General information, not legal advice. Wage-and-hour rules vary by country and, in the US, by state; confirm your position with counsel before changing schedules.
“The four-day week” is three different offers wearing one label. A 32-hour week cuts a full day of work. A 36-hour week trims an hour a day, or works four nines. A compressed 40-hour week (4x10) keeps every hour and moves them. The evidence, the risks and even the overtime law treat the three differently, so choose deliberately.
The three models side by side
| 32 hours (4×8) | 36 hours (4×9 or shorter days) | 40 hours compressed (4×10) | |
|---|---|---|---|
| What changes | Hours and days both fall | Hours fall slightly | Only the pattern changes |
| Pay in most trials | 100% maintained | Usually maintained, sometimes debated | Unchanged |
| Evidence base | Strongest: UK, US/Ireland, Iceland follow-through, 2025 multi-country study | Iceland’s 35–36 hour trials; many collective agreements | Long shift-work history; thinnest wellbeing evidence of the three |
| Main risk | Output dips if work is not redesigned; coverage gaps | Gains too small to notice; 9-hour days strain some roles | Fatigue on long days; childcare and commute clashes; benefits read as schedule-only |
The pattern in the research is consistent: wellbeing gains track real hour reductions. The large 2025 study found improvements tied to how much hours actually fell, which is why 32 hours shows the strongest effects, 36 a gentler version, and 4x10 mostly a longer weekend rather than a lighter week.
Benefits and risks of each
32 hours. Benefits: the full evidence package applies here: burnout, stress and sick days down in the big trials, retention up, output maintained where redesign happened. Risks: it is the biggest operational jump. Without a stop-doing list the same work is squeezed into less time, intensity rises, and the model gets blamed for a redesign failure.
36 hours. Benefits: a softer landing. Four hours is easier to find in meetings and low-value work, coverage is simpler, and Iceland shows it can stick nationally. Risks: expectations. Staff promised “a four-day week” who get 4×9 may feel short-changed, and nine-hour days create their own problems (see overtime below). Some employers pair 36 hours with pro-rata pay debates the 100-80-100 model deliberately avoids.
40 hours compressed. Benefits: no pay or output arithmetic to solve, and the three-day weekend is real. Risks: ten-hour days are tiring, error-prone late in the day, and hostile to school runs. It is also the model most likely to be granted instead of a reduction and resented for it. Treat it as a scheduling option, not the destination.
Exempt versus hourly: where overtime actually bites
Classification decides whether the clock legally matters at all.
Exempt (salaried) staff. Properly classified exempt employees are paid a set salary regardless of hours, so a shorter week changes no pay calculation. Two traps: first, salary must not be docked because the office closes on the fifth day, since improper deductions can undermine the exemption itself. Second, invisible overtime: exempt staff absorb the squeezed work at 9pm, so measure actual hours, not scheduled ones.
Non-exempt (hourly) staff. Every hour counts and must be recorded, including work on the day off that the employer “suffers or permits”. Overtime then depends on thresholds:
- Federal US rule of thumb: overtime at 1.5× is owed for hours over 40 in a workweek. On that test, 32 and 36-hour schedules sit safely under the line, and a 4×10 at exactly 40 hours triggers no federal overtime either. The danger is drift: swapped days, cover shifts or a busy week pushing someone past 40.
- Daily-overtime states change the maths. California-style rules pay overtime after 8 hours in a day (double time after 12), so 4×9 and 4×10 schedules can owe daily overtime unless a formal alternative-workweek election (a two-thirds secret-ballot vote, filed correctly) is in place first. Several other states and many collective agreements have daily or seventh-day rules.
- Regular rate means regular rate. Non-discretionary bonuses and differentials feed the overtime rate, and record-keeping duties do not shrink with the week.
Net effect: compressed 4x10 is legally simple in weekly-threshold jurisdictions and paperwork-heavy in daily-overtime ones; reduced 32/36-hour models are legally simple almost everywhere but need the strictest guard against off-the-clock creep, because unpaid hours worked on the fifth day are both a wage claim and a failed pilot.
Actionable steps
- Name the model in writing: hours per week, pay treatment, and which days, team by team. Never announce “four-day week” without the number.
- Map your overtime exposure by jurisdiction and classification before choosing day lengths; in daily-overtime territory, price 4×9 and 4×10 honestly or run the election process properly.
- For exempt teams, add an actual-hours check (pulse question or sampled diaries) so intensity creep is visible.
- For hourly teams, define how day-swaps and cover are approved so nobody drifts past 40 hours, or over a daily threshold, by accident.
- Match expectations to the model: sell 36 hours as a step, 4x10 as flexibility, and reserve the full wellbeing claims for real reductions.
Sources
- US Fair Labor Standards Act overtime framework (29 U.S.C. §207; DOL guidance on the 40-hour workweek threshold and the exempt salary-basis tests, federal salary level $684/week after the 2024 rule was vacated; verify current figures).
- California Division of Labor Standards Enforcement: daily overtime after 8 hours, double time after 12, and alternative workweek schedule election requirements.
- Iceland shorter-hours trials (2015–2019) at 35–36 hours and subsequent collective agreements; UK 2022 pilot; 4 Day Week Global US/Ireland pilot; Nature Human Behaviour (2025) multi-country study linking gains to actual hour reductions.
Take back day five. Day5Group helps organisations transition to a 4 day work week. More about Day5Group · All blog articles