After the VAT change, what would count as evidence of its impact?

01

The answer pointed backwards

Asked what assessment had been made of VAT’s impact on private schools, the Department for Education’s answer on 11 June 2026 referred to the Treasury’s Tax Information and Impact Note. That is a consequential distinction. The note, updated in November 2024, set out the policy’s expected effects before implementation.12

Its existence is evidence of an official assessment. It is not, by itself, a post-implementation evaluation showing what subsequently happened or how much of an observed change the policy caused. The question therefore remained open in a more precise form: which later evidence could test the expectations?12

The Commons Library’s 26 June briefing provides a further development in that discussion. It brings together the policy history, forecasts and emerging pupil evidence while identifying limits to causal explanation. Our earlier market report examined the populations counted by different censuses. This feature follows the next question: even when the population is correctly defined, what can a change in that population tell us about the tax? A sector total can establish a movement without establishing the decisions behind it.3

02

Three questions too often become one

The fiscal question concerns the revenue generated and any offsetting public costs. The school question concerns the change in income and expenditure for individual providers. The family question concerns the price, affordability and service received. These are connected, but none can stand in for the others.

A school could retain its pupils while absorbing some cost and reducing its financial headroom. Another could raise its total charge and see fewer new entrants. Both could contribute tax receipts. Describing either situation as the single national result would discard important information.

The Treasury impact note contains forecasts rather than a ledger of realised receipts. The Library briefing also records a later upward revision to projected yield, linked to the earnings assumption used to project fees. A revised forecast is new information about the model; it is not equivalent to new evidence that schools retained every pupil or that families faced no pressure.23

Likewise, a forecast of fewer pupils over several years cannot be evaluated simply by comparing one year’s net headcount change with the full long-term estimate. The time horizon and the counterfactual must match.23

03

The missing pupils are not a single flow

A lower school roll can result from fewer arrivals, more departures, larger leaving cohorts, or a combination. A family that never applies does not appear in a withdrawal survey. A child who switches from boarding to day attendance changes the business mix without necessarily leaving independent education. A pupil who transfers between independent schools changes the fortunes of both providers while leaving the sector count unchanged. These are mechanisms to investigate, not estimates of how much any mechanism contributed in June.

The Library’s discussion of the English data explicitly recognises several plausible influences, including demographics, wider household costs, fee trends and international demand. It also notes that the published boarding totals do not reveal all underlying movements. That restraint matters for leadership.3

If a school mistakes a smaller relevant birth cohort for a failure of its proposition, it may spend heavily on the wrong response. If it dismisses an affordability problem as demography, it may ignore a change that families are already making. Better interpretation requires more than a convenient national explanation.3

04

The school-level evidence that adds something

The most informative local account follows comparable entry cohorts through enquiry, application, offer and acceptance, then records the eventual attendance pattern and net fee contribution. Its purpose is to explain where a change occurred, not to manufacture a precise attribution to VAT from small samples.

Reasons supplied by families deserve careful treatment: they can reveal recurring concerns but may be incomplete, affected by how the question was asked, or shared by too few respondents to support a broad claim. Publishing the method is part of publishing the result.

Price comparisons also need a consistent basis. HMRC’s guidance distinguishes the treatment of a package from genuinely separate supplies. A headline tuition figure can therefore be a poor proxy for the family’s total bill or the school’s retained income.4

Our recommended analysis would identify the included service, tax-inclusive payment and any assistance on a comparable basis. It would then ask what changed in the offer. A lower net contribution accompanied by higher enrolment is a different strategic outcome from a higher contribution with fewer pupils, even if both are described as resilience.4

05

Public benefit needs its own evidence

The connection to public education introduces another boundary. In a January 2026 letter to the Scottish Parliament, the Exchequer Secretary explained that VAT receipts are not earmarked in a way that permits particular additional spending to be traced to this policy.5

That does not disprove the government’s fiscal argument. It means a claim that a specific school improvement was paid for by a particular private-school VAT receipt would require evidence the general statement does not supply. Revenue forecasts, spending settlements and service outcomes should remain distinct parts of the account.5

By June’s end, a defensible assessment could acknowledge real financial pressure and emerging changes while avoiding claims of causation that the available evidence did not support. The next useful development would connect a consistent pupil population, an explicit comparison period and a credible account of what would otherwise have happened.

For individual schools, the immediate value lies in understanding their own family decisions and financial response. For this knowledge bank, returning to the story should mean adding observed receipts, cohort flows or a published evaluation, rather than relabelling an earlier forecast as proof.

Leadership questions

Questions that turn the development into a governing conversation.

What we are monitoring next

Publication is the beginning of the watch.

• Published tax receipts distinguished from forecast revisions.
• Pupil-flow evidence and comparable intake cohorts.
• Any explicit post-implementation evaluation and its counterfactual.
• A credible evaluation identifying causal effects, or observed fiscal and pupil-flow evidence on a consistent basis, could change the assessment substantially.

Continue the evidence trail

Read across the historical record

• After the tax change: the financial decisions behind the independent-school fee ↗

The continuing record

How the record develops

• The independent-school market enters a more local phase ↗

  Returns to the evaluation question using the government’s later answer without converting monitoring into causal proof.

• The December VAT appeal: the court timetable and the school decisions being made now ↗

  Adds the separate, verified legal timetable to the earlier question of how policy effects should be evaluated.