Will the national funding formula support ‘excellence for all’?

.

This post explores tensions between the ‘excellence for all’ objective now underpinning national education policy and the proposed design of the national funding formula for schools.

It is framed as an exploratory discussion document. If a convincing counter-argument emerges I am ready to adjust my views.

.

Setting the scene

‘Doing Differentiation Differently?’ mentioned in passing a shift in national policy and rhetoric, from ‘no child left behind’ to a substantively different ‘excellence for all’ position.

‘No child left behind’ was a hangover from the Coalition government which briefly survived the change of administration, whereas ‘excellence for all’ – aka ‘education excellence everywhere’ – deliberately puts clear blue water between this administration and its predecessor.

It is identified as one of three core objectives in the DfE’s Single departmental plan: 2015 to 2020:

‘Educational excellence everywhere: every child and young person can access high-quality provision, achieving to the best of his or her ability regardless of location, attainment and background.’

This apparently innocuous statement has profound implications, because we might reasonably expect it to apply consistently across the full spectrum of education policy.

It is an objective, not an aspiration. It is essentially guaranteeing a universal entitlement for learners to access the education they need to maximise their achievement.

It applies without reference to:

• Prior attainment, so high attainers have the same, equal right to access such provision as those achieving at or below national performance benchmarks, wherever those benchmarks may be set. 

• Background, so disadvantaged high attainers have the same, equal right to access such provision as disadvantaged (and advantaged) learners with lower prior attainment. 

• Location, so it should operate regardless of the learner’s place of residence and educational setting (as long as it is state-funded).

The final element is significant. It would not be sufficient to make such provision in the broad vicinity of the learner through an exclusively structural intervention – for example by creating a fresh tranche of selective 16-19 free schools or even schools for the gifted as floated by James Frayne of Policy Exchange.

That may be part of the package however.

Nor would it be sufficient to rely on the intervention regime for failing and coasting schools, though that may also be part of the package.

The Single Departmental Plan again:

‘Now we want to see that educational excellence extended to every corner of the country. To do that we will ensure that all schools are fairly funded and where they are failing or coasting and not pushing every child to reach their potential, we will step in to turn the school around.’

This post is principally about that reference to the funding formula. I’ll return to it in a moment.

.

A 2016 White Paper

A blend of free schools, intervention and funding reform won’t be sufficient to take this trick. The Government’s as yet unfulfilled manifesto commitment says:

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

The Plan also refers to:

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

The creation of those opportunities – available in (or from) every state-funded primary and secondary school – is the missing link. Secretary of State Morgan has herself declared.

.

.

I need to resurrect my ‘Proposals for a 2015 White Paper: Most Able’ and change the year in the title.

For it seems increasingly likely that a Schools White Paper will appear imminently. One hopes it will not be confined solely to explaining how another Prime Ministerial commitment will be realised:

‘I want every school in the country to have the opportunity to become an academy and to benefit from the freedoms this brings.’

There is a raft of other business – the national funding formula, admissions reform, ‘compulsory EBacc’, Year 7 resits, reception/KS1 baseline, STEM, most able – and, rather than relying entirely on a sequence of one-off consultations, the Government needs to show how these knit together in a convincing policy narrative.

We need to understand how the full gamut of policy consistently supports the three central objectives in the new plan, ‘education excellence everywhere’ amongst them.

We also need to appreciate the seamless connection with parallel policy in the higher education and life chances white papers being overseen by BIS and DWP respectively.

.

The nub of the argument

Within the stage one consultation document on a schools funding formula, one of the four proposed building blocks is ‘additional needs’.

The consultation proposes using three factors in combination to allocate funding for additional needs:

• Socio-economic deprivation

• English as an additional language (EAL) and

• Low prior attainment

There will be no requirement to spend specific amounts on learners with these needs, but the adoption of specific proxy factors within the formula will send implicit messages to that effect. If the formula is properly transparent, schools will be able to see the weightings and will be inclined to apply them.

This post argues that the proposal to use low prior attainment for this purpose is inconsistent with the wider objective of promoting ‘educational excellence everywhere’, enabling every learner to access high quality provision regardless of prior attainment.

Further it argues that, historically, this factor was introduced to stand as a proxy for ‘high incidence low needs SEN’, enabling the identification of a ‘notional SEN budget’. The high needs consultation expects to eliminate that notional budget in due course, so raising the question whether the proxy is still needed.

Even if additional costs associated with ‘high incidence low needs SEN’ must still be accommodated within the formula, it ought to be feasible to use a specific measure rather than a proxy. If this creates a perverse incentive to inflate the incidence of SEN, it should be feasible to introduce efficient, light-touch moderation to control it.

The only remaining justification for including a low prior attainment factor would be if the costs of supporting low attainers ‘to achieve to the best of their ability’ are necessarily higher than the costs of supporting other learners to do so – and if those differential costs are not attributable to deprivation (or EAL). That is contestable, for the reasons given below.

. 

The National Funding Formula consultation

On 7 March 2016 the Government published two consultations with deadlines for response of 17 April, a little less than six weeks ahead:

• Schools national funding formula: Government consultation – stage one intended to establish the underpinning principles, the elements from which the formula is constructed and the factors included within it.

• High needs funding formula and other reforms: Government consultation – stage one intended to clarify principles, proposals and options for the high needs formula, but also the approach to ‘high incidence low needs’ SEN in the schools national funding formula.

Responses to these first consultations are to be published before second stage consultations are launched. These will address the weighting of factors within the schools and high needs formulae and, in the case of the latter, ‘the detail of the other proposed improvements’.

A clutch of background documents was also published on 7 March:

• Schools and high needs funding reform: The case for change and consultation summary

• School revenue funding: Current funding arrangements

• National funding formula: Equality analysis

• High needs funding consultation: Technical note

• Research on funding for young people with special educational needs: Research report (originally published in July 2015)

There was a Written ministerial statement and a press release and PUSS Sam Gyimah also gave a speech on 10 March which set out the Government’s vision:

‘…we must approach education with the aim to deliver a level playing field for all pupils. It must be an approach in which all our schools, teachers and, most importantly, pupils can reach their full potential.

‘It means a funding system that is wide-eyed to factors that impact educational success – be that special educational needs, disability or economic disadvantage.’

‘Our most important role is to ensure the right level of funding supports each and every pupil. Of course, schools and areas must be funded adequately, but our most important principle is to get the right level of funding to each pupil. We will achieve this by aligning our funding principles to educational needs.’

.

The case for a low prior attainment measure

The schools formula consultation document proposes seven principles to underpin the formula. It must: support opportunity, get funding straight to schools, be fair, efficient, transparent, simple and predictable.

Two principles are particularly relevant here. The gloss on ‘support opportunity’ says:

‘The funding system should support schools and local authorities to extend opportunity to all pupils. Reliable and robust data about pupils’ characteristics should determine the resources their schools receive, in a consistent way around the country. The funding system should support educational excellence everywhere.’

The gloss on ‘fair’ is:

‘Each school’s funding should directly relate to the characteristics of the pupils in it. Variations in funding should be due to differences in pupil characteristics and specific circumstances – not due to historic allocations based on out-of-date data. We do not believe it is possible to achieve true fairness while we have 152 different local school funding formulae. Instead, the great majority of each school’s funding should be determined by a single, national formula which uses objective and robust measures of pupil and school characteristics.

There is no reference to perverse incentives within the principles.

The four ‘building blocks’ proposed for the formula are: per-pupil costs, additional needs, school costs and geographic costs.

The rationale for the ‘additional needs’ block is stated thus:

‘We believe that the national funding formula should use data about pupil characteristics to target funding to schools so that they can better support pupils who are likely to face additional barriers in reaching their full potential. Pupils with additional needs can particularly benefit from additional teaching, specialist intervention or materials, extra pastoral support, involvement in multi-agency working and many other types of support from their school. A funding formula that only allocated a flat amount to each pupil would make it extremely difficult for schools to support fully all pupils to reach their potential, and deliver educational excellence everywhere.’

Despite a general objection to reliance on historical practice, the consultation document proposes to select these factors from those currently used in local formulae. The national formula should utilise the minimum number to be clear and understandable yet also accurate.

These factors should:

• Be linked to significant costs in schools

• Be based on accurate school-level data

• Not be associated with perverse incentives to increase funding and

• Be tied to pupil characteristics wherever possible.

Three factors are proposed for additional needs: socio-economic deprivation, EAL and low prior attainment. They are the same three used to allocate additional funding in 2015 (see below).

Socio-economic deprivation is properly associated with additional needs. It is also a (crude) proxy for low prior attainment (two causes are named) and for SEN.

The inclusion of EAL is more debateable, though it may indicate additional needs and provide a second, even cruder proxy for low prior attainment. (The consultation says an alternative and more accurate measure of English language proficiency would be preferable and is actively being explored.)

Alongside these two proxies, it is also thought necessary to include a direct measure of low prior attainment. The document says:

‘We believe that it is important for a national funding formula to adjust schools’ funding in response to the attainment characteristics of their pupils.’

But it provides no further justification for adjustment in favour of low prior attainment, presumably relying on the preceding statement that it is associated with additional needs and barriers, and especially the additional support to overcome them that carries an additional cost.

These needs and barriers cannot be associated with deprivation or EAL, since otherwise there would be ‘double counting’. Maybe they are associated with other undeclared additional needs (SEN for example) which can be readily disaggregated, most likely because they affect non-deprived, non-EAL low attainers.

Evidence of the impact of low prior attainment on future attainment is discussed with reference to defunct national benchmarks:

‘In 2015, only 7% of pupils who did not reach level 4 at key stage 2 went on to achieve the expected standard at key stage 4 (5 A* to C including English and maths), compared with 52% of pupils who achieved level 4 and 91% of pupils who achieved above level 4 at key stage 2.’

There is recognition that these are about to change.

For secondary schools the formula will rely on the as-yet-undefined ‘new expected standard’, which cannot be confirmed until summer 2016.

For primary schools it might use the reception baseline, or it might not:

‘In our response to the primary assessment and accountability consultation in 2014 we signalled our intention to use the reception baseline assessment to trigger low prior attainment funding. As changes to primary accountability are finalised and take effect, we will keep the primary low prior attainment factor and its data source under review. We will publish further proposals for revising the low prior attainment factor for primary schools in due course.’

Such uncertainty is not entirely consistent with the insistence on ‘robust measures’.

.

Development of the ‘high incidence low needs SEN’ proxy

The proxy is a relatively recent invention.

An April 2011 consultation on the rationale and principles of a national funding formula invited respondents to suggest the pupil factors it should contain.

The consequent proposals reveal that, under the school finance regulations then in force, schools were permitted to allocate funding to schools on the basis of 15 criteria including prior attainment (and also ‘gifted and talented’). The consultation proposes rationalisation.

It proposes a factor for additional educational needs, most probably confined to deprivation. Analysis of underachievement patterns does not support the inclusion of additional factors for EAL and ‘Under Performing Ethnic Groups’. However, consideration will be given to additional funding for EAL, confined to the first few years that the learner is within the school system.

Low attainment is not mentioned. But it is necessary to introduce a ‘general assumption about the notional funding for low cost SEN within the mainstream funding blocks’. A sum of £6,000 is proposed:

‘This would mean that pupils whose needs cost less than £10,000 in total (the £6,000 additional needs and £4,000 basic) would not, for the purpose of the national formula, be deemed to be high need and the assumption would be that the funding would be found from the mainstream funding blocks.’

A 2012 consultation ‘School funding reform: Next steps towards a fairer system’ claims that there are in fact 37 possible formula factors for schools block funding and proposes reducing them to 10, these to include: deprivation, looked after children, ‘low cost high incidence SEN’ and EAL (for three years after the pupil enters the system).

The justification for the inclusion of LAC is that:

‘Looked after children often need additional support to help them to achieve… The nature of their care arrangements means that these children may not be eligible for FSM and so would not attract additional funding on that basis. Whilst they might attract some funding through the prior attainment factors, support for this group of pupils often spans wider than academic achievement and can include services such as pastoral care and liaison with social services.’

These ‘prior attainment factors’ are associated exclusively with low cost high incidence SEN. The document says local authorities currently use a variety of allocation methods:

‘For example, some use a complex combination of proxies such as FSM, pupil mobility and prior attainment. Others use a separate SEN lump sum to capture a wide range of SEN need in a particular school.’

It proposes simplification:

‘We have considered a range of measures to identify these pupils. We expect that the notional SEN budget will include an identified element of the basic entitlement and, in many cases, of deprivation funding. But we do not think this provides a complete answer. However, whilst there is no ‘perfect’ way of identifying pupils with low-cost SEN, we are not keen on allowing a measure which is based on direct identification of pupils as having SEN as this can be subjective…

Prior attainment, if used appropriately, can provide us with a good proxy for many SEN pupils not identified through a deprivation measure…’

The criteria that determine what constitutes a good proxy are not discussed. The obvious riposte to the point about subjectivity is that, if direct identification is robust enough for the school census, surely it is robust enough for the funding formula too.

The point is restated in similar terms in ‘School Funding Reform’ (June 2013):

‘In 2013-14, local authorities were able to target funding to schools for pupils with low cost, high incidence special educational needs (SEN) partly through the optional prior attainment factor. We acknowledged that there was no perfect way of identifying pupils with low cost SEN but that prior attainment provided us with a reasonable proxy for some kinds of SEN. As prior attainment will not identify all low cost SEN, we invited local authorities to use this factor alongside other factors (such as deprivation, for example) in order to identify a notional SEN budget for all mainstream schools in the local authority area.’

So ‘good’ has been downgraded to ‘reasonable’.

This document confirms retention of an EYFSP measure for primary, even though:

‘Some…local authorities were concerned that the EYFSP measures development rather than attainment and could be an unreliable measure to identify children who would need additional support to attain well.’

For secondary, it announces a shift from a measure based on KS2 L4 in English and maths to KS2 L4 in English or maths, which counts in roughly double the proportion of learners (20% rather than 10%).

Some respondents:

‘…suggested that the Department should use deprivation factors alone as a proxy, or that we should consider a system where schools were allocated funding based on the actual number of pupils with a statement of SEN.’

The 2014 consultation on the allocation of an additional £350m to the least fairly funded areas, proposes a minimum funding level for the core amount per pupil (age weighted pupil unit) and four characteristics: deprived background, LAC (still), EAL and low prior attainment.

This is the first occasion on which low prior attainment is not mentioned in the context of notional SEN budgets, so the document is unclear whether it is being included in its own right or as a proxy for SEN.

The consultation response records that 48% favoured the proposed characteristics while 42% were opposed and 10% were unsure. But it says:

‘We selected the characteristics we think matter most to the attainment of pupils…We remain convinced that the proposed characteristics are those that should underpin the fair funding of pupils and schools – a view shared by a majority of those responding to this question.’

.

‘High incidence low needs SEN’ in this consultation

The ‘Research on funding for young people with special educational needs’ (ISOS, July 2015) argues that prior attainment and deprivation ‘are two of the most powerful factors in predicting SEN’ but suggest targeting might be improved through the inclusion of an additional factor: a 0-15 disability living allowance claimant measure.

It reports doubts in the system about a proxy driven approach:

‘…neither schools nor local authorities feel that it is possible to use proxy factors to model the incidence of SEN with 100% accuracy. The situation is exacerbated by the fact that schools’ attitudes towards, and aptitude for, meeting the needs of children with SEN is very varied, leading to an uneven distribution of pupils with SEN. The funding challenge is that any formula-based method of allocating resources will not be able to reflect differences in the size of a school’s SEN population where these are driven to a large extent by the culture of the school in question, rather than underlying demographic and contextual factors that can be measured, such as deprivation or prior attainment.’

Further:

‘When speaking to schools and local authorities we have heard on a number of occasions that there are forms of SEN which do not correlate well with deprivation and do not result in lower cognitive ability and therefore do not necessarily lead to lower prior attainment. Some of the examples that have been given to us include sensory impairments, physical disabilities, complex emotional and mental health needs and high-functioning autism. Logically these are the types of need which may not be adequately reflected in current formula allocations.’

Indeed:

‘During the course of our research the argument was put forward by a large number of schools that the only really fair way to fund SEN would be to construct a system in which the full cost of support followed the child. Schools which proposed this advocated using actual numbers of children with a top-up or reported as requiring SEN support on the school census as a variable in the funding formula rather than proxy measures such as deprivation and low prior attainment. They recognised that this could lead to inflationary pressures, with more children being identified as requiring SEN support in order to access higher levels of funding, but felt that this could be overcome by putting in place adequate moderation mechanisms.’

The authors reject this approach as bureaucratic and expensive and are unable to suggest an efficient and workable compromise. Instead they propose the additional proxy, but also suggest that:

’…the DfE should consider removing notional SEN budgets from the funding system for mainstream schools’.

The new stage one consultation documents propose to retain the notional SEN budget ‘in the short term’ pending further work to develop an alternative approach.

The association with the proxy is mentioned:

‘‘In the current funding system, the low prior attainment factor is commonly used by local authorities as the main factor for determining how much of the core funding a mainstream school receives is notionally intended to support its SEN provision (the ‘notional SEN budget’)’

but we are again left in the dark about whether that is the intention this time round.

Low prior attainment also forms part of the proposed high needs formula which is built around five factors including deprivation and low attainment.

The high needs document argues that ‘there is a strong correlation between low attainment and SEN’. It proposes as a primary measure pupils not achieving KS2 L2 in reading and, as a secondary measure, pupils not achieving 5+ GCSEs at A*-G at KS4 ‘or equivalent standards as changes are made’.

.

Argument

Returning to the three arguments against the low prior attainment measure:

.

• It stands as a potentially unnecessary proxy for low needs SEN.

The question arises whether such a proxy is needed at all when the notional budget is dispensed with. Assuming it remains necessary, far more serious consideration should be given to the possibility of using actual data on the incidence of SEN drawn from the school census.

It should be perfectly feasible to monitor through a computerised process the SEN sections of schools’ census returns, flagging any schools where SEN numbers increase significantly, or continue to rise year-on-year. Human moderation of a sample of these schools would cost a tiny fraction of the devolved budget for SEN.

If existing SEN classifications are deemed too fragile and ‘subjective’, that begs two questions. First, why are they judged sufficiently reliable for the school census but not for the funding formula? Second, if it is really necessary, why not develop and introduce a more trustworthy measure, as with EAL?

.

• It is contestable that schools incur higher costs in educating low attainers that are not accounted for by the parallel deprivation and EAL measures.

It is conceivably more expensive to make appropriate educational provision for low prior attainers, when that provision is not caught by the other two measures.

There are two ways of coming at this issue: what counts as ‘achieving to the best of one’s ability’ (outcomes); and whether low prior attainers are a distinct special case for additional needs funding once the effects of deprivation and EAL have been discounted (inputs).

Beginning with the first, If the definition rests on a standard national benchmark – say a given Attainment 8 score or KS2 scaled score – then those who have furthest to travel to achieve it may require additional ‘catch up’ support.

But this argument becomes more tenuous if the measure is progress rather than achievement of a benchmark. It might just hold under the former assessment system, if one accepted that all learners should have been expected to make no more than two levels of progress from KS1-2 and three levels of progress from KS2-4.

That line is advanced by Policy Exchange:

‘Despite the changes to the accountability system, which seek to recognise progress as well as attainment, low prior attaining pupils will still require additional support to make equivalent gains (Datalab 2015).’

The reference is to this Datalab piece which demonstrates the relationship between expected progress and prior attainment. But, even under the old system, it was standard in most schools to expect relatively more progress from higher attainers. The national expectation was ‘at least’ 2/3 LOP.

Theoretically at least, Progress 8 is designed to bring about a radical adjustment consistent with ‘excellence for all’. It compares each learner’s outcomes against those of others with the same prior attainment. As such it is explicitly designed to encourage schools to support all their pupils, regardless of their proximity to national benchmarks.

Even if it is judged not to do so successfully, that does not undermine the policy intention.

Shifting to inputs, can one argue that provision for low attainers is typically more costly than provision for high attainers, for example? And can one still do so once much of the cost has been ‘explained’ by the other measures?

I am doubtful.

High attainers also need extra support to ‘achieve to the best of their ability’.

How are schools to pay for those ‘new opportunities to stretch the most able’? There will be an additional cost attached, regardless of whether they are provided in-house or purchased from an external provider.

These are additional needs applying to a minority of learners and therefore unsuitable for inclusion in the universal per-pupil sum.

.

• There is a tension with the wider ‘excellence for all’ objective

Applying a low prior attainment measure in the proposed national funding formula might have passed master when the watchwords were ‘no learner left behind’. Now that ‘excellence for all’ is a central policy objective it really behoves us all to think again.

If there really is no alternative to using low prior attainment as a crude proxy for SEN – which is by no means proven – the justification is simply that and it ought to be articulated as such.

If the measure is being applied because there are perceived to be additional costs associated with the education of low attainers that cannot be swept up by deprivation or EAL proxies then it would be useful to see some evidence to support that contention.

Even if there is an evidence base derived from present practice, the architects of the formula should appreciate that it represents the status quo rather than the bright new future represented by ‘excellence for all’.

Retaining such a measure sends signals that are directly at odds with that position. While not bound to allocate funding in line with the composition of the formula, those preferring ‘no child left behind’ will understand this as endorsing their standpoint.

And that would undermine the achievement of ‘excellence for all’.

So I recommend that additional needs within the formula contains two factors only: deprivation and a specific measure of ‘high incidence low needs SEN’.

If there is any case for the additional recognition of low prior attainment, it should be matched by a measure for high prior attainment, in recognition of the ‘neglect’ acknowledged by the 2016 schools white paper.

.

TD

March 2016