The data are not as supportive as claimed.
On February 21, a consortium of advocacy groups and academics released a report on a program piloting four-day workweeks for employers in the United Kingdom from June to December 2022. In a press release, the authors described the pilot program as a success and claimed that it “demonstrate[d] the benefits of reduced-hour, output-focused working.” Major news organizations, such as NPR and CNN, amplified these talking points. Although the report does seem to contain some positive data on this issue, a careful review of the report raises questions about the results.
First, it is important to understand who is behind the study. The team includes (1) Autonomy, an economics research organization; (2) 4 Day Week Global, a non-profit dedicated to promoting the concept of a four-day workweek; (3) Dr. Juliet Schor, a professor at Boston College and advocate of reduced work schedules; and (4) Dr. David Frayne and Professor Brendan Burchell from the University of Cambridge, researchers with a focus on labor and employment matters. Thus, at least some of the authors have a predisposition to achieving results that demonstrate the value of a four-day workweek.
Second, the report, its results, and the underlying methodologies have not been peer reviewed.
Third, the program and its results seem likely to suffer from selection bias. The press release touts the program as having taken place at “over 60 companies.” However, many of the companies that agreed to take part in this program were workplaces likely to be politically or socially aligned with the goals of the program, or workplaces in which a four-day workweek would not be disruptive to operations. Indeed, the report says that “initially, 70 companies had signed up to take part in the pilot” but nine then bowed out for various reasons. (p. 17, n. 20) Next, although 61 organizations participated in the pilot, only 44 to 51 provided data for the report. (p. 15) It seems quite possible that the organizations that responded to the survey were those that had a positive experience with the program. It would be important to know why 10 to 17 organizations did not provide data.
The employee data on this point are even more troubling. Only 58 percent of the 2,900 employees covered by the pilot provided survey responses at the conclusion of the study. Again, those who chose to respond were likely to be those who supported the program.
Fourth, the reduction in work is not what it seems. The report describes a variety of ways in which the employers altered their schedules to create a “four-day workweek”: some closed all operations on Fridays; some divided their staffs’ days off between Mondays and Fridays; some continued to work a five-day week, but only with four days’ worth of work; and, some adjusted schedules so that the workload was spread over a year. (pp. 20-21) Nevertheless, the average number of days worked decreased from 4.86 to 4.52 (p. 32), only a third of a day’s reduction. Moreover, the average weekly workload decreased from 38 to 34 hours (id.), a reduction of only about half a day.
Fifth, the effect on company performance is unclear. The researchers’ methods here seem particularly confusing. (See p. 29.) In one bucket, the researchers analyzed the revenue of 23 companies (weighted by size) and found an average increase in revenue of 1.4 percent from the beginning to the end of the trial. This result is well below the rate of inflation, which has been around 10 percent in the U.K. since the pilot began. For another 24 companies, the researchers averaged the companies’ revenues before and after the pilot, and found a size-weighted average percentage increase of 35 percent. The researchers do not explain how the companies were split between the two analyses. There is no obvious reason why both analyses could not have been run for all companies. The underlying data here too would be important to provide for verification.
The researchers otherwise do not appear to have made any effort to measure and evaluate companies' outputs. The pilot was only six months long, and thus a negative effective on revenue may certainly lag until a later date.
That said, the employees who responded to the survey do seem to have liked the change. On average, employees reported mild but statistically significant improvement in stress, job satisfaction, mental health, and sleep. (p. 36-38) The biggest positive changes were reported in employees’ ability to manage their family and household duties. (p. 39) Perhaps not surprisingly, 96 percent of respondents expressed a preference for a four-day workweek. (p. 45)
In an ongoing tight labor market, a four-day workweek may become a recruitment tool used by some employers. Some organizations may find that they operate more efficiently on such a schedule. However, this report does not clearly make the case for the adoption of a four-day workweek. Everyone would benefit from the results of a longer, more representative trial that is qualified and scrutinized appropriately in a peer-reviewed publication.