The Working Time Regulations 1998 (WTR) entitle a full-time worker to a minimum of 5.6 weeks’ (28 days) paid annual leave.  Part-time workers are still entitled to 5.6 weeks, but this is usually expressed in terms of days or hours as a pro-rata proportion of the full-time entitlement where the part time worker works a fixed number of days each week or month.

The annual leave entitlement under the WTR is expressed in weeks, likewise holiday pay is expressed in terms of a ‘week’s pay’.   For casual staff who may work occasionally at discrete times throughout a year, and for those who work part of the year only, employers have often expressed and calculated holiday entitlement as a percentage of total annual hours worked, typically 12.07% (the Percentage Approach). This derives from the fact that the standard working year is 46.4 weeks (that is, 52 weeks less the statutory 5.6 weeks holiday entitlement) and 5.6 weeks is 12.07% of 46.4 weeks.   Indeed, ACAS guidance used to suggest this approach.  Is this correct?

Invariably not.  In the case of Harpur Trust v Brazel [2022] UKSC 21 the Supreme Court has held that the 5.6 weeks’ annual leave entitlement should not be reduced pro-rata in this manner for so called ‘part year’ workers (i.e., those who are employed for the whole year but do not work every week). The case is complex and in fact focussed on the calculation of holiday pay itself, but the principle was established.

Mrs Brazel was a visiting music teacher at a school run by Harpur Trust (Harpur).  She worked part-time under a permanent contract on a zero-hours basis with an entitlement to 5.6 weeks’ paid holiday, which she was required to take during school holidays.

Mrs Brazel generally worked school term times only, and when she did work, her hours varied.  She was paid only for the hours she worked.  By agreement, Mrs Brazel was paid holiday pay at the end of each term. Her actual holiday pay however was calculated based on 12.07% of Mrs Brazel’s earnings for the previous term.

Mrs Brazel complained she was being underpaid holiday and brought an employment tribunal claim arguing that the 12.07% Percentage Approach bore no relation to the calculation required by the WTR and that holiday pay should be calculated according to the Employment Rights Act 1996, i.e., the level of average earnings over the 12-week period immediately before holiday is taken, with weeks which no remuneration is earned being ignored.  (This has since changed for statutory holiday pay purposes to average remuneration over the previous 52 weeks (or the number of complete weeks for which the worker has been employed, if that period is less than 52 weeks), with earlier weeks being brought into the calculation if there are weeks where no remuneration was earned, up to a maximum of 104 weeks before the calculation date.)

The tribunal dismissed this claim finding that a pro rata principle should apply for part-time workers who worked fewer than 46.4 weeks per year so that payment was capped at 12.07% of annualised hours.

However, the Employment Appeal Tribunal and the Court of Appeal disagreed and held that there was no pro-rating requirement in the WTR for part-year workers.  The Court of Appeal held that attempting to build into the WTR pro-rating for part-year workers would make the scheme entirely different and Parliament had made a policy decision not to do this when drafting the WTR.  This meant that Mrs Brazel’s holiday pay represented about 17% of her annual earnings in fact.

Harpur appealed to the Supreme Court, but the appeal was dismissed. In doing so it has comprehensively rejected the Percentage Approach and this should no longer be relied on.

Provided a worker remains under contract, they have an entitlement to 5.6 weeks’ paid holiday in each year, regardless of the amount of work actually done. This means we have a system where non-working weeks are included for calculating accrued holiday entitlement but ignored when calculating holiday pay itself.


There were two elements to Mrs Brazel’s part time working (1) she did not work a full working week, and (2) for large parts of the year she did not work for Harpur at all.  It is only the second type of part-time working (“part-year working”) which this case was concerned with.

However, the decision has far reaching implications as it affects all workers/employees who have a permanent contract but who do not have normal working hours.

For example, a casual or zero-hours worker engaged under a contract for the whole year but who works varied hours and from time to time will still have holiday entitlement of 5.6 weeks. The pay for this will not be reduced by virtue of their non-working periods, with holiday pay being calculated according to a ‘week’s pay’ under the Employment Rights Act rather than the 12.07% Percentage method.

This does lead to anomalies, as a worker with no fixed hours or days of work who works, say two or three days per month on average, will have the same holiday entitlement as a full-time worker and more holiday than a part-time worker who has a fixed working pattern; it could also result, in some cases, that they will receive more remuneration through holiday pay than hours worked.  Whilst this might not have been the intention of Parliament, this is the effect of the legislative drafting.

The case has led to claims being brought already, and a failure to modify holiday pay arrangements based upon the Percentage approach, creates an ongoing risk of exposure.

As a result of the Supreme Court judgment the government wishes to address the disparity so that holiday pay and entitlement is proportionate to the time workers spend working.  The proposal is to amend the 52-week reference period, so that weeks in which no remuneration is received are included. The government has issued a consultation paper to understand the implications of the judgment on different sectors.  The consultation closes on 9 March 2023.

For advice on issues surrounding this judgment please contact a member of the Employment Team