For a while, the four-day week was a pilot programme with a press release. It is now a clause in real employment contracts, and the details of that clause matter more than the headline.
Most trials that converted to permanent arrangements kept pay the same and cut the working week to around 32 hours. That is the version everyone talks about. The contracts behind it are more varied, and some are considerably less generous than the slogan suggests.
Three models wearing the same name
The first model is a genuine reduction in hours at full pay. Staff work four days instead of five, with the same salary and the same output expectations. This is the version that produces the strongest wellbeing results, and the one that is hardest to sustain in roles with fixed coverage.
The second model compresses the same 40 hours into four longer days. It is often sold as a four-day week, but it is really a compressed schedule. Employees get an extra day off and pay for it with longer shifts, which can be exhausting for parents and people with long commutes.
The third model is a hybrid: shorter days, fewer of them, or a nine-day fortnight. It is more flexible and easier to fit around customer demand, but the gains in rest time are smaller.
What the contracts actually specify
The contracts that hold up tend to be specific. They state the number of hours, the days of the week, how overtime is calculated and what happens during peak periods. They also define coverage: who answers the phone on the fifth day, and whether that is a rotating duty or a separate role.
Vague arrangements tend to collapse. If the contract says "four days subject to business needs", the fifth day quietly returns during the first busy month. Employees who accepted a lower salary for fewer hours can find themselves working the old schedule for less money.
A four-day week is not a perk. It is a scheduling system, and like any scheduling system it has to say who covers Friday.
The coverage problem
Customer-facing work is the hard case. A hospital, a call centre or a bus network cannot close on Friday because the office staff have the day off. The solution is usually staggered rotas, which spread the benefit unevenly and can create resentment between teams.
Some organisations hire additional staff to cover the fifth day. That is expensive and only works when the productivity gain or the retention benefit is large enough to pay for it. Others accept a slower response on the off day, which is a service decision as much as an employment one.
Who it helps, and who it leaves out
The strongest evidence comes from knowledge work: software teams, consultancies, marketing departments and back-office functions. These roles have some slack, and a shorter week forces teams to cut meetings and focus on the work that matters.
The evidence is thinner for hospitality, retail, healthcare and manufacturing, where output is tied to hours on site. A four-day week can still work there, but it usually requires more staff, not just better meetings.
There is also a class dimension that the pilot reports rarely discuss. Workers who can compress their week into four long days are often those without caring responsibilities. Those who cannot may end up excluded from the best roles, or forced to choose between the schedule and the family.
What to ask before signing
Read the hours clause, the overtime clause and the coverage clause. Ask what happens during the busiest month, and whether the fifth day is guaranteed or conditional. Ask whether the arrangement is contractual or a policy that can be withdrawn.
If the answer is "we will see how it goes", treat it as a trial, not a term of employment. Trials can end.
Key facts
- Most permanent schemes keep pay and cut hours to around 32.
- Compressed schedules are not the same as a shorter week.
- Coverage is the main obstacle in shift-based industries.
- Vague contracts tend to revert during busy periods.
- Knowledge work has the strongest evidence; hospitality and healthcare the weakest.
The bottom line
The four-day week is becoming a real employment model, but only where the contract is explicit about hours, pay and cover. The pilots proved it can work in some workplaces. They did not prove it works everywhere, and the sectors still excluded deserve honest answers rather than another case study from a software company.
For another look at who absorbs the cost of a public service decision, read who pays when a city removes free parking.
