Calculating holiday for term time or casual workers

Under the Working Time Regulations 1998 (WTR), all workers are entitled to 5.6 weeks’ holiday each year. The WTR make no provision for pro-rating this entitlement for employees who only work for part of the year, for example during term time only.

Under the relevant legal provisions, holiday pay for someone with no normal hours of work is based on their average pay over a 52-week reference period immediately before the holiday (or a 12 week reference period prior to April 2020). You ignore weeks when there was no pay and bring earlier weeks into account.

Calculating holiday entitlement and pay for casual workers with no set hours is notoriously difficult. Rolling up holiday pay into an individual’s hourly rate is currently unlawful. Before the litigation in this case brought the topic into the spotlight, Acas recommended saying that casual workers accrue statutory holiday entitlement at the rate of 12.07% hours worked. This method aims to ensure that casual workers build up holiday entitlement in the same proportion to hours worked as a standard worker. It is not, however, referenced in the WTR.

The history of the case: employer used 12.07% method

Lesley Brazel was engaged by The Harpur Trust, which runs Bedford Girls’ School, on a permanent zero-hours contract as a “visiting” music teacher. Mrs Brazel worked term times only. Her working hours varied, to cater for the changing number of pupils requiring tuition each week, and she was paid an agreed hourly rate each month for the hours worked in the previous month.

Mrs Brazel took her annual leave during the three school holidays when she was not teaching. The Trust made payments for her leave at the end of each school term, calculated as one-third of 12.07% of her earnings in the previous term.

Mrs Brazel claimed that she was being underpaid for her holiday. She asserted that she was entitled to payment for 5.6 weeks’ holiday in accordance with the legal rules for workers with no normal hours of work. She said that her holiday pay should be based on her average pay during working weeks and that the 12.07% method had no basis in law. The Trust argued that there had to be some principle of pro rating so that holiday entitlement and pay reflected hours actually worked.

The Employment Tribunal dismissed Mrs Brazel’s claim, but the Employment Appeal Tribunal and Court of Appeal both agreed with Mrs Brazel. The Trust appealed to the Supreme Court (SC), and Acas withdrew the 12.07% guidance.

Supreme Court ruling: 12.07% method is incorrect

The SC agreed with previous appeal decisions and has ruled that:

• The EU Working Time Directive, on which the WTR are based, generally calculates holiday by reference to work actually carried out by the worker. However, EU law does not prevent the UK from making more generous provision for holiday for part-year workers. It is also not discrimination against full-time workers, as EU law only covers less favourable treatment of part-time workers. This means that the amount of leave to which a part–year worker under a permanent contract is entitled is not required to be proportional to that of a full–time worker.

• The answer therefore depends on interpretating the wording of domestic law. Looking at the WTR, the amount of leave to which a part–year worker under a permanent contract is entitled must not be pro–rated in this way. The correct calculation is the “calendar year” method set out in the WTR – i.e. average pay over a reference period, which does not include weeks in which there was no pay.

• The alternative methods of calculation put forward by the Trust were rejected as not complying with the WTR. These were the 12.07% method, and a more complex “worked year” method. The SC said that Parliament had made a policy choice that holiday pay for irregular hours workers should be calculated in accordance with average pay over a reference period.

• The alternatives put forward by the Trust were also rejected because they would involve complicated calculations for those working irregular hours, requiring all employers and workers to keep detailed records of every hour worked even if they were not paid at an hourly rate (although in our view this is not a compelling point as employers already tend to keep detailed records for workers with irregular hours).

• Although this meant workers in Mrs Brazel’s position received holiday pay representing a higher proportion of their annual pay than those working regular hours, a slight favouring of workers with a highly atypical work pattern is not so absurd as to justify the wholesale revision of the statutory scheme.

Where does this leave the law on holiday entitlement and pay?

Workers under permanent contracts are entitled to 5.6 weeks’ holiday per year, regardless of the amount of work they do. Pay must be calculated in accordance with the legal provisions for holiday pay and not the 12.07% method.

In practice, this does not impact “classic” part-time workers with set hours, for example those on a 3-day standard week. You can still calculate their holiday entitlement as a reduced number of days compared to a 5-day week worker because in practice they will end up with 5.6 weeks of holiday at their normal rate of pay.

The significant impact of the ruling is on people who are under a permanent contract for the whole year but actually work for less than a full year, such as term time only workers. As argued by the Trust, this could lead to some “absurd” results. For example, an exam invigilator retained on a permanent contract who works only one week of the year, earning £1,000 for that work. The invigilator would be entitled to 5.6 weeks' annual leave. Holiday pay for this leave must be calculated based on their average pay over the last 52 weeks – but ignoring any weeks in which they received no pay. So, they would receive £5,600 holiday pay for a job that paid only £1,000. This seems to be rather more than the “slight” favouring of atypical workers referred to by the SC, but the SC dealt with this by saying that general rules sometimes provide anomalies in untypical cases, and that it would be unusual for this kind of arrangement to be a permanent contract.

The better news is that this does not necessarily affect all term time workers. Many such workers will have normal working hours and receive a regular salary throughout the year, including during the holidays. The legal provisions for holiday pay in such cases simply require you to maintain normal salary during periods of holiday. It should generally be possible to argue that such workers are already paid in compliance with the rules, as they continue to be paid their normal salary for 5.6 weeks of holiday.

The ruling does, however, potentially affect all casual workers with no normal working hours operating under permanent contracts. It completely rejects the 12.07% approach. Even if workers in these situations work fewer than 46.4 working weeks in the year, you must nonetheless provide 5.6 weeks of holiday and pay holiday pay for those 5.6 weeks on the basis of average weekly pay over the last 52 paid weeks.

It is very hard to translate this holiday entitlement for casual workers into days or hours. The current government guidance says “employers may wish to calculate average days or hours worked each week based on a representative reference period, although the [WTR] do not expressly provide for this". Unfortunately, the SC did not engage with arguments about how to do this in its judgment. This makes it very difficult (if not impossible) to work out what should be paid to a casual worker who wishes to take one day of holiday.

What next for employers: solutions and prospects of legal reform

If you have people under contract for the whole year but they work short of a whole year, or if you use the 12.07% method for casual workers on permanent contracts, then it would be sensible to get your contracts looked at and start considering your options.

Some employers may now find themselves facing significant claims for underpaid holiday from irregular workers on permanent contracts where the 12.07% method has been used. However, it is worth noting that this was also the position under the earlier appeal court decisions, claims for holiday pay based on deductions from wages are limited to 2 years of back pay, and the argument from the Bear Scotland case that a break of more than 3 months between payments will break any chain of deductions is still available (this was doubted by the SC in the recent Pimlico Plumbers decision but not overruled).

Employers may also want to review contracting options more widely. Generally speaking, the risks from this judgment arise when you have people under contracts for long periods of time when they are not working. Returning to the exam invigilator example, they might be better hired as freelancers or on a temporary contract which ends after each annual week of work. Employers could consider migrating away from ongoing umbrella contracts and operating a system whereby each engagement terminates at end of each assignment, and then simply make a payment of accrued but untaken holiday on termination of each engagement as set out in the contract – although it would be unfortunate if this ruling has the effect of deterring permanent contracts at a time when policymakers are looking to promoting greater stability through the Good Work agenda.

Perhaps the better solution would be if the government reformed the law to cater for the realities of modern flexible working relationships.

Now may be a good time for the government to consider allowing rolled up but separately itemised holiday pay, as long as it genuinely represents a supplement to normal pay. This is prohibited by pre-Brexit ECJ caselaw which remains binding on the UK but the government could now legislate to permit the practice. Rolled up holiday pay was notably recommended as a possible option back in 2017 by the Taylor Review. Legislating to permit this practice would also make it easier for the proposed Single Enforcement Body to enforce holiday pay for vulnerable workers, which is intended to be a core remit of that body once it is set up.

Harpur Trust v Brazel – judgment available here.