment and accountability regime, I judge that it continues to apply.

Arguably there is a case for adjustment to bring within scope statistics describing fair access gaps in admissions to selective higher education. But that would add another layer of complexity which would militate against wider understanding and acceptance of the excellence gaps concept.

So I have decided to keep fair access separate while also including relevant national data in the statistical section below.


Disadvantaged high attainers in educational discourse 

The social mobility plan recognises that pupil premium should be targeted more effectively towards disadvantaged high attainers (calling them ‘more able/most able disadvantaged pupils’) but it does not connect this with the case for closing substantial excellence gaps at every key stage.

It does hint briefly that such socio-economic gaps exist at different levels, though without developing this statement – and, in doing so, risks further complexity by confusing ability and attainment:

‘Disadvantaged pupils of all abilities are more likely to underperform even in otherwise strong areas and strong schools.’

I will not pursue further here the terminological and definitional shortcomings which bedevil this field, for fear of further extending an already ambitious piece. They have been discussed at length several times in my previous posts, most recently here and they will reappear in my proposals for further action below.

My principal contention is that the failure to close substantial and persistent excellence gaps is pivotal to both the educational standards and social mobility narratives.

Yet much of the wider discourse associated with closing socio-economic achievement gaps is predicated on the false assumption that disadvantaged learners are, by definition, low attainers.

And it is too readily assumed that closing gaps between advantaged and disadvantaged learners and closing gaps between high and low attainers is one and the same thing when they are quite different.

There is certainly considerable overlap between these two pursuits. But a significant minority of disadvantaged learners are also high attainers.

(The higher the attainment hurdle, the smaller the minority but, in some senses at least, the fewer the number, the more significant they become.)

They do not benefit from strategies designed to close the gap between low and high attainers. At the same time they routinely underperform relative to their equivalently qualified yet more advantaged peers.

I have previously drawn attention to aspects of this Government’s education policy – particularly the weightings in the national funding formula and adherence to certain orthodoxies of the ‘maths mastery’ movement – which seem focused explicitly on closing high-low attainment gaps, to the detriment of disadvantaged high attainers.

The Education Endowment Foundation gives substantially more weight to the impact of interventions on lower and middle attainers – witness its material relating to setting and streaming –  presumably because they form the substantial majority.

Indeed, the EEF has almost completely neglected this important minority, even though this needn’t be a zero sum game.

But there is inconsistency. All this happens even though the Government’s wider education strategy, formulated initially for the white paper Education Excellence Everywhere (March 2016), includes as one of its three strategic goals:

‘…every child having access to high quality provision, achieving to the best of his or her ability regardless of location, prior attainment and background’.

More recently this has been replaced by commitments to ‘help everyone reach their potential’, such as this from the latest DfE single departmental plan:

‘Our purpose is to help create a country where there is social mobility and equality of opportunity by providing excellent education, training and care, and to help everyone reach their potential, regardless of background.’

One must assume that ‘background’ stands proxy for prior attainment, as well as location and all other relevant variables. Certainly that is the fundamental justification for the recognition given to disadvantaged high attainers in the social mobility plan.

Meanwhile, the parallel discourse on social mobility frequently features statements to the effect that it must push beyond support for disadvantaged high attainers to enter selective higher education and, subsequently, professional careers.

Some argue that the very idea of social mobility is imbued with elitism, focused as it is on opportunities for people to ‘better themselves’, to increase their incomes and so succeed only in the narrow terms defined by the prevailing elite.

Any attention given to academic ‘high fliers’ is similarly perceived as narrow and elitist: social mobility relates equally to technical and vocational routes it is argued – and to those with more modest attainment. (I do not mean to imply that vocational routes are inappropriate for high attainers).

Quite right too, but there is no justification, in extending the educational context of social mobility, for diverting scarce resource away from disadvantaged high attainers pursuing the most demanding academic routes. 

As my statistical analysis demonstrates, access to the most selective universities remains extremely unfair, in the sense that disadvantaged high attainers are so poorly represented.

Even where numbers of disadvantaged students are increasing, the access gaps between them and more advantaged students are typically continuing to grow, because the latter’s numbers are increasing at a faster rate.

One might argue that our consistent failure to resolve this problem, despite significant investment over many years, requires that that we redouble our efforts, rather than parking it in the tray marked ‘too difficult’.

In pursuing social mobility we must continue to strive to keep excellence and equity in equilibrium.

We must also understand that support for disadvantaged high attainers and fair access to the most selective universities are two ends of the same policy issue, demanding co-ordinated action across the schools, post-16 and HE sectors alike.

The fundamental flaw in both strands of discourse – relating to education standards gaps and social mobility respectively – is the refusal to accept that disadvantaged learners have an equal right to support, regardless of their prior attainment. 

There is a widespread false assumption that high attainers, whether disadvantaged or not, are already ‘succeeding’ and need relatively less challenge and support to remain on their positive trajectory.

But there is no justification for prioritising lower attainers over higher attainers independently of their socio-economic background – to do so is to penalise unfairly the life chances of disadvantaged high attainers. 

The arguments for closing excellence gaps

It follows from the discussion above that the generic arguments for closing socio-economic achievement gaps, apply equally to disadvantaged high attainers:

  • The goal of education should be to provide all learners, including disadvantaged learners, with the opportunity to maximise their educational potential, so eliminating what has been called ‘the soft bigotry of low expectations’.
  • Schools should be ‘engines of social mobility’, helping disadvantaged learners to overcome their backgrounds and compete equally with their more advantaged peers.
  • International comparisons studies reveal that the most successful education systems can and do raise attainment for all and close socio-economic achievement gaps simultaneously.
  • There is a strong economic case for reducing – and ideally eradicating – underachievement attributable to disadvantage.

But there is additional justification for targeting support specifically at disadvantaged high attainers:

  • An exclusive or predominant focus on gaps at the lower end of the attainment distribution is fundamentally inequitable and tends to reinforce the ‘soft bigotry of low expectations’.
  • Disadvantaged learners benefit to some extent from successful role models – predecessors or peers from a similar background who have achieved highly and are reaping the benefits.
  • Moreover, success at the very top of the distribution (for example disadvantaged students entering Oxbridge) has a symbolic value far in excess of the number of direct beneficiaries.
  • An economic imperative to increase the supply of highly-skilled labour places greater emphasis on the top end of the achievement distribution.
  • Some argue that there is a ‘smart fraction’ tying national economic growth to a country’s stock of high achievers. There may be additional spin-off benefits from increasing the supply of scientists, writers, artists, or even politicians!
  • The most highly educated disadvantaged learners are least likely to confer disadvantage on their children, so improving the proportion of such learners may tend to improve inter-generational social mobility.

Seven core policy principles

Following publication of the selection green paper Schools that Work for Everyone (September 2016) I proposed a framework of seven core principles to inform national policy on the education of disadvantaged high attainers.

  • All learners have an equal right to an education that meets their needs, regardless of background and prior attainment. No learner should be denied access to suitable learning opportunities tailored to their needs.
  • Pursuing the twin priorities of raising standards and closing advantaged-disadvantaged attainment gaps means giving relatively greater priority and dedicating relatively more resources to learners from disadvantaged backgrounds, regardless of prior attainment.
  • There is no equivalent justification for prioritising or disproportionately resourcing low attainers over high attainers, or vice versa. A disadvantaged high attainer should attract more priority and more resources than an advantaged low attainer.
  • Attainment gaps must not be closed by holding back improvement amongst learners from advantaged backgrounds. Both advantaged and disadvantaged learners should continue to improve, the latter at a relatively faster rate.
  • Impact will be maximised through the efficient use of scarce resources. That means minimising deadweight costs, investing in standards over structures so capital costs are avoided and, wherever feasible, targeting support at the learners rather than their schools.
  • Education tailored to background and prior attainment is not always most efficiently and effectively provided through universally inclusive learning opportunities (whether or not provided in the same institution). Some opportunities will be restricted to the subsets of learners most likely to benefit. This requires selection in some form.
  • Selection operates at many different levels and in many different contexts, including between-school, within-school and within-class selection. None is inherently superior to the others and they may operate independently or in combination. Key considerations when establishing the optimal pattern of selection include need, flexibility, capacity, impact and efficiency.

Note that the principles recognise the contribution that selection – in its various forms – may make to the effective education of disadvantaged high attainers.

There is, however, a warning that investing in new buildings, new (selective) institutions, is unlikely to be the most efficient use of scarce resources.

The selection green paper unhelpfully conflated arguments for increased selection and strategies to increase the number of quality school places.

It is critical to understand that selection may be on the basis of both high attainment and socio-economic disadvantage, so avoiding the risk that such provision will eventually become dominated by the ‘sharp-elbowed middle classes’.

There are also risks associated with artificially broadening the definitions of high attainment (emphasising equity at the expense of excellence) and disadvantage (for example by including ‘ordinary working families). Both dilute impact by spreading the available resource too thinly.

Not only does selection take many different forms, it is also widely used in many jurisdictions as an instrument of differentiation. We should not forget that Singapore, which leads all international comparisons studies, relies quite heavily on different forms of selection to achieve those outcomes, nor is Shanghai untainted in this respect.

The key question is how to adopt different forms of selection – perhaps combine them – in such a way that: the right cohort is selected; there is sufficient flexibility to respond to changing circumstances; and the benefit to those selected is maximised while any negative impact on those not selected is minimised. Existing sub-optimal practice often falls down in one or more of these respects.

But, as hinted above, selection (in its broadest sense) need not be on the basis of high attainment – it can be (and often is) applied on the basis of many different variables including low attainment, socio-economic disadvantage, SEN, EAL and so on. Those who complain about it only tend to do so when high attainers are the beneficiaries

In an ideal scenario, it should be feasible to select simultaneously every learner in a cohort for an additional high quality educational opportunity tailored specifically to their needs. Everyone is selected.


I extended an open opportunity to debate this framework, to propose additions and amendments, since we will not make sustained and lasting progress unless there is broad consensus on the underpinning principles.

Since no comments were forthcoming, they are presented here in their original format. I remain ready to discuss them and, if necessary, revise them. The commentary above is designed to provoke further discussion.


Statistical analysis

All efforts to present the statistical evidence for excellence gaps are hamstrung by one glaring omission in the Government’s school accountability framework: while there are identified high attainment thresholds at KS2 and KS5, there is no proper equivalent at KS4.

This makes it extremely difficult to demonstrate the full extent of the problem at GCSE – and so construct a clear and consistent narrative across the different phases of the educational ‘life cycle’.

And, because the latest published material is more substantive at KS2, the treatment below is inevitably unbalanced. It will be updated from time to time and take in new statistical evidence as that becomes available


KS2 excellence gaps

In Why isn’t pupil premium closing excellence gaps? (February 2016) I reviewed how KS2 excellence gaps had changed between 2013 and 2015, under the previous accountability regime.

To summarise:

  • Across all assessments, gaps between disadvantaged and other learners were almost invariably larger at KS2 L5+ than the equivalent gaps at the ‘expected level’ (KS2 L4+)
  • Whereas gaps at the ‘expected level’ were typically narrowing steadily, if slowly, the L5+ excellence gaps were mostly unchanged
  • So any relative improvement in the performance of disadvantaged learners was not extended to disadvantaged high attainers – and indeed their position was worsening relative to disadvantaged learners at the ‘expected level’.

For the purposes of this post I have undertaken fresh analysis under the new assessment regime, comparing 2016 and 2017 excellence gaps.

This is based on the underlying data published alongside SFR62/2016 and SFR69/2017 respectively. I have also drawn on the 2017 primary performance tables.

The analysis reveals that:

  • Between 2016 and 2017 the percentage of all KS2 learners achieving the overall ‘higher standard’ (a scaled score of 110+ in the reading and maths tests and ‘working at greater depth’ in the teacher assessment of writing) has increased substantially, by 3.3 percentage points, from 5.4% to 8.7%.
  • Improvements in the percentage of all KS2 learners achieving the ‘higher standard’ in each separate assessment have typically been even more impressive: up 5.9 percentage points in the reading test (to 24.7%), up 6.1 percentage points in the maths test (to 22.7%) and even up 2.9 percentage points in the problematic writing TA (to 17.7%). In the parallel grammar, punctuation and spelling (GPS) test, the increase is even greater – 8.5 percentage points (to 31.0%).
  • Improvements, even on this scale, might have been expected, demonstrating the ‘sawtooth effect’ in action. Yet they mask expanding excellence gaps, regardless of whether they are between FSM and all other learners, or between disadvantaged (‘ever 6’ FSM and looked after) and all other learners.
  • Chart 1, below, compares performance at the composite ‘higher standard’ and in each assessment (including the GPS test) by FSM learners and all others. It shows improvement in success rates across the board – for FSM and non-FSM alike – and, in some cases (such as the composite measure) the proportional improvement is noticeably larger for the FSM population. But, because their starting point is so much lower, the excellence gap grows regardless.
  • On the composite measure, FSM excellence gaps have increased from 4.5 percentage points in 2016 (1.6% versus 6.1%) to 6.9 percentage points in 2017 (2.8% versus 9.7%). So a 2.4 percentage point increase and, in proportional terms, a substantial change for the worse.
  • Turning to the individual assessments, the FSM excellence gap for the reading test has increased by 2.6 percentage points (from 12.1 points to 14.7 points), the FSM excellence gap for writing TA has increased by 1.8 percentage points (from 9.3 points to 11.1 points), the FSM excellence gap for the maths test has increased by 3.2 percentage points (from 10.7 points to 13.9 points) and the FSM excellence gap for the GPS test has increased by 2.7 percentage points (from 12.6 points to 15.3 points).


Chart 1: Percentages of FSM and all other learners achieving the ‘higher standard’ in the 2016 and 2017 KS2 assessments.


  • Chart 2, below provides the equivalent analysis for the comparison between disadvantaged learners and all others. It shows that, for the composite measure, disadvantaged excellence gaps have increased from 5.0 percentage points in 2016 (2.0% versus 7.0%) to 7.3 percentage points in 2017 (3.7% versus 11.0%). So a 2.3 percentage point increase, very slightly less than for FSM.
  • As for the individual assessments, the disadvantaged excellence gap for the reading test has increased by 2.6 percentage points (from 12.1 points to 14.7 points, the disadvantaged excellence gap for writing TA has increased by 1.7 percentage points (from 9.6 points to 11.3 points), the disadvantaged excellence gap for the maths test has increased by 2.9 percentage points (from 11.4 points to 14.3 points) and the disadvantaged excellence gap for the GPS test has increased by 1.7 percentage points (from 13.0 points to 14.7 points).


Chart 2: Percentages of disadvantaged and all other learners achieving the ‘higher standard’ in the 2016 and 2017 KS2 assessments


  • In sum, the excellence gaps on every one of these assessments has grown – by between 1.8 and 3.2 percentage points for the FSM gaps and by between 1.7 and 2.9 percentage points for the disadvantaged gaps.
  • According to the underlying data, in 2017, just 2,581 FSM pupils achieved the overall higher standard. Only 5% of all those achieving the higher standard were FSM. Similarly, only 6,925 disadvantaged pupils achieved the higher standard, some 13.3% of all those achieving that outcome.
  • According to the 2017 primary performance tables, approximately 6,000 schools had no disadvantaged learners at all achieving the composite RWM higher standard and, of those schools, some 620 were shown as being rated ‘outstanding’ by Ofsted.
  • By comparison with the excellence gaps described above, gaps at the ‘expected’ level (once again calculated to one decimal place from the underlying data) show for the RWM measure a slightly increasing gap for FSM pupils and a slightly reducing gap for disadvantaged learners. All gaps for the individual assessments fall within the range + 0.9 (a 0.9% increase in the gap) to – 1.5 (a 1.5% reduction in the gap). In all cases the situation is better at the ‘expected level’ than at the ‘higher standard’, as Chart 3 illustrates.


Chart 3: Percentage point changes in FSM and disadvantaged gaps at the KS2 ‘expected level’ and the ‘higher standard’ between 2016 and 2017


  • The most startling discrepancies are apparent in the FSM gaps on the maths and GPS tests, and in the disadvantaged gaps on the reading and maths tests. In each case, the slightly closing gaps at the ‘expected level’ contrast markedly with the significantly increasing gaps at the higher standard. This suggests that, across the board, schools have been far more successful in closing gaps lower down the attainment distribution then they have been at the higher level.
  • And this over a period when the Government’s preferred indicator of socio-economic attainment gaps – the Disadvantaged Attainments Gaps Index – shows a small improvement.


KS4 excellence gaps

By comparison, KS4 data is comparatively thin on the ground:

  • SFR03/2017 (Table CH4b) showed numbers of pupil premium and other pupils achieving different combinations of average grades in the English and maths EBacc pillars in 2016. Figures are also given for several preceding years but there are significant comparability issues. In 2016, 14,568 pupils achieved an average A* grade (in maths and English language or English literature), of which 976 (6.7%) were disadvantaged and 13,592 (93.3%) not. Of the 80,209 achieving an average A grade, only 8,476 (10.6%) were disadvantaged and so 89.4% were not. Changes to GCSE grades will compromise any comparison with equivalent data for 2017.
  • The Social Mobility Commission’s 2016 State of the Nation Report uses a ‘5+ GCSEs at A*/A with at least grade C in English and maths’ measure, finding that 5% of FSM students and 17.5% non-FSM students achieve this. Substantial regional variation is highlighted, with the FSM success rate varying between 9.6% (London) and only 2.8% (East Midlands).
  • In 2015 transition matrices were published comparing the performance of disadvantaged and other students in the previous year, so 2014. My analysis showed that there was a 24 percentage point gap between the proportion of disadvantaged and all other students with KS2 L5a in English who went on to achieve an A* grade at GCSE (31% versus 55%). For maths, the equivalent excellence gap was 20 percentage points (31% versus 51%). Between 2011 and 2014, this gap remained fairly stable for maths (from 19 to 21 percentage points) but in English it had doubled, from just 12 percentage points in 2011.
  • A Cambridge Assessment report (January 2017) analyses GCSE A*/A performance in 2015, including by broad socio-economic background, using IDACI data to divide the population into three equally-sized groups. The analysis includes independent school candidates and deprivation data was missing for a significant minority of the population, but this makes it possible to derive excellence gaps of a sort. For example, 13.1% of those achieving 10+ A* grades were in the high deprivation group (for whom there was IDACI data), rising to 15.7% of those achieving 5+ A* grades, 16.2% of those achieving 10+ A*/A grades and 19.9% of those achieving 5+ A*/A grades.
  • Although not strictly a high attainment measure, it is worth noting in passing the Government’s commitment to increasing the proportion of disadvantaged students entering the EBacc. The consultation response Implementing the English Baccalaureate (July 2017) confirmed an ‘ambition’ to radically increase take-up – and the social mobility plan hints at further support targeted at disadvantaged students. There is no equivalent target (or even ‘ambition’) to increase success rates and close EBacc achievement gaps, yet that must be a desired outcome. In 2015 just 9.9% of FSM students achieved the EBacc, rising to 10.2% in 2016. The gap between FSM and all other students remained unchanged at 16.7 percentage points. For disadvantaged students the gap closed marginally, from 18.1 percentage points to 18.0 percentage points. The imminent 2017 data might be expected to show a substantial increase in FSM/disadvantaged success rates and a substantial fall in the concomitant socio-economic achievement gaps, but the changes to GCSE grades, initially for English and maths, will once more create comparability issues. The introduction of an EBacc average points score measure in 2018 should provide a helpful additional means of isolating socio-economic achievement gaps. This data will be shared with schools in 2017 but not published.

The dearth of GCSE data is directly attributable to the failure to introduce a suitable high attainment measure in the secondary accountability framework. I can find no rational justification for this omission, or any explanation why it was deemed more essential at KS2 than at KS4.

For excellence gaps to be monitored effectively, this situation must be rectified. The most obvious solution is to introduce an Attainment 8 ‘higher standard’ threshold of 70+, equivalent to an average GCSE grade of A/7.

Alternatively, SFRs might report the proportion of disadvantaged and all other learners achieving an Attainment 8 average point score, of 7+, following the new EBacc methodology noted above.

The limited dataset set out here suggests a pattern similar to that at KS2: excellence gaps are substantial and are typically increasing, so failing to reflect the limited progress shown by the secondary disadvantaged attainment gaps index.

Hence disadvantaged high attainers lose out, not just in relation to advantaged high attainers, but in relation to lower attaining disadvantaged students as well.


KS5 excellence gaps 

KS5 excellence gaps tend to be less substantial because all students are comparatively high-attaining, having been expected to reach required performance thresholds at KS4 in order to continue into KS5.

Nevertheless, performance differences at the highest levels remain substantial: 

  • A DfE statistical ad hoc note A level Attainment Characteristics (March 2017) provides data for students who were FSM three years prior to completing KS5 in 2016. Amongst the majority (82.2%) for which there was matched data, 4.9% of the FSM students achieved 3 A grades or better at A level, compared with 11.0% of all other students, giving an excellence gap of 6.1 percentage points.
  • The underlying data shows the trend, from 2011 to 2016 for FSM students achieving this outcome and from 2013 to 2016 for ‘ever6’ FSM students. On the FSM measure the largest gap was 6.6 percentage points in 2012. The smallest gap was 5.4 percentage points in 2015, so the 2016 outcome was something of a reversal. The ‘ever6’ measure tells a similar story, with the smallest gap being 10.0 percentage points in 2015, compared with a 2016 gap of 11.4 percentage points. This is illustrated in Chart 4, below.


Chart 4: Percentage point gaps (FSM and other students ‘ever6’ FSM and other students) achieving 3 A grades or better at A level, 2011-2016


  • A DfE FoI response (August 2016) provided a slightly different snapshot of the 2014 A level performance of high-attaining students who had been FSM at the end of KS4. It stated that 4.2% of the FSM population achieved 3 A* grades or better at A level, compared with 9.2% of all other students (excellence gap of 5.0 percentage points); 3.6% of the FSM population achieved AAB or better, all in facilitating subjects, compared with 7.9% of all other students (excellence gap of 4.3 percentage points); and 5.7% of the FSM cohort secured AAB grades or better, at least two in facilitating subjects, against 12.5% of all other students (excellence gap of 6.8 percentage points).
  • A Cambridge Assessment statistical report Candidates awarded the A* grade at A level in 2016 (December 2017) uses the same methodology to compare outcomes for relatively disadvantaged students as deployed for the GCSE study referenced above. There is a warning that, since deprivation data was available for less than 60% of the population, results cannot be considered representative. Nevertheless, the study shows that just 0.85% of the high deprivation cohort achieved grades of A*A*A* or better, compared with 2.23% of the low deprivation group. The comparable percentages for AAA or better were 6.12% and 12.88% respectively. An equivalent study of the 2014 cohort suggests that all these percentages were slightly higher, but the gaps were similar in size.
  • UCAS end of cycle data for 2017 provides the A level point scores (A* = 6, A = 5 and so on) by POLAR3 quintile of UK-domiciled applicants (excluding Scotland) who are aged 18 and hold at least three A levels. The data is available for every year since 2010. The percentage of all students scoring 18 points and drawn from quintile 1 (most deprived) has varied between 3.1% and 4.4% throughout the period, reaching its high point in 2017. The corresponding percentages for quintile 5 (least deprived) have varied between 51.3% and 53.6%. Excellence gaps have changed very little.

More data should become available shortly because the 2017 performance tables include separate indicators against all the headline measures for disadvantage – ie for students who were in receipt of pupil premium in the final year of KS4.


HE destinations excellence gaps

Here the twin problems are limited consensus on the most appropriate deprivation indicators and the shortcomings of the various measures of HE selectivity.

With those ‘health warnings’ in mind:

  • I routinely give top billing to DfE’s KS5 destination data, widely ignored in the HE sector because it relies on FSM and pupil premium deprivation measures. The Government typically quotes the increasing proportions of disadvantaged students progressing to Russell Group and/or high-tariff universities but, while these have increased year-on-year since 2011, so that Russell Group FSM numbers are up 170% and high-tariff numbers up 112% compared with 2011, the same cannot be said of Oxbridge. Although the overall improvement is respectable, at 83% compared with 2011, progress is once again faltering, as the table below illustrates. A belated surge in 2015 has gone into reverse in 2016, but all these numbers remain desperately low. All published figures are rounded up to the nearest five.


2011 2012 2013 2014 2015 2016
FSM to Oxbridge 30 50 50 50 65 55


Disadvantaged numbers were published for the first time in 2015. There were 90 disadvantaged students entering Oxbridge in that year, rising slightly to 95 in 2016.

  • The excellence gaps methodology is tricky to apply in respect of Oxbridge admissions because the percentages of successful candidates are fractions of one percent, even for the non-disadvantaged populations. In 2016, for example, 0.14% of the FSM population was successful, compared with 0.75% of the non-FSM population. The gap between the two, 0.61 percentage points, is fairly typical of the last five years, during which this gap has remained between 0.58 and 0.65 percentage points. The raw numbers explain the issue more clearly.
  • The distinctions are clearer for Russell Group and high-tariff (aka ‘top third’) admissions, as Chart 5 shows. I have calculated the gaps to two decimal places using the underlying data to clarify the scale of the changes year-on-year. The chart shows that FSM gaps have been moving in the right direction since 2014. To give an example of what is being shown here, in 2016, 6.33% of the FSM cohort made it to Russell Group universities according to the KS5 destination indicators methodology, whereas 12.78% of the non-FSM cohort did so, giving a difference of 6.45 percentage points. The emerging trend for disadvantaged gaps, first published in 2015, is not quite so positive.


Chart 5: Percentage point gaps between FSM and non-FSM and disadvantaged and non-disadvantaged success rates for students admitted to Russell Group and High Tariff Universities, 2012-2017


  • The UCAS End of Cycle Report 2017 (December 2017) is also somewhat less positive. The UCAS data explorer facility shows that, since 2012, the difference between the entry rates for FSM and non-FSM students to high-tariff universities has increased year-on-year, from 6.075 percentage points in 2012 to 7.187 percentage points in 2017. This is because the methodologies are different – and because improvements in the FSM entry rate are consistently outstripped by improvements in the non-FSM rate.
  • There is a similar pattern in the relationship between the entry rates to high-tariff universities from POLAR3 quintile1 and quintile 5 respectively. The gap grew annually from 2012 to 2014, rising from 16.346 percentage points to 17.394 percentage points. There was a marginal harrowing in 2015, but the gap continued to expand in 2016 and 2017, peaking at 17.782 points in 2017. UCAS prefers to highlight alternative measures that paint a more positive picture, such as the proportional increase in the entry rates for each quintile (better for quintile 1) and the fact that students from quintile 5 are now only 5.5 times more likely to enter a high-tariff institution than students from quintile 5, down from 5.9 times as likely in 2016.
  • UCAS also now foregrounds its new MEM (Multiple Equality Measure) metric, which reflects a range of variables including both POLAR3 and FSM. The gap in entry rates to high tariff universities from MEM group 1 (the most deprived) and MEM group 5 (the least deprived) has increased year-on-year since 2012, from 20.6 percentage points to 21.65 percentage points. But the end-of-cycle report prefers to draw attention to the ratio between these entry rates – and that has fallen from 10.4 in 2016 to 9.8 in 2017!

As is so often the case, different indicators contribute to different narratives. The Government and highly selective universities both have a vested interest in frontloading those that support a narrative of continued improvement, often relying on frequent repetition to impose their version of events.

But a ‘balanced scorecard’ approach would force all of us to confront more clearly the areas in which there has been least success.


Qualitative evidence – Ofsted

In The most able students (June 2013), Ofsted examined the performance of higher-attaining students in non-selective secondary schools.

One of the key findings was that:

‘Inequalities between different groups of the most able students are not being tackled satisfactorily. The attainment of the most able students who are eligible for free school meals, especially the most able boys, lags behind that of other groups. Few of the schools visited used the Pupil Premium funding to support the most able students from the poorest backgrounds.’

The main text added:

‘Pupil Premium funding was used in only a few instances to support the most able students who were known to be eligible for free school meals. The funding was generally spent on providing support for all underachieving and low attaining students rather than on the most able students from disadvantaged backgrounds.’

In an update published two years later (March 2015), Ofsted complained of insufficient progress.

One of its main findings was:

‘Schools visited were rarely meeting the distinct needs of students who are most able and disadvantaged. Not enough was being done to widen the experience of these students and develop their broader knowledge or social and cultural awareness early on in Key Stage 3. The gap at Key Stage 4 between the progress made by the most able disadvantaged students and their better off peers is still too large and is not closing quickly enough.’

It found that only about one-third of schools surveyed were ‘using pupil premium effectively to support the disadvantaged most able pupils’.

There was a recommendation that Ofsted itself should:

‘…make sure that inspections continue to focus sharply on the progress made by students who are able and disadvantaged’.

The following year HMCI’s monthly commentary (June 2016) again bemoaned limited progress:

‘What is most depressing is that the brightest children from disadvantaged backgrounds are the most likely not to achieve their full potential. The most able children in receipt of pupil premium funding still lag well behind their more advantaged peers. They are also less likely to be entered for the English baccalaureate (EBacc) than other bright pupils and when they are entered, are less likely to achieve it.’

HMCI’s 2015/16 annual report (December 2016) noted:

‘For pupils who are both most able and from a disadvantaged background, the quality of teaching and the determination of a school to stretch and challenge these pupils is essential if they are to realise their potential…

…In those non-selective schools with the smallest proportions of the most-able disadvantaged pupils, only 48% of these pupils made expected progress. By contrast, in non-selective schools where the proportions of most-able disadvantaged pupils were highest, 69% made expected progress…

…Areas with selective schools represent some of the areas with the highest proportions of most-able pupils reaching top grades at the end of secondary school. However, some of these areas have very substantial differences in achievement when family background is taken into account.’

Ofsted’s concern at slow progress contrasts with some evidence of teacher complacency.

The NFER Teacher Voice Omnibus (January 2017) showed that 18% of teachers were ‘very confident’ in their school’s ability to stretch the most academically able pupils, while a further 59% were ‘fairly confident’.

Such misplaced confidence has to be tackled head on if we are to make significant progress.


Towards a set of key excellence gap indicators

It would be helpful to adopt a smaller set of ‘excellence gap key indicators’ to enable progress to be monitored annually.

This involves some difficult choices – and no indicator is perfect – but I propose the following:

  • Percentage point gap between disadvantaged and all other learners achieving the composite higher standard (RWM) and comparison with this gap in previous years. 
  • Percentage point gap between disadvantaged and all other learners achieving a new Attainment 8 higher standard of 70+ and comparison with this gap in previous years. 
  • Percentage point gap between disadvantaged and all other learners achieving AAA grades or better at A level and comparison with this gap in previous years. 
  • Percentage point gap between disadvantaged and all other learners entering high tariff universities and comparison with these gaps in previous years. 
  • Numbers of FSM and disadvantaged students entering Oxbridge and comparison with previous years. 
  • Percentage of a. primary schools, b. secondary schools, c. post-16 institutions and d. MATs where the relevant percentage point gaps are higher than the national gaps listed above.

These six indicators might also form the basis of an ‘excellence gap index’ showing outcomes for different geographical areas, so as to inform the geographical dimension of an improvement strategy.

The Social Mobility Commission’s existing Social Mobility Index is almost entirely devoid of excellence gaps data: the only concession is the inclusion of a ‘progression to high-tariff universities’ measure.

DfE’s collection of Education statistics by LA district and pupil disadvantage (October 2017) is not much better, though it does include the percentage of learners achieving the composite KS2 higher standard.

The unknown quantity here is the likely impact of the consultation on free school meal eligibility under Universal Credit – especially how this will change the preferred measures of socio-economic disadvantage and so impact on trend analysis.


The evolution of Government policy on excellence gaps


Before the selection green paper

In 2015 the Tory Manifesto promised (page 35) to:

‘… make sure that all students are pushed to achieve their potential and create more opportunities to stretch the most able.’

The DfE’s Single Departmental Plan (February 2016), intended to cover 2015 to 2020, referred to work already under way:

‘…embedding rigorous standards, curriculum and assessment, for instance by…creating more opportunities to stretch the most able’

The White Paper Educational Excellence Everywhere (March 2016) singled out the most able as one of four groups ‘neglected by the previous curriculum and accountability system’.

There was a generic commitment:

‘We will ensure that all schools can stretch their…most academically able pupils by increasing the focus on, and supporting approaches aimed at, boosting their attainment’

There were also three more specific undertakings:

  • Action to ‘ensure that the pupil premium is used effectively in all schools, for all children – including the most able’
  • Introduction of a new core ITT framework that would ‘now include a specific focus on stretching the most academically able pupils and cutting edge evidence on how these pupils can be challenged and stimulated to achieve the very highest standards’ and
  • ‘To identify and spread what works for the most able, we will investigate, fund and evaluate approaches to help the brightest students in state schools to fulfil their potential. This could include developing new, prestigious challenges and competitions at key stages 2 and 3, like the UK Mathematics Trust challenges. These would be open to all able pupils – wherever they go to school. And they could aim to provide both additional content that stretches these children beyond their school’s regular curriculum; and opportunities for further stretch outside school. We will also look for opportunities to extend the approaches used in Isaac Physics and the Cambridge Maths Education Project to other A level subjects.’

Implementation was tokenistic at best.

A paragraph was included in Effective Pupil Premium Reviews (May 2016) a guide prepared by the Teaching Schools Council, but this was merely a statement of the problem rather than any advice about effective practice:

‘Effective schools also recognise that disadvantaged pupils are not a homogenous group and employ targeted approaches for groups or individuals facing particular barriers. For example, whilst the pupil premium has focused many schools’ attention on raising the attainment of low performing pupils, more able disadvantaged pupils are at risk of underachievement too. Analysis by the Sutton Trust shows that many disadvantaged pupils who are high performing at key stage 2 fall badly behind their peers by key stage 4. This underachievement is also reflected in the low proportions of disadvantaged pupils progressing to higher ranked universities after key stage 5. Ofsted has highlighted a lack of support for more able disadvantaged pupils, particularly during key stage 3, as an area that many schools need to address.’

Note that the social mobility plan admits that ineffective use of pupil premium remains a problem in some schools.

When it appeared A framework of core content for initial teacher training (ITT) (July 2016) managed no more than a single sentence stating the blindingly obvious:

‘Providers should equip trainees to be able to inspire and provide extra challenge for the most able pupils.’

There was no reference whatsoever to disadvantaged high attainers.

The reasons for this unhelpful volte-face were never explained: I gave up trying to get a clear answer from the chair of the working group.

To date the third commitment has generated nothing at all, unless one counts the Mandarin Excellence Programme, launched in September 2016, but that was not badged as such.

(Moreover, Underground Mathematics, the successor of the Cambridge Maths Education Project, lost its Government funding in September 2017.)


The selection green paper

At this point radical proposals to enhance between-school selection were set out in the green paper Schools that work for everyone (September 2016) on the basis of the flimsy evidence already quoted above.

It is unnecessary to rehearse at length the proposals within the green paper, so this section is confined to consideration of what might emerge in the Government’s response, once it is eventually published.

All we have to date is confirmation of action to improve disadvantaged admission rates to the existing 163 grammar schools.

The GSHA agreement is reportedly delayed by the reshuffle. But something is not quite right with this story, since the GSHA claims to have submitted the draft in September 2017 and been requested to make only minor revisions in response to DfE feedback.

One assumes that publication is being held back until the rest of the green paper response is ready. The quality of the agreement will depend on how substantive and binding on member schools the GSHA commitments. This could easily prove a damp squib.

At least one PQ reply has implied that, beyond the work on admissions to existing grammar schools, there will be no new policy initiatives to strengthen selection.



But recent suggestions that the Prime Minister aims to wrest back the initiative on improving education standards – and indeed the appointment of a new Secretary of State regarded as more supportive of selective schools –  throws this rather up in the air.

Any reforms requiring legislation will almost certainly be placed on the back burner. So there will not be a new tranche of secondary grammar schools, on top of the 163 already in existence.

That also rules out super-selective grammar schools and pupil premium grammar schools – both ideas floated when the green paper was published.

There are anyway strong logistic and resource arguments against this approach: building new schools is very expensive, diverts resource into capital and away from teaching and learning, and inevitably creates a ‘postcode lottery’.

But there is little to stop creeping expansion via ‘satellite’ institutions and straightforward PAN increases. An annual £50m expansion fund was announced in the 2016 Autumn Statement and seems likely to be confirmed. Recent BBC analysis has shown that grammar schools have been expanding regardless.

Rather than reversing this trend, opposition interests might focus more constructively on ensuring that all expansion is conditional on substantive improvements in admission rates for disadvantaged learners.

Those who oppose selective schools on ideological grounds would do well to steer clear of advocating outright abolition because most of the existing 163 have strong local reputations and powerful support networks. There are very few votes in closing them down or changing their selective character.

But there are important questions to be asked about the comparative performance of secondary grammar schools. Some are consistently excellent, but others are rather less successful, especially when it comes to closing excellence gaps. Scrutiny of the latter should increase.

Any growth in the number of selective institutions is likely to be concentrated in the 16-19 phase, although the MAT centres of excellence wheeze might supply a back-door route to new 11+ institutions which do not themselves constitute a separate school. The MAT Development and Improvement Fund might be applied to this end. Any experiments of this nature require careful piloting and full independent evaluation.

At 16-19, more specialist maths schools were once again promised in the Industrial Strategy, though it remains to be seen whether enough selective universities – historically extremely wary – are prepared to sponsor them.

We can also expect renewed interest in selective generalist 16-19 settings, on the Harris Westminster/LAE model – and possibly extension of the specialist model into other subject areas. But it is important to bear in mind the inefficiency of small specialist institutions – and understand that they will not necessarily suit all disadvantaged high attainers.

These institutions will no longer find it as easy to offload students at the end of Year 12 when they are not on target for sufficiently high A level performance. And high admission rates for disadvantaged learners must always be a ‘non-negotiable’, especially if sponsoring universities wish to divert access agreement funding to this end.

When it comes to between-school selection, more widespread development of grammar streams in non-selective schools appears a front runner, since there are several influential advocates.

The problem here is that the evidence against generic streaming is strong (far stronger than against setting) – and thus far the advocates of grammar streams have not been able to demonstrate that their practice is successfully avoiding the many pitfalls. Once again, extended pilots and robust independent evaluation are essential.

There might also be interest in further promotion of setting, but it would be foolish to adopt a fixed position ahead of the EEF-funded study, which may also open up further discussion of additional, alternative models of ability grouping and within-class selection strategies.

Ultimately the best way forward is a holistic approach. The Government should commission a thorough review of all forms of selection, collating the national and international evidence for and against different methods, this to inform a few extended pilot schemes to test the viability of some of the more promising strategies.

This should eventually generate meaningful guidance for national policy-makers, local authorities, MATs and schools.

When it comes to disadvantaged high attainers, the expectation must be that the provider can produce reliable evidence that the preferred approach will:

  • Directly (even exclusively) benefit them, rather than average attainers, learners from ‘ordinary working families’ or ‘the sharp-elbowed middle classes’
  • Consistently and significantly reduce excellence gaps and
  • Avoid artificially depressing the performance of lower and middle attaining disadvantaged learners (quite possibly by making appropriate alternative selection opprtunities available to them).


The social mobility plan

This selection policy package must complement (rather than replace) the Future Talent Programme described in Greening’s social mobility plan: a decision to bin that initiative so soon after announcing it would be extremely hard to justify.

But there is certainly scope to strengthen the plan, insofar as it supports disadvantaged high attainers.

There is the small matter of that January 2017 PQ reply




The Future Talent Programme, as set out in the social mobility plan, will not support the ‘most academically able pupils across the full range of state schools’. It will not support the most academically able from disadvantaged backgrounds across the full range of state schools either.

Even if it can be said to extend across all state-funded, non-selective schools, which is a moot point, the only concession to the primary sector is the reference to primary-secondary transition.

The gap might be filled by further activity to: ‘investigate, fund and evaluate approaches to help the brightest students in state schools to fulfil their potential’, as promised in the 2016 white paper.

This might overlap with plans for a ‘curriculum fund’, first announced in the 2017 Tory Manifesto:

‘We will introduce a curriculum fund to encourage Britain’s leading cultural and scientific institutions, like the British Museum and others to help develop knowledge-rich materials for our schools’.

Greening’s final press release (6 January 2018), though mostly featuring support for English and literacy, committed to:

‘a new £7.7 million curriculum fund – delivering on a manifesto commitment – to encourage the development of high quality teaching resources by organisations, including by leading cultural and scientific institutions. These resources will help teachers deliver the government’s new curriculum while reducing workload and giving them more time to focus on what they do best – teaching.’

But there was no further detail – and that funding won’t stretch very far.


So, to summarise, there is still a substantive gap between Government commitments and Government policy announcements in support of high attaining learners, especially those from disadvantaged backgrounds.

Increasing the number of 11+ grammar schools is almost certainly not part of the solution, but enhancing one or more forms of selection might be.

Taken together, a carefully designed selection policy package and the Talent Fund Programme go part of the way towards meeting those commitments, but significant further action is needed.


The positioning of key organisations

Before turning to detailed analysis of the social mobility plan, it is necessary briefly to look across a range of Government agencies to give a clearer, more complete picture of the policy context at this point.



More recent developments at Ofsted are not encouraging – with all signs pointing to far lower priority being given to disadvantaged high attainers.

This was a personal mission for HMCI Wilshaw, but HMCI Spielman’s 2016/17 annual report (December 2017) associates shortcomings in the education of the most able exclusively with weaker schools and MATs, and makes no explicit reference to the disadvantaged most able.

This despite the fact that the social mobility plan, published only days later, points out that:

‘Disadvantaged pupils of all abilities are more likely to underperform even in otherwise strong areas and strong schools.’

In March 2017 I challenged Ofsted over its prioritisation of ‘most able education’ and did not receive an encouraging reply. This needs to change.



Social Mobility Commission

As the statistical analysis illustrates, the Commission has in the past made some useful interventions in respect of disadvantaged high attainers and fair access to the most selective universities.

But these were almost completely ignored in its final report: State of the Nation 2017: Social Mobility in Great Britain (November 2017) which relied on the flawed social mobility index to present an exclusively geographical and generic analysis.

The Board’s subsequent mass resignation is also difficult to interpret.

Ostensibly it had more to do with failure to pursue the whole government social mobility agenda that Prime Minister May had announced with so much fanfare – and with suggestions that the Commission itself had been starved of resources.

Chairman Milburn wrote in his resignation letter:

‘…I have little hope of the current government making the progress I believe is necessary to bring about a fairer Britain. It seems unable to commit to the future of the commission as an independent body or to give due priority to the social mobility challenge facing our nation…

…Individual ministers such as the secretary of state for education have shown a deep commitment to the issue. But it has become obvious that the government as a whole is unable to commit the same level of support. It is understandably focused on Brexit and does not seem to have the necessary bandwidth to ensure that the rhetoric of healing social division is matched with the reality. I do not doubt your personal belief in social justice, but I see little evidence of that being translated into meaningful action.’

Does this suggest that May’s social mobility plan was originally prepared as part of a wider cross-government strategy?

The plan itself states:

‘We are under no illusion that these issues can be tackled quickly. Nor, importantly, can they be tackled by education alone. It will require a long-term, sustained commitment across government and beyond. This plan – focusing on the role of education in improving social mobility – will therefore sit at the core of the wider work of this government to spread equality of opportunity across the full range of its activity.’

This begs the question where and when this ‘wider work’ will materialise.

Milburn goes out of his way to commend Greening’s commitment to social mobility, yet the mass resignation, less than two weeks before she published her plan, can hardly be regarded as a ringing endorsement of it.

The Board will already have seen and commented extensively on the draft plan. On Greening’s departure, Milburn was reported as saying that was ‘a loss as far as social mobility was concerned’ but, as far as I can establish, neither he nor any other Board member has made any public comment on the plan itself.

The Commission will not oversee the Greening plan – that will be done by ‘a national Social Mobility Partnership Board’ with invited external experts, one of whom will be the new Commission Chair.

Ultimately one is forced to the conclusion that the Commission’s impact has been marginal at best and that – if and when if it is reconstituted – it will continue to have very limited impact.



Offa has signally failed to make sufficient progress in strengthening fair access to the most selective universities, mainly because its default methodology is inadequate for this purpose.

Offa’s strategic guidance on developing universities’ 2018/19 access agreements (February 2017) followed the selection green paper in promoting increased collaborative work with schools and colleges to raise attainment amongst disadvantaged and under-represented groups.

But it stopped some way short of insisting that they pursue academy sponsorship or establishing new free schools, including specialist 16-19 maths free schools.

The guidance also notes:

‘We will have the greatest expectations of those institutions with the highest entry requirements and those that have the furthest to go in widening participation, because they require the highest prior educational attainment in their applicants.’

The associated press release has Offa’s Director, Ebdon, drawing these two elements together:

‘For some time, a number of universities – especially those with the highest entrance requirements – have told me that there’s a limit to what they can do to improve fair access because people from disadvantaged areas secure – on average – lower entrance grades. I’m afraid this argument just doesn’t hold water. It is precisely because there are lower rates of attainment in disadvantaged areas that universities must work in close partnership with schools to raise attainment.’

By June 2017, Ebdon was calling explicitly for more progress at Oxbridge specifically:

‘Do I think there’s fair access at Oxbridge? Well, obviously not…

If you ask me, ‘Should they be doing more?’, the answer is yes, obviously, because they have so few students from [the most disadvantaged groups], so few students on free school meals, so few students from different ethnic minorities…So yes, they certainly should be doing more, and that’s my job, to make sure that they do do more…

…Oxbridge need to make a decision…They have a series of decisions to make and I actually am legally not allowed to interfere with the admissions process. But I wish they would recognise potential more than they currently do.’

But progress in these (and other) universities continues to lag because the access agreement methodology is too blunt to secure more than marginal annual adjustments.

The official line on progress with fair access to Oxbridge was set out in a PQ reply from November 2017.




The final paragraph describes the new OfS regime, which will rely predominantly on the same blunt methodology.

Much is made of the new

‘…Transparency duty requiring higher education providers to publish application, offer, acceptance, drop-out and attainment rates of students broken down by ethnicity, gender and socio-economic background’

But it is extremely hard to see how this will help to achieve the step-change required.

The OfS consultation on its regulatory framework (October 2017) describes a model of risk-based intervention which is said to apply to social mobility (although details are extremely sketchy).

It includes this gnomic paragraph:

‘With access and participation plans the OfS may wish to consider in the future, as part of its risk-based approach to regulation, whether to introduce further innovation and flexibilities according to the level of risk posed by providers to the OfS’s expectations on access and participation.’

The combination of access and participation plans, transparency requirements and risk-based regulation is simply not up to the task of increasing substantially fair access for disadvantaged high attainers to universities with the most demanding entry requirements.

One very promising policy adjustment – the Scottish decision to work with universities to develop a system-wide set of access thresholds – has been almost completely ignored on this side of the Border.

But more systematic cross-phase reform remains essential. There is no shortage of expenditure, or of providers striving to improve matters, but they are doomed to failure because there are major shortcomings on both the demand and supply sides in this market which require urgent attention – and some degree of regulation too.


EEF/Sutton Trust

Part of the EEF’s role is:

‘…generating evidence of what works to improve teaching and learning, funding rigorous trials of promising but untested programmes and approaches.’

It and the Sutton Trust are also jointly designated as a ‘what works centre for education’, responsible for:

‘…gathering, assessing and sharing the most robust evidence to inform policy and service delivery’.

The EEF is supposedly ‘dedicated to breaking the link between family income and educational achievement’, but has consistently avoided any work focused explicitly on the needs of disadvantaged high attainers.

Then in April 2017 rumours circulated that the EEF and the Sutton Trust were jointly commissioning a literature review of support for highly able learners in secondary schools, especially those from disadvantaged backgrounds, this to inform ‘funding decisions and intervention design’.

It is unclear why this opportunity never appeared in EEF’s published list of research tenders.

The research was due for completion in June 2017, but it seems no contract was awarded, probably because the deadline was impossibly tight. Sources suggest that the work was subsequently begun but, seven months on, nothing has been published.

The Sutton Trust’s own research record in this field is best described as ‘limited’. Excellence gaps have not featured.

The Trust has lobbied for a ‘highly able fund’ – clearly the inspiration behind the Talent Fund Programme – but this recommendation has been tacked on to its published research with no proper explanation of how one derives from the other, and no detail about how the fund would operate.

The Trust’s website says:

‘Our Global Gaps report shows that while England’s highest achievers consistently score above the OECD average across English, maths and science, bright but poor pupils lag behind their better-off classmates by around two years and eight months.

The government should create a new ‘highly able fund’ to trial the most effective approaches to improving attainment and aspiration for the most able comprehensive school students.

This fund should then be used to develop an evidence base of effective approaches for highly able pupils and ensure training and development for teachers on how to challenge their most able pupils most effectively.

Additionally, the government must make a concerted effort to lever in additional support from universities and other partners with expertise in catering for the brightest pupils.’

Until now no effort has been made to establish the baseline – what evidence is already available. That task ought to have been undertaken by the EEF some time ago, drawing on published data and research (much of it analysed in detail on this blog and its predecessor).

The derivation of the talent fund commitment is made explicit in Secretary of State Greening’s launch speech (14 December 2017):

‘And for those bright children from less well-off backgrounds who need a bit of extra help to fulfil their potential, I’m announcing a £23 million Future Talent programme, something I know the Sutton Trust has called for, for some time.’

One suspects that neither the Trust nor the Government were particularly interested in canvassing wider views to inform the latter’s policy development.


Across the board then, sustained commitment to improving the attainment and progression of disadvantaged high attainers has been wanting.

For excellence gaps to be closed, all parts of the education system have to work collaboratively, systematically and in a sustained fashion to that end. This is especially true at the centre: there is no leeway for any central organisations – others just as much as those named above – to soft-pedal in addressing this priority.


The Talent Fund Programme

The social mobility plan refers to:

‘…a new £23 million Future Talent Programme to trial approaches and present clear recommendations on ‘what works’ to support the most able disadvantaged children, particularly during key stage 3 when they so often fall behind’.

This is slightly different to the Sutton Trust’s recommendation in that:

  • It relates exclusively to disadvantaged high attainers
  • It is not apparently confined to those attending comprehensive schools – so leaving scope to incorporate partnership activity involving the 163 grammar schools (but see immediately below)
  • It prioritises KS3

The full justification for the Programme, as set out in the plan, is as follows:

‘More broadly, our historic investment to support the attainment of disadvantaged children through the pupil premium – coupled with the step change in the quality of evidence that schools can draw on through the EEF – is increasingly showing real reward. However, there remains further to go to embed best practice across the system, with some schools not yet using this funding effectively. This can be especially true for their more able disadvantaged pupils, who fall back compared to their more affluent peers, even after a promising start. And crucially, disadvantaged pupils of all abilities are less likely to be entered for the core EBacc subjects that can open doors to later opportunity, despite evidence from the Sutton Trust that this can help to close the attainment gap.’

On 14 December 2017 initial tender documentation for a Future Talent Fund was published on Contracts Finder.

This does not explain the distinction between the Future Talent Fund and the Future Talent Programme, as described in the plan. There is scope for the Programme to extend beyond the Fund, but it is clear that the £23m budget is devoted exclusively to the latter.

The purpose of the tender is to identify a contractor to manage the Fund. The summary description defines this role as including:

‘…warming the market for potential projects; assessing bids and recommending to DfE the funding of projects; managing approved projects; procuring independent evaluation, and disseminating the findings.’

At first sight this seems exactly the role of the EEF, yet previously Government funding has been directed to the EEF without a competitive tender.

The attached document provides a little more detail, specifying that the task is:

‘…to test and evaluate ‘what works’ to support the most academically able, disadvantaged students in non-selective, state-funded secondary schools in England to fulfil their potential. The Future Talent Fund will build an evidence base on good practice for this cohort, by trialling projects in schools, identifying clear delivery mechanisms for provision, and encouraging more evidence-led interventions for spending Pupil Premium funding.’

So the Fund (if not the Programme) is after all confined to ‘what works…in non-selective schools’. This places it firmly in the context of the January 2017 commitment ‘to support the most academically able pupils across the full range of state schools, particularly those from the most disadvantaged backgrounds’.

There is a presupposition that the evidence base underpinning support strategies for disadvantaged high attainers in non-selective secondary schools is entirely lacking. There is no built-in requirement to undertake research to establish the baseline.

There is also an assumption that the projects trialled will be fully replicable, fully scalable and completely sustainable, with or without additional funding, though the implication is that Pupil Premium will meet all costs.

There is no further clarification of the phrase ‘support…to fulfil their potential’, so these projects need not be attainment-raising, and may not therefore contribute – even indirectly – to the closing of excellence gaps.

The document also indicates that the KS3 focus extends to ‘the transition between key stage 2 and key stage 3’.

The predominant focus on KS3 reflects the Ofsted publications cited above, but it is unduly narrow. The remainder of the social mobility action plan covers four successive ‘life stages’ and the analysis above shows that excellence gaps are well-established by the end of KS2.

Additionally, there will be particular emphasis on ‘category 5 and 6 areas (based on the Social Mobility Index)’, though this is not further explained. The footnote refers to the ‘achieving excellence areas’ methodology, even though none of the standards-based indicators is directly relevant to the most able disadvantaged, let alone to excellence gaps.

The nuts and bolts of the tendering process are also noteworthy:

  • The tender was launched on 14 December 2017 and closed on 19 January 2018, so allowing only 36 days for the entire process, much of it spread across the Christmas holidays. The number of working days available to potential bidders is minimal.
  • A two-hour briefing session for prospective suppliers took place on 5 January 2018. The Q and A from this event was not published on Contracts Finder – the supporting document says:

‘All future information will be issued to potential suppliers via the Bravo Solutions e-procurement platform.’

This is different to the Redimo e-procurement service advertised on the DfE website. The terms and conditions associated with the platform appear to restrict participants from sharing this information more widely. I could not find OJEU documentation on TED.

  • A 2-year contract will be awarded, beginning on 1 April 2018, with a possible 2-year extension ‘contingent upon supplier performance and the availability of funding’. Only £18m is guaranteed by the Government, the remaining £5m to be raised by the contractor ‘from a variety of alternative sources, but not from state-funded schools’. The contractor will receive £2m of the Government funding.

The ratio between Government funding for activity and for the contractor’s running costs appears unusually generous to the contractor, which will be paid £1 for every £9 of funding administered.

The tight deadlines and the secrecy bear all the hallmarks of an artificially restricted competition. The procedure described seems inconsistent with the spirit – if not with the specific terms – of current government guidance on procurement transparency.

It is hard to escape the conviction that this competition has not been designed to attract the widest possible field of potential contractors, but specifically to narrow the field, quite possibly to a single potential contractor.


The wider social mobility plan

The Talent Fund Programme appears in a section of the plan devoted to ‘closing the attainment gap in school’.

The broader design of the plan foregrounds area-based interventions under the strapline ‘no community left behind’.

What is described as ‘broad place-based targeting’ will permeate all dimensions of the plan – witness the reference to ‘category 5 and 6 areas (based on the Social Mobility Index)’ in the tender documentation for the Future Talent Fund.

Alongside this there will also be:

‘…more intensive and coordinated support for a limited number of areas with entrenched challenges, through the Opportunity Areas programme’.

It is unclear why Opportunity Areas are not mentioned explicitly in relation to the Talent Fund.

There are 12 Opportunity Areas to date. The document refers to upcoming:

‘…plans for how we will go further and extend the impact of our Opportunity Areas to other parts of the country’

but it does not commit specifically to increase the number of Opportunity Areas or the budget available to support them.

The Opportunity Areas initiative is progressing very slowly indeed.

The first six were announced in October 2016 and did not publish their delivery plans until 12 months later. The remaining six were announced in January 2017 and delivery plans were published in January 2018.

None of the first tranche of six published delivery plans features support for disadvantaged high attainers (though, in fairness, half do mention associated National Collaborative Outreach Programme projects). Targets are invariably expressed in terms of supporting more learners to achieve the expected standard.

Only one of the second tranche of delivery plans features this population – that for Stoke on Trent – including a focus on ‘high ability’ children at the KS2/3 transition point, so following the lead in the wider social mobility plan.

Underpinning this geographical dimension, the plan defines four ‘life stage ambitions’, which will apply in the geographical priority areas but also nationally.

Policy commitments are brigaded under a series of ‘challenges’ applying within each of these age-related headings:

  • Early years: Ensuring more disadvantaged children are able to experience a language rich early environment; improving the availability and take-up of high quality early years provision by disadvantaged children and in challenging areas; and improving the quality of early years provision in challenging areas by spreading best practice.
  • Schools: Improving the quality of teaching in challenging areas and schools; Improving the school improvement offer in more challenging areas; and Supporting pupils from less advantaged backgrounds of all abilities to fulfil their potential.
  • Post-16: Creating high quality technical education options to improve the choice for young people at age 16; investing in the further education sector; and ensuring young people from disadvantaged backgrounds access the highest quality provision.
  • Careers: Collaborating with businesses large and small to widen opportunity, and drive up local skills and productivity; improving the quality and availability of good careers guidance and experiences, targeting ‘career cold spots’; and ensuring those in lower paid work are able to re-train to move into more rewarding careers.

The fundamental problem with this segmentation is that support for disadvantaged high attainers is confined to the schools section and, as noted above, restricted exclusively to KS2/3 transition and KS3.

Fair access to the most selective universities is mentioned under the final challenge in the post-16 section, but foregrounds the existing National Collaborative Outreach programme and the transparency duty.

The only new components are:

  • Further ‘behavioural insights trials to identify how to encourage applications to the highest quality institutions and courses, including the most selective universities’. These follow the previous work published in March 2017.

Otherwise there is nothing specific to disadvantaged high attainers elsewhere in the document. Yet there is an obvious need to mainstream and co-ordinate such activity throughout the educational lifecycle.

Some sort of co-ordinating function is essential to ensure that meeting the needs of disadvantaged high attainers is integral to all relevant policy commitments, new and continuing, set out within the plan, especially:

  • Embedding support for emerging high attainers in early language and literacy support, including the role of the new national network of English Hubs, and in work to develop best practice in the reception year.
  • Securing support in the planned ‘extended programme of early career support and development across core areas of practice’ for teachers (a consultation document was published in December 2107), in projects supported by the TLIF and SSIF, in the extension of the maths hub network, in Curriculum Fund projects, in independent school and university partnerships and in improving the effectiveness of pupil premium spending.
  • Acknowledging and responding to their needs when developing all new technical and vocational routes and in all aspects of careers education.

Further work is necessary to identify progress measures which should include dedicated indicators for closing excellence gaps, drawing on the proposed suite outlined above.


The 10-point plan revisited

A year ago I set out a 10-point reform plan, in response to the January 2017 PQ reply, outlining ‘the best ways to support the most academically able pupils across the full range of state schools, particularly those from the most disadvantaged backgrounds’. This drew heavily on other work over the past few years.

Here it is, revised and updated to reflect these recent developments:

One: Prepare, consult on, refine and adopt a set of core principles to inform the Government’s approach to the education of (disadvantaged) high attainers. Develop these core principles into a succinct guidance document for schools and an updated set of non-prescriptive quality standards following the model previously published. Incorporate the substance of these standards into the Ofsted inspection regime.

Two: Develop an associated glossary of (disadvantaged) high attainment which: distinguishes clearly between ability and attainment; incorporates the excellence gaps concept; distinguishes between high attainment and high prior attainment; distinguishes the different measures of disadvantage; and distinguishes the appropriate measures of high attainment and high prior attainment at each key stage. Identify an appropriate end-of-KS4 high attainment benchmark (eg Attainment 8 score of 70+) and introduce it into the KS4 accountability regime, measuring excellence gaps against it. Ensure that there is full consistency of approach across Ofsted inspection and all performance tables. Encourage all schools and colleges to adopt this approach. Define and publish annually a suite of national excellence gap indicators.

Three: Audit – and regularly re-audit – all existing and planned education policy strands to ensure they are consistently supportive of (disadvantaged) high attainers and fully consistent with the core principles above. Focus particularly on a subset of ‘non negotiables’ – those critical and substantive policy interventions with the capacity to bring about lasting change. Ensure that all four ‘life-stages’ and all area-based dimensions of the social mobility plan reflect the needs of (disadvantaged) high attainers. Update all Opportunity Area delivery plans to  follow suit, ensuring that each incorporates specific, measurable, timebound improvement targets. Make continued funding conditional on achieving those targets. Ensure that there are similar targets for the social mobility plan as a whole and that progress against them is reported annually.

Four: Extend the remit of the Talent Fund Programme so that it covers provision for disadvantaged high attainers across KS1-KS5, with particular emphasis on KS2-KS4. Prioritise fully scalable, cost-effective and sustainable teaching and learning strategies that will close excellence gaps. Commission a fresh literature review to draw together existing evidence of effective practice worldwide. Give the EEF full responsibility for this work and hold it accountable for effective delivery. Develop and periodically update guidance on how best to spend pupil premium to meet the needs of disadvantaged high attainers.

Five: Commission from EEF a review of the full spectrum of approaches to between-school and within-school selection and how these might be used/combined to improve support for all (disadvantaged) high attainers, in such a way that other learners are not unfairly denied additional learning opportunities appropriate to their needs. Consult on, refine and adopt core principles for system-wide policy development, for local authorities, MATs and schools; provide succinct guidance for schools and MATs based on those principles.

Six: Make all expansion of selective school places (including at selective 16-19 institutions) conditional on the institution meeting demanding fair access targets to increase the admission rate of disadvantaged high attainers. Start from the presumption that at least one new place in every four (25%) is reserved for learners eligible for pupil premium (or, in the case of 16-19 institutions formerly eligible for pupil premium).

Seven: Use an annual £50m pupil premium topslice to introduce a national selection programme for all high-attaining disadvantaged learners from KS3 through to HE entry. Make this open access during KS3 with participation thereafter based on prior attainment and continuing good progress. The measure of success will be places secured on the most competitive and demanding HE courses, especially at Oxbridge. Strongly encourage universities to make contextual and unconditional offers to participants and provide matched funding. On the demand side, require schools and colleges to follow a standard needs-assessment, commissioning and review process to secure a coherent programme of additional learning and support opportunities for their participating students. On the supply side, require the full range of providers (universities, commercial and third sector) to position and advertise their services within a common provision framework. This will bring much-needed coherence to a fragmented market without requiring heavy regulation or a centralised government-run programme.

Eight: Support the further development of enrichment and attainment raising activities for disadvantaged high attainers to fit within this common provision framework, encouraging MATs, selective schools, independent schools and universities to pursue this route to attainment-raising rather than sponsoring academies and founding new free schools. Dedicate part of the new Curriculum Fund to this purpose.

Nine: Introduce a national network of some 50 high attainer hubs with a remit to help close excellence gaps and introduce peer-led ‘high attainer reviews’ on the pupil premium review model. Fund the hubs to pursue both national and local improvement priorities and to work with parallel maths and English hubs on common priorities. These should include a review of ‘maths mastery’ practice and associated guidance to ensure that high attainers’ needs are fully met. Ensure that selective school outreach is integrated within these hubs so that, as far as possible, it is available nationwide.

Ten: Invite tenders for an independent national research centre based in a leading university. It would specialise in publishing data and research relating to high attainment, especially the closing of excellence gaps, and fair access to the most selective universities. It would also undertake programme evaluations and develop guidance. It would publish an annual review assessing Government progress against the national indicators.



The persistent underachievement of disadvantaged high attainers at all key stages, as identified through excellence gaps analysis, has been neglected for far too long.

The recent history of policy implementation has been wanting and, although there is new provision within the Greening social mobility plan, the Talent Fund Programme is too narrowly focused and will have only limited impact.

Only co-ordinated, coherent, system-wide reform will bring about significant and lasting improvement. There is scope to apply some forms of between-school and within-school selection as part of this reform programme, subject to robust safeguards.

But the Government’s approach to selection-based reform, thus far, has been haphazard and not backed up by meaningful evidence. It needs to review the full panoply of selection options and so establish a more consensual way forward.

The 10-point plan, above, places considerable emphasis on the publication of succinct national guidance documents and the pursuit of clearly-defined national objectives – two potentially effective improvement strategies which recent governments have neglected for no good reason. It is time to redress the balance.

The plan is designed to circumvent my ‘Eight Types of Ambiguity’ which, taken together, explain why a laissez faire approach to improving the education of (disadvantaged) high attainers – where each school is almost entirely free to define its own effective practice – is doomed to failure.

A few exceptional schools can work in this way, but most need a ‘flexible framework’ to guide their approach. The notion of a fully school-led, self-improving system is still some way off, if indeed it can ever be achieved completely. Meantime, a greater onus falls on key central organisations – and the centre should not shirk this responsibility.

This analysis has also highlighted two examples of procurement processes which have fallen some way short of best practice. Fully competitive procurement exercises which strive for maximum transparency will normally produce the best results. Anything that falls short risks being interpreted as sharp practice, or even cronyism.

As the social mobility plan recognises, effective implementation places heavy reliance on partnership and collaboration across the full range of interested parties.

Similar principles also apply to effective policy design: it behoves policy-makers to reach out beyond the usual suspects when seeking advice to inform their decisions. The usual suspects do not have a monopoly on wisdom – and may be too wedded to the pursuit of their own agendas.



January 2018

2 thoughts on “How should England close its excellence gaps?

  1. On the SMC, we’ll find out more at the Education Select Committee hearing on Tuesday 23rd

    We know for a fact that Milburn did not resign – the Government, after months of indecision due to No 10/Greening infighting, informed him his appointment would not be renewed. He was able to frame it as a principled mass-resignation due to the Department for Education’s incompetence in not communicating this decision publicly as soon as they informed him of it.

    I suspect the others followed mainly due to wanting to move on from their unpaid Commission roles after 5 years and partly due to annoyance at how the Commission had been sidelined since the 2015 General Election, with no requests for advice on social mobility (it would not have had any formal input into the Social Mobility Action Plan), a narrowed role and an inability of Government to appoint new Commissioners to replace those whose terms expired in 2015: it was limping along with 6 commissioner vacancies for its final 2 years. It was very nearly abolished in 2015 and only saved as political cover for repealing much of the Child Poverty Act 2010, including the child poverty targets, requirement to produce a child poverty strategy and disbanding the Child Poverty Unit. If the Government had a decent majority and time for the primary legislation required, it would abolish it and replace it with a non-independent Ministerially-directed Social Mobility Unit along the lines suggested by Robert Halfon.

    Can a reconstituted Commission have a direct impact on government policy? Perhaps under a new Conservative-affiliated Chair who is seen as a philosophical fellow traveller by the Prime Minister, and with a shift away from its “set-piece headline-grabbing accountability” mode of operation to one of providing under-the-radar constructive advice and a move away from DfE to Number 10 (lots of turf wars with the officials once Greening established her own Social Mobility Unit, plus it would have greater clout in Whitehall and help persuade the PM that social mobility is about more than education for teenagers and that she needs to broaden her strategy).


Leave a Reply

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

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

Facebook photo

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

Connecting to %s