When it comes to changing habits, losing something you've been given may be a more powerful incentive than getting a gift in the first place.
The latest in a series of studies from behavioral economists at the University of Pennsylvania reported an example on Monday: promising people a monthly pot of money to increase their physical activity, but then deducting some every day they missed the goal led to a 50 percent jump in successful days - far better than simply paying them to exercise more.
More than 80 percent of large employers offer some kind of incentive to get workers to take up healthier habits, on the theory that this will improve well-being and reduce health-care costs.
But the trend has grown faster than research into what works best. Commonly offered "hidden" incentives, such as lower insurance premiums the next year, don't "engage" employees enough for them to change their lives, said Mitesh S. Patel, lead author of the new study and one of 50 faculty affiliated with Penn's Center for Health Incentives and Behavioral Economics.
With a spate of studies in top medical journals, the center has become the national leader in research to find incentives that work.
Its latest, published in Annals of Internal Medicine, compared incentives to increase physical activity among overweight and obese adults. The goal was 7,000 steps per day (the U.S. average is 5,000) for 13 weeks. After the incentives ended, the 281 participants, all Penn employees, were tracked for an additional 13 weeks.
They were randomly assigned to four groups:
A "gain incentive" paid $1.40 for each day the goal was achieved.
A "lottery incentive" offered the possibility of a daily $50 reward that could be collected only if the goal had been met the previous day.
A "loss incentive" allocated $42 a month up front but deducted $1.40 for each day the goal was not reached. (The dollar amounts for each type of incentive were designed to be roughly equal.)
A control group received daily feedback from a smartphone app, as did everyone else.
All three incentive groups did better than the controls. But the "loss incentive" produced a far better improvement: 50 percent more of that group's members took at least 7,000 steps a day.
The design of each incentive was shaped by earlier research, much of it from the center at Penn.
Members of the "loss" group were assigned $42 to start because "a lot of people don't want to put down their own money," said Patel, the study team's leader and a physician. Checks for that amount minus the daily deductions were mailed each month to draw people's attention to their winnings.
Even though participants were not using their own funds, Patel said, the setup "made it like the money was theirs to lose."
Patel, whose research interests combine behavioral economics and wireless technologies, had previously found that smartphones might have more potential than other devices to improve health.
The tech industry has had high hopes that "wearable devices" such as the Fitbit would lead to better health while generating big profits. But it's expensive to buy another device, and a hassle to learn how to use it, keep it charged, and remember to carry it.
"Two-thirds of Americans already have smartphones and take them with them," Patel said.
His new study relied on Moves, one of the earliest smartphone-based activity trackers.
But ease of use - participants set the app to run in the background and went about their business - likely accounted for a finding that was not part of the study's goal.
Three months after the incentives ended, 96 percent of participants were still getting daily feedback on their performance via email, text, or voice call.
In the eight years since it was founded, the Center for Health Incentives and Behavioral Economics has published dozens of studies on how to influence behavior and health. It is one of only two in the field - the other is at Harvard - funded by the National Institutes of Health.
The center's director, Kevin G. Volpp, a coauthor of the new exercise study, takes particular pride in findings that had an immediate impact in the real world.
A study published last year in the New England Journal of Medicine examined how CVS Health employees responded to different financial incentives to quit smoking.
It led to a program that offers workers who smoke hundreds of dollars but also requires that they deposit $50 to start - because the study found people do better when they have skin in the game.
Several years ago Volpp did a project with insurance giant Humana that examined Americans' comprehension of health coverage plans. No surprise, most of us don't have a firm grasp of insurance concepts.
"The neat thing about this study was it found that people understand co-payments better than any other kind of cost-sharing," Volpp said. Based on that, the authors recommended a simplified plan, which the insurer designed.
"Humana Simplicity uses only co-payments. Most plans have 100-plus pages of benefits," Volpp said. "Humana created a description 11/2 pages long."