The Supreme Court of the United Kingdom has published its long-anticipated decision in Harpur Trust v Brazel confirming that paid holiday for part-year employees/workers on permanent contracts must not be pro-rated.

The Supreme Court decision of Harpur Trust v Brazel provides much-needed clarity as to how paid holiday should be calculated for those employees and non-employed workers¹ who work irregular hours but who are on permanent contracts, such as term-time only and zero hours contracts (referred to as “part-year workers”).

The Supreme Court dismissed the employer’s appeal, in a decision that is favourable to part-year workers and means that employers should revisit, if they haven’t done so already, how they calculate paid holiday for such part-year workers.

This update does not consider paid holiday for part-time employees and non-employed workers who work throughout the year but for only part of the week.

Background

Employees and non-employed workers are entitled to paid holiday (please see our note earlier this year). 

Under the relevant regulations, employees and non-employed workers in the UK are legally entitled to take at least 5.6 weeks (equivalent to 28 days for full-time employees working a typical 9-5 Monday to Friday schedule) paid holiday per holiday year. This is easy enough to implement when it comes to full-time employees and non-employed workers but has caused confusion when it comes to more flexible working arrangements.

For part-year workers, many employers have historically pro-rated paid holiday entitlements using the “12.07% method”. This was set out in (now withdrawn) ACAS guidance, which said that the holiday entitlement of 5.6 weeks is equivalent to 12.07% of hours worked over a year. This was seen as a useful shorthand for the full calculations needed under the legislation.

However, the Court of Appeal in Harpur Trust v Brazel held that the “12.07% approach” is legally incorrect and paid holiday entitlement should not be pro-rated. The Court of Appeal said that an employee or non-employed worker is entitled to 5.6 weeks paid holiday in each holiday year and employers should calculate how much to pay for this holiday by first determining an average week’s pay in accordance with the Working Time Regulations and Employment Rights Act 1996 (at the relevant time, this was based on an average over the 12 weeks ending with the start of the leave period, ignoring weeks in which no pay was received) and then multiplying that average by 5.6 (the “Calendar Week Method”).

Decision

The Supreme Court has confirmed that the Court of Appeal was correct – paid holiday should not be pro-rated for part-year workers. Instead, it should be calculated using the Calendar Week Method, even if in practice this approach may lead to anomalous results for some part-year workers, who may be entitled to receive proportionally more holiday pay than full-year employees and non-employed workers.

This decision will have a significant impact on employers who engage part-year workers and who have been pro-rating holiday pay using the well-known 12.07% method. Such employers will now need to reconsider how they have been calculating paid holiday entitlements and change their policies to reflect this decision. This may lead to an increased ongoing cost for holiday pay for employers with part-year workers, some of whom may indeed be entitled to receive paid holiday that is a greater proportion of their pay than full-year employees and non-employed workers. 

Employers may also find themselves facing claims for potentially significant amounts of underpaid holiday pay.

Paid holiday for those with non-standard working hours is now an even more complex and tricky area. It is an area ripe for reform and simplification: much of the legislation predates the expansion of atypical working that is now commonplace – leading to anomalous results that were arguably not intended when the legislation was drafted. In this instance, the Supreme Court was no doubt bound by the inflexible statutory language it needed to interpret in this case.