Do Return-to-Office Mandates Work? What the Evidence Says

Evidence grade: Strong. Several independent studies with credible causal identification point the same way. This grade describes the published evidence. It is not a prediction about your workforce and not a result RetainScore has produced.

Strong, and pointing the way most mandates bet against

This is the best-evidenced question on the site, and the grade needs reading carefully. Strong does not mean return-to-office mandates work. It means several independent studies with credible designs agree on what happens to retention when flexibility is taken away: more people leave, the people who leave skew senior, skilled and female, and no study has found the performance gain the mandate was supposed to buy.

If "work" means bring people back to the building without losing them, the evidence says no. If it means something else, the mandate's owner should be able to say what, because the retention cost is the part that has been measured.

Two randomized experiments, ten years apart

Most retention questions have no controlled trial behind them. This one has two, both run by Nicholas Bloom's group at the same Chinese travel company under two different names, both randomizing by birthday so that nobody chose their own condition.

In 2010 and 2011, Ctrip call centre staff who had volunteered to work from home were randomly assigned to home or office for nine months. Attrition among the home workers was 17% against 35% for the office group: "the total attrition rate in the treatment group (17%) was less than half of that in the control group (35%)". Performance rose rather than fell.

In 2021 and 2022, Trip.com ran the version that matters more to today's argument: 1,612 engineers, marketing and finance staff, randomized between a full office week and a hybrid week with Wednesdays and Fridays at home, for six months. The paper, published in Nature in 2024, reports that "hybrid working improved job satisfaction and reduced quit rates by one-third". Two years of subsequent performance reviews found that "hybrid working did not affect performance grades over the next two years of reviews", and promotions did not differ either.

Neither experiment tested a mandate directly. They tested the thing a mandate removes, and found that removing it costs roughly a third to a half of the quits it would otherwise prevent. The direction is the same either way.

What happens when a mandate actually lands

Two observational studies follow real mandates at real companies, and both land where the experiments predict.

Van Dijcke, Gunsilius and Wright matched resume histories to the return-to-office mandates at Microsoft, SpaceX and Apple in 2022. Against synthetic comparison firms, tenure fell, and the fall was concentrated among the longest-tenured staff: "We estimate a reduction in counterfactual tenure that increases for employees with longer tenure". The seniority distribution shifted toward junior positions, and "These senior employees left to direct, large competitors".

Ding, Jin, Ma, Xing and Yang widened the lens to 54 technology and finance firms in the S&P 500, tracking three million employment histories on LinkedIn. After a mandate, "the average turnover rate for RTO firms increased by 14 percent after RTO mandates", measured against each firm's own baseline. The increase was not evenly spread: "female employees, more senior employees, and employees with higher skill levels are more likely to leave RTO firms". Hiring got harder at the same time. Vacancies took about 23% longer to fill, around twelve extra days, and hire rates fell about 17%.

That skew is the finding to sit with. A mandate does not lose a random slice of the workforce. It loses the people with the most options, which is the same population every other page on this site describes as the hardest to replace.

What the mandate was supposed to buy

The retention cost would be worth paying if mandates delivered what their announcements promise. Ding and Ma looked for that in S&P 500 firms with mandates, using Glassdoor reviews and financial results before and after each announcement, and found that "RTO mandates hurt employee satisfaction but do not improve firm performance". Job satisfaction and views of management fell. Financial performance and firm value did not move.

The hybrid experiment says the same thing from the other side. Two years of performance reviews and promotion decisions showed no difference between the office and hybrid groups. The performance case for a mandate is, so far, an assertion.

The honest boundaries

• The experiments are about hybrid, not fully remote. The Trip.com trial tested two days at home out of five, and its authors say the result "might not replicate to all other hybrid work arrangements". Nothing here is evidence for a workforce that never comes in.
• Home working has a cost the experiments also measured. In the 2015 study, "conditional on performance, it reduced rates of promotion by about 50%", and once given a free choice, half of the home workers went back to the office. Flexibility is not free for the employee either.
• The early trial used volunteers. Its participants were "Call center employees who volunteered to WFH", so it says little about people who would rather be in the building.
• The mandate studies are large tech and finance firms, and short. The three-firm study could see only one quarter: its authors write that "we are only able to study these mandates' effects one quarter after their implementation". The 54-firm study is explicit that "we cannot draw causal inferences based on our setting". A regional employer with a local labour market may see less movement, because its people have fewer places to go.
• Intention surveys are not turnover. Gallup's 2023 panel found three in ten hybrid and six in ten fully remote workers "extremely likely to leave an organization if not offered at least some degree of remote flexibility". That is what people say. The studies above are what they did, and the gap between the two is usually large.

This page replaced one of ours

Until today this URL held a page titled Flexible and Remote Work as a Retention Strategy. Its direction was right, and its evidence was a single Gallup survey of stated intentions. Writing this grade meant reading the experiments and the mandate studies properly, and the honest thing was to replace a thin page with the evidence rather than leave both up.

If you are deciding on a mandate

• Name the population before you announce. The cost falls on the people who can leave. Count your long-tenured, senior and highly skilled staff who are currently hybrid or remote, and treat that list as the exposure.
• Expect the skew, not the average. A 14% rise in turnover sounds survivable until it turns out to be your most experienced engineers and your most senior women.
• The hybrid middle is where the evidence is. Two days a week at home is the design that has been tested and found to hold performance while cutting quits. A five-day mandate and a zero-day policy are both outside what anyone has measured well.
• Measure quits by tenure band, before and after. That is the number the studies moved, and it is the one you can check against your own payroll rather than against a survey.

Work through the retention diagnostic to see how flexibility sits alongside the other factors the research finds predictive, or read the companion grade on whether retention bonuses work, the intervention most often reached for once a mandate has started costing people.