Trials of a reduced working week with unchanged pay have run in several countries, with enough participants and enough duration to produce real data.
What the results show
Consistently: high rates of companies continuing after the trial, improved self-reported wellbeing, reduced sick leave, and revenue that held steady or grew.
Those findings have replicated across trials, which is more than most workplace research manages.
The caveats that matter
Participating companies volunteered. Organisations that sign up for a productivity trial are not representative – they are ones that already suspected they had slack and were willing to restructure.
Most participants were knowledge-work businesses where output is not tied to hours present. Results from manufacturing, healthcare and retail are far thinner, and those are the sectors where the change would be hardest.
And most trials measured self-reported productivity rather than objective output, because objective output is genuinely hard to measure in office work.
What actually changed inside companies
Meetings, overwhelmingly. Fewer, shorter, with fewer people. That single change accounts for a large share of the recovered time in most reports.
Which raises an uncomfortable question about how much of a normal week was already unproductive.
Where it does not transfer
Anywhere output scales directly with hours. A shop open six days needs staffing six days. Reducing individual hours there means hiring more people, which is a cost rather than an efficiency.
The realistic conclusion
A genuine improvement for a specific category of work, oversold as a universal solution. The interesting finding is not the four-day week itself but how much slack the five-day week contained.