The four-day week is an operating-model bet you can't see

A four-day week is usually sold as a scheduling change. It is better understood as a bet: that a meaningful share of the working week was never load-bearing, and can be removed without the work suffering. Win the bet and you have cut waste. Lose it and you have crammed five days of fragmented work into four, then pushed the overflow into people's evenings. The awkward part is that most organisations running the experiment cannot tell which of those two things happened.

The well-being evidence, at least, has stopped being in doubt. The largest study to date tracked nearly 2,896 employees across 141 organisations in six countries against a control group of about 300, and found consistent improvements in burnout and in mental and physical health, with the largest gains where hours fell most.¹² The control group showed no comparable change, which is what lifts this above the usual pilot anecdote. Retention is firmer still: the UK pilot saw the number of staff leaving fall 57 per cent and 90 per cent of organisations choose to continue, with sick days down 65 per cent.³ On well-being and retention, the case is made.

Why it is an operating-model bet, not a calendar change

For an operating leader the interesting question is not whether people feel better. It is why productivity holds when it holds. Here the researchers behind the largest study are direct about the mechanism: companies that kept output up did it by cutting low-value or questionable activities, meetings above all, rather than by driving people to work harder in less time.⁴ A compressed week, in other words, does not make anyone faster. What it does is force the organisation to delete coordination overhead, which is a structural change dressed up as a timetable one.⁵

You can see the dependency in where the model succeeds and where it stalls. Gains cluster in coordination-heavy knowledge work, the kind of role where low-value meetings can be stripped out. They thin in shift-based, customer-facing and partner-dependent jobs, whose hours are already tied to demand.⁵ Iceland's large public-sector experiment, covering more than 2,500 workers, found productivity neutral to positive, with the compressed pattern forcing sharper prioritisation and fewer meetings.⁵ The pattern is consistent: the four-day week pays out to the degree there was coordination waste to remove, and not otherwise.

The blind spot, and what would actually answer the question

And here the evidence base has a soft centre, exactly where operators need firmness. Productivity is the single hardest outcome to measure objectively. Well-being effects come through far more clearly than output effects, and the data carries a honeymoon effect plus selection bias from volunteer companies already inclined to succeed.⁶ Even the headline study never measured productivity directly. Workers self-reported it, with 52 per cent saying they had become more productive on fewer hours, which is suggestive at best.² So firms finish the trial with good instruments for the perk and poor ones for the bet. Revenue holds, survey scores rise, and the two things that would actually settle it, focus time and after-hours creep, go unmeasured.

Two objections are worth taking head on. The first: the well-being case is now strong enough that demanding granular instrumentation before adopting a four-day week over-engineers a change that already works. Concede it fully. The point here is narrower. Well-being is well-measured; the operating-model claim is not, and a firm can adopt the four-day week on well-being grounds while staying blind to whether it is structurally sustainable. The second objection comes straight from the psychology literature: productivity is genuinely hard to measure, so claiming any signal can adjudicate a trial overstates the data.⁶ Agree, and narrow the claim accordingly. Workstyle signals are not a productivity gauge. What they read is the mechanism, the change in meeting load, interruption cadence, deep-work windows and evening activity, baselined before the trial and tracked through it. Treat them as leading indicators of whether coordination waste fell or merely moved, never as a verdict on output.

That distinction is also why the measurement has to be trust-based to be worth anything. During a high-stakes schedule experiment, the incentive to look productive peaks, so surveillance-style tools get gamed hardest exactly when the numbers matter most. A consent-based model that reads aggregate patterns and shows employees their own data first takes away the incentive to perform, because there is nothing to be caught not doing. So the honest version of the four-day-week question was never 'did revenue hold'. It is the sharper one a board should ask before it congratulates itself. Did we remove the coordination, or did we just hide it in the evenings?

Footnotes

1. Fan, W., Schor, J. B., Kelly, O., & Gu, G. (2025). Work time reduction via a 4-day workweek and its effects on workers' well-being. Nature Human Behaviour. (Working paper: Worktime Reduction and Workers' Well-Being, ASA.) https://www.asanet.org/wp-content/uploads/2023/12/Worktime-Reduction-and-Workers-Well-Being-2023-12-16-4DW_wellbeing.pdf ↩

2. Business Insider. (2025). Thousands of workers tried 4-day workweeks, and the results are clear. https://www.businessinsider.com/workers-try-four-day-workweek-burnout-mental-health-research-productivity-2025-7 ↩ ↩²

3. UKRI / ESRC. (2023). Making the case for a four-day working week (UK pilot; Burchell, Cambridge & Frayne, Salford). https://www.ukri.org/who-we-are/how-we-are-doing/research-outcomes-and-impact/esrc/making-the-case-for-a-four-day-working-week ↩

4. Boston College. (2025). A four-day work week? BC News. https://www.bc.edu/bc-web/bcnews/nation-world-society/sociology/-study-pilots-four-day-work-week.html ↩

5. Future of Work Media. (2026). Four-day work week results: what the 2026 data says about productivity, retention and wage pressure. https://www.future-of-work.net/four-day-work-week-results-what-the-2026-data-says-about-productivity-retention-and-wage-pressure ↩ ↩² ↩³

6. Abrams, Z. (2025). The rise of the 4-day workweek. Monitor on Psychology, 56(1). American Psychological Association. https://www.apa.org/monitor/2025/01/rise-of-4-day-workweek ↩ ↩²