The National Funding Formula for Schools
Who gets what and why, and how has it changed things?
One thing we are bad at in Westminster is going back and looking at whether policies had the effects people hoped they would. This article is about something I was tangentially involved in - the National Funding Formula for Schools (NFF).
Here’s how much we spend per pupil in different parts of the country. There are massive variations.
Leicestershire, Central Bedfordshire and Wokingham are 10% below the national average per pupil. On the other hand there are a bunch of inner London authorities that are 40-50% above the national average. The overall story is essentially cities vs the rest:
London in general sees the best school results too. Though the one isn’t necessarily the cause of the other, it’s very striking that we give quite a lot more money to the places that today, are doing best. Funding certainly doesn’t seem to target low achievement much: quite a lot of those lower performing areas are also lower funded.
How did we get here?
The National Funding Formula for Schools
The variations we see in funding above exist under the National Funding Formula for Schools, established seven years ago, but they mainly predate it.
Prior to 2018/19 funding for schools was a hodgepodge of historic allocations - a postcode lottery in which the amount of funding per school varied massively, depending on random historical accidents. Similar schools with similar pupils in similar places could get quite different amounts of money.
It took several attempts to agree to the creation of a formula at all. Labour had promised a formula in 2003, but never created one. We had another go in the Coalition years but couldn’t agree. I remember Michael Gove and Dom Cummings coming out of David Cameron’s study looking unhappy after one attempt.
Having a clear formula was bound to create winners and losers, and older officials at DFE were still traumatised by the 2003 schools funding crisis.
Back in 2003, although overall funding was going up in real terms, changes to allocations caused a lot of churn, and the Education Secretary was forced to apologise after chaos in schools. One long-serving official described it to me as the equivalent of “sinking the ship on a completely calm sea.”
That experience is one reason the creation of a NFF took a long time to agree, and also explains why the NFF baked in some of the variations and priorities that were there before: to minimise churn.
The formula was introduced at the local authority level. LAs get a notional amount for each school, but are able to move money around between schools themselves, using local formulas to smooth out the transition. Over time LAs have converged more onto the national formula, and an increasing share of schools have been given the actual amount the national formula suggests.
After several false starts and years of discussion there were attempts to take the next step a moves towards a “direct” formula which would end this local redistribution: in the 2023 consultation response set out how the last government saw this working out, with a gradual tightening of requirements. It stated that: “we expect to have moved to the direct NFF within the next five years – that is, by the 2027-28 funding year at the latest.”
How does the NFF work?
Totting up how much we spend on schools is not straightforward. First, there is what you might think of as the “core” schools budget (the Dedicated Schools Grant, or DSG). This is around £56 billion in 2024/251.
The biggest element of this is the Schools Block of £45.3 billion. But there’s also £10.4 billion for high needs pupils and a smaller sum (about £390m) for services for schools. All three have their own formula. (Early Years provision is also included in DSG but that’s for another piece.)
Outside the DSG is the Pupil Premium, which is another £2.9 billion this year, and again has its own formula outside the National Funding Formula. In many years there have been other one-off grants which are then rolled into the formula2.
Crucially, the block with the highest growth under the formula has been in High Needs. Between 2018/19 and 2024/25 Schools Block spending grew £11.6bn (35%) while High Needs Block spending grew £4.3bn (70%). The recent Budget added another £1bn to High Needs, so it is likely to continue to grow faster.
The Schools Block
The Schools NFF determines the allocation of that core £45 billion. The diagram below from DFE explains it. There is a certain amount for each pupil (blue). Then top ups for various things to do with pupils (in pink), like deprivation, the prior achievement of pupils, EAL and how often pupils move mid-year (mobility). There are then adjustments at the school level (orange) for the costs of running a small school (the lump sum) and other premises costs. Finally, there’s an area costs adjustment (yellow) and then some elements things to prevent too much turbulence and to pull up the very lowest funded schools (green).
If we stack up the numbers for 2024/25 you can see what matters most. About 75% of the money is going through the basic per pupil element. 17.8% goes through the additional needs elements, of which 10.2% is for income deprivation, 6.4% is low prior attainment, and 1.1% is English as an Additional Language. 6.3% goes through the lump sum. About 2.5% of the money is moved by area cost adjustments3. Things like the minimum pupil amount introduced under Boris are a very small part of the picture, so don’t drive much convergence.
I remember debates at the time about other metrics that could have been used. In particular, you could look at final attainment. Some groups of low prior attainment pupils make much more progress than others - particularly those with English as an Additional Language.
This, in addition to the particular measures of low income that have been chosen (Free school meals and IDACI) tend to push money towards urban areas. Coastal and shire areas score poorly on this basis, even where school achievement is very low4.
The pupil-level measures that are in the NFF tended to reduce the amount of change the formula brough about, as they effectively baked in a number of the priorities that the previous Labour government had pushed at different times.
But over time patterns of under-achievement have radically changed, with London in particular climbing the league tables. In the 80s the Inner London Education Authority was seen as a bastion of the “loony left” and London schools were dire. Today they are the best in the country. Even in 1997 22 out of 32 London boroughs were below average at GCSE, last year only 3 were.
This is partly due to some great London-based academy chains like ARC and Harris, but also due to demographics: London has become an expensive place to live and has many more graduate parents than anywhere else.
But the resource differences are big too, and aren’t just eaten up by higher salaries and costs in the capital. If you prefer to think in terms of people rather than cash, in England there are 20.8 pupils for every primary school teacher, but in inner London there are 18.4.
The High Needs Block
As noted above, the High Needs Block has seen the fastest growth in funding over the period that the NFF has been in place.
Unlike the schools formula, just over a quarter of the money in the High Needs block is determined by purely historic patterns of spending in 2017/18.
Like the schools formula it contains the same measures of deprivation.
It adds an element for child health and numbers in special schools.
However, unlike the schools formula, it does contain an element which directly targets low final achievement, not low prior attainment.
The High Needs element accounts for a bigger part of the DSG in London - in inner London it is about a quarter of the total, compared to a fifth nationally. It tends to make the funding gap between London and the rest bigger, so other things equal its faster growth would tend to boost relative funding levels in the capital.
What effect has the NFF had?
The NFF is a massive compromise, incorporating multiple principles, and as a result there is no particularly clear logic to the variations in per pupil spending if you look through one lens at a time.
Variations between authorities outside London don’t obviously reflect massive differences in the cost of living: why is Leicestershire 10% lower than neighbouring Lincolnshire?
But I am particularly struck that they don’t reflect school achievement - or at least today’s patterns of school achievement.
At present overall there is no connection between achievement and funding level. If anything there is a slight positive correlation: achievement and funding are both high in London.
Strip out London and there is a slight connection. But places with similar costs and achievement levels still have unbelievably different rates of funding. Buckinghamshire and Dudley get similar amounts of money, but radically different results. Same for Shropshire and Hertfordshire.
In general, many of the areas which combine low funding and low achievement are shire and coastal areas (North East Lincolnshire, Cumberland, Shropshire). Urban areas (Birmingham, Manchester, Bristol) tend to be above the curve.
This lack of a strong link between low achievement and higher funding is odd, given that a big chunk of the NFF aims to boost funding for pupils with characteristics that are correlated with low achievement. But these things are only partly correlated, and (at least in the main Schools Block) there is no element directly targetting low achievement.
The factors used in the disadvantage section (English as an additional Language, Low prior achievement, the particular measures of disadvantage used) tend to have a strong urban focus, and for this reason the NFF tends to mean pupils from ethnic minority groups are in schools with higher funding per head.
The chart below shows the average Schools Block plus Pupil Premium funding per pupil for different groups. Between the highest and lowest groups the differences are substantial - for example the average Bangladeshi-heritage pupil in secondary school sees per pupil funding nearly a fifth greater than the average white British pupil (£8,218, compared to £6,884).
Is NFF closing funding gaps?
At least at the level of the Local Authority, the changes in the seven years since it came in have been pretty slight.
Changes reflect both changing demographics in different areas AND also the effects of the formula and tweaks to the formula. For example, levels of deprivation in London on Labour’s preferred measure (the Index of Multiple Deprivation) have been going down, which should relatively reduce funding.
Here’s a map of the change in funding compared to the national average:
And here’s a scatterplot. The differences between the best and worst funded per pupil have been slightly compressed over time - the line of dots is slightly steeper than the 45 degree line in red.
But the differences are not large - most authorities are in roughly the same sort of place as when the formula came in. In general, authorities which saw larger downward convergence to the national average remain substantially above it.
I should declare an interest in so far as our kids are at schools in Leicestershire, and it is not clear why our funding should be so much lower than places with much higher achievement - including many places outside London with similar costs.
Over the period the NFF has been in place for, the authorities with lower achievement have been very marginally more likely to see funding increases compared to the national average, but only a little. Outside London there is no clear pattern.
At the level of the region, the effect of the pupil premium increases total per-pupil funding in the North East about 2% compared to the national average and increases it about 1% in Yorkshire and the West Midlands. In London the high needs block tends to increase funding relative to the national average. Overall, at the regional level the patterns of funding are little changed between 2018/19 and 2024/25, with London much higher (though a little less so) and other regions very marginally higher.
Conclusions
After decades of trying, it was a huge achievement to create a proper funding formula. As well as being fairer and more accountable, this also enables other reforms and gives greater certainty to schools.
The schools formula has to juggle many demands. For example, I have always been keen on the lump sum, which helps small local schools and village schools stay open, with wider social benefits.
At present the formula is pretty weak at targetting additional funding to lower-achieving areas, and if we include London it doesn’t do so at all. The creation of the NFF baked in quite a lot of historic decisions and patterns which don’t reflect today’s patterns of underperformance.
One reason for this is that the Schools Block does not target low achievement directly. The High Needs block does, so we already kind of crossed that issue of principle. But I can see why people would be nervous about “rewarding failure” if you tied higher funding to lower school performance.
However, you could add a metric which looked at low outcomes in the area. Philosophically, the justification for having measures like EAL, lower prior achievement, deprivation etc in the formula is that you are trying to level up opportunity where it is lowest. But as the scatterplots above show, their impact is only to make funding very weakly correlated to low opportunity. You could instead use a measure that targets low performance and fewer opportunities directly.
You can argue for higher or lower levels of effort to level up opportunity. But it seems to me that however much effort you want to put into this, you should want efforts to be more on target than they are now.
Under the NFF there has been a marginal compression in the variation in funding levels in different places. But this probably more reflects falling relative deprivation in London than anything else.
In general, the NFF is not changing the pattern of spending at the level of the local authority much, and measures like the minimum per pupil funding element are marginal. It remains a good thing and allows people starting new schools to have some clarity on how much money they might get. But to bring low funded/low achieving places up closer to average funding the formula would have to change more than it has.
Dedicated Schools Grant (DSG) minus Early Years Funding.
Partly because of the pandemic, Ukraine and inflation, there has been a tendency to add one off grants with different (non-NFF) allocation formulas and then roll them into core funding. For example, in 2024/25 there is a further:
£900m for the Teachers Pay Additional Grant
£1.25 billion for the Teachers’ Pension Employer Contribution Grant
£1.1 billion for the Core Schools Budget Grant.
You can’t really add this up as it is applied to the basic amount, pupil and premises factors. For simplicity in the chart I have subtracted it from the basic amount to give a sense of how important it is compared to other factors, though this isn’t quite right.
There is a whole other blog to be written on how we measure deprivation and limits to opportunity in different places. If you emphasise measures like low pay and low school achievement more you’d get a different picture to the IMD, which has a strong urban skew.