by Gneezy, Uri
Finally, we hope you’ll come to a new understanding of how incentives can be used as a way of framing questions and gathering insights that are not only interesting, but important and useful.
We hope you enjoy the adventure.
CHAPTER ONE
How Can You Get People to Do What You Want?
When Incentives (Don’t) Work and Why1
If you want people to do what you want, incentives can be incredibly handy. When you were little and your mom promised you a toy for cleaning your room, you probably cleaned your room. And if you didn’t clean it the next week, she took away the toy until you did. Much of what we learn from the time we can say our first words is largely based on an application of carrots to reward and sticks to punish. Negative incentives in the form of punishments and fines can steer people away from undesirable behaviors. Positive incentives—often in the form of monetary enticements—can cause people to move mountains, clean up their acts, and do the “right” things.
But incentives are trickier than they seem. They are sophisticated tools, and they don’t always operate the way we think they will. Before putting an incentive scheme into place, you first need to understand how it works and then use it to understand why people behave as they do. Once we understand what people value and why, we can develop effective incentives and use them to change our kids’ behavior, motivate employees, attract customers, and even convince ourselves to do things. Field experiments are a powerful tool to understand how and why incentives work.
In some cases, incentives can even backfire, causing people to behave in the opposite way you would expect them to.
This lesson hit home some years ago when Ayelet (Uri’s wife) and Uri were late in picking up their kids from day care. Ayelet and Uri had enjoyed a beautiful day at the beach in Tel Aviv, a nice lunch, and good conversation that caused them to lose track of time. It was almost four o’clock, and they had less than fifteen minutes to pick up their daughters from the day-care center a good half an hour away. When they finally got there, their girls greeted them like excited puppies. Then they saw Rebecca.
Dear Rebecca. She was a kind, warm woman, the owner, principal, and matriarch of the day-care center. For years, she had worked hard and saved her money until she had enough to open her own center, situated in a beautiful old suburban house, about twenty minutes from Tel Aviv. Each room was colorful and filled with light, and the kids shrieked with happiness in the play yard. Rebecca had hired a dream team of teachers to look after the little ones, and the center quickly gained a reputation as one of the best in the city. She was proud of her center, and for good reason.
But when she saw Uri and Ayelet, she pursed her lips.
“I’m so sorry we’re late,” Uri ventured. “The traffic. . . .”
Rebecca nodded her head. She said nothing as Uri and Ayelet scooped up their daughters. What was she thinking? They knew she had to be upset, but how upset was she? It was hard to figure out, as Rebecca was always so nice. Uri and Ayelet felt awful for showing up late, wondering whether she might even treat their kids a little more poorly because of their tardiness.
Rebecca gave Uri and Ayelet a little insight into how she felt about their tardiness when, a few weeks later, she announced that her day care would begin charging a NIS10 (about $3) fine to parents who picked up a child more than ten minutes late. In saying this, she made it clear exactly how bad it was to be late: $3.
So how did Rebecca’s incentive work? Not very well. Since she only charged $3 for coming late, Uri and Ayelet figured that was a pretty good deal for some extra day care. Next time they were at work or enjoying a day at the beach and knew they would be late, they didn’t drive like crazy to get to the center as soon as possible. After all, they didn’t have to face the wrath of Rebecca. Now that she’d imposed a $3 late fee, they would gladly pay it and continue with what they were doing without worrying or feeling guilty.
The experience with Rebecca and her fine for tardiness inspired us to work, together with Aldo Rustichini, with ten day-care centers in Israel to measure the effect of a small fine on late-coming parents over a period of twenty weeks. First we measured what happened when there was no fine. Then, in six of the centers, we introduced a flat $3 fine for parents who were more than ten minutes late. As you might have guessed by now, the number of parents who came late increased drastically. Even after the day-care centers removed the fine, the number of parents coming late remained higher in the centers that had initially introduced it.2
So what was going on? When Rebecca imposed the fine, she changed the meaning of a late pickup. Before the fine, parents operated under a simple unspoken agreement, according to which arriving on time was “the right thing to do” for the sake of the children, Rebecca, and her team.
Yet that contract with Rebecca was incomplete. It said that parents should pick up their kids by four in the afternoon, but it didn’t specify what would happen if they failed to do so. Would Rebecca and her teachers be content to remain with the children until all the parents arrived? Or would Rebecca and her team be upset and treat the children badly as a result? We just didn’t know.
But once Rebecca introduced the fine, the meaning of the agreement between the parents and the teachers changed. The parents realized that they didn’t have to drive recklessly through traffic to arrive on time. Furthermore, Rebecca set a clear price—a low one, but a price nonetheless—on lateness. Accordingly, being late no longer involved breaking any tacit agreement. The teachers’ overtime simply became a commodity, like a parking space or a Snickers bar. The market-based incentive completed the contract: now everyone knew exactly how bad it was to be late. If you were Rebecca, you would quickly realize that imposing a fine was a lot less effective than a simple guilt trip.
Changing meaning in this way turns out to be a big deal. Let’s say that you are the parent of a teenager. You talk with your child about drugs in the hope of convincing her that taking them is bad. If you’re lucky, she listens to you. But if you have suspicions, you might demand that she take a drug test. How does this sort of demand change your relationship with your teen? You are no longer just a parent; you are also a cop. And your teen might now focus on finding ways to cheat the drug test instead of questioning the use of drugs in general.
Negative incentives in the form of day-care fines and drug tests change meaning, but of course rewards change meaning too. We all assume that offering people money will get them to do what we want. But let’s say you go into a bar after work. You meet someone attractive, and you sense the feeling is mutual. You buy each other drinks and have an interesting conversation. After a while, you say, “Hey, I really like you! Want to come back to my place?” Who knows? You might get lucky. But what will happen if you add, “I’m even willing to pay you $100”? You’ve completely changed the meaning of the interaction and insulted the other person by effectively turning him or her into a prostitute. By adding a monetary value to your interaction, you’ve essentially destroyed what might have blossomed into a nice relationship.
The Devil’s in the Details
The rub of the episode with Rebecca is that if you are going to use incentives, you have to make sure that they really work. In fact, if you use incentives that involve money, you’d better be pretty careful about the details, because incentives can easily change our perception of the relationship.
Consider the following two scenarios involving a policy aimed at encouraging people to recycle soda cans.
Scenario 1: Let’s say you live in a place where people aren’t paid to recycle soda cans. On a freezing morning you see a neighbor carrying a large bag, full of cans, on her way to the recycling center.
Scenario 2: Your town has changed its policy. Now people can receive a five-cent reward for each recycled soda can. You see your neighbor carrying a large bag of soda cans to the recycling center.
What do you think of your neighbor in Scenario 1? In Scenario 2?
In the first scenario, you probably th
ink that your neighbor is an environmental steward—a citizen of high character, doing her part for the environment.
But once the small, five-cent-per-can reward is in place, you might think that she is either cheap or really down on her luck. “Why,” you might ask yourself, “is she going through so much effort for such a small compensation? Is she a miser?”
In fact, the five-cent incentive might have changed the meaning of what your neighbor thinks she is doing. Before the policy changed, her can-collecting was all about protecting the environment. But after the change, she might be aware that she looks cheap or desperate. “What’s next,” she might say to herself, “Dumpster diving? In that case, my can-collecting ain’t worth it.” Given this shift in her self-perception, she might eventually stop recycling.
Another example of how using money as an incentive can backfire took place during Israel’s widely publicized “donation days.”3 Every year, high school students go door-to-door to collect donations for a charitable organization supporting, say, cancer research or aid to disabled children. On average, the more houses the students visit, the more money they collect.
Our experimental objective was to determine whether a monetary incentive would increase the amount collected and, if so, how much money it would take to maximize the students’ performance. So we divided 180 students into three different groups (none of the participants knew that they were taking part in an experiment). The first group listened to the leader talk about the importance of the donations to the charity, explaining that the charity wanted to motivate them to collect as much money as possible. For the second group, the leader added that each student would receive a token 1 percent of the amount he or she collected (we made it clear that the bonus would not come from the donations). This 1 percent added an external monetary motivation to the intrinsic do-good one. The third group was told they would be given 10 percent of the donated amount.
The group that garnered the most donations was the one receiving no payment at all. Basically, this group wanted to do good for others. But apparently, when the monetary compensation was introduced, the students in the other two groups stopped thinking about the good they were doing, and concentrated instead on a simple cost-benefit calculation with respect to their monetary payments. The group offered the 10 percent payment came in second. Those offered 1 percent garnered the fewest donations. Why? Because in this case, the money didn’t support the intrinsic, do-gooding incentive—instead, like Rebecca’s day-care fine, it crowded out the higher motivation. That is, the money became more relevant than the desire to do good.
When you’re deciding whether to motivate someone, you should first think about whether your incentive might crowd out the willingness to perform well without an incentive (to help the environment by recycling soda cans, to help with cancer research, and so on). Crowding out could occur because of a change in the perception of what you are doing, or because you have insulted the person you are trying to encourage or discourage. When you decide to take the incentive route, you should make sure that the incentive is large enough to reap gains. Think of an incentive as a price. If you charge a lot (for example, if Rebecca had charged late parents, say, $5 per minute, as occurs in some places in the United States), people will be more likely to behave the way you want them to. So, the moral of the story is to either pay enough, or don’t pay at all.
Cash, in the end, really isn’t king; some things can’t be bought. Rewarding people on the basis of what they really value—their time, their self-image as good citizens, even candy—is often much more motivating than just slapping down, or taking away, a couple of bills. In short, not all incentives are created equal.4
I’ll Have What She’s Having
Incentives can influence behavior in other strange ways too. Consider, for example, what happens in an episode of the sitcom Friends when all the friends go out to dinner at a nice restaurant. Monica, Ross, and Chandler, who make decent livings, order full-course dinners with all the trimmings, but Rachel, who doesn’t earn much money, only orders a side salad. Phoebe, similarly short in her bank account, orders a cup of soup, and Joey, who is no trust-fund baby either, chooses a miniature pizza. When the bill arrives at the end of the meal, Ross announces they will split it—and the final tally comes to $33.50 a piece. A pall falls over the table. “No way,” Phoebe resentfully retorts. So much for the nice evening out with friends.
Splitting the bill makes a lot of sense on the surface: after all, sitting around figuring out exactly who ate what and how much sales tax they owe is an unpleasant way to end an otherwise nice experience. Indeed, in some cultures it’s considered pretty gauche to do so. In Germany, diners will figure out the price of their individual bills to the last cent, and no one feels bothered. But in Israel, and in many places in the United States, such behavior is considered rude. When a group of people jointly enjoys a meal at a restaurant, there is often an unspoken agreement to divide the bill equally. So how does splitting the bill really affect behavior?
We conducted a study to see what would happen when different groups of diners—students who didn’t know each other—were faced with different ways of paying the bill.5 We divided our participants into three types of groups and changed the way they paid for the bill. In one case, six diners (three men and three women) paid individually; in the second, they split the bill evenly. In the last case, we paid for the whole meal. How did the payment scheme affect what each person ordered?
Now, imagine you are one of six students going to lunch in our experiment, and you are told that you are going to split the tab with the other five. You’re pretty hungry, so you order a lobster roll ($20), a side of fries ($3.50), and a beer ($5). The person sitting next to you isn’t very hungry, so she just orders a salad ($8) and an iced tea ($2.50). After you all eat your lunch, you and a few others at the table decide to top it off with a piece of pie ($4) and a cappuccino ($5.50), whereas others abstain.
Then the waiter comes along and delivers the bill: the total comes to $125, including tax, and tip, which means each of you pays $25. This is no problem for you because, if you had paid individually, your share would have come close to $40. But it is a problem for the woman who only ordered $10.50 worth of food.
It turns out that the way you split the bill actually affects what you order. We found that people ate the most when we footed the bill for the whole meal. No surprise there. But when it came to the bill-splitting group, people tended to order more expensive items than they did when each person paid for his or her own meal. You have to wonder about the people who “ordered up.” They weren’t “bad” people who took advantage of others; they just reacted to the incentives they were facing. After all, for every extra dollar they ordered, they had to pay only one-sixth of the cost. So why not order the $20 lobster roll, if all you have to pay for is less than $4 extra? Of course, there are no free lunches (apart from those in our experiments). Someone has to pay for the other $16 for the lobster roll.
This is an example of a “negative externality”—that is, someone else’s behavior that affects your well-being. Let’s say you’re a non-smoker, and a smoker sitting near you decides to light up. He enjoys his cigarette, but you are also “consuming” his smoke. The guy smoking has bestowed a negative externality on you. Simply put, the party consuming the good is not paying all of its cost. In a bill-splitting situation, the person enjoying the large, expensive lunch while others consume less is doing the same thing. People simply react to the incentives they are facing.
What Works?
Throughout this book, we look at significant issues such as discrimination, gender and education gaps, charitable fundraising, and business profitability. The lesson that recurs is this: incentives shape outcomes. But it’s crucial to set them up right and to finely tune them to match the underlying motivations of people.
Consider, for example, what it might take to get people to lose weight. The last decade has seen a dramatic rise in obesity in the United States. And
obesity is a major risk factor for heart disease, diabetes, and other health problems. Can incentives help people get control of their weight?
After yet another holiday season of too much food and drink—all those Christmas cookies, fried Hanukah latkes topped with sour cream, that New Year’s Eve champagne-and-caviar extravaganza—you look in the mirror, stand on the scale, and see that you’ve reached what might be generously referred to as “critical mass.” You have to cinch your belt more loosely. You feel pretty guilty about this, and you vow to slim down.
The local gym is offering a discount on an annual membership, so you forgo the pay-as-you-go option of $10, and sign a year-long contract. If you are like many people, you go to the gym a few times in January, less frequently in February, and not so often after that.6 You have several reasons (excuses?) for absenting yourself: you lack the time; you are too embarrassed to show up in your Spandex with that belly roll; you’re out of shape anyway, so you aren’t able to exercise vigorously; maybe you just plain don’t like sweating. Because you don’t go to the gym more than a few times, your decision to pay the annual membership fee costs you a lot more money than simply using the pay-as-you-go option would have.
Your failure to stick to an exercise routine once you’ve signed up for an annual membership could be because you were too optimistic—you truly believed you were going to exercise much more than you did in the end. Another, more sophisticated, explanation is that you have “played a game with your future self.” That is, you have a hunch that you may be less willing to exercise in the future. You know that with the current pay-as-you go option, you have a choice. You imagine, for example, that you could use the $10 for a one-time visit, or you could go to a movie instead of the gym. You have a feeling you’d choose the movie. So you pay the annual membership now in order to reduce the perceived cost in the future. If you pay now, you figure, then later on you will not give your lazy, future self another reason (saving the $10) to not go exercise.
