Holiday pay for those with irregular hours

The Harpur Trust v Brazel

A zero hours, term-time employee’s holiday pay should not have been capped at 12.07% of her annualised hours as suggested in ACAS guidance but rather calculated using the 12-week averaging method in the Working Time Regulations.


Ms Brazel worked as a part-time music teacher for a school and she did not work every week. Her zero hours contract entitled her to the full-time equivalent of 5.6 weeks’ annual leave, and she had to take her leave during school holidays. The school calculated her holiday pay at 12.07% of hours worked in a term (12.07% being reached by dividing 5.6 (full-time equivalent) by 46.4 (the total number of weeks in a year less the 5.6 weeks’ holiday)), paid in three instalments at the end of each term, relying on the relevant ACAS guidance (see p. 6 of Holidays and Holiday Pay). Ms Brazel, supported by her union, believed that her holiday pay should rather have been calculated using her average weekly earnings over the 12-week period immediately before her holiday was taken (WTR, reg. 16 and s. 221-224 of the Employment Rights Act 1996).

At the tribunal, the school successfully argued that the statutory entitlement of 5.6 weeks holiday pay should be pro-rated where an employee had worked fewer weeks than a standard 46.4 week working year (deducting 5.6 weeks of statutory leave from 52 weeks). Ms Brazel’s entitlement was then pro-rated using the percentage of 12.07% of her term pay. As the school year varied between 32 and 35 weeks a year, this resulted in her holiday pay entitlement being lower than if her holiday pay had been calculated by averaging her normal rate of pay over the 12 weeks prior to the holiday being taken (excluding out-of-term weeks) – the approach outlined in the WTR. Applying the WTR calculation, a part-time worker on a 32-week year, would be entitled to 17.5% of term-time pay in holiday pay (46.4/32 x 12.07%). Consequently, if the WTR calculation was applied, Ms Brazel would receive proportionately more holiday pay than a full-time employee with 5.6 weeks paid leave. To prevent full-time workers being treated less favourably than part-time workers and to avoid an unjustified bonus for term-time only workers, the tribunal found that words should be read into WTR to the effect that where a worker has no normal hours and works fewer than 46.4 weeks per year, holiday pay should be capped at 12.07% of annualised hours. Ms Brazel appealed.

EAT and Court of Appeal decisions

Allowing her appeal, the EAT held that Ms Brazel’s holiday pay should have been calculated using the 12-week averaging method in the WTR. There was, said the EAT, no requirement to pro-rate the leave entitlement of part-time employees (which is what the 12.07% formula achieves), whether to avoid a ‘windfall’ for term-time only workers or to avoid full-time employees being treated less favourably than part timers. The EAT recognised that this might produce anomalies such as to favour an employee in Ms Brazel’s position who didn’t work throughout the year, but in its view the legislation was unambiguous.

The tribunal had overlooked the overriding principle that part-time workers were not to be treated less favourably than full-time workers. There is no principle to the opposite effect. To read into the WTR, as the tribunal had done, was to completely misinterpret the reasoning and purpose of the statutory scheme on paid annual leave. The school appealed.

Its appeal was dismissed. The Court of Appeal, upholding the EAT’s decision, used the term ‘part-year worker’ for those employed all year round but not working the whole year. The court rejected the school’s argument that a pro rata principle should be applied to the accrual of leave for ‘part-year workers’; EU law did not require leave to be reduced pro rata, and it wasn’t necessary to apply a pro rata principle to the accrual of leave under the WTR. All the WTR require is working out a week’s pay (under s. 221-224 of the Employment Rights Act) and multiplying it by 5.6. The court wasn’t persuaded by the argument that such an approach was a boon to those on permanent contracts but who are only required to work for a few weeks a year – if it was, then so be it.

While the legal arguments were complicated, the Court of Appeal summarised the position in one sentence: 'The WTR do not provide for the kind of pro-rating for which the [school] argues and which underlies the application of the 12.07% formula in the case of a part-year worker. The exercise required by reg. 16 and the incorporated provisions of the 1996 Act is straightforward and should be followed’.

Link to judgment:


This decision will produce an anomaly that those who work part time might end up receiving more for their holiday than those who work full time. It is quite likely that many schools and colleges apply the 12.07% in accordance with the ACAS guidance and this will now have to be revisited. However, it is not just teachers who are affected; all zero-hours employees who do not work a full year must have their holiday calculated in this way and not by the application of the 12.07% formula. So, for instance, employees contracted to work at specific festivals in a year but who do not work in the intervening period.