The AI Journal

The five day working week has achieved something remarkable: it has become so familiar that we rarely ask whether it is any good.

For most office workers, the arrangement is treated almost like a law of nature. Monday through Friday is work. Saturday and Sunday are not. A company that changes this arrangement is considered innovative, progressive or unusually generous. Yet the evidence emerging from recent experiments suggests that the burden of proof may belong on the other side. If an organisation can produce the same valuable work in fewer hours, there is little reason to preserve the longer schedule simply because it is conventional.

The strongest version of this argument is not that every occupation should immediately move to a four day week. A hospital, restaurant, factory or emergency service cannot necessarily operate according to the same schedule as a software company or an accounting firm. Nor should a four day week mean compressing forty hours into four ten hour days. The proposition is narrower: for knowledge work organisations, a genuine reduction in working time, with no corresponding reduction in pay, should become the default experiment.

The distinction between fewer days and fewer hours matters. A compressed week can simply make each working day longer and more intense. The model worth testing is closer to thirty two hours over four days, while maintaining pay and expecting organisations to redesign work so that unnecessary activity disappears.

There is now enough evidence to take this seriously. A 2025 study published in Nature Human Behaviour examined 2,896 employees across 141 organisations in Australia, Canada, Ireland, New Zealand, the United Kingdom and the United States. Organisations underwent a period of preparation before a six month four day week trial. Compared with twelve control companies, participating workers experienced improvements in burnout, job satisfaction, mental health and physical health. The researchers also found that reductions in working hours were associated with improvements in wellbeing, with reduced fatigue and sleep problems among the mechanisms involved (Fan et al., 2025).

This does not prove that every company will become more productive after removing a working day. It does establish something more useful: reducing working time is not inherently incompatible with organisational performance.

Earlier experiments point in the same direction. Iceland’s large scale trials, conducted between 2015 and 2019, reduced working hours to thirty five or thirty six per week without reducing pay for roughly 2,500 workers. The analysis reported that productivity and service provision remained broadly stable or improved, while indicators including stress, burnout and work life balance improved. The subsequent expansion of shorter hours across Iceland makes the experiment particularly interesting because it moved beyond a small collection of enthusiastic companies (Autonomy and Alda, 2021).

The United Kingdom provides another important test. Its 2022 trial involved sixty one organisations across different sectors. Research associated with the experiment found that companies could reorganise work around the shorter schedule, with benefits reported for employee wellbeing and organisational outcomes. The subsequent evidence is important because the experiment did not end when the formal trial ended: many participating organisations continued with the arrangement. UK research also found that reducing working time required practical changes to meetings, communication, workflow and priorities rather than simply telling employees to work faster (Burchell and Frayne, 2024).

That last point gets to the real reason the four day week is interesting. The argument is not fundamentally about giving people Fridays off. It is about discovering how much of the conventional workday is actually necessary.

Knowledge work is particularly vulnerable to confusing activity with output. Meetings create calendars full of evidence that work is happening. Email produces a visible stream of responsiveness. Long hours create the appearance of commitment. But none of these measures necessarily corresponds to useful work. A shorter week forces an organisation to ask uncomfortable questions. Which meetings can disappear? Which approvals are unnecessary? Which recurring reports are unread? Which tasks exist mainly because they have always existed? Which decisions could be made by one person instead of five?

In that sense, a four day week can function as an organisational audit.

Research examining the psychological mechanisms behind shorter working weeks offers another reason for optimism. A 2024 review in Trends in Cognitive Sciences argues that improved performance under reduced working time can arise through recovery and motivation: workers who are better rested and more motivated may produce better work during the time they are actually working (Rae and Russell, 2025). A separate scoping review of four day week research similarly concludes that the evidence is promising while emphasising that outcomes depend heavily on how the model is implemented (Jahal, Bardoel and Hopkins, 2024).

The objection is obvious, and it deserves to be taken seriously. A shorter week may work for a software company with flexible deadlines, but what about organisations whose customers expect continuous service? What about hospitals, schools, shops, manufacturers or small businesses that cannot simply close for an extra day? And even inside an office, might employees simply experience the same workload compressed into fewer hours, producing greater stress rather than less?

Yes. This is the strongest objection to treating the four day week as a universal policy, and it is why the argument should not be universal. A four day week is not a magic number. Some organisations need five or seven days of coverage. Some jobs are constrained by physical presence. Some teams may discover that reducing hours genuinely harms service. And even knowledge workers may respond badly if management treats the shorter week as a demand for the same volume of work at greater intensity.

But these objections do not undermine the narrower claim. They strengthen it. The point of making the four day week a default experiment is precisely to test whether a particular organisation can make it work. A failed experiment is useful information. Continuing indefinitely with a five day schedule without testing an alternative is not neutrality; it is simply an unexamined assumption.

The important safeguard is that the experiment must measure outputs rather than appearances. Organisations should track customer service, revenue, quality, deadlines, employee turnover and other meaningful performance measures alongside working hours and wellbeing. They should not declare success merely because employees like having an extra day off, nor declare failure because some traditional activity has disappeared.

The five day week once made sense as a social compromise between industrial production and human limits. But the nature of work has changed. In many knowledge economies, the scarce resource is no longer the number of hours a person can physically remain at a workplace. It is attention, judgement, creativity and sustained concentration.

Those resources are not necessarily maximised by keeping people at work longer.

The evidence does not justify claiming that four days will always beat five. It justifies something less dramatic and more important: we should stop treating five days as the scientifically established answer. For knowledge work, the four day week deserves to become the baseline experiment, with organisations required to discover whether the fifth day is actually buying them anything.

If many can produce equal or better results with less time, the question will no longer be why some companies offer a four day week. It will be why everyone else still needs five.