We’re doing vacation wrong, Neil Pasricha and Shashank Nigam write in this compelling Harvard Business Review article. The better method might be enforcing mandatory vacation periods, something the authors tried out in a small but promising experiment.

Pasricha, author of The Happiness Equation and director of The Institute for Global Happiness, and Nigam, CEO of SimpliFlying, an aviation strategy firm, drew inspiration from a TED talk called “The Power of Time Off” given by designer Stefan Sagmeister. Sagmeister takes an entire year off every seven years, and Pasricha and Nigam tested this mandatory sabbatical model at SimpliFlying, which has about 10 employees.

Staffers were required to take a scheduled week off every seven weeks, and they didn’t get to decide when they went. (The company staggered vacation times so the whole office wasn’t OOO at the same time.) Though that may seem a little controlling, the authors underscore how the imposed structure helped coworkers and clients communicate more effectively about projects and status updates and ensured employees actually left the office.

To keep people from checking in when they should have been relaxing, the company created a financial punishment for people who tried to contact the office via email, Slack and the like. If you reached out, you wouldn’t get paid for your week off. (Talk about incentive to relax.) The experiment lasted for 12 weeks, with managers rating employees’ productivity, creativity and happiness using a five-point scale both before and after the mandatory time off. They found that, on average, happiness levels rose 25 percent, productivity increased 13 percent and creativity went up 33 percent.

The experiment was designed to help fix what the authors describe as a broken vacation system. And it’s a bigger issue than people not taking their vacation time (which many employees, especially Americans, don’t do)—it’s a complex combination of logistic hassles to actually get away from the office, the pressure to stay plugged in while you’re gone and piles of work to deal with before and after your supposedly restorative and relaxing time off.

But the flipside to “traditional” vacation models, offering unlimited vacation a la companies like Netflix or Twitter, isn’t the right approach either, the authors argue. When people are given unlimited vacation days, office peer pressure kicks in. You don’t want to seem like the person who’s always gone, so the resulting social stigma might result in no one taking time off at all, despite theoretically having all the time in the world.

While the authors point out that their experiment was small and preliminary, the findings align with employees’ own reflections on their time off, which included finally finding time to do a variety of “bucket list” items like learning a language or traveling to a new location, which makes sense in terms of returning to the office more inspired than when they left.

The design wasn’t perfect—they found that the one-week-to-seven-weeks ratio was too frequent, for one—but this early research suggests that “vacation systems are broken and aren’t actually doing what they’re advertised to do,” Pasricha and Nigam write. If you go away on vacation and don’t get rest and replenishment, “you didn’t get the benefit of creating space.”

The most important thing for employees and employers is that people actually leave the office to unplug and recharge. Whatever system works best for you and your company is the one you should use, but this experiment suggest that there’s more than one way to establish an office vacation policy that works.

Read more on the Harvard Business Review.