Return-to-Office Mandates Are Spreading. The Evidence for Them Is Thin.
Amazon now requires five office days a week, but research on turnover and hybrid work suggests strict attendance rules cost more than they deliver.
In September 2024, Amazon chief executive Andy Jassy told employees that the company would require them in the office five days a week from January 2, 2025, up from the three days a week it had required since May 2023. "We want to operate like the world's largest startup," he wrote in the memo, according to news reports at the time. The decision put one of the largest employers in the United States firmly on one side of a debate that has run since 2020.
The office itself is not the problem. Shared space remains useful for onboarding, for intensive collaboration and for the informal contact that helps teams cohere. What is harder to defend is the mandate: a blanket rule that ties every role, every week, to a fixed number of days at a desk, regardless of what the work requires. The research published since the pandemic suggests that such rules carry real costs for employers, and that the gains they promise are difficult to find in the data.
Where attendance stands
Office use has recovered, but not to its old level. Kastle Systems, which tracks badge swipes in commercial buildings, recorded its highest utilization since the pandemic during 2025, according to a December 2025 review by the real estate publication Propmodo. Average occupancy across major U.S. markets still sat mostly in the mid-50 percent range, with some cities passing 60 percent on peak days, below 2019 levels.
The same review found that about 70 percent of companies now require some in-office time, with three days a week the dominant standard. Most employers, in other words, have settled on hybrid arrangements. Full-time mandates such as Amazon's remain the exception, which makes them a useful test of whether more attendance produces better results.
What the research shows
The most rigorous evidence comes from a randomized experiment at Trip.com, the Chinese online travel company. Stanford economist Nicholas Bloom and his co-authors placed about 1,600 employees in engineering, marketing, accounting and finance on either a full-time office schedule or a hybrid schedule with two days a week at home. The results, published in Nature in June 2024, showed that resignations fell by 33 percent among non-managers on the hybrid schedule. The researchers found no measurable effect on productivity or on promotions.
A separate study points to what happens when firms move the other way. Researchers at the University of Pittsburgh, including associate professor Mark Ma, examined 54 companies in the S&P 500 that introduced return-to-office mandates between April 2020 and June 2023. According to Fortune's account of the working paper, employee turnover rose by an average of 14 percent after the mandates took effect. Departures were concentrated among women, mid-level and senior managers and employees with more skills listed on LinkedIn. The time needed to fill vacancies rose by about 23 percent, and hire rates fell by 17 percent.
Taken together, the two studies suggest that strict attendance rules can push out experienced staff and make them harder to replace, while flexible schedules retain people without visible damage to output.
Why mandates persist
If the evidence points in one direction, the persistence of mandates needs explaining. Part of the answer is visibility. A 2022 Microsoft Work Trend Index survey of about 20,000 people in 11 countries found that 87 percent of employees said they were productive at work, while 85 percent of leaders said hybrid work made it hard to be confident that their staff were productive. Microsoft called the gap "productivity paranoia." When managers cannot see work happening, attendance becomes a proxy for it.
Imitation plays a role as well. The Propmodo review cited founder surveys indicating that more than half of companies had changed their policies in response to what peers were doing. A policy adopted because competitors adopted it is not necessarily one that serves a particular business, its staff or its customers.
A narrower role for the office
None of this argues for closing offices. The Trip.com experiment kept employees in the office three days a week, and its authors describe hybrid work as beneficial for both companies and workers. The office works best as a tool used deliberately: for onboarding new hires, for planning sessions and for the kind of collaboration that genuinely benefits from being in the same room.
Measuring output rather than presence is the harder management task, but it is the one supported by the evidence. Companies that define clear objectives, give teams discretion over where routine work gets done and reserve office days for activities that need them are likely to keep more of their experienced people than companies that count badge swipes.
What it means for where people live
Flexible policies also change housing decisions. Employees with two or three remote days a week, or with fully remote roles, can increasingly spend weeks or months in another city without changing employers. That has added to demand for furnished housing let by the month rather than by the year, typically in cities with good transport links and space to work from home.
For those weighing such a move, RentRemote lists furnished apartments for monthly stays in cities including Barcelona and Lisbon, set up for people who work from home.