The four-day work week: what the trial data actually shows

Between 2015 and 2024, several large-scale trials of a four-day work week — with no reduction in pay — have run in Iceland, the United Kingdom, Spain, Portugal, Belgium, and Japan. The trials varied in design and scale but share enough methodology to produce a coherent picture. The picture, after a decade, is more positive than skeptics expected and more conditional than advocates often present.

1. The Iceland trials

Between 2015 and 2019, the Reykjavík City Council and the Icelandic government ran trials covering more than 2,500 workers — roughly 1% of Iceland's working population. Workers moved from 40-hour to 35-36 hour weeks at full pay, typically by taking Friday off. The trial was assessed by independent researchers at the Association for Sustainable Democracy.

The headline finding: productivity, measured by output and service-quality metrics, was maintained or improved in most workplaces. Worker wellbeing, measured by self-report and physiological indicators, improved substantially. The trial was deemed sufficiently successful that the Icelandic union federation negotiated permanent reductions in working hours for roughly 86% of the working population (Haraldsson & Kellam, 2021).

2. The UK trial (2022)

The largest single trial: 61 UK companies, ~2,900 workers, six months. Companies negotiated their own implementation (most chose to give workers Friday off; some used staggered schedules). Productivity was assessed by company self-report alongside worker wellbeing measures.

Results: 92% of companies continued the four-day week beyond the trial. Worker self-reported burnout decreased significantly. Revenue, on average, increased slightly relative to the same period the previous year. The company-level economic findings should be taken with some caution — companies opting into the trial were probably not representative — but the wellbeing findings were robust (4 Day Week Global, 2023).

3. The conditions that determine success

Where four-day weeks have worked, they have shared features:

Output-oriented metrics. Work measured by output rather than hours. Knowledge work and services where productivity can be assessed without tight time-attendance correlation.

Genuine reorganization. Successful companies redesigned meetings and processes. They cut low-value work rather than compressing all the existing work into four days.

Strong manager buy-in. Manager skepticism is the most reliable predictor of trial failure.

Where it has been less successful:

Manufacturing with continuous-process requirements. Hard to redesign without coverage trade-offs.

Sales roles measured by customer-contact hours. Compressed schedules tend to lose contact opportunities.

Companies that simply gave the day off without reorganizing. Productivity tended to compress into other days but with rising stress.

4. The skeptic's points

Several economists have raised reasonable cautions:

• The companies in trials are self-selected and probably overrepresent workplaces where the change is feasible
• Trial periods are relatively short; longer-term productivity effects are not yet measurable
• Productivity gains may come partly from cutting genuinely valuable activities (training, exploration, weak-tie networking) that show up as costs later

These caveats are real. They suggest the four-day week is promising and conditional, not universally applicable.

5. The summary

For knowledge work in companies willing to reorganize: the evidence supports the four-day week as a viable alternative to the five-day standard, with measurable wellbeing benefits and no clear productivity cost. For other contexts the evidence is thinner.

The default labor norm of the five-day week is increasingly looking like a historical accident rather than a productivity optimum. Whether that translates into widespread change depends on factors — political pressure, labor market tightness, manager willingness — that are not primarily about the data.