The Secret to Weight Loss, According to Behavioral Economics
Is there any way to upset the set point?
Posted Dec 13, 2017
Losing weight is incredibly difficult. People who have put on unwanted pounds need all the help they can get to trim their waistlines. Will behavioral economics come to their rescue?
Behavioral economics builds on the insights of economics on the rational ways people respond to incentives by incorporating principles of psychology, especially ones relevant to the unconscious or even irrational forces that influence our behavior. For example, a straightforward economic approach to helping people lose weight would, say, pay people $25 for every pound they lose. According to economic theory, if you choose the right price, you should be able to get everyone to be as trim as you want them to be (barring some kind of health condition that prevents weight loss). You see, according to economic theory, at some point the benefits of money will outweigh the harms of having to lose weight–the pain of dieting and, god forbid, exercise, therefore causing a rational person to diet and/or exercise in order to lose weight and make the money.
By contrast, a behavioral economics approach might tweak that financial incentive, to trigger less rational psychologic impulses. For example, Kevin Volpp and colleagues at the University of Pennsylvania have conducted a slew (even a gaggle?) of studies where they present people with financial incentives in the form of lotteries. In one study, some people got a small amount of money every time they took their medications while others instead got enrolled into an actuarially equivalent daily lottery (with the same average amount of money paid out per person per day). People randomized to be in the lottery group were more likely to take their pills. The psychology of the lottery increased the economic impact of the financial incentive.
In a couple recent studies, Volpp and colleagues tested these behavioral economic interventions out on overweight and obese people. They ran experiments focused either on weight loss or exercise, randomizing people to one of several incentives.
In their exercise study, Volpp’s team told people they should try to walk at least 7,000 steps per day. They randomized people to one of four groups, all of which got daily feedback on how many steps they had walked.
Control Group: Only received daily feedback
Gain Group: Received $1.40 each day they met the goal
Loss Group: Lost $1.40 each day they failed to meet the goal
Lottery Group: Eligible to win either $5 or $50 each day they met the goal
Which group do you think walked the most? It was the loss group, who were so motivated to avoid losing money they walked more than everyone else. Here is a picture (Figure 2) of the results.
The other two intervention groups appeared to do a little better than the control group, but these improvements were not statistically significant. The people in the loss group, however, had a legitimate improvement. Building on one of the most durable findings in behavioral economics–the psychologic power of loss-aversion–the studies showed that the prospect of losing money hurts enough to motivate behavior.
Sadly though, as the figure shows, once the interventions stopped, people quickly reverted to their earlier behavior. Three months of walking every day weren’t enough to create a walking habit, it seems. More on this problem later. But first let me tell you about the second study, one that tried to incentivize people to lose weight.
Volpp’s group randomized people in the weight loss study to one of four groups:
- Control Group: No financial incentive
- Delayed Premium Adjustment: Lose 5 percent of weight, and the next year they’d save $550 on their insurance premiums
- Immediate Premium Adjustment: Lose 5 percent of the weight and they’d immediately receive equivalent reductions in their premiums
- Daily Lottery: These people were given a daily weight target. When they met that target, they were eligible to win either $10 or $100, once again actuarially equivalent to the premium adjustment groups.
As it turns out, people in the lottery group were way more likely to weigh themselves regularly than others, given the importance of daily weighings in making them eligible for their payments. Here is a picture of that result:
Once again, this picture shows that behavioral interventions often fade quickly. Sigh…
As for weight loss, none of the intervention groups were any more likely to meet the 5 percent weight loss target than the control group, with fewer than one in five people meeting that goal. Here is a picture of that result, shown in the last row of data:
What’s going on here? How come these lotteries failed, even though earlier research had shown lotteries to be successful at encouraging people to take their pills?
Volpp and colleagues offer an explanation. In their earlier studies, when people didn’t take their pills they were still enrolled in the lottery. If they won that day they received a message explaining how much money they would have won that day had they taken their pills. This “desire to avoid regret” can be very motivating. But in the weight loss study, Volpp’s group did not have access to people’s work schedules, and therefore could not tell whether people failed to weigh in because they failed to weigh in, or, instead, because they had the day off from work. Therefore, they didn’t send messages telling them that they missed out on the lottery for failing to weigh in.
The only problem with this explanation is that in the first study, the 7,000 steps study, Volpp’s group did give people feedback on whether they would have won the lottery on any given day had they met their goal. And still, the lottery failed.
Clearly we have more to learn about when lotteries do and do not motivate behavior change.
But I do think we can draw a lesson on how to use behavioral economics to lose weight. Unfortunately, the lesson is a disappointing one. Behavioral economics has not proven itself up to the task of helping people lose weight and maintain that weight loss. That is no criticism of behavioral economics. As a behavioral scientist, I frequently draw upon the insights of behavioral economics in my own research.
The real problem here is not economics or behavioral economics or psychology. It is obesity. Losing weight is insanely hard to do. And keeping weight off is even harder.
My guess is that no combination of behavioral interventions will ever be up to the task of helping adults lose weight and keep that weight off. Instead, we probably need to wed these behavioral interventions to some kind of biologic treatments, some kind of medical interventions that affect metabolism or that influence people’s neural reward systems. We need to find ways to help people not only lose weight, but that also help them avoid regaining that weight. It is time to shift our efforts from behavior change to habit formation, from weight loss to sustained weight loss!
*Previously Published in Forbes*