Should we pay people to break bad habits or to form good ones? Does it matter whether we call this payment a reward or a penalty? How about asking people to bet on themselves, letting them win money if they succeed and lose money if they fail?
CVS and researchers at the University of Pennsylvania explored these questions in a smoking cessation study. The study randomly assigned about 2,500 CVS employees and their friends and relatives into five groups. Two groups were offered an $800 reward if they quit smoking. Participants in the first group were enrolled as individuals and participants in the second group as six-member teams. Two other groups were asked to place a bet on themselves by risking $150. If they quit smoking, they got their $150 back plus an additional $650 bonus, more than quadrupling their money. If they continued to smoke, they lost $150. Once again, participants were enrolled in one of these betting groups as individuals and in another as six-member teams. The fifth group was the control group. This group was offered a regular smoking cessation program with free information and smoking cessation aids but without a financial reward or penalty.
The researchers wanted to answer the following questions. Will people be willing to bet on themselves and sign up for a program that required them to make a deposit up-front if they knew they might lose this money? Who would be more likely to quit smoking: people who hoped to get a reward, people who made a bet on themselves that could go either way, or people who were offered information and help without a financial incentive? And does it matter whether people do any of this as individuals or in teams?
The researchers found that it didn’t matter whether people participated as individuals or as teams. The only thing that mattered is whether participants were in the reward groups, the betting groups, or the control group.
Many fewer people were willing to bet that they would quit smoking if there was a possibility of losing money than were willing to try to quit in order to get a reward. Less than 14 percent of those randomly assigned to betting groups agreed to participate in the study. Over 90 percent of those assigned to reward groups agreed to participate. The answer to the first question is that people prefer to try for a reward without risking their own money.
The answer to the second question is more complicated. More people in the reward group quit smoking than in the betting group. After 12 months, more people in the reward group remained tobacco-free than either in the betting group or in the group that got free information and quitting aids without a financial incentive. However, in percentage terms, more of those who bet on themselves quit smoking, even after the researchers adjusted the results for the possibility that those in the betting group were different from those in other groups. Because a larger percentage of people in the betting group quit smoking, the researchers concluded that the possibility of losing $150, what economists call “loss aversion,” motivated people to quit more than a reward did. They explained the results this way:
As a result of this study CVS redesigned its smoking cessation program so that everyone who wants to participate must now bet on their own success or, as some prefer to call it, “have some skin in the game.” CVS reduced the amount of the bet to $50 and raised the amount that can be won to $700. Nevertheless, there is no longer an option to participate in a program that just offers a reward or information and support. People who sign up for CVS’s smoking cessation program must pay $50 and will lose this money if they do not pass a nicotine drug test at six and 12 months.
Smoking is the leading cause of preventable death and illness. There is no question that people improve their health when they quit smoking. CVS stopped selling tobacco products because, as a health company, it was unwilling to make money from products that damage health. This is admirable. CVS’s desire to help its employees quit smoking is admirable as well.
However, I question the new design of the smoking cessation program. Study participants made it clear that they much prefer a program that offers a reward to a program that requires them to place a bet. They signed up in much larger numbers and quit smoking in larger numbers than those assigned to a betting group. Importantly, more people in the reward group remained tobacco free after 12 months than those in the other groups. In fact, there was no difference at 12 months between those in the betting group and those offered information and support without a reward. Nevertheless, CVS redesigned its smoking cessation program to force all employees into a betting regime. In doing so, the company relied on the statistical analysis showing that betting programs would have been more effective “among people who would have accepted either” a reward-based program or a betting program. Although the study clearly demonstrated that this is not the case, CVS decided to treat everyone as if they were indifferent between getting a reward and placing a bet.
Cass Sunstein, one of the economists who popularized the notion of “nudging” people’s behavior by framing their choices, wrote an editorial about this study. He said, “The challenge is to find a way to nudge people to enroll in such [betting] programs. If that challenge cannot be met, reward programs are much better bets.” CVS met the challenge by taking away participant choice altogether. The company is no longer nudging participants. It is simply telling them that they have to pay $50 up-front in order to participate in the smoking cessation program. CVS did this because, as an employer, it can.