Transcription complète de Episode 059 - The Four-Day Week: What the Largest Trial Ever Actually Found

INTRODUCTION Welcome to PsychChat, where we share tips on navigating the complex world of work. I'm your host, Dr. Austin Tay. If this is your first time here, hit subscribe - it means you don't miss the next one. And if you've been listening a while, a review genuinely helps. It's how people who need this find it. Ask any HR director what happens when you cut a working day and most give you the same answer. People panic about the workload. Or they quietly compress five days of stress into four and call it a win. Wen Fan and her colleagues at Boston College just ran the largest four-day week trial ever conducted: two thousand eight hundred and ninety-six employees, a hundred and forty-one companies, six countries; expecting to find some version of that story. They didn't find it. What they found instead is more interesting, and considerably more useful, than anything you've seen shared on LinkedIn under a photo of someone at the beach on a Friday. PART 1 — The Trial Nobody Expected to Work Fan, Schor, Kelly and Gu published this in Nature Human Behaviour in 2025, and it's the real thing - not a self-selected pilot with a press release attached. Two thousand eight hundred and ninety-six employees across a hundred and forty-one organisations in Australia, Canada, Ireland, New Zealand, the UK, and the US. Twelve control companies that didn't make the change, tracked alongside the rest, so we're not just comparing people to their own hopeful memories of how things used to be. Here's the part everyone skips past. This wasn't 'cut a day and see what happens.' Every participating company got roughly eight weeks before the trial started to actually redesign how they worked - streamlining meetings, redistributing tasks, deciding what simply stops happening. Then, and only then, did the six-month trial begin. Full pay. No reduction. That detail matters more than the headline, and I'll come back to it. The results: burnout fell from 2.83 to 2.38 on a five-point scale. Job satisfaction rose from 7.07 to 7.59 out of ten. Mental health and physical health both improved. None of that shifted in the twelve control companies. And critically, the researchers went looking for the mechanism, not just the outcome. Three things mediated the improvement: better self-reported work ability, fewer sleep problems, less fatigue. Fan told Nature something I think is the most honest line in the whole write-up. She expected that if people were still delivering the same output in less time, they'd simply work faster and burn out doing it. That's not what happened. Stress went down, not up. Which tells you the four-day week isn't a productivity hack. It's a recovery mechanism that happens to leave productivity intact. Now, the caveats, because I'd rather give you those than let a vendor give you the sales version later. Companies opted in — this wasn't randomised. Everything was self-reported. People who wanted an extra day off may well have reported feeling better than they strictly were, in the hope of keeping it. The authors say as much themselves and call for a randomised trial. I'll take honest self-report data over confident nonsense any day, but hold that caveat lightly in your other hand. Quick gut check. If your organisation announced a four-day week tomorrow, would you actually work four days or would Friday just quietly relocate into Thursday night? PART 2 — The Four-Day Week That Isn't Here's where I get pedantic, because the term "four-day week" is doing a lot of unearned work in the current conversation. Two completely different interventions hide under one label, and conflating them is how you end up with a policy that makes people worse off and a leadership team that concludes the whole idea doesn't work. Model one is what Fan and colleagues studied: total hours actually come down, roughly twenty percent, pay stays the same, and the organisation redesigns the work first. Model two is what a lot of companies quietly mean when they say they've "gone four-day": the same total hours, compressed into four longer days. Ten-hour shifts instead of eight. Nobody's workload changed. The calendar just got rearranged. A systematic review published this year in the International Archives of Occupational and Environmental Health (Bernstrøm, Alves, Houkes, Lillebråten and Nilsen, 2025) looked specifically at that second model across twenty longitudinal studies. The results were mixed, which is itself the finding worth sitting with. Compressed schedules improved shift satisfaction in several studies. They also increased sickness absence. The reviewers were blunt: nobody has yet worked out when, for whom, or under what conditions a compressed schedule helps rather than harms. There's no consensus effect, positive or negative - there's a genuinely unresolved literature. So when a company tells its staff, or its shareholders, that it's adopted a four-day week, ask which one they mean before you congratulate anyone. Model one is a wellbeing intervention with real data behind it. Model two is a scheduling change dressed up in the language of a wellbeing intervention, and depending on the role, it can go either way. I'll be honest about the sardonic bit here, because it's earned rather than performed: the version of this that gets the LinkedIn post and the awards submission is almost always model two. The version with the eight weeks of unglamorous workflow redesign beforehand - the actual mechanism that made Fan's trial work, doesn't photograph nearly as well. PART 3 — Why the Redesign Is the Point There's a second finding buried in Fan and colleagues' data, matters more than the headline burnout figure. Both company-level and individual-level reductions in hours were associated with wellbeing gains - but individual-level reductions, where someone had genuine personal control over how their hours came down, were associated with larger gains. The company-level cut alone wasn't enough. Control over it mattered.