Levelling Down: Against VAT on Schools
Argues that VAT on private school fees raises less than advertised and shrinks the market for schooling, and that fairness should come from improving state schools.
Public policy requires striking a balance between equity and efficiency. Policies that promote fairness can reduce total welfare, and policies that maximise efficiency can lead to unequal outcomes. The introduction of value added tax on private education is a clear manifestation of this tension. Private education is viewed by some as a positional good – something valued largely because others do not have it – which should therefore be taxed to raise revenue and promote equality. However, such an argument overlooks the larger economic function of private education in society.
This essay explores the economic consequences of introducing VAT on private education. It examines human capital effects, behavioural responses and tax incidence through supply and demand analysis. With these factors considered, it argues that VAT on private education is economically inefficient and reduces welfare.
The Importance of Private Education
To many, private schools are highly exclusive institutions. In fact, as Daniel Lilley has documented for Civitas, only around six to seven per cent of pupils in the UK attend one – though the proportion roughly doubles in the sixth form. But this perception ignores private education’s role as a social and economic institution.
From a human capital perspective, Gary Becker argued that education is an investment which increases productivity, innovation and long-term economic growth. These benefits extend beyond the individual and affect society at large. Statistics reveal that private education is typically of higher quality: research for the Sutton Trust by Rebecca Montacute and Carl Cullinane found that pupils at private schools achieve higher grades on average and are disproportionately represented at the most selective universities. Private schools also tend to have better facilities and smaller classes – Lilley’s figures give the independent sector roughly nine pupils per teacher against eighteen in the state sector. Discouraging enrolment in private education therefore reduces human capital investment.
Furthermore, private schools often engage in activities that indirectly benefit the broader education system, including teacher training, curriculum innovation and the setting of performance benchmarks. Even students in the state sector benefit from competitive pressure: research presented to the Royal Economic Society found that the presence of high-performing private schools raises achievement for private and state school pupils alike. VAT, by thinning that sector, indirectly hinders national schooling.
Private schools also absorb substantial educational demand. A report by Oxford Economics for the Independent Schools Council found that independent schools saved the taxpayer around £3.5 billion a year by educating pupils the state would otherwise have had to fund – a sum equivalent to roughly 3.5 per cent of total state spending on education. That relieves pressure on the state sector and frees money to be spent elsewhere.
By providing diversity of provision, private education also enables families to choose environments that best suit their children’s needs – particularly, though this is rarely acknowledged in the debate, for children with special educational needs whose requirements the state sector has struggled to meet. Although that option is available to only a small minority of pupils, it offers a degree of educational choice that would otherwise be unavailable, and VAT can only decrease the number of families who can reach it. Since private education generates a range of positive externalities – benefits enjoyed by people other than the fee-payer – reducing its consumption through taxation generates unintended social costs.
Consequences of VAT
Even at a non-technical level, VAT produces a chain of economic effects beyond revenue collection. Since 1 January 2025 private school fees have carried VAT at the standard rate of twenty per cent, and the cost of private schooling has risen sharply as a result. The Independent Schools Council’s census recorded average day fees rising from £6,021 a term in January 2024 to £7,382 in January 2025, an increase of 22.6 per cent including VAT, against 8.4 per cent the year before. HMRC had assumed schools would raise fees by around ten per cent on average, absorbing much of the tax themselves. They could not.
When fees rise by more than a fifth, some families inevitably withdraw their children. These are rarely the wealthiest; they are middle-income families who have made great financial sacrifices to access private education, and who sit closest to the margin at which the sums stop working. The government forecast that around 35,000 pupils would move into state schools, with HMRC putting the long-run figure at 37,000. The Department for Education’s census recorded about 11,000 fewer independent school pupils in the policy’s first year, and by the Independent Schools Council’s 2026 census the sector stood at just over 526,000 pupils – roughly 30,000 below its pre-VAT peak and its lowest level in almost a decade.
Supporters of the policy read those numbers as vindication: a fee rise of over twenty per cent has produced an enrolment fall of only a few per cent, which implies that demand is relatively inelastic and that the revenue base is secure. That reading is too quick. Aggregate demand for private education looks inelastic because it is dominated by families for whom fees were never the binding constraint, and by pupils already deep into a course of study whose parents will pay almost anything rather than move them before an examination. The response is visible instead at the entry points, where families are making a fresh decision: the 2026 census recorded falls of over six per cent in Year 1, Year 3 and the start of the sixth form, and over five per cent in Year 7. The sector is not being abandoned. It is failing to be joined. That is a slower process than an exodus and, for exactly that reason, a more durable one.
When these students move into the state system, the burden on public resources increases. This leads to larger class sizes, higher teacher workloads and greater pressure on school infrastructure. In effect, although the government gains tax revenue from VAT, it also incurs new educational obligations. VAT revenue is therefore partly offset by new public costs.
With reduced demand, private schools may also be forced to cut investment in teachers and facilities, and in some cases to close entirely. At least 65 independent schools shut during 2025, of which 51 were mainstream schools and 14 special or alternative providers. It is fair to note that this is close to the sector’s ordinary annual rate of closure, and that falling birth rates are shrinking the school-age population regardless of tax policy; the government had itself budgeted for roughly 100 closures over three years above normal turnover. The point is not that VAT alone destroyed these schools, but that it removed the financial margin that small preparatory and specialist schools relied on to survive a demographic squeeze they were already facing. The result is reduced educational diversity and competition, which currently encourage high standards across the sector.
Interpreting the Tax Incidence Diagram
To formalise these consequences, the policy can be analysed through a standard microeconomic model of taxation using supply and demand. In this framework, private education is treated as a market in which families demand schooling services while private institutions supply them.
VAT enters the model as a per-unit tax on consumption. It creates a wedge between the price consumers pay and the price suppliers receive, represented on the diagram by a leftward shift of the supply curve. The new equilibrium involves a higher market price, a lower effective price received by schools, and a lower equilibrium quantity of private education.
This aligns with the behavioural effects described above. The withdrawal of families from private education corresponds to the contraction in equilibrium quantity, and the extent of that contraction depends on how price-sensitive demand turns out to be. Where demand is elastic, a tax produces a proportionally larger fall in quantity than the rise in price, and substitution towards state schooling is significant; where it is inelastic, the fall is smaller and the revenue larger. The model does not settle the question – it simply shows what hangs on it.
Tax Incidence
A significant implication of this model is that the burden of taxation is not borne solely by consumers. Peter Mieszkowski’s work on the theory of tax incidence established that the economic burden of a tax is shared between buyers and sellers according to the relative elasticities of demand and supply. Part of the tax is therefore absorbed by private schools through reduced revenue. In practice, the split is visible in the accounts: schools raised fees by roughly the full value of the tax, but simultaneously increased means-tested bursary spending to retain families who would otherwise have left, which is absorption of the burden by another route. The schools with the least room to do either – small preparatory and specialist schools with thin reserves and no endowment – are the ones that have reduced provision or exited the market.
Fiscal Offset
A further inference from the model concerns effects beyond the private education market. When students exit private schools they are absorbed by state-funded institutions, creating a fiscal offset. Although VAT increases government revenue, it simultaneously increases public expenditure obligations, so that each pupil transferring into the state sector represents an additional marginal cost to the budget. Net fiscal gain equals VAT revenue minus the cost of additional state provision.
This is not a hypothetical adjustment; it is built into the official numbers. The Department for Education put the gross yield at about £1.725 billion a year. The Institute for Fiscal Studies, netting off the cost of educating the pupils who move, put the medium-to-long-run gain at £1.3–1.5 billion. The gap between those figures is the fiscal offset, and it is not small.
Some may point out that many state schools have spare capacity, meaning the marginal cost of educating an additional pupil is relatively low, and the government has emphasised that 83 per cent of primary schools and 77 per cent of secondaries have at least one unfilled place. However, this argument ignores the fact that spare capacity in the state sector is not evenly distributed. There is often little slack in the areas with the highest demand for school places. There are many spare places in parts of London, for example, while schools in Kent are heavily oversubscribed. The real marginal cost of absorbing additional pupils may therefore be much higher than national averages suggest, and higher still for pupils with special educational needs, whose provision the state sector is already struggling to fund.
This relocation of responsibility for educational provision from private agents to the government explains why the expected revenue gains from VAT are often overstated. Effective tax yield is reduced by both behavioural responses and compensatory public spending requirements.
Deadweight Loss
The greatest implication of the supply and demand framework is the existence of deadweight loss – the disappearance of mutually beneficial exchanges that no longer occur once a market is pushed away from its socially optimal equilibrium. This loss of economic efficiency and welfare is not transferred between agents; it is eliminated entirely from the market. Nobody gains it.
The total deadweight loss increases with the elasticity of demand, which is why the composition of the response matters more than its headline size. A market in which most existing families stay but fewer new ones join is one in which the losses accumulate quietly, year on year, at the margin. The deadweight loss triangle represents the school closures, the reduced competition and the lost educational diversity described above, since these exchanges no longer occur.
Crucially, this loss is not merely financial but a reduction in human capital formation. That the education system can provide less of a student’s preferred education indicates a fall in allocative efficiency. In other words, this deadweight loss is a loss of potential productivity, innovation and long-term economic growth.
Conclusion
Each of the qualitative arguments presented earlier maps onto the model. The reduction in educational choice corresponds to the fall in quantity exchanged. The decline in school investment corresponds to reduced producer surplus. The increased financial burden on families corresponds to reduced consumer surplus. The overall reduction in the value of education corresponds to the deadweight loss triangle.
Twenty months of evidence have not produced the collapse the policy’s opponents warned of, and it would be dishonest to pretend otherwise. What they have produced is quieter and harder to photograph: fees up by more than a fifth, a sector at its smallest in a decade, entry cohorts shrinking by six per cent a year, and a net yield materially below the headline figure once the state sector’s new obligations are paid for. These are the costs a deadweight loss triangle is drawn to represent, and they compound.
Taken together, this framework demonstrates that VAT on private education is more than a fiscal adjustment. If education is human capital formation, then this policy is ultimately a choice between short-term redistributive gains and long-term societal progress. Although it may be justified on grounds of equity, true equality should be pursued through levelling up rather than levelling down – ensuring that improvements to fairness in the state sector are not achieved by reducing opportunities elsewhere.
