The best-documented test of the four-day week to date, a UK pilot of 61 companies run from June to December 2022 and coordinated by researchers at Cambridge and Boston College with the advocacy group 4 Day Week Global, ended with 56 of the 61 companies still on the shorter schedule — and 18 making it permanent, per the pilot's published results. Company revenue over the trial held essentially flat, rising 1.4 percent on average weighted by headcount, while resignations fell 57 percent and sick days dropped by roughly two-thirds against the same period a year earlier. Those numbers come from self-reported company data, a qualification the researchers themselves state plainly.
How were the trials run, and who was in them?
The UK pilot recruited companies mostly from knowledge-work sectors — marketing, software, consulting, charities — with a median headcount under 50. Each participated for six months, moving staff to 100 percent of pay for roughly 80 percent of hours, with the commitment to hold output at prior levels. Participants reported monthly revenue, headcount, and staff well-being metrics, compiled in the published academic analysis released in February 2023.
The design is the honest limit of the findings. These were volunteers, not a random sample; a company confident it could survive the change was likelier to enroll, a selection effect the published report acknowledges. What the trial establishes is that the schedule held up in the firms that tried it — not that it would hold up everywhere.
What happened to the workers, on the numbers?
The measured changes were consistent across the well-being instruments the researchers used. Sick days fell about two-thirds; staff turnover fell 57 percent across the trial period; self-reported burnout and anxiety declined on validated scales, and 71 percent of employees reported lower burnout at the end. The published report also records the harder edge: some workers in customer-facing roles reported tighter, more intensive workdays, and some managers described the first two months as the hardest reorganization of their careers.
The finding worth keeping is the quiet one. Revenue did not collapse. For a workforce debate usually argued in absolutes, a flat line is the whole story.
What did the companies that quit the trial say?
The five companies that returned to five days cited readiness and workload mechanics rather than the concept, per the published report — insufficient time to restructure workflows, and in one case an upcoming restructuring unrelated to the trial. The exits are too few to generalize from, and the report does not.
How do the results compare across the other pilots?
The UK trial sits in a family of similar efforts, and the pattern repeats:
| Pilot | Scale | Continued after trial | Reported revenue effect |
|---|---|---|---|
| UK, 2022 | 61 companies, ~2,900 staff | 56 of 61 | +1.4% weighted average |
| Iceland, 2015-2019 | ~2,500 workers, public sector | Made permanent for most | Service hours maintained, per national analysis |
| US & Ireland, 2022 | ~70 companies per published reports | Majority, per organizing group's report | Reported stable or higher |
The Iceland trial, run by Reykjavík's city government and national authorities and analyzed in a 2021 report, is the largest public-sector test: working hours fell without a measured drop in service output. The US-Ireland figures come from the organizing group's own report and have not been peer-reviewed to the same standard — a distinction this analysis preserves.
What remains unsettled?
Manufacturing, healthcare, and hourly-shift work are barely represented in any published trial, and the economics of coverage — hiring a fifth shift — are a different problem from compressing knowledge work. The documented record establishes that the four-day week survived where it was tried, and that the firms trying it were self-selected. Whether it generalizes is not yet a question the evidence can answer, and the strongest reading of the record waits for a trial that did not recruit its own believers.
For more context, read What the WARN Act Actually Owes Workers When a Plant Closes.