Can a four-day week preserve productivity?

The promise sounds like a hard equation: deliver 100% of the work in 80% of the time for 100% of the pay.

Some organizations have done something close to that. In large coordinated pilots, working hours fell while employers often reported stable performance, and the smaller set of firms sharing revenue data did not show a clear collapse. Employees also reported better health and less burnout.

But the evidence is not a clean productivity trial. Firms volunteered, work was redesigned before the clock was shortened, objective output was rarely comparable across companies, and a large peer-reviewed study focused on wellbeing rather than business production.

So the answer is conditional. A four-day week can preserve reported performance inside some prepared organizations. Current evidence cannot promise the same output in every sector or prove that the missing day caused it.

Four days and fewer hours are not the same policy

Four-day week can describe at least two very different changes:

• a compressed schedule, such as four 10-hour days with 40 hours unchanged;
• reduced working time, often around 32 hours, with no reduction in pay.

The recent pilots that drive the strongest claims generally tested the second idea. Even within those pilots, implementation varied. Some firms closed on one day. Others staggered days off, annualized the reduction or made the shorter week conditional on operational demand.

That variation is not noise to erase. It is part of the intervention. A customer-service team that staggers coverage and a design studio that closes on Friday are not running the same operating model.

The UK reporting subset showed no aggregate revenue decline

The 2022 UK program enrolled 61 organizations and around 2,900 workers for six months. Before the trial, companies received workshops, coaching and support to redesign meetings, communication and work processes.

The company-level result was encouraging but not universal. Fifty-six of the 61 organizations said they would continue immediately after the pilot; three paused, and two extended or modified their experiments. On 0-to-10 management ratings, average perceived company performance and productivity were both 7.5.

Revenue provides a harder, if still imperfect, check:

Those figures do not establish a causal productivity gain. Different firms reported different metrics, no common objective output measure existed, and economic conditions could change revenue independently of working time. The large year-over-year number may include recovery, inflation, growth or client mix. The report itself tells readers to focus on broad tendencies.

The defensible reading is narrower: among the companies able and willing to report these data, revenue did not reveal an obvious six-month breakdown while hours fell. That is meaningful feasibility evidence, not proof that every firm became 25% more productive.

Germany added controls but still lacked broad objective output

A 2024 German pilot selected 51 organizations across varied industries. Six withdrew before implementation, leaving a 45-organization cohort. One did not start within the reporting window, one was excluded for reducing pay and two exited after two months, leaving 41 organizations in the reported analysis. Unlike the UK program, parts of the German study included employees or units that remained on their previous schedule as controls. Work-time reductions among the 41 also varied: 34% cut 20%, while 46% reduced hours by 10% or less.

Employees moving to a four-day model reported higher productivity at the final wave than controls: 7.69 versus 6.83 on the study’s scale (p=.002). Work pace increased, while reported workload did not change significantly. Managers also rated the model’s influence on productivity positively on average. The employee survey had substantial attrition: responses fell from 643, including 108 controls, at baseline to 332, including only 39 controls, at the final wave.

The objective ledger was much thinner. Only 12 organizations supplied usable revenue and profit figures; neither changed significantly against the same period in 2023. Only two provided individual objective performance data, too little for the report to publish as a result.

Stable financial output with fewer hours is compatible with higher labor productivity. It is not identical to it. Revenue and profit contain prices, demand, capital and market conditions as well as employee output. The German authors explicitly said their data could not support a definitive conclusion about organizational performance.

Two organizations exited after two months. One large organization faced external economic pressure that the report says was unrelated to the four-day-week model. The second cited economic reasons, but the report says the exact cause was unclear and notes incomplete leadership support. Among the 41 analyzed organizations, 73% planned to continue, extend or make the model permanent, 20% did not plan to continue and 7% were undecided. Half of the non-continuers said they might reconsider later. Those outcomes belong in the result because adoption friction is part of whether a work design scales.

A large peer-reviewed study answered a wellbeing question

In 2025, researchers published a large peer-reviewed analysis of the coordinated trials: 2,896 employees in 141 participating organizations in a cross-national sample, compared with 285 employees in 12 US control organizations.

Company-level averages of employee-reported hours fell from 39.21 to 34.01 per week. At the individual level, reported hours fell from 38.98 to 34.26. Burnout, job satisfaction, mental health and physical health improved in the trial group, unlike in the controls. Larger individual hour reductions were associated with larger wellbeing gains.

That is important evidence for the employee side of the equation. It is not an objective productivity result. The study used self-reported work ability as a proposed mediator, and its outcomes were self-reported wellbeing measures. It did not compare standardized business output before and after the intervention.

The control group also does not turn the program into a randomized trial. Companies chose whether to join, the 12 controls were a small and non-representative US group interested in the model, and participating firms were often small and unusually supportive of flexibility. The authors state that selection may overestimate the true wellbeing effect. The main comparison covered six months, with additional 12-month robustness data. The authors declared no competing interests; the participant data were restricted, with public ICPSR availability planned for 2027.

Work redesign carries part of the result

The pilots did not simply remove Friday and observe what happened. Companies first reviewed meetings, handoffs, communication norms, priorities and tools.

That creates two interpretations:

1. reduced time may improve recovery, focus and retention;
2. preparation may remove low-value work that could have been removed under a five-day schedule too.

Both can be true. It also means a four-day week should be evaluated as an organizational redesign, not as a calendar trick. If the fifth day disappears but meetings, approval queues and coverage demands do not, work may simply intensify or spill into nominal time off.

The German increase in work pace makes that risk visible. Better wellbeing at six months does not guarantee that pace is sustainable over several years.

What the evidence changes

The next useful trial would randomize implementation where practical, pre-register one operational output per organization, track quality and errors alongside quantity, preserve customer coverage, and follow both adopting and exiting firms for more than a year.

Primary sources

• Fan et al., 2025, Work time reduction via a 4-day workweek
• UK Four-Day Week Pilot Results, 2023
• German Four-Day Week Pilot Study Report, 2024

Independent editorial summary. The authors are not affiliated with LifeScore.