Schools have faced pressure to make efficiencies in the last three years. The coalition government initially protected per-pupil funding until 2015/16, but schools experienced annual declines in funding between 2015/16 and 2017/18. The growing share of pupils with special educational needs has put further pressure on schools

Schools have managed these pressures through the public sector pay cap, which has kept teacher wages down, and by becoming more productive. There are more pupils per teacher in secondary schools, while pupil attainment has been at least maintained.

The Johnson government’s school spending announcement earlier this year will relieve these immediate pressures. By 2022/23, per-pupil funding will be almost back to 2009/10 levels in real terms. Some of this money will be used to address clear pressures – such as boosting starting salaries to address teacher recruitment problems – but it is not clear what exactly the government expects schools to deliver with the money.

As of January 2019, there were almost 17,000 state primary schools and almost 3,500 state secondary schools in England.[1] Together, they educated 6.73 million (m) pupils aged from 5 to 15 – 650,000 more than in 2010.* Most secondary schools (75%) and approximately one third of primary schools (32%) are academies – state schools that do not have to follow the national curriculum and have greater control over their admissions and budgets.[2] The rest are ‘maintained’ schools, whose funding flows through local authorities.

This chapter focuses on pupils aged 5–15 in mainstream academies and maintained schools. All references to schools refer only to pupils of this age in these schools, unless stated otherwise.**

*These figures do not include pupils in state special schools (116,013), pupil referral units (16,065) or independent schools (522,035). They also do not include 16–18 education, including school sixth forms.

** We focus on pupils in mainstream schools because funding for special schools and pupil referral units is “driven more by the needs of individual pupils” and spending on 16–18 education followed a different trend to spending on pupils aged 5–15; principally, there have been bigger per-pupil spending cuts. See Belfield C and Sibieta L, Long-Run Trends in School Spending in England, Institute for Fiscal Studies, 2016, p. 13.

Schools’ spending has risen by 7.2% in real terms since 2011/12

Spending by schools (£bn, 2018/19 prices)

Funding for schools is a complicated area. It includes a number of streams –  local authority funding, the pupil premium, capital spending – and, in the case of academies, there is a shortfall of good data. This makes it hard to calculate changes.

In 2018/19, day-to-day school spending (all spending excluding capital spending such as building new schools) in England was £38.3 billion (bn). Average school spending per pupil increased by 1.1% between 2011/12 and 2015/16. However, there was a disparity between spending in primary and secondary schools. On average, primary spending per pupil increased by 7.2% in real terms between 2011/12 and 2015/16, while secondary spending per pupil fell by 2.9%. After 2015/16, the growth in pupil numbers outpaced spending growth, and average per-pupil spending fell by 4.1% in real terms in both primary and secondary schools.

The government does not know exactly how much school spending initially rose, because the Department for Education (DfE) does not hold good-quality data on academy school spending before 2011/12. This gap means that our 2009/10 and 2010/11 numbers underestimate school spending – and therefore overstate the overall increase in spending that has occurred since 2010/11.*

Between 2011/12 and 2017/18 – the years for which consistent data is available – school spending rose by 4% in real terms, but is difficult to calculate. The UK Statistics Authority has recommended that the DfE publish regular official statistics on school funding,[3] though it has not committed to doing so.[4]

Since 2011, a fraction of school spending described above has come from the pupil premium grant, which is paid to schools based on the number of children that attend who: receive free school meals; are looked after by the local authority; or have, or had, a parent in the armed forces.[5]

This money (which totalled £2.4bn, or 5.6% of schools funding,** in 2018/19) is intended to improve disadvantaged children’s attainment. It continues a longer trend of targeting spending towards schools where pupils have greater levels of disadvantage, which was particularly notable during the 2000s. The introduction of the pupil premium funding compensated for real-terms cuts to other school grants under the coalition government.[6]

Schools must account for how they have used this funding to improve educational outcomes for disadvantaged pupils. They have to publish pupil premium statements online explaining how they spent the money[7] – although there is no hard ringfence on the pupil premium grant.

While some of the grant has been spent on specific interventions to help disadvantaged students, such as hiring dedicated teaching assistants, some schools have used it for other purposes. A 2019 survey of teachers conducted by the National Foundation for Educational Research (NfER) found that 27% of secondary school leaders used the pupil premium funding to “plug gaps elsewhere in the budget”.[8]

Our school spending figures include the pupil premium grant, but exclude some wider spending on schools such as spending on school sixth forms and local authority spending on services for schools. Taking these into account, the Institute for Fiscal Studies (IFS) estimates that total spending on schools fell by 8% between 2009/10 and 2018/19, primarily due to cuts to local authority services for schools.[9]

Such local authority spending, which includes sessions with educational psychologists and transport, fell by 57% in real terms per pupil across this period.[10] Spending cuts may have resulted in new financial pressures for schools, as they may now pay for some services out of their own budgets that were previously provided for them (or they may no longer provide these services). This may have increased pressures on schools – although we cannot assess how deep or widespread these are.***

Spending on school sixth forms also fell, by 30% per pupil.[11] However, we exclude this from our analysis because our focus is on pupils aged 5–15; but it will have had an impact within secondary schools because resources such as staff and buildings are often shared across age groups. In 2018/19, the government spent £2bn on school sixth forms,[12] compared to £38.3bn on primary and secondary schools.

*There were 203 academy schools in 2010 (out of a total of 20,301 state-funded primary and secondary schools) so the difference between spending by non-academy schools in 2009/10 and 2010/11 (shown in the figure above) and spending by all schools will not be vast.

**5.6% of the combined total of the schools block grant, central school services block grant, high-needs block grant, teachers’ pay grant and pupil premium funding. See Education and Skills Funding Agency, Dedicated schools grant (DSG): 2018 to 2019, 2019; Education and Skills Funding Agency, Teachers’ Pay Grant: September 2018 to March 2019 allocations, 2019; Education and Skill Funding Agency, Pupil Premium: Allocations and conditions of grant 2018 to 2019, 2019.

***Some of the spending cut reflects maintained schools converting to academies. Academies receive funding from the DfE to cover the cost of services that would have previously been provided centrally by their local authority. See Belfield C, Sibieta L, Long-Run Trends in School Spending in England, Institute for Fiscal Studies, 2016, pp. 11, 18.

Growth in primary pupil numbers is feeding through into secondary schools​

Pupil numbers in primary and secondary schools (ages 5–15, millions)

The numbers of primary and secondary school pupils have been rising for 10 and four years respectively – but the headline figure should be treated with some caution.

There are now 10.5% more pupils in mainstream state primary and secondary schools than there were in 2009/10. Most of that increase has been in primary schools – there were 18.1% more primary school pupils in 2018/19 than in 2009/10. But the number of secondary school pupils has also been growing since 2014/15** and, at 2.9m, is 3.8% higher than in 2009/10. These figures on pupil numbers may, however, understate the increase in demands placed on schools.

One reason for this is that the share of pupils who receive additional support for special educational needs (SEN) has recently increased. Between 2009/10 and 2016/17, the share of pupils with an SEN statement or Education, Health and Care (EHC) plan – which entails schools, councils and clinical commissioning groups providing additional support[13] – held steady at 2.8%. But this figure rose over the last two years to 3.1%. In January 2019, 271,165 pupils had an SEN statement or EHC plan, almost 30,000 more than in January 2017.[14]

But we should be cautious in interpreting these figures because there is a relationship between SEN statements and EHC plans, the amount of support children receive in school and the funding that schools receive. There are, therefore, a range of conflicting incentives on parents, schools and local authorities to have children classified in this way.

Parents have an incentive to push for their children to have statements or plans because of the entitlement to resources that entails.[15] As schools’ core budgets have been squeezed, the attraction for parents of ensuring their child benefits from these additional resources has increased. A 2018/19 Education Select Committee inquiry found that parents “do not get much help until they get an EHC plan, which basically means that the lower level of SEN support is not working for a lot of children”.[16]

The share of pupils receiving lower-level SEN support (without a SEN statement or EHC plan) fell from 21.1% in January 2010 to 14.9% in January 2019,[17] which may also have strengthened parents’ incentives to push for a SEN statement or EHC plan.

The reverse of this is that local authorities are incentivised to slow down or not award plans to avoid the requirement to provide schools with additional funding for pupils who need more than £6,000 of support a year.[18] The first port of call for these funds is local authorities’ high-needs block grant. If that is exhausted, local authorities can in principle transfer up to 0.5% of their schools block grant to top up the high-needs block grant. In practice, however, the government has given some local authorities permission to transfer more than this.*

This lets local authorities channel more money to schools with disproportionately large numbers of SEN pupils by cutting funding for other schools where they have the agreement of their local schools forum – a statutory group of local education providers who decide how to allocate local authority school budgets.[19]

As a result of these diverging incentives, parents and guardians have registered more appeals against local authority decisions with the Special Educational Needs and Disability Tribunal. In 2018/19, the tribunal received 6,374 cases, up from 3,397 in 2009/10. The number of cases rose particularly sharply after 2015/16.[20],** The percentage of cases where the tribunal ruled in favour of the appellant – that is, the parent or guardian – rose slightly from 86.3% (2014/15) to 88.6% (2017/18),[21] suggesting that the recent rise has not been due to an increase in weaker cases.

The incentives that schools face are unclear. On the one hand, they are responsible for funding the first £6,000 of additional provision out of their ‘notional SEN budgets’.[22] They face additional financial pressure if more of their pupils require support up to this level, discouraging them from accepting pupils with additional needs.[23] On the other hand, any support required above £6,000 a year will be met by the local authority from the high-needs block grant, meaning they will receive additional funding, which schools may be able to use for the benefit of more than just the pupil in question.

Whether or not the share of pupils who need additional support is increasing, the rise in children receiving statements or plans represents an additional financial pressure that schools and local authorities have borne.

*In 2019/20, the government allowed 22 councils to move more than 0.5% of their schools’ block grant to the high-needs block. See Gibb N, ‘Schools: Finance: Written question – 228138’, Parliament.UK, March 2019, 

**This partly reflects the expansion of eligibility for SEND (special educational needs and disability) resources under the 2014 Children and Families Act. The rate of appeal – how many decisions are appealed as a percentage of all decisions issued – remained constant between 2014 and 2018, the only years for which we have data, at 1.6%. See Ministry of Justice, ‘SEND Tribunal tables: statistics on the appeal rate to the SEND Tribunal’, 2019.

Growth in teacher numbers has not matched growth in pupil numbers

More so than other public services, schools’ main expense is staff. Three quarters of school spending pays for staff costs;[24],* most of that goes on teaching staff.

Approximately 65% of total school spending goes on teachers, supply teachers and teaching assistants. This is true for both primary and secondary schools, as well as academies and maintained schools, although primary schools spend a slightly larger share of their budget on teaching assistants.

Trends in teacher numbers since 2011/12, the earliest year for which there is comparable data for maintained schools and academies, have differed in primary and secondary schools.

Pupil–teacher ratios in primary and secondary schools

The number of primary school teachers rose broadly in line with pupil numbers until 2016/17 and is now 13.1% higher than in 2010/11.[25] This increase levelled off after 2016/17, while pupil numbers continued to rise, which caused the pupil–teacher ratio in primary schools to rise very slightly to 20.8 in 2018/19, up from 20.5 in 2011/12 (the earliest year for which comparable data is available).

The number of secondary school teachers has fallen; this fell at roughly the same rate as that of pupils between 2010/11 and 2014/15,[26] keeping the pupil–teacher ratio steady at around 15:1. But since 2014/15, pupil numbers have risen, while teacher numbers have continued to fall. This has resulted in a greater pupil–teacher ratio in secondary schools; there were 16 pupils to every teacher in 2018/19.

On average, teachers are working similar hours now to 2010. According to the Office for National Statistics’ UK Labour Force Survey – the longest methodologically consistent time-series of teacher working hours – primary school teachers, on average, are working slightly more hours than they did in 2010, and secondary school teachers slightly fewer.[27] Overall, teacher working hours have remained broadly stable between 46 and 48 hours per week over the last 20 years.[28]

A separate OECD Teaching and Learning International Survey has shown that secondary school teachers reported working similar hours between 2013 and 2018. On average, full-time secondary school teachers reported working 49.3 hours a week during term time in 2018, a slight increase from 48.2 hours in 2013, with the largest rise in time spent providing pupil guidance or discipline.[29]

Teachers are not the only staff that work in schools. There has been a rapid increase in the number of teaching assistants employed in primary schools (up 29% since 2011/12), though this is mirrored by a steep decrease in the numbers in secondary schools (down 13% in the same period),[30] reflecting the different funding trends in primary and secondary schools.[31]

That means an overall 7% rise in teaching staff – teachers and teaching assistants – from 2011/12, though the total number peaked in 2016/17 and has been falling ever since.

*For the latest years for which we have data, maintained schools (2017/18) spent 75% of their budgets on  staff while academies (2015/16) spent 71%. These should not be taken as equivalent because academies and maintained schools submit data organised into different categories.

Schools are finding it harder to recruit and retain staff

There are growing problems in both the training and retaining of teachers – especially in secondary schools – that are storing up problems for schools across the country. These are not, however, easily identifiable.

On the surface, there are few workforce problems in schools: staff vacancy rates are low and have not changed much since 2009/10. In 2018/19, there were only 371 recorded vacancies in primary and nursery schools, and 1,222 temporarily filled posts – just 0.6% of teachers. Though the vacancy rate in secondary schools has more than doubled since 2010/11, it remains at only 1%.[32],* But the vacancy rate fails to capture growing problems in training and retaining teachers.

Recruitment to postgraduate initial teacher training relative to subject targets

The number of people training for a teaching qualification each year fell from 39,010 in 2009/10 to 32,248 in 2016/17, before rising slightly to 34,420 in 2018/19. The government has missed its recruitment targets every year since 2011/12, when consistent data became available. Last year, 14% of target places for trainee teachers were unfilled** – the biggest shortfall since 2011/12.

As was the case in 2017/18, primary schools met their recruitment target and there was only a shortfall in secondary school trainees in 2018/19.*** Only four secondary subjects met their recruitment targets in 2018/19: biology, English, history and PE. In contrast, design and technology fell 75% short of its target (872 trainees); physics fell 53% short (644 trainees); and computing fell 27% short (193 trainees).

Teacher entry and leaving rates

Note: Figures as of year to November 2011 – year to November 2018.

Teachers continue to leave the profession at comparable rates to 2011. The equivalent of 9.8% of the workforce (42,073 teachers) left between November 2017 and November 2018, a slight increase on 9.7% in the year up to November 2011. The number of joiners fell each year after 2014, contributing to a narrowing gap between entrants and leavers. Following the decrease in leavers last year, this gap widened again. The small number of net joiners has not been enough to keep up with the growth in pupil numbers.

There are also signs that school working conditions may be driving teachers to leave before retirement age.[33] The proportion of leaving teachers who went out of service (that is, who left the teaching profession but did not retire) rose from 60% to 85% between 2011 and 2018.

The retention rate among the newest teachers has also been falling. Each cohort of newly qualified teachers**** since 2010 has seen a larger share of teachers leave within three years than the one before. Of those teachers who entered the profession in 2015/16, 27% left within three years, compared with 22% of those who entered in 2010/11.[34]

As a result of difficulties recruiting and retaining enough permanent staff, schools have turned to supply teachers to plug gaps. School spending on supply staff rose faster than overall staff spending between 2012/13 and 2016/17,[35] the only years for which we have consistent data for both academies and maintained schools. School spending on supply teaching staff increased by 10.6% in real terms, from £1.17bn to £1.24bn (in 2018/19 prices), between 2012/13 and 2016/17.

Schools are spending a large part of this with supply agencies, rather than using cover supervisors or former teachers,[36] implying that they are increasingly reliant on agency staff to plug recruitment gaps.[37] Almost two thirds (64.6%) of school spending on supply teachers was spent on agency staff in 2015/16, up from 52.5% in 2011/12.[38] In interviews with supply agencies and schools, the National Institute for Economic and Social Research found that schools were increasingly using agency staff to fill long-term placements and cover positions they had been unable to recruit to, rather than for short-term assignments.[39]

*The published figures are taken from a snapshot in November and as such may underestimate underlying vacancies. Contributors to a 2017 Education Select Committee inquiry into the school workforce reported that, by the time this snapshot is taken in November, headteachers may have found alternative ways to plug gaps, such as covering lessons with non-subject specialist teachers. See Education Select Committee, Inquiry: Recruitment and retention of teachers, 2017,

**In 2018/19, the DfE recorded that the number of trainee teachers recruited was 9% (2,971) short of the  target. That figure, however, allows for over-recruitment in some subjects to make up for under-recruitment in others. Comparing trainee numbers to subject targets, 4,465 targeted trainee places were not filled last year – meaning the government was 14% short of targets.

***In 2018/19, primary school trainees were over-recruited by 423, although the primary target was missed between 2011/12 and 2014/15, and in 2016/17.

****Teachers who gained their qualified teacher status, took on a teaching post the same year and were still in that post in November.

Exam results have been improving, and schools have been doing more

The headline figure on which schools are judged is, more often than not, exam results. And school attainment has at least been maintained and may even have improved since 2010. This is one of the most easily comparable ways we have to measure the quality of schools.* Although these results do not capture everything that schools do, primary and secondary pupils must sit national, standardised assessments at the end of both school stages. In years where grading criteria remained consistent, pupil attainment rose, particularly in primary schools.

Pupil attainment at the end of primary school (key stage 2) and secondary school (key stage 4)

Across all periods for which we have consistent data, the proportion of children achieving the national expected standard in reading, writing and maths at the end of primary school** has been rising – though this may be as much due to teachers getting used to the new curriculum and assessments as a fundamental rise in attainment.

The share of pupils achieving level 4 or above in reading, writing and maths rose from 64% in 2009/10 to 80% in 2014/15. Between 2016/17 and 2017/18, the percentage of pupils reaching the expected standard rose from 63.9% to 65.2%.***

Secondary school pupil attainment has been on a similar trend, but has not risen as quickly. Between 2009/10 and 2012/13, the proportion of pupils achieving grades  A* to C in English and maths GCSE rose from 56% to 61%, and then remained at 59% between 2013/14 and 2015/16. The proportion of pupils gaining a ‘pass’ of 4 or above in English and Maths – roughly equivalent to A* to C[40] under the new system – rose slightly between 2016/17 and 2017/18, from 61.0% to 64.0%.

Changes to exam criteria and which subjects count in performance table measures make it difficult to judge how pupil attainment has changed across the whole period. Before 2013/14, re-sit grades were included in the secondary school figures; since then, only the grades from the first time each exam was taken have been counted.

Further, recent changes to assessment in primary schools (in 2015/16) and secondary schools (starting in 2016/17) introduced totally new grading schemes.

The OECD provides an additional method for assessing trends in attainment through its Programme for International Student Assessment (PISA) scores – a standardised international test of pupil attainment conducted every three years. In contrast to the school assessment results outlined above, these show no major changes: between 2009 and 2015, the mean average science scores for England fell from 515 to 512; maths remained at 493; and reading rose from 495 to 500.****

But teaching children is not the only thing schools do. Schools also undertake other activities – from providing extra-curricular activities to meeting parents, visiting pupils’ homes and contacting and co-ordinating with other agencies involved in serving children. Since September 2013, state-funded schools have also had a statutory duty to provide careers advice,[41] leading to an unavoidable increase in staff workloads. The amount of time teaching staff spend on these other activities has probably increased, although the impact is hard to quantify.

Teachers now refer more pupils to children’s social care, which could suggest they are spending more time supporting vulnerable pupils. The number of referrals rose by 39% between 2013/14 and 2017/18, the only years for which we have consistent data (see Chapter 5, Children’s Social Care).

In addition to these permanent changes, teachers have also faced numerous one-off demands that have required them to adapt working practices, notably: the new national curriculum introduced in September 2014[42] 

new GCSE specifications introduced in 2015.[43]

Aside from attainment, the other main measure for the quality of schools is Ofsted inspection results, which appear to be improving. Between August 2010 and March 2018, the proportion of schools rated Outstanding or Good at their last inspection rose from 64% to 76% in secondaries, and from 67% to 87% in primaries.[44]

But Ofsted ratings are less useful for measuring changes in the quality of schools.***** In 2012, the grading scale for primary schools was updated, with the category previously described as Satisfactory being replaced by Requires Improvement.[45] There is some evidence that this has made inspectors less likely to rate primary schools in this category – rating them instead as Good.[46]

Furthermore, since 2012, Ofsted has changed the way it inspects schools – it now re-inspects schools with lower ratings more frequently and does not routinely re-inspect Outstanding schools.[47] This means that, at least initially, the share of Good or Outstanding schools is likely to increase because the lowest-ranked schools are being inspected more often and so have less scope to fall down the grading scale.

*As a measure of school quality, attainment is inferior to contextual metrics that identify the progress pupils have made, as it does not account for differences between different cohorts of students. The government has recently introduced a progress metric for Key Stage 4 – ‘Progress 8’ – but this is a relative metric, so cannot be used to judge overall change over time.

**At the end of Key Stage 2.

***2017/18 attainment is not fully comparable to 2016/17 due to changes in the teacher assessment frameworks for the written component of these tests.

****The different trends in attainment and PISA results are not necessarily inconsistent because the former measures the percentage of children above or below a certain threshold while the latter measures mean average attainment.

*****There is some evidence that cuts to Ofsted’s budget led to reductions in how often it met its statutory school re-inspection targets. See Brackwell L, Acton C, O’Riordan A, Parrett M, Taylor A, Cockburn I, Hughes C, Pein N, Thorius A, Turner M, Ofsted’s Inspection of Schools, National Audit Office, 2018, pp. 7–8.​

Have schools become more efficient – and if so, can that be maintained?

Since 2010, schools have made savings on both teachers’ pay and – to a limited extent – the cost of goods and services they buy.

Teachers have been asked to do more, while being paid less in real terms. But this does not look sustainable – a fact that the government seems to have acknowledged in its announcement of substantial real-terms increases in per-pupil funding over the next three years. As noted, schools face growing recruitment and retention problems, with the government consistently missing recruitment targets for teacher trainees.

Schools were not, initially, under the same pressure as some other services covered in this report – up to 2015/16, schools spending rose in line with pupil numbers. Nonetheless, schools cut their costs over that period by holding down staff pay, which now makes up three quarters of schools spending. The public sector pay cap, by limiting the growth of spending on staff in schools, was the largest single saving schools made; teacher pay was frozen in 2011/12 and 2012/13 and pay awards were limited to below-inflation increases until July 2018.[48]

The cap helped constrain pay growth, although other new costs added to schools’ pay bills. In particular, in 2015/16 employer (i.e. schools’) contributions to the teachers’ pension scheme rose from 14.1% of salary to 16.4%,[49] rising again, to 23.6%, in September 2019.[50],*

The education think tank the NfER estimates that, on average, real-terms teacher annual pay fell by 12% between 2009/10 and 2015/16.** As a result, teacher pay has moved out of step with comparable private sector occupations. By 2017/18, the average earnings of a 21–30-year-old teacher (outside London) were 12% lower than those of people of the same age working in professional private-sector employment (also outside London), compared to just 2% lower in 2011/12.[51]

Beyond the pay cap, there is limited evidence that schools have made savings on the other goods and services they buy. Real-terms per-pupil spending on some goods and services fell notably between 2009/10 and 2016/17, including ICT learning resources (15.5%), energy (15.2%) and non-ICT learning resources (7.6%).[52]

But it is not possible to disentangle whether this is because schools bought goods and services more cheaply or just reduced how much they bought. Buying fewer of some goods and services may be efficient – for example, energy – but buying fewer learning resources represents a real reduction in inputs that may end up reducing school attainment.[53]

Schools did not achieve the DfE’s aim to save £1bn on back-office and procurement spending between 2010 and 2015.[54] The department introduced its School Financial Health and Efficiency programme in January 2016 with guidance and tools – expanded in 2018[55] – to help schools make savings, but there are few signs this has led to major savings. DfE surveys suggest that more teachers are using these resources – a 2018 survey of teachers found that 18% and 68% used the department’s efficiency metric and benchmarking data respectively,[56] up from 3% and 23% in 2017[57] – but this has yet to show up in school spending figures.

In 2017/18, the department piloted the introduction of school resource management advisers (specialist financial advisers employed to help schools find efficiencies), which it claims identified £35m of savings and ‘revenue generation’ opportunities. But the breakdown between savings and revenue generation – and how far these are genuine efficiencies, i.e. cost reductions that will not reduce the quantity or quality of school output – is not clear.

The biggest single change since 2010 that could have affected school efficiency was the expansion of academies under the then education secretary, Michael Gove. This policy was not solely intended to increase school efficiency, but the white paper that laid the ground for academies’ expansion stated that one goal was to “create a system in which schools are better able to raise standards”.[58]

However, the evidence on whether academies improved pupil attainment and school quality faster than comparable maintained schools is inconclusive,[59] suggesting expanding academisation had a limited impact. There is some evidence that ‘sponsored’ academies (poorly performing secondaries that were taken over and run by sponsors as academies under Labour) improved performance slightly faster than similar maintained schools.[60] But it is not clear whether this will hold true for ‘convertor’ academies (already high-performing schools that became academies), which formed most of the expansion after 2010.

There may be scope for further efficiencies in non-staff spending,[61] but this only accounts for a quarter of overall school spending. As such, buying goods and services for less will have a smaller impact on overall spending than savings on staff spending.

The DfE estimates, for example, that its new free service to advertise teacher vacancies could save £75m a year in recruitment costs[62] if all schools recruit only through the website – but that equates to just 0.2% of school spending in 2017/18.

Schools have mainly met financial pressures by making staff work more productively. Both primary and secondary schools have become more productive, but the biggest improvements have been in secondary schools. Both pupil–teacher ratios and average class sizes rose notably in secondary schools;[63] the average secondary school teacher is now teaching one more pupil than in 2009/10.

The number of secondary school teaching assistants also fell substantially, by 13.1% between 2011/12 and 2018/19,[64] while teachers have also taken on a greater role in other areas such as careers advice and pastoral care. Trends in pupil attainment suggest that this increased workload on teachers has not yet affected the quality of teaching in schools.

Schools are facing growing problems recruiting and retaining teachers. As a result, the government announced pay increases*** for teachers in July last year. Analysis from the Gatsby Foundation and Education Policy Institute (EPI) suggests that further financial incentives could help retain teachers in shortage subjects,[65] such as maths and physics (for which the DfE began testing retention payments in May 2019[66]) – as these are subjects for which teachers’ qualifications allow them to command higher salaries elsewhere.[67]

But increasing pay may not be enough on its own. There is some evidence that pay is not the sole cause of – or solution to – schools’ recruitment and retention problems. For most subjects, teacher surveys suggest that working conditions and flexibility are the biggest barriers.

A National Audit Office (NAO) survey of school leaders found that workload was most often considered the key barrier to retaining teachers (it was mentioned by 67% of respondents).[68]The NfER found that people leaving teaching took an average 10% pay cut[69] – and that 79% of teachers were satisfied with their pay.[70] Of those who left teaching, there was a 20% increase in part-time working,[71] suggesting that more teachers would like to work part-time[72] but that schools struggle to accommodate this.

There is, however, some evidence that schools can under some circumstances accommodate requests for part-time working: of those who return to teaching, just over 60% work full-time, compared to over 90% of the rest of the teaching workforce.[73]

Taken together, the available evidence suggests that schools will need to change working practices in order to mitigate recruitment and retention problems. The DfE introduced a teacher recruitment and retention strategy to address some of these problems – such as creating a new job-share website to help teachers find job-share partners – in January 2019.[74] But it is too early to judge whether this strategy will improve teacher retention.

Despite the department’s previous efforts to reduce workload, secondary school teachers reported working slightly longer hours in 2018 than in 2013. If the strategy is successful, we would expect teachers to report working fewer hours in the next OECD survey in 2021 – and for retention rates to duly increase in the next few years.

*The department has funded the higher school contribution rate for 2019/20. See Department for Education, ‘Teachers’ Pension Scheme protected to ensure it remains among most lucrative’, 2018,

**This does not account for changes in the composition (age and experience) of teachers. See Hillary J, Andrade J, Worth J, Teacher Retention and Turnover Research Update 4: How do teachers compare to nurses and police officers?, National Foundation for Educational Research, 2018, p. 7.

***In 2018/19, teachers on the lower pay bands (around 40%) received a 3.5% increase, with smaller increases of between 1.5% and 2% for more senior colleagues.

Have these efficiencies been enough to meet demand?

Unlike the other public services covered in this report, schools cannot ration the provision of their main service, which is providing children with access to lessons during the school term,* as every child has a statutory right to a place at school and the teaching that entails.** Without legislative change or a significant rise in independent education,*** the number of pupils taught by state schools will always rise in line with the number of school-age children.

There is, however, some evidence that schools have reduced the scope of school education by narrowing the curriculum. A 2018 survey of school governors and trustees carried out by the National Governors’ Association and the Times Educational Supplement found that around 5% of primaries had reduced the number of qualifications they off  red and 10% had reduced the number of subjects on offer, rising to over 40% and 50% respectively in secondary schools.****,[75]

However, it is not possible with the available data to disentangle whether this reduction in subjects is due to financial pressure or other changes that have occurred at the same time to school accountability and performance table measures.

Percentage of maintained schools with a cumulative deficit

Rising cumulative deficits in schools (when a school has a negative balance on its revenue account*****) are the clearest evidence that they have not been able to improve their efficiency enough. Schools appear to have had to spend consistently more than they received to provide services.

The average cumulative deficit among maintained secondary schools rose from £192,000 in 2009/10 to £484,000 in 2017/18.[76] This increase was driven by an increase in the number of schools with a deficit, rather than a few schools with growing cumulative deficits. The percentage of maintained secondary schools with a cumulative deficit rose from 18% in 2009/10 to 30% in 2017/18.[77]

It is difficult to compare trends in academies because the DfE does not publish data on individual academies’ cumulative deficits over time.[78] The NAO calculated that cumulative deficits in single academy schools are lower on average than in maintained schools, though they rose in a similar way between 2011/12 and 2014/15. The proportion of secondary single academy schools with a cumulative deficit increased from 3.2% to 6.1% over this period.[79]

In 2017, the DfE began publishing figures for the proportion of academy trusts (charitable bodies that run one or more academy schools) that have cumulative deficits. However, this data counts trusts with multiple schools as equivalent to single academy trusts. The proportion of academy trusts with a cumulative deficit rose slightly from 5.5%[80] in 2015/16 to 6.4% in 2017/18.[81]

Cumulative deficits arise if schools consistently run in-year deficits – that is, if they spend more than they receive in a given year. In-year deficits are volatile, and some may be for planned spending such as capital projects, but the overall trend has been for the proportion of maintained schools that are spending more than they receive each year to increase. The EPI found that 48% of maintained primaries and 60% of maintained secondaries ran in-year deficits in 2017/18, compared with 33% and 30% respectively in 2010/11.[82]

The proportion of primary and secondary academies with in-year deficits was slightly lower than among primary and secondary maintained schools in 2016/17,[83] the latest year for which we have data for both academies and maintained schools. The likelihood of an academy having an in-year deficit was generally lower the larger the size of the trust it belonged to.

The trend in the prevalence of in-year deficits among academies initially followed maintained schools – the NAO calculated that the proportion of secondary single academy trusts which spent more than their income rose from 38.8% to 60.6% between 2012/13 and 2014/15.[84] But since then the proportion of academies running in-year deficits has fallen. Comparing academies for which data is available in all years, the EPI found that the proportion of academies running in-year deficits fell slightly between 2014/15 and 2016/17, from 39% to 31% in primary academies, and from 58% to 51% in secondary academies.[85]

The increasing prevalence of cumulative deficits, and in-year overspending in maintained schools, presents a challenge to the government. While variation in spending and educational outcomes implies there could be some efficiencies left to find in some schools, the way individual schools are using their budgets suggests the scope for savings is limited. Some may not have the skills to manage budgets more efficiently, or may be constrained from changing how they use staff by the size and structure of their buildings.

The DfE’s theoretical estimates of the efficiencies possible have not been matched with a practical understanding of how, and at what pace, schools are able to make them.[86] The department must analyse whether and how individual schools could, in practice, reduce the inputs they use or reconfigure how they use them in order to save money without reducing the quality of teaching or pupils’ educational outcomes.

Schools might be able to operate more efficiently by increasing class sizes, cutting the number of teaching assistants or reducing opening hours, but the government has not yet been explicit about how far it thinks these changes are either possible[87] – or politically acceptable.[88]

*A 2018 survey of school governors and trustees found that reducing school opening hours was the least common response to financial constraints, with fewer than 3% of school governor respondents reporting that their school had reduced hours. See Holland F, School Governance in 2018: An annual survey by the National Governance Association in partnership with TES, National Governance Association, 2018, p. 30.

**The NHS constitution sets out similar targets for GPs and hospitals, but demand is less predictable so it provides a minimum ‘floor’ of services rather than a set standard. See Department of Health and Social Care, The NHS Constitution for England, 2015,

*** The number of children in independent schools has risen slower than overall pupil numbers, at only 1.1% since 2009/10. See Department for Education, ‘Schools, Pupils and their Characteristics’, 2019,

**** The accuracy of this is doubtful, however, because it was a self-selecting sample of volunteers, who may not be representative of school governors and trustees. See Holland F, School Governance in 2018: An annual survey by the National Governance Association in partnership with TES, National Governance Association, 2018, p. 4.

*****This would normally be described as debt, although school cumulative deficits are not owed to any third parties.

How will demand for schools change?

Pupil numbers are predicted to rise – but not quicker than the promised increase in school funding. The new three-year settlement, set out by the Johnson government in September 2019, will result in school funding rising by 14.2% in real terms (10.3% per pupil in real terms) between 2018/19 and 2022/23.[89]

If the government chose to hold per-pupil funding constant in real terms between 2022/23 and 2023/24, it would need to allocate schools a further real-terms budget increase in 2023/24, leaving total school spending in that year 14.5% higher in real terms than it was in 2018/19 (see the table below).

The Institute for Fiscal Studies estimates that the increases promised up to 2022/23 would reverse almost all of the cuts to per-pupil spending that have occurred since 2009/10.[90] This funding boost is a substantial change in direction following the 4.1% real-terms cuts to per-pupil spending between 2015/16 and 2018/19.

The DfE estimates that there will be 4.4% more pupils in primary and secondary schools in England in 2023/24 than in 2018/19.* Schools will also face some unavoidable additional cost pressures over the next few years – such as the increase in their contribution to teachers’ pensions schemes, which rose in September 2019. But the size of the planned increase in spending means that schools should be able to use this money to improve performance and address growing workforce pressures as well.

Projected demand and spending on schools


Projected increase in demand by 2023/24 4.4%
Spending scenario Confirmed schools settlement Meet demand
Change in real-terms spending by 2023/24 14.5% 4.4%
Spending in 2023/24 (2018/19 prices) £42.3bn £38.6bn
Projected gap (2018/19 prices) -£3.7bn -

Source: Institute for Government calculations. See Chapter 13, Methodology.

The government has not specified exactly what it expects schools to achieve with the extra money they have been given, although it has said it will introduce more support for the “most challenging” schools[91] and increase teacher starting salaries.[92] It has also indicated that it will judge the impact of the extra money by looking at changes in the proportion of pupils receiving the expected standards at the end of Key Stage 2 (by the age of 11), the proportion of 16-year-olds who receive a grade 4 or above in Key Stage 4 English and maths exams, and the three-year teacher retention rate.[93]

Setting clear goals for the additional money is a sensible approach to maximising the value of the spending boost, although a blanket increase in teacher starting salaries is unlikely to be good value for money, even if it does look like an attractive selling point.[94] Some subjects, for instance – such as history and PE, which have exceeded recruitment targets every year since 2011/12 – do not face immediate recruitment problems that require a salary boost to solve.

*See Chapter 13, Methodology