The four-day week: what the trials found and who is actually doing it

The four-day week has been “about to go mainstream” for most of a decade. It has not gone mainstream. It has, though, moved from a stunt to a settled option that a meaningful number of employers offer, that several governments have piloted, and that a large body of trial data now describes in some detail. If you are wondering whether it could work where you work, or whether to take a job that offers it, that data is worth understanding, because it says something more specific than “people liked having Fridays off”.

This piece explains the different models (they are not interchangeable), what the main pilots actually measured and found, which kinds of organizations have made it stick, and how to put a proposal together that a skeptical manager might accept.

The models, because the label hides a lot

“Four-day week” covers at least four arrangements with very different consequences.

100-80-100. One hundred percent of pay, 80 percent of the hours, in exchange for a commitment to 100 percent of the output. A 40-hour week becomes 32. This is the model the major pilots tested and the one campaigners mean when they use the term. It is a genuine reduction in working time.

Compressed hours. The same 37 to 40 hours squeezed into four longer days, typically four tens. Pay and hours are unchanged. This is common in healthcare, manufacturing, policing and some public services, and Belgium wrote a right to request it into law in 2022. It suits some people well and exhausts others, particularly those with caring responsibilities that do not compress.

Flexible fifth day. Employees work four days but are available or on light duty on the fifth, or the fifth day is a “no meetings, work if you must” day. Often a halfway house adopted by firms nervous about full 100-80-100.

Staggered or rota-based. The organization stays open five or more days, but each employee works four, with days off rotated across the team. This is how customer-facing businesses and shift-based workplaces do it, and it is the model that scales beyond small professional firms.

There are further variants (nine-day fortnights, annualized hours with summer Fridays off, reduced hours with a small pay cut), but when you see a claim about the four-day week, first ask which of these is meant.

What the main trials did and found

Iceland, 2015 to 2019. The Reykjavik city council and the national government ran trials covering around 2,500 public-sector workers, in offices, preschools, hospitals and social services. Hours were cut, usually from 40 to 35 or 36, at full pay. Researchers reported that productivity and service provision held steady or improved in most workplaces, and wellbeing and stress measures improved. The trials led to unions negotiating permanent reduced-hours arrangements covering a large majority of the working population. This is the most consequential result anywhere, though note it was a reduction to a 35-hour week rather than a strict four-day, 32-hour week.

New Zealand, 2018. The trust and estate company Perpetual Guardian ran an eight-week 100-80-100 trial that attracted worldwide attention. The company reported unchanged output and improved engagement and made the arrangement permanent. It was a single employer with a few hundred staff, and its founder became a prominent campaigner, so treat it as an early demonstration rather than proof.

Microsoft Japan, 2019. A one-month trial in August that closed the office on Fridays, reported as a productivity gain of around 40 percent measured by sales per employee. It was widely quoted and rarely qualified: one summer month, in a country where August is already slow, with meetings cut and shortened at the same time. The company did not adopt it permanently.

United Kingdom, 2022. The largest coordinated pilot to date, run by 4 Day Week Global and the think tank Autonomy with researchers from Cambridge and Boston College. About 60 companies and roughly 2,900 employees moved to a 100-80-100 model for six months. The participating firms were mostly small: marketing agencies, software companies, consultancies, a fish-and-chip shop, a brewery, a few charities and manufacturers. The findings, published in early 2023, were that company revenue was broadly flat to slightly up over the period, sick days and resignations fell, and employees reported lower stress and burnout and better sleep. The large majority of companies continued after the trial and many made it permanent. Follow-up work a year later found most were still on four days.

Other pilots. 4 Day Week Global ran parallel pilots in the US, Ireland, Australia and South Africa with similar designs and broadly similar results. Portugal ran a government-backed pilot in 2023. Germany ran one in 2024 involving a few dozen firms. Scotland’s government piloted the model in a small number of public bodies. South Cambridgeshire District Council in England ran a well-documented trial that, despite political opposition, reported improvements in recruitment and most service measures and was made permanent.

What the trials measured, and the caveats

The pilots mostly tracked three things: business metrics chosen by each company (revenue, output, customer satisfaction, error rates), staff wellbeing surveys before and after, and HR data on sickness absence and turnover. On those measures, the results are consistent and positive.

The caveats are real, and anyone proposing a four-day week should know them before a finance director raises them.

Companies chose to join. Firms that expected it to fail, or whose work obviously could not compress, did not sign up. The participants skew small, knowledge-based and already flexible. Results from a 30-person design agency say little about a 3,000-person hospital.

Six months is short. Some of the productivity gain came from a one-off clear-out of wasted meetings and processes that any change program might have produced. Whether output holds at year three is less documented, though the follow-ups so far are encouraging.

“Productivity” in knowledge work is hard to measure, and the studies leaned on self-report and company-chosen metrics. The Iceland results, which drew on public-sector service data, are more solid on that front.

Some firms reverted, usually because customer demands or growth made coverage hard, or because the intensity of compressing five days of work into four wore people down. That last point deserves emphasis: a badly run 100-80-100 week can become a compressed week in disguise, and the burnout it was meant to relieve returns. Our piece on recognizing and recovering from burnout describes the warning signs.

Who is actually doing it

Outside trials, the pattern is clear. The employers that have adopted a genuine four-day week permanently are concentrated in a few types: small and mid-sized professional services and creative firms (agencies, consultancies, software studios), technology companies with strong recruiting competition, a scattering of manufacturers that reorganized shifts, some charities and nonprofits, and a small number of public bodies, mostly local government. Several UK firms, including a bank and a number of agencies, have run on four days for years and use it explicitly as a hiring tool.

Governments have largely stopped short of mandating it. Belgium gave workers a legal right to request a compressed four-day week. Some emirates in the UAE moved public-sector staff to a four-and-a-half or four-day week. Tokyo’s metropolitan government introduced a four-day option for its staff. Bills to shorten the standard working week have been introduced in the US Congress and in several state legislatures without passing. The UK government has not legislated, though it has said employers are free to adopt it, and the trend in the UK public sector is toward local experiments rather than national policy.

Where it does not appear, and probably will not in the 100-80-100 form: hospitals and care homes, retail and hospitality, transport, construction sites, and any workplace where output is closely tied to hours on the floor. Those sectors can and do run compressed and rota-based four-day patterns, which are a different bargain.

Which industries suit it

The work that compresses well shares a few features. Output is measured by results rather than presence. Much of the week is meetings, email and coordination that can be cut. Customer contact can be scheduled or covered by a rota. The team is experienced enough to manage its own time.

The work that resists it: continuous operations, roles where response time is the product, work with hard daily volumes (a claims processor handling a fixed queue), and teams so lean that one absent day means nothing gets covered. In those settings the honest options are compressed hours, a rota, or a nine-day fortnight, all of which are worth proposing but which are not a reduction in working time.

How to propose one at your workplace

Managers say no to four-day weeks for predictable reasons: coverage, fairness between teams, client perception, and the fear that output drops and nobody can prove it either way. A proposal that answers those four in advance has a real chance, especially as a pilot.

Start with the model. Propose 100-80-100 if your team’s work suits it; propose a rota-based version if it does not. Do not propose a compressed week unless people actually want ten-hour days.

Set the pilot terms. Three to six months, a fixed start and review date, and an explicit statement that the company can end it at the review if the metrics are not met. A pilot with an exit is far easier to approve than a policy.

Name the metrics before you start. Pick three the business already tracks (tickets resolved, revenue per head, project milestones hit, customer response times) and one wellbeing measure (a short survey or sickness absence). Agree what “success” means in numbers.

Solve coverage on paper. Show the calendar: who is off which day, how urgent requests are handled, what clients are told. For a customer-facing team, propose staggered days rather than a closed Friday.

Propose the productivity changes that pay for the day. Meeting-free blocks, a cap on meeting length, fewer status updates, clearer handovers. This is the part most proposals skip and the part that determines whether the fifth day is genuinely freed or just quietly reabsorbed.

Offer a written proposal of a page or two, then ask for a meeting. If the first response is “not now”, ask what evidence would change the answer and offer to gather it. If your employer already allows hybrid work, the same case-building approach applies, and our guide to asking for remote or hybrid work covers how to frame a proposal in your manager’s terms.

A realistic expectation

For most people, the near-term outcome is not a 32-hour week at full pay. It is a pilot in one team, a compressed or staggered pattern, or a job move to one of the firms that already offers it. If the last of those appeals, ask in interviews how long the arrangement has been in place, whether it is written into contracts, and what happened during the last busy period. An employer that has run four days for three years through a bad quarter is offering something real. One that announced it last spring is still running a trial, whatever the job ad says.