Term-time staff: 52-week average or 12.07% after Harpur Trust and the 2024 reforms
For leave years beginning on or after 1 April 2024, a term-time worker who meets the legal definition of a part-year worker accrues holiday at 12.07% of the hours worked in each pay period, and is paid for it either on a 52-week average of normal pay or as a 12.07% rolled-up uplift on each payslip. For leave years before that, the Supreme Court's 2022 decision in Harpur Trust v Brazel applies: 5.6 weeks of paid leave a year with no pro-rating for weeks not worked. An exposure check has to apply the right rule to each leave year.
Two rules, one dividing line
Harpur Trust v Brazel was about a music teacher paid only for the weeks she taught. The Supreme Court held on 20 July 2022 that her statutory entitlement was 5.6 weeks of leave a year, paid at her average weekly pay, and that the employer could not reduce it in proportion to the weeks she worked. The widely used 12.07% method was, for workers like her, unlawful at the time.
The government then amended the Working Time Regulations. From leave years beginning on or after 1 April 2024, a worker who meets the new definition of a part-year worker or an irregular-hours worker accrues holiday at 12.07% of the hours worked in each pay period, with holiday pay calculated either on the previous 52 paid weeks or by rolled-up pay. The gov.uk guidance explains the figure: 5.6 weeks of leave against 46.4 working weeks is 12.07%.
Who the new definition catches
Acas gives the test for a part-year worker. The contract must:
- say the worker is required to work only part of the year;
- include periods of at least a week when they are not required to work and are not paid for; and
- be in place all year round, including when they are not working.
The middle condition does most of the damage in schools and trusts. A teaching assistant on an annualised salary, paid one twelfth every month including August, is paid during the weeks they do not work. Whether that person is a part-year worker is a question of how the contract is written, not of how the rota looks, and the answer decides which calculation applies.
What an exposure check has to do
| Leave year | Rule | Entitlement | Pay basis |
|---|---|---|---|
| Began before 1 April 2024 | Harpur Trust v Brazel | 5.6 weeks, no pro-rating | Average of the previous 52 paid weeks |
| Began on or after 1 April 2024, worker meets the definition | 2024 regulations | 12.07% of hours worked per pay period | 52-week average, or 12.07% rolled up on each payslip |
| Began on or after 1 April 2024, worker does not meet the definition | Ordinary rules | 5.6 weeks | Normal pay for the first four weeks, basic for the rest |
Three groups of staff, two dates, and a definition that turns on contract wording. Most underpayments in this sector come from applying one rule to everyone, usually the convenient one, across both periods.
What normal pay must include
For the first four weeks of leave (and for all of a part-year worker's 12.07% entitlement) the regulations require normal pay: payments tied to the tasks in the contract, payments for status such as length of service or qualifications, and other payments such as overtime that were paid regularly in the 52 weeks before the calculation. A term-time worker with regular paid overtime whose holiday was paid at basic rate is underpaid even if the entitlement was right.
What to send for a snapshot
Payroll exports for the last 52 paid weeks, the contract wording on working weeks and pay, and the leave records since 6 April 2026. The £295 snapshot recomputes a sample of term-time and irregular-hours staff under the rule for each period and gives the range in writing. It is a calculation by a chartered accountant, not legal advice, and where the contract wording needs a lawyer it says so.
Common questions
Is every term-time employee a part-year worker?
No. Acas sets out three conditions: the contract says the worker is required to work only part of the year, there are periods of at least a week when they are not required to work and are not paid, and the contract is in place all year round. A term-time worker whose salary is spread evenly across twelve months is paid in the non-working weeks, so the second condition may fail and the 12.07% method may not apply.
Which rule applies to a leave year that started on 1 September 2023?
The old one. The new accrual and pay rules apply to leave years beginning on or after 1 April 2024, so a September-to-August leave year moved to the new rules on 1 September 2024. Anything earlier is calculated under Harpur Trust v Brazel.
Does the 52-week average skip the summer?
Weeks with no pay are left out and earlier paid weeks are brought in, going back up to 104 weeks to find 52 paid ones. A term-time worker's average is therefore an average of working weeks, which is the point of the method.
Get the number in writing
A fixed-price snapshot recomputes a sample of your irregular-hours staff and gives you the exposure range in five working days.
Request the snapshotRelated
Sources
- UK Supreme Court: Harpur Trust (Appellants) v Brazel (Respondent), UKSC 2019/0209, judgment 20 July 2022.
https://www.supremecourt.uk/cases/uksc-2019-0209 - Department for Business and Trade, GOV.UK: Holiday pay and entitlement reforms from 1 January 2024 (definitions, the 12.07% accrual method, the 52-week relevant period, rolled-up holiday pay).
https://www.gov.uk/government/publications/simplifying-holiday-entitlement-and-holiday-pay-calculations/holiday-pay-and-entitlement-reforms-from-1-january-2024 - Acas: Irregular hours and part-year workers: building up holiday (what a part-year worker is; 12.07% accrual).
https://www.acas.org.uk/irregular-hours-and-part-year-workers - Acas: Irregular hours and part-year workers: holiday pay (the 52-week average).
https://www.acas.org.uk/irregular-hours-and-part-year-workers/calculating-holiday-pay