The entitlement to 5.6 weeks’ paid annual leave in the Working Time Regulations (WTR) is fairly straightforward to operate for full-time employees, working five days a week and paid a fixed salary. The employer just needs to let them take five weeks and three days off work, and carry on paying them their salary. Ever since the seminal judgment of the CJEU in BECTU we have known that the right in the parent Directive applies to all workers, without exception, from day one. But what does a right, expressed in weeks, mean for those, often misleadingly referred to as ‘atypical’ workers, who work only some days a week, only some weeks of the year or under a zero hours contract? What is the link, if any, between the time or hours worked and the period of annual rest?
The Harpur Trust v Brazel is the latest in the series of cases grappling with these problems. In its judgment today, the Supreme Court unanimously rejects the employer’s argument that the statutory leave entitlement of a worker who works irregular hours or only some weeks of the year is reduced as a function of her part-time working. Mrs Brazel’s entitlement was not to be calculated on the basis of 12.07% of the annual hours she worked. Rather, in common with every worker, she was entitled to the full entitlement, of 5.6 calendar weeks’ annual leave. Because Mrs Brazel did not work normal hours, pay in respect of that period of leave must be calculated in accordance with the formula for a “week’s pay” set out in s.224, so that it was based on her average pay in the 12 weeks prior to taking leave, ignoring weeks for which she was not paid anything.
The case is important for the leave entitlement of every worker in the UK. Had the Harpur Trust’s arguments succeeded, employers, workers and ETs would have faced the spectre of needing to determine the hours worked by every worker to calculate the annual leave entitlement. No doubt it would have required records of hours worked. Even for those paid by the hour, pay is not a surrogate for actual working time, as any hourly-paid tutor will tell you. What are now usually one-hour hearings in the ET to establish the amount due on termination under regulation 14 WTR risked being transformed into detailed forensic examinations of the actual hours worked over the leave year, with serious consequences for access to justice. Instead, after Brazel every worker receives the same, fixed entitlement to 5.6 weeks’ leave. The only pro rata adjustments for those working part-time or irregular hours are those which result from the complicated rules on a “week’s pay”, applying to pay not the entitlement.
Mrs Brazel was employed by the Bedford School as a Visiting Music Teacher, giving individual tuition to pupils. She taught only during term times, which ranged between 32 and 35 weeks a year. She had no guaranteed hours – in that sense, she was on a “zero hours” contract – and was paid an hourly rate for the teaching she delivered. It was not in dispute that she took her leave during the school holidays. But what was the extent of her entitlement to leave? Purporting to follow the previous iteration of the ACAS guidance on holidays, the School decided she was entitled to a paid holiday entitlement based on 12.07% of the hours she worked across the year. In each holiday, it therefore paid her 12.07% of the hours she worked in the preceding term, paid at her normal hourly rate. The result was that she received less leave, and therefore less pay, than if she were paid for 5.6 calendar weeks’ leave at the statutory rate of a “week’s pay”.
The 12.07% figure was always a mathematical abstraction, based on the proportion statutory leave bears to the working year in the case of a hypothetical full-time worker. It has never figured in the Directive or the WTR, both of which express leave in weeks. Nevertheless, Mrs Brazel’s challenge to the employer’s method failed before the ET. Drawing support from Greenfield and Hein, in both of which the CJEU indicated that the period of leave in the Directive was to be calculated in accordance with the days or hours worked by a worker, the School argued that WTR were intended to confer entitlements to leave which were adjusted pro rata to the amount of work done by a worker. A full-time worker got 5.6 weeks’ leave; a part-timer, working half the weeks, got half the leave. The ET accepted this argument, holding that WTR were to be interpreted as including a pro rating exercise, which could be achieved by using the 12.07% formula.
Both the EAT and Court of Appeal, where UNISON intervened in support of Mrs Brazel, considered that the EJ below was wrong. In dismissing the employer’s appeal today, Lady Rose and Lady Arden (with whom Lords Hodge, Briggs and Burrows agreed), the Supreme Court agreed. Probably the fundamental point from the judgment is that the School’s method of calculating holiday pay, based on the amount of hours in fact worked, amounted to the replacement of the existing legislative scheme with a new and different scheme. It therefore went beyond the scope of legitimate statutory interpretation.
The first plank in the single judgment delivered by Lady Rose and Lady Arden is that the Directive only lays down minimum standards. While in cases such as Greenfield and Hein the CJEU had said that the period of annual leave governed by the Directive must be determined by reference to periods of actual work – what Counsel for the School labelled the “conformity principle” – the Directive (and the Part-time Workers Directive) expressly permits Member States to grant more generous entitlements to workers. It followed from this that the issue was one of domestic statutory interpretation: the radical Marleasing principle, which requires the a compliant interpretation of domestic law, so far as possible, where it falls short of the result required by EU law, was not engaged.
So the focus of the judgment is on the interpretation of the WTR. The Court decides the domestic rules were inconsistent with the employer’s “conformity principle”. WTR presupposed that a “week” for the purpose of leave meant a calendar week – that is, seven days. If WTR intended that the 5.6 weeks’ leave should be calculated on the basis of the amount of time worked, they would adopt a formula to achieve that result. Moreover, where WTR reduce the entitlement to 5.6 weeks’ leave, such as in the first leave year, they do so as a function of the time elapsed in the leave year, not the hours worked by the worker: see regulations 13(5), 13A(5), the formula in regulation 14 and regulation 15A. The various hypothetical examples proposed by the School in its submissions, designed to show why a 12.07% was rational, in fact demonstrated that its interpretation of WTR in fact amounted to the substitution a new and complicated scheme, fundamentally inconsistent with the methodology of WTR, its adoption of the rules on a week’s pay and the policy choices made by the legislator.
The principal practical effect of the Supreme Court judgment is that the leave entitlement in WTR, of 5.6 weeks, is fixed and universal. It is not linked to the amount of work in fact done by the worker. It cannot be adjusted downwards, save where WTR expressly provides for reductions to the entitlement (e.g. as to the amount of leave which can be taken in the first leave year under regulation 15A, introduced in the wake of the BECTU judgment). The only pro rata discounts are those implicit in the rules on a “week’s pay” in ERA, under which pay is linked to the amount in fact paid to a worker. But even here the overarching requirement is that pay in respect of annual leave corresponds with normal remuneration received while working: see British Airways v Williams and British Gas v Lock. The domestic rules on a “week’s pay” ordinarily achieve that result. Where those rules are more generous to the worker than the Directive requires, it is the domestic rules which must be applied, as now adjusted by the amendments made in 2018 to put in place a 52-week averaging period. Where the rules are less generous, they must be adjusted under the Marleasing obligation, preserved after Brexit by s.5 of the European Union (Withdrawal) Act 2018. Full leave at full pay is the resounding message after Brazel.
The Supreme Court has not eliminated every uncertainty, no doubt disappointing “those hooked on the hopeless quest for completeness” (per Mummery LJ in NHS Leeds v Larner). There will still be some practical issues to be ironed out in calculating the leave entitlements of those working part-time or irregular hours, such as when they ask to take a day’s leave. The price of legal certainty and a detailed code may be some anomalies around the margins, but these were recognised right from the outset of the introduction of the rules on a “week’s pay” in the Contracts of Employment Act 1963. But these are small matters compared to the bewildering factual and legal issues employers, workers and ETs would have had to confront if the Supreme Court, under the guise of interpretation, had decided WTR involved a system based on the progressive accrual of leave as function of the hours of “work” (itself a concept of great uncertainty). Just imagine doing a holiday pay calculation for the claimant in the Smith v Pimlico Plumbers case recently decided by the Court of Appeal.
Lastly, it is to be hoped that the judgment puts paid once and for all to lawyers referring to the “accrual” of leave under WTR. Leave doesn’t accrue save, perhaps, under regulation 15A in the first leave (and even then the regulation only says it is “deemed” to accrue). The entitlement to leave is just there, to be taken by everyone, from day one of each leave year. And every worker, after Brazel, is entitled to 5.6 weeks’ leave each leave year. But we all knew that anyway just from reading WTR, didn’t we?
Michael Ford QC and Mathew Purchase QC, instructed by Shantha David of UNISON’s Legal Department acted for the intervenor, UNISON.
William Meade (Senior Clerk)