Harpur Trust v Brazel

This appeal concerned how statutory paid annual leave is to be calculated for part-year workers who remain under a continuing contract but work only in some weeks of the year. The key statutory materials were the Working Time Directive (Council Directive 2003/88/EC), the Working Time Regulations 1998 (in particular regulations 13, 13A and 16) and the definition of a "week's pay" in section 224 of the Employment Rights Act 1996 as applied by regulation 16. The court held that the correct domestic method (the "Calendar Week Method") treats the weeks in which the worker is employed but receives no remuneration as ignored only for the purpose of calculating a week’s pay under section 224, and does not permit reducing the number of weeks of leave in regulations 13/13A by excluding non-working weeks.

Accordingly, the Supreme Court rejected the employer’s proposed alternative methods (the "Percentage Method" and the "Worked Year Method") and concluded that the Working Time Regulations do not implement entitlement by reference to time actually worked (the conformity principle) so as to prorate the statutory 5.6 weeks for part‑year workers. The Court further held that that construction is compatible with retained EU law: the WTD permits member states to provide more favourable rules than the Directive and the WTR, properly construed, give a more generous domestic entitlement.

Background and parties. The appellant Harpur Trust employed the respondent Mrs Brazel as a visiting music teacher who worked only during school terms and was paid only for hours taught. Mrs Brazel’s contract treated her as entitled to 5.6 weeks’ paid annual leave to be taken during the school holidays. She claimed unlawful deductions for underpayment of holiday pay in respect of the period 2011–2016. The Trust changed its method of calculation in September 2011 and paid less under what it called the Percentage Method.

Nature of the claim and procedural history. The respondent brought a claim under Part II of the Employment Rights Act 1996 for unlawful deductions from wages (holiday pay underpayment). The Employment Tribunal dismissed her claim; the Employment Appeal Tribunal allowed her appeal; the Court of Appeal [2019] EWCA Civ 1402 dismissed the Harpur Trust’s appeal; permission to appeal to the Supreme Court was granted and the Supreme Court heard the appeal.

Issues framed.

• Whether, and to what extent, the statutory entitlement to 5.6 weeks’ paid annual leave in the Working Time Regulations should be calculated by reference to the weeks actually worked by part‑year workers (the "conformity principle" under CJEU case law), so that non‑working weeks reduce leave entitlement.
• Whether domestic provisions (regulations 13, 13A and 16 WTR and section 224 ERA 1996) should be construed to permit the employer’s Percentage Method or the Worked Year Method rather than the Calendar Week Method.
• Whether the domestic construction that gives a part‑year worker the full 5.6 weeks (calculated by using the section 224 reference for a week’s pay) is compatible with retained EU law.

Reasoning and decision. The court accepted that the CJEU has generally developed a "conformity" or accrual approach — that entitlement to paid leave is often determined by reference to periods of actual work — but it held that the Working Time Regulations must be construed by reference to their text and overall scheme. Regulation 16 expressly incorporates sections 221–224 of the Employment Rights Act 1996 for calculating a "week’s pay" and section 224(3) already provides for ignoring weeks with no remuneration when computing average weekly pay. By contrast, regulations 13 and 13A set the duration of leave in weeks and contain express time‑apportionment rules (for starting part way through a leave year and for the first year of employment). On ordinary principles of statutory construction, the absence of any provision excluding non‑working weeks from the computation of the number of weeks’ leave shows Parliament did not intend entitlement to be prorated by reference to weeks not worked. The court therefore rejected the employer’s alternative methods as inconsistent with the WTR and unduly complex. Finally, the court held that this domestic construction is compatible with retained EU law because the WTD sets minimum entitlements and permits member states to adopt more favourable measures.

Wider context. The Court noted the practical anomalies that can arise in atypical work patterns but considered them insufficiently absurd to displace the statutory scheme and stressed that domestic law may be more generous than the Directive.

Appeal dismissed. The Supreme Court held that the Working Time Regulations, properly construed, do not require prorating the 5.6 weeks’ statutory annual leave of a part‑year worker by excluding weeks in which no work is performed; regulation 16 incorporates section 224 ERA 1996 for calculating a week’s pay but there is no statutory basis for the employer’s Percentage or Worked Year methods, and the domestic construction is compatible with retained EU law.

Employment Tribunal (Bury St Edmunds) dismissed the claim (15 January 2017); Employment Appeal Tribunal (HHJ Barklem) allowed Mrs Brazel’s appeal; Court of Appeal dismissed the Harpur Trust’s appeal [2019] EWCA Civ 1402; permission to appeal to the Supreme Court granted (19 June 2020); Supreme Court judgment [2022] UKSC 21 dismissing the appeal.