Budget 2026 Policy Recommendations

22 September 2026

The National Foundation for Educational Research welcomes the opportunity to suggest policy recommendations ahead of HM Treasury’s Budget 2026. 

Our representation focusses on education policy issues, specifically revenue implications for the Exchequer and wider macroeconomic implications. It draws on recent NFER research, principally our 2026 series of annual reports on the education workforce, The Skills Imperative 2035 programme, analyses of pupil rolls and research into SEND. 

You can skip to recommendations about each subject below:

Departmental spending envelopes for this period were set at the 2025 Spending Review. Our proposals are therefore weighted towards measures that improve the return on money already committed, that use data collections already in place, and that bear on labour supply, tax receipts and welfare spending over the medium term. 

Education workforce 

Delivery of the Government’s education ambitions depend on the sufficiency and quality of the workforce. In 2026, NFER published four reports examining the current state and trends of the early years workforce, school teachers, school support staff and further education teachers. 

School teachers 

Recruitment and retention of school teachers has been improving in recent years. In 2023/24, recruitment to secondary initial teacher training (ITT) represented 48 per cent of the target. Last year secondary recruitment reached 89 per cent of target, and our forecast for 2026/27 – based on applications made and accepted to February 2026 – is 86 per cent. The teacher leaving rate has fallen from 10.6 per cent in 2016/17 to nine per cent in the most recent data, and the exit rate of first-year early career teachers was 10.3 per cent, the lowest since the data began. 

This improvement is likely to be driven by weakness in the wider labour market. UK unemployment rose from 4.2 per cent in summer 2024 to 5.2 per cent in autumn 2025, and economy-wide vacancies fell from 1.3m in spring 2022 to 700,000 in late 2025, below the pre-pandemic level of 800,000. This matters in terms of fiscal planning: if the labour market recovers, teacher supply is likely to deteriorate again, and more spending on pay, bursaries or retention payments would be needed to sustain current levels of supply.  

The most cost-effective instruments to support recruitment and retention, which are available, are targeted ones. NFER’s evaluation of bursaries, using data across many years of bursary changes, found that they produce sustained changes in supply and are more cost effective than alternatives such as pay increases and retention payments.

For 2026/27 the Government reduced several bursaries by as much as £21,000 and eliminated some entirely, while retaining maths, physics, chemistry and computing at £29,000. Our analysis of DfE data suggests lower recruitment in the affected subjects than would otherwise have been expected. 

Our modelling of routes to the Government’s teacher supply pledge illustrates the scale of the difference. Meeting the target through pay increases alone was feasible but would have required increases of nearly ten per cent per year for two years, at £4.9bn annually from 2026/27 onwards. Bursary increases delivered around 2,500 additional teachers by 2027/28 at roughly £600m per year, and early career retention payments around 2,600 at roughly £300m per year, concentrated in shortage subjects.  

DfE has proposed an increase in teachers’ pay of 6.5 per cent over three years. Our analysis suggests this may mean teachers’ earnings fail to keep pace with wider earnings growth. DfE has separately said that at existing funding levels schools cannot afford more than 2.7 per cent over the next two years. 

The Government should: 

  • consider bursary increases for languages and arts subjects that are below their ITT targets, to support implementation of the revised national curriculum and its changes to accountability measures
  • aim to maintain the competitiveness of school teachers’ pay by matching the growth in average earnings outside teaching, and fund schools to deliver it. 

Further education teachers 

The competitiveness of FE teacher pay has fallen faster than that of any other education workforce group. Our analysis of Annual Survey of Hours and Earnings data suggests FE teacher pay was 20 per cent lower than school teacher pay in 2025. In 2023/24 there were around 3,000 vacant FE teaching posts, with four per cent of roles vacant in general FE colleges.

These vacancy rates vary by region, reaching 6.5 per cent in Yorkshire and The Humber against two per cent in the East Midlands. FE teachers are also less likely than similar workers to agree their job offers good opportunities for career progression, (46 per cent for FE compared to 58 per cent for similar workers in 2024). 

The 16 to 18 population is growing and there is a rising demand for FE college places. The Post-16 and Skills White Paper sets out an £800m increase in funding for 16 to 19 education in 2026/27, which the Institute for Fiscal Studies calculates as a 2.5 per cent real-terms increase in spending per student.

A pay gap between FE and school teachers of this size is likely to constrain what the £800m can achieve, because FE providers compete for staff not only with schools but with the industries in which those staff already have expertise. 

However, this evidence base is weaker than for school teachers, and this is itself a value-for-money problem. The latest available data on FE teacher retention comes from DfE analysis of Teacher Pension Scheme data running only to 2017/18. Without current retention data, neither DfE nor the Treasury can assess whether the £800m is likely to impact the staffing problem it is intended to address. 

The Government should: 

  • provide funding that allows FE providers to pay teachers more, prioritising this within the post-16 settlement given the scale of the gap and the expected growth in demand.
  • publish regular data on FE teacher retention as soon as it can, either through the Further Education Workforce Data Collection (FEWDC) or, failing that, using pension scheme data as it has done before. The FEWDC is already in place, so the administrative cost of this is low. 

Early years workers 

Growth in the early years workforce appears to have stalled. DfE reported that the workforce grew by 39,559 between summer 2021 and summer 2024, an increase of 12 per cent. However the number of staff increased by just 628, or 0.2 per cent, between summer 2024 and summer 2025. DfE has estimated that the workforce would need to increase by 35,000 staff between December 2023 and September 2025 to meet the needs of the expanded entitlement. Our analysis indicates that the South West, West Midlands and North East may be worst affected by potential staffing shortfalls. 

The proportion of staff with a level 4 or above qualification remains at around a quarter of the workforce, while 15 per cent of staff in school-based settings, 20 per cent in group-based providers and 24 per cent of childminders are qualified only to level 2 or below. Pay compression linked to increases in the National Living Wage is likely to be reducing the pay premium associated with higher qualifications, which weakens the incentive for lower-qualified staff to train further. 

On value for money, the evidence on small-scale financial incentives is discouraging. An evaluation of the £1,000 payment available to new recruits in eligible areas concluded that the pilot did not result in a measurable increase in the number of applicants per vacancy, though the evaluators noted that a lack of provider engagement and low awareness among potential applicants may have limited its effectiveness. 

The Government should: 

  • ensure that funding rates support early years providers to offer competitive wages to recruit and retain staff, with a pay structure that rewards higher qualification levels.
  • design the new early years professional register so that it produces individual-level workforce data, and commission a representative individual-level workforce survey.  

School support staff 

Support staff make up more than half of the school workforce. NFER analysis suggests that almost one in five support staff left the school system between 2023/24 and 2024/25, a rate that has been rising for three years and has reached its highest level since the data began in 2011/12. In our 2025 survey of school leaders, around three in four leaders in both primary and secondary schools said they found recruiting teaching assistants difficult or very difficult, a higher proportion than said the same of teachers.

In 2024, 40 per cent of support staff agreed they have good opportunities for career progression in their current job, against 55 per cent of similar workers, and 15 per cent said they would like more work at the current level of pay. 

The School Support Staff Negotiating Body is expected to make its first decisions for the 2027/28 financial year. DfE already holds School Workforce Census data on support staff entrants, retention and pay that it does not publish. 

The Government should: 

  • add a measure of support staff shortages to the School Workforce Census, and publish the statistics on entrants, retention and pay that DfE already has the data to produce. The marginal cost of publishing existing data is small.
  • consider, with the SSSNB, whether low pay levels are driving support staff recruitment and retention challenges, and what can be done to improve career progression opportunities, which fall within the body’s remit. 

Post-16 education and skills 

The Skills Imperative 2035, a five-year NFER research programme funded by the Nuffield Foundation, identified what the labour market will look like by 2035, what ‘Essential Employment Skills’ (EES) will be most needed to do these jobs, and who is most and least likely to be affected by changes in jobs and skills requirements. Its findings have implications on growth, on employment and on the medium-term path of tax receipts and welfare spending. 

Around 12 million people in England work in occupations that are declining, including administrative, secretarial, customer service and machine operative roles. If the current rate of decline continues, between one and three million jobs in these occupations could disappear by 2035.  

Some 3.7 million workers in England already have substantial deficiencies in the six Essential Employment Skills we identify as most needed across the future labour market: collaboration; communication; creative thinking; information literacy; organising, planning and prioritising; and problem solving and decision making. This is projected to increase to around seven million by 2035 unless action is taken. 

The fiscal consequences to this are that a shortage of skilled employees at the top of the labour market is likely to constrain employers’ capacity to grow, while fewer opportunities at the lower end raise the risk of worklessness. Both imply lower tax receipts and higher welfare spending.

The distributional and locational effects are uneven: workers in high-risk occupations are disproportionately at either end of the age distribution, working part-time and located outside London and the South East, and around half hold no qualifications beyond GCSEs. 

Most people who will be in the labour market in 2035 are already working today, so reskilling cannot be set aside in favour of education reform alone. Public investment in adult learning fell by 38 per cent between 2010/11 and 2020/21, leaving the system underfunded and fragmented. Reinvigorating the adult skills system, including through a step change in funding, should be a national priority.  

At the same time, more young people need to leave education able to compete for entry-level roles in growth occupations. Uptake of T Levels was less than three per cent of publicly-funded qualifications for 16 to 18 year olds in 2024/25, and colleges face challenges with staffing, specialist equipment and delivery capacity. In 2019, only 10 per cent of adults aged 18 to 65 held a Level 4-5 qualification as their highest achievement, compared with 20 per cent in Germany and 34 per cent in Canada.

The proportion of the employed UK population with a Level 4 or above qualification is projected to rise from 48 per cent in 2020 to 61 per cent by 2035, but that band is dominated by degrees. England has a gap between Level 3 and degree level, whereas comparable economies have an established route. 

Methods that require no additional spending include recognising Essential Employment Skills explicitly within a knowledge-rich curriculum, promoting a common skills framework, and building training on how to develop these skills into initial teacher training and the Early Career Framework.  

On apprenticeships, our research has consistently found that the barriers facing younger and more disadvantaged applicants are financial and structural rather than motivational.

Our interviews with small and medium-sized enterprises found that the minimum apprenticeship wage may be insufficient for some young people to live on, and that low wages act as a barrier to both recruiting and retaining young apprentices, particularly those from disadvantaged backgrounds.

Travel costs compound this, especially where a young person's local area does not offer the apprenticeship they want and they must travel to reach a more specialised opportunity. 

Minimum English and maths entry requirements for most intermediate and advanced apprenticeships may also be preventing young people from starting at all, which matters given the proportion of disadvantaged pupils in England who do not achieve grade 4 or above in both GCSE English and maths. This evidence dates from 2022 and 2023 and predates subsequent changes to apprenticeship minimum wage rates, so the specific rates cited in it are superseded, but the mechanisms it identified remain in place. 

Withdrawal is another issue. Our analysis shows that apprentices aged 16 to 17 are more likely to leave early than those aged 18 to 20, and that learners who already hold a Level 3 qualification are eight percentage points less likely to withdraw than those with no prior qualifications.

Learners from disadvantaged backgrounds, those with lower prior attainment and those facing additional barriers are disproportionately more likely to withdraw. Those who do appear to face lasting economic consequences: three years after leaving, they have significantly lower wages and are more likely to be unemployed than peers who completed, which suggests they are not leaving for better-paid work elsewhere. Non-completion is therefore not a neutral outcome for the Exchequer – it carries both the cost of part-delivered training and a subsequent earnings and employment penalty. 

The same analysis indicates that pay is not the whole answer here. A ten per cent increase in first-year apprentice pay is associated with only a small reduction in withdrawal risk, which suggests pay rises alone are unlikely to solve the completion challenge.

This does not weaken the case for reviewing the minimum apprenticeship wage for under-18s, which our earlier research addressed as a barrier to young people starting an apprenticeship rather than to finishing one, but it does mean the wage should not be presented as a lever for completion. 

The Government should: 

  • reinvigorate the adult skills system with a step change in funding focused on retraining workers in declining occupations, and reduce the financial and time barriers adults face in accessing training
  • review the appropriate level of the minimum apprenticeship wage for under-18s, and extend the 16-19 bursary fund so that it can be used to fund travel costs for apprentices from disadvantaged backgrounds, which would reduce the geographic barriers that limit access to specialised apprenticeships outside urban centres 

Pupil numbers 

Falling pupil numbers reduce the total cost of the schools system while raising the risk of deficits in individual schools.  

DfE’s own projections show that primary pupil numbers will fall by five per cent, equivalent to approximately 185,000 pupils, between January 2025 and 2028, with a further drop of approximately 115,000 expected between January 2028 and 2030.

That amounts to an overall projected fall of seven per cent, which is equivalent to the number of pupils in over 1,000 average-sized primary schools. Secondary pupil numbers are set to peak in 2026, with a decline of 97,000 forecast by 2030. 

Nationally, this creates scope for the Government to reallocate funding without reducing the amount the average pupil receives. At school level the arithmetic works differently. In 2025/26, schools receive at least £4,955 for every primary pupil. A drop in pupils at a school may not be large enough to allow for its leaders to cut a class and save the staffing cost, but their funding would be reduced regardless. 

In the 2023/24 financial year, 13.2 per cent of primary schools were running an overall deficit, with substantial variation by location and size: 20 per cent in London against 9.3 per cent in the East of England, 15.9 per cent in the North East and 14.7 per cent in Yorkshire and the Humber; and 16.2 per cent among small primary schools of up to 150 pupils against 8.3 per cent among schools with 450 or more. Our survey of over 350 senior leaders in June 2025 found 71 per cent of primary and 65 per cent of secondary leaders anticipating an in-year budget deficit in 2025/26. 

Our scenario modelling illustrates that a two per cent fall in pupils would have increased the proportion of local authority maintained primary schools running an overall deficit in 2023/24 from 14.8 per cent to 19.3 per cent. Falls of five and ten per cent would have taken it to 26.8 and 40.8 per cent respectively, although these are upper bounds, as schools may take cost-cutting steps where they can. They indicate that the scale of decline now projected is capable of moving a large number of schools into deficit. 

Schools are already adjusting, and the adjustments have consequences elsewhere in the Government’s programme. Between 2017/18 and 2024/25, teaching assistant full-time equivalents fell by 15.8 per cent in the ten London local authorities most affected by declining pupil numbers, against a national trend of 5.7 per cent growth.

Given the role teaching assistants play in supporting pupils with special educational needs and disabilities, and the Government’s intention that mainstream schools support a wider range of needs, reductions of this scale in the areas losing pupils fastest warrant attention before the SEND reforms are implemented. 

The Government should: 

  • publish a range of scenarios alongside its pupil projections, rather than a single central forecast, which would enable clearer planning for future funding options at national and school level. 
  • treat the savings arising from falling rolls as a reallocation decision rather than an automatic reduction in the schools budget, given the evidence above on how unevenly the losses fall across regions and school sizes. 

Special educational needs and disabilities (SEND)

The Schools White Paper sets out an expectation that every local mainstream school should meet a wider range of need, with legislation planned through the proposed Education for All Bill. NFER’s research on the distribution of pupils with SEND across mainstream schools, funded by the Nuffield Foundation, suggests this ambition will be difficult to realise on the current pattern of provision and funding. 

In 2024/25 more than 1.7 million pupils in England were identified as having SEND, around one in five, against 1.2 million in 2015/16. The rise in the proportion of pupils with an Education, Health and Care plan (EHCP) has been most acute, nearly doubling from 2.8 per cent to 5.3 per cent of pupils in state provision, while the proportion receiving SEN Support rose more gradually from 11.6 per cent to 14.2 per cent. Over half of pupils with EHCPs, at 56 per cent, are now in mainstream schools, up from 49 per cent in 2015/16, with the sharpest increase in mainstream primary schools. 

In 2024/25, primary schools in the top quartile for EHCP rates had on average six times the rate of those in the lowest quartile, equivalent to 17 pupils per school against three. Among secondary schools the ratio was five times, at 54 pupils per school against 14.

The gap has widened: between 2018/19 and 2024/25 the difference in EHCP rates between schools at the 25th and 75th percentiles grew by one percentage point in primaries and 0.8 percentage points in secondaries. Our report, drawing on a survey of 800 SENCOs and school leaders alongside national data and school case studies, found that fewer than one in five schools report they can meet the needs of all pupils on their roll. 

The distributional and locational picture matters for how any additional resource should be targeted. Pupils with SEND are considerably more likely to be from disadvantaged backgrounds: 44 per cent of pupils with EHCPs and 40 per cent of those on SEN Support are eligible for free school meals, against 22 per cent of pupils with no identified SEND.

Schools with persistently high EHCP rates relative to their catchment are more likely to serve deprived communities, particularly at primary, where 31 per cent fall in the most deprived quintile against nine per cent in the least deprived. Regionally, primary EHCP rates ranged from 4.3 per cent in London to 2.8 per cent in the West Midlands.  

Schools must cover the first £6,000 of support for any pupil with SEND, with the remainder of provision specified in an EHCP met by the local authority. That notional threshold has not increased since 2013, despite the substantial rise in the cost of specialist provision over the same period. Senior local authority officers we interviewed reported that some schools pursue EHCPs partly to access funding not otherwise available.  

Only 46 per cent of EHCPs were issued within the statutory 20-week period in England in 2024, a fall of 14 percentage points since 2021. Among tribunal-decided outcomes, almost all, at 99 per cent, include at least some elements decided in favour of the parent. Assessment backlogs and appeals absorb capacity. 

This interacts directly with the pupil number pressures set out above. Local authority officers reported that financial pressures caused primarily by falling rolls had led schools to cut back on early intervention and staff training, which they perceived to have led to needs escalating and more children requiring EHCPs.

It sits alongside a measured change in the same direction: between 2017/18 and 2024/25, teaching assistant full-time equivalents fell by 15.8 per cent in the ten London local authorities most affected by falling pupil numbers, against national growth of 5.7 per cent, and DfE’s own survey found that among primary leaders taking action in response to falling rolls, 89 per cent reported that teaching assistants were supporting a greater number of pupils with SEND.

Early intervention capacity in mainstream schools is being reduced by demographic pressure at the same time as the reforms rely on mainstream schools absorbing more need. 

Our evidence also indicates that inclusion bases are unlikely to resolve this on their own. Many schools report that these arrangements can strengthen expertise, but provision and integration are currently uneven and many schools face significant staffing and resource pressures in delivering them.

Staff and parents also report that some pupils require specialist provision to be supported safely and effectively. Any settlement that assumes mainstream inclusion replaces specialist places is likely to underestimate the cost. 

The Government should: 

  • review funding so that it better reflects differing levels of need across schools, making inclusion sustainable rather than dependent on goodwill. For example, the notional £6,000 threshold schools must meet before local authority top-up applies has been unchanged since 2013
  • ensure inspection and accountability systems reward inclusive practice and meaningful progress for pupils with SEND
  • monitor the distribution of pupils with SEND locally and take action where school intakes appear persistently unrepresentative of local need. 
Back to all policies