Five days a week. That’s the new return-to-office policy at Jeff Bezos’ Washington Post and at Amazon.com Inc., the retail giant now run by his designated successor Andy Jassy. It’s also the RTO recommendation Elon Musk and Vivek Ramaswamy have announced for the U.S. federal government.

It’s a terrible way to manage talent. It seems obvious that these unpopular policies are a way to spur employees to quit, in a kind of “self-deportation” version of layoffs — but without severance payouts, health insurance continuance or WARN Act notification periods. That may sound appealing to some employers, but they should recognize which workers are most likely to leave voluntarily: those with the most sterling credentials, the most in-demand skills, and the best alternative employment options. In other words, the top-tier talent employers are usually desperate to retain.

Jassy has denied that his RTO policy is a backdoor layoff, but the head of Amazon Web Services told staff that if they don’t want to work full-time in the office, “That’s OK, there are other companies around.”

A leaked Post memo shared on BluSky takes the same tone: “If an employee decides they do not want to return to The Post on a 5-day a week office schedule, we understand and will accept their resignation.”

Similarly, Musk and Ramaswamy, whom President-elect Donald Trump has named to spearhead a new government efficiency effort, laid out their thinking in a Wall Street Journal op-ed. A five-day-a-week schedule could result in a “wave of voluntary terminations that we welcome,” they wrote. Ramaswamy told Tucker Carlson that the plan could get perhaps 1 in 4 federal workers to quit.

And Musk has repeatedly slammed those who work from home, announcing the end of remote work at Twitter (now X) in his very first email to staff. Although he later backtracked and has told some staff to work remotely after closing offices — one of his first acts at the company was to stop paying rent at several locations — he’s made no secret of his view that any form of WFH is lazy and even morally suspect.

He’s not the only executive to assume that those who aren’t working five days in the office are less valuable or less committed. Malcontents. Slackers. Dead weight.

But these assumptions are dead wrong. The employees likeliest to stay after the imposition of a strict, five-day in-office mandate are not necessarily the most talented or the most committed, but those convinced they have no better employment options; those close to retirement; and those who live nearest the office. That’s not exactly strategic talent management.

The employees who will quit in the face of a strict RTO policy are the very ones who are the most employable — the mid-career stars who will find it easiest to get another job. Remote and hybrid workers tend to be more highly educated, wealthier, and to do the kind of work that isn’t tied to a specific location; that is, the kind of work that would be easy to do for a rival company with a more appealing set of policies.

I am no fan of mass layoffs — I believe the data that suggests they are nearly always counterproductive and corrosive — but if an organization must cut staff, it would be far better to make strategic choices about which departments should go, rather than simply let the people with the most options walk out the door.

The idea that five days a week in the office is better rests on several false assumptions that simply are not backed up by the data. There are both upsides and challenges to fully remote work; hybrid arrangements, on the other hand, show only upsides.

Nevertheless, there are some CEOs — a minority, but a vocal one — who keep saying remote and flexible work just doesn’t work; studies keep proving those assertions wrong. Faced with facts that don’t confirm their beliefs, even executives who usually demand rigorous evidence for major decisions simply decide to throw the data out. (Bezos in particular made Amazon synonymous with data-driven insight, and brought to the Post more attention to customer analytics.)

Here’s just some of the information they’re choosing to ignore: A 2024 meta-analysis of over 100 research papers on remote and hybrid work found such arrangements had a positive impact on “job satisfaction, organizational commitment, perceived organizational support, supervisor-rated performance, and turnover intentions.”

Sometimes such studies are criticized because of the possibility that the employees who want to work remotely may be different in some way than those who work in the office; more responsible, more self-motivating, or with more family commitments. To allay those concerns, Stanford’s Nick Bloom and his colleagues conducted a randomized experiment where employees were assigned to work in the office five days a week, or told to come in three days a week, based only on their birthdays. The randomized hybrid arrangement was so successful — reducing turnover and increasing satisfaction among both employees and managers — that the skeptical company decided to expand the program to all employees.

By ignoring such evidence in favor of a strict in-office policy, Musk, Ramaswamy, Jassy and Bezos are essentially daring their most credentialed, in-demand workers to quit. That’s a game of chicken they’re ultimately going to lose.

Sarah Green Carmichael is a Bloomberg Opinion columnist and editor. Previously, she was an executive editor at Harvard Business Review.