The Supreme Court ('SC') in Harpur Trust v Brazel has discredited the commonly used method of calculating holiday pay for part year workers at the rate of 12.07% of hours worked, where those workers are engaged under a permanent (whole-year) contract. The SC held instead that the Working Time Regulations 1998 ('WTR') clearly set out an entitlement 5.6 weeks’ leave and that since the WTR do not permit this entitlement to be qualified in any way, that entitlement applies, regardless of the number of weeks actually worked during the year.
Under the Working Time Regulations 1998 all workers (including employees):
- are entitled to take a minimum 5.6 weeks’ paid leave per year;
- accrue annual paid leave in the years they start or end their employment at a rate of 1/12th of that entitlement for each month they are engaged by their employer in that year; and
- are entitled to a week’s pay for a week’s holiday, to be paid at the time that they take it (except at the end of their employment, when any accrued but untaken leave can be paid in lieu).
The way in which a week’s pay is calculated is set out in sections 221 to 224 of the Employment Rights Act 1996 (ERA) and is different for workers with normal working hours and those without:
Normal working hours: a week’s paid leave is their normal pay for their normal working week. For a full-time worker this will be five days’ pay and for a part-time worker this will be the number of days they work per week (e.g. 3 days’ pay if they work 3 days a week).
No normal working hours: a week’s paid leave is calculated by working out the worker’s average weekly pay over a 52-week reference period immediately before the holiday. This method is known as the ‘calendar week method’. The pay reference period only includes weeks for which the worker was actually paid. Weeks for which the worker was not paid because they did not work that week are ignored and earlier weeks are brought forward into the reference period instead (up to a maximum of 104 weeks prior to the holiday in question).
Therefore, the method used for workers with no normal working hours relies on doing a (potentially complex) calculation each time the worker takes any holiday. To make the process simpler, it had become a relatively established practice for employers to calculate holiday pay for workers with irregular hours by using what is known as the ‘percentage method’ which results in workers accruing holiday at the rate of 12.07% of the hours they actually work. This method came about because the standard working year is 46.4 weeks (i.e. 52 weeks minus 5.6 weeks’ leave) and 5.6 weeks is 12.07% of 46.4 weeks.
The WTR are silent on how to calculate holiday for ‘part-year’ workers, who only work during certain weeks of the year, but have a permanent contract for the whole year. Such workers include, for example, those who work term-time only or on a seasonal basis (such as working at holiday resorts). Therefore, the ‘percentage method’ is also widely used for these types of workers, to calculate holiday accrual in proportion to work done.
However, although this seemingly logical approach complies with EU law (which allows for accrual of paid leave based on worked hours) and also with previous ACAS guidance; it conflicts with the WTR. This is because accrual under the WTR is based on the length of time a worker is engaged, not on how much work is done.
The SC has now clarified the correct approach.
Ms Brazel was engaged by The Harpur Trust (the Trust) to work at Bedford Girls' School, as a visiting music teacher. Ms Brazel was engaged on a permanent employment contract that continued for the whole year. However, she worked term-time only. She was paid monthly at an agreed hourly rate for the hours she had worked in the previous month.
Ms Brazel was required to take her paid leave during the three school holidays, which the Trust paid for in three equal payments. These were calculated by using the percentage method (i.e. by multiplying Ms Brazel’s agreed hourly rate by 12.07% of her total hours worked during the previous term).
Ms Brazel considered that she was effectively receiving less than 5.6 weeks’ paid leave and brought a deduction from wages claim in the ET. She said that the Trust should not use the percentage method as it conflicted with the WTR. Ms Brazel’s position was that she should be paid for a full 5.6 weeks’ leave using the calendar week method under s224 ERA, even though she did not work a full calendar year.
The Trust argued that (in accordance with EU law) paid holiday entitlement should be pro-rated and accrue in line with time spent working.
The Employment Tribunal ('ET') dismissed Ms Brazel's claim. However, the Employment Appeal Tribunal ('EAT') and Court of Appeal (COA) both agreed with her and the Trust appealed to the SC.
Supreme Court decision
The SC dismissed the Trust’s appeal.
It held that it was wrong and not in accordance with the WTR to use the percentage method as it meant that if Ms Brazel worked fewer hours in a particular term she would accrue less holiday. This would (incorrectly) reduce her entitlement to below 5.6 weeks paid holiday per year. This entitlement should not be based (or pro-rated) on hours worked and pay should be calculated using the calendar week method under s. 224 ERA (ignoring non-working weeks).
The SC acknowledged that this could lead to a windfall for some part-year workers. In fact, the greater the number of non-working weeks, the more their paid holiday entitlement would be (in proportion to time spent working).
For example, an exam invigilator is engaged by a school on a permanent (year-round) employment contract to work three weeks per year (at 40 hours per week). She earns £800 per week. She is entitled to 5.6 weeks’ holiday per year. Pay for a week’s leave will be £800 (as all non-working weeks are ignored). She will be entitled to £4,480 holiday pay (5.6 x £800) which is more than she receives in salary for the three weeks she works during the year.
However, in spite of this and the fact that part-year workers will likely receive proportionately more holiday pay than full-time workers, the SC said the calendar week method was the correct one to ensure WTR compliance.
There is no higher appeal court and the government has not (currently) indicated that it intends to change the law. Therefore the final legal position is that workers engaged under a permanent contracts are entitled to 5.6 weeks' holiday per year, regardless of the amount of work they do in that year. Their holiday pay must be calculated in accordance with the WTR and ERA and not using the percentage method. Therefore, this is an important decision for employers who use the percentage method to calculate holiday pay, who would be well advised to change their approach if they want to avoid the potential to accrue unpaid holiday pay.
Generally speaking, as this decision only applies to part year workers who are permanently employed, this case arguably has a relatively narrow application in practice. That said, it is likely to affect other workers too, for example casual (or zero-hours) worker who work under an ‘umbrella’ contract (whereby individual assignments are connected by a continuing contract). We recommend that employers carry out an audit, to assess their potential liability and options for mitigating against these.
Ultimately, the SC has discredited the percentage method. The only reliable rules on how to calculate paid holiday entitlement are set out in the WTR. These clearly only allow for holiday accrual to be pro rated in the first and last years of employment (and then only on the basis on time spent engaged by the employer, rather than hours worked) and require holiday pay to be calculated using the calendar week method under s224 ERA.
To minimise its holiday pay liability when using the calendar week method employers can:
- ensure that workers have only a few (or no) non-working weeks by spreading the workload across the year to ensure that some work is performed every week;
- use a fixed-term contract to cover only the period of an individual assignment, rather than engage part year workers on permanent contracts; and
- engage contractors who are genuinely self-employed and have no holiday entitlement.
Even if employers continue to use the percentage method, limiting the number of non-working weeks will be key to reducing the risk of underpayments. This will narrow the difference between the amount of holiday pay calculated using the percentage, rather than calendar week, method (by reducing the average weekly pay figure). To check this difference, employers could perform a regular reconciliation calculation (say quarterly and/or at the end of a contract). If there is a significant difference, employers could make any required additional payment to remove it.
Remember, this decision should not affect:
- part-time (rather than part year) workers who have normal working hours throughout the year; or
- those who work term-time only, but are paid in 12 equal monthly instalments.
It is likely that these workers will already be paid in compliance with the WTR, as they continue to be paid their normal weekly salary during their 5.6 weeks of holiday.
That said, employers may well face questions from these workers (and others with normal working hours) if they find out that they are being paid proportionally less than their part-year worker colleagues. The fact that this disproportionate outcome is the law is unlikely to stop such workers from feeling aggrieved.
Employers may also face pressure from workers and unions to recalculate holiday pay in the past and reimburse any underpayments or face claims for unlawful deductions from wages. These claims can be brought within 3 months of the last underpayment in a series of deductions or on termination of employment and can include up to two years' back pay from the date of the claim.
If you would like our help with any of the issues in this bulletin, please do contact us.