by Gneezy, Uri
What kinds of incentives would it take to get these two under-achievers to succeed in school? Would paying Urail and Kevin, or their parents, to perform make sense? Before rejecting this idea out of hand, let’s consider how we usually get people to do what we want them to do. To get people to recycle more or to buy environmentally friendly cars, we reward them with financial incentives. Could paying students to perform better work, too?
When we took our idea about paying students for performance to the school board of Chicago Heights, it was met with something close to scorn. After all, most grown-ups agree that students should learn for the sake of learning. But the brutal fact is this: millions of kids in public schools just do not see it that way. As we pointed out to the school board, kids should clean up their rooms, but they don’t. They should brush their teeth and always obey their parents, but they don’t. They should eat fruit instead of cookies, but they don’t. And they should like to learn, but often, they don’t.
The realism of our argument hit home when a school board member cited research showing that extrinsic incentives, like money, can crowd out intrinsic incentives like enjoying learning and doing well in school. (Sound familiar? He was basically citing some of the psychology and economics research, including our own, we discussed in Chapter 1.) We agreed on the importance of intrinsic incentives and with the spirit of those studies, but we swiftly replied that when there is nothing to crowd out, money talks. We were met with a sigh from the board members who, after all, knew that their district was in dire straits. They grudgingly admitted that they were willing to try anything that stood half a chance of succeeding.
Because monetary incentives in education are so controversial, it’s not yet fully known how best to direct them.3 Our first idea was to shift the incentives to the near term: rather than paying students at the end of the semester or year for good performance, we would pay them closer to the moment of achievement, thus satisfying their desire for immediate gratification. (As we said above, behavioral economists have shown that many people respond much more dramatically to incentives paid earlier rather than later.)
Our second behaviorally based idea was to use a lottery to pay the students. A lottery is a terrific behavioral testing tool, because human beings tend to overweigh low-probability events. For example, the chance of winning a state Powerball lottery is typically lower than one in a million, but people love to play anyway, in part because they believe the odds are better than they really are. (In reality, you’re more likely to be struck by lightning than to win a Powerball lottery in most states.) We thought that if we could offer rewards through a lottery, where the prize was large but the chances of winning were small, the rewards would seem more relevant; students might overweigh their chances of winning the lottery, inducing them to try harder.
The final idea was an obvious, whack-on-the-side-of-the-head one that stems from trying to figure out what goes into the education “production function”: use incentives to get parents involved, and see how their involvement affects their kids’ performance. We figured that paying parents would not only work, but that doing so would tell us more about the most effective ways to increase achievement. Getting parents more involved in helping their children study might help siblings, too. After all, once parents start working with one child, it’s only fair to do the same with the others.
We just had one problem, and it was a big one. Pulling off a field experiment to test these ideas would not be easy. And, it would cost large sums of money that we simply didn’t have.
The Griffin Gift
Right around this time—in the spring of 2008—we received a serendipitous phone call from a couple of philanthropists, Kenneth and Anne Griffin. Kenneth Griffin is the founder of Citadel, one of the world’s largest hedge funds, and he and his wife were interested in our research. They were looking for help in setting up a charitable foundation, and wondered whether we could possibly meet with them to discuss our work. We had no idea the call would change our lives.
We drove to the Citadel building in downtown Chicago, a gigantic, steel-and-glass tower, with 1.4 million square feet of office space, set right in the economic center of the city. After passing through the marble-walled lobby, we entered the elevator and pressed the button for the thirty-seventh floor. Our ears popped, and we felt a bit nervous. The elevator hushed open, and a nice receptionist ushered us into a tastefully decorated conference room. She offered us coffee, and we waited.
When the Griffins entered the room, our first take was that they looked like one of those gorgeous couples whose nuptial photos you find gracing the pages of the New York Times’ Sunday Styles section. Kenneth, good-looking and incisive, is a brilliant entrepreneurial type; a product of public schools, he learned all about trading within the confines of his college dorm room. Anne, a French native who speaks five languages, is, like her husband, also a product of a public school system, and her mother was a teacher.
We had little idea what we had gotten ourselves into. Most well-intended, wealthy donors we knew wrote a big check for research and, with a flourish of the pen, said something like “You can talk about your results at my next dinner party.” But the Griffins were different.
We launched into some theories of behavioral economics, summarized a bit of our research, and walked through our ideas about what kinds of incentives might work with the Chicago Heights school kids. As we talked, their eyes lit up.
Even though a few hours of the Griffins’ time is probably worth tens of thousands of dollars, they spent a long time meticulously working through our experimental ideas, surprising us with their knowledge and insight. “Why do you think people overweigh small probabilities?” they asked us. “Why do you think so many young people don’t think graduating matters?” Both Kenneth and Anne grilled us and sharpened our ideas with thoughts of their own. Like us, they wanted whatever interventions we came up with to be scalable, firmly grounded in theory, and cost-effective.
The Griffins soon became our full research partners. They passionately believe in bettering America’s public education system, understanding that doing so is the only way to improve people’s lives and the economy in general. They wanted to get up to their elbows in interventions that could help urban kids overcome the education gap and raise American educational standards in general.
By the time we left the room, we were convinced of one thing: had they gone the academic route, both Ken and Anne would have been our research equals and then some. We departed with a solid experimental design on our hands, and within twenty-four hours, the Griffins gave us the initial $400,000 we needed to run the experiment.
Before we entered the room that day, Ken and Anne knew how they wanted to change the world; we were fortunate enough to show up in the right place at the right time. Suddenly we understood how Columbus must have felt when Queen Isabella gave him the resources to find the New World. We not only found donors, but two new friends; indeed, our new colleagues would help us tackle one of the most important problems facing America today.
The Voyage to Public Schools
One day, a kind, slim redhead with the gentle demeanor of a good high school counselor called Kevin Muncy into her office. Her name was Sally Sadoff, and at the time she was one of our graduate students administering our experiment.4 When Kevin came in, Sally smiled broadly at him. “How’s it been, starting in a new school?”
“I like it. It’s really easy.”
“The classes are really easy? Let’s see what your report card says.” She scanned Kevin’s awful grades. “So Kevin,” Sally asked kindly, “what do you need to improve on?”
“Everything.”
“So you probably want to know what you can get if you meet the new monthly achievement standards—no unexcused absences, no daylong suspensions, and grades of C or higher in all your classes. Right?” She pulled out a folder and handed it to him.
Kevin opened the folder. “Fifty bucks?”
Sally smiled. “And you get $50 e
very month as long as you keep your grades up!”
“I think a lot of people will start doing their homework then.”
“But what about you?”
Kevin began to dream a little. “What could I do with $50 a month? I could pay for my skateboards. Get sponsors and clothes and stuff until I graduate.” On hearing about this incentive, Kevin’s mom doubled it: if he raised his grades to the monthly standards, he could earn $100 a month.
But there were more incentives, and Sally made a big deal of them. In fact, we pulled out all the stops we could think of. At the end of each month during the eight-month program, the kids all lined up in the school cafeteria for free pizza and the big payout. Each one was called up to a table where Sally and the other researchers looked over their grades and talked with them. If they (or their parents, depending on the experimental treatment) won the cash payout, they walked away grinning—and not just because of the cash.
Even more fun was the big, suspenseful bingo-ball-style lottery. Each month, we drew ten names. If a student who met the standards criteria won, he or she (or their parents, depending on the treatment) would take home the grand prize: $500 in cash (as well as a giant Ed McMahon–style pseudocheck) plus a ride home in a white, chauffeur-driven Hummer stretch limo, complete with comfy leather seats, tiny blue and green interior lights, TV consoles, ice compartments, and all the other trimmings. When Urail King saw the limo, he went wild. “Oh my G-o-d!” he shouted. “This is awesome! Oh yes, yes, you are getting straight As from me! Take me home, Jenkins!”5
If the kids didn’t meet the monthly standards, Sally and the other researchers would make suggestions for catching up. The researchers even gave the students reminder calls during the month to ask them how they were doing in their classes. And, of course, the parents encouraged the kids and worked with them, too. After all, who would not want their kid to win the grand prize?
So how did the students and their parents respond to all these expensive incentives? Given the wiring of teenage brains (“I want what I want now”), was it too much to ask the students to wait a month to receive their rewards?
Our overall results showed interesting gains.6 We estimated that the program helped about 50 borderline students out of the 400 in the experimental group to meet the ninth-grade achievement standards. Among the students who were on the brink of failing, we figured the program had increased achievement by about 40 percent. Happily, these students continued to outperform their un-incentivized peers after the program ended in their sophomore year. In fact, our estimates suggested that about forty kids who would otherwise have dropped out would receive their diplomas because of our program. (We also found that students’ performance increased slightly more if their parents, rather than they, received the reward.)
Given that every additional year of secondary schooling increases lifetime earnings by 12 percent, offering such students an incentive during their freshman year seemed to be a clear, cost-effective intervention. If you also count the fact that the kids spent their time in school, rather than dropping out and hanging around the streets, the program was that much more successful. We had found a way to reach a slice of the kids on the brink—but only a slice.
Reframing Achievement
Tom Amadio was impressed with these results, but he pressed a question on a different front, beyond keeping kids in school: Could we increase the test scores of his students? After all, test scores are important door-openers, and are tied to future outcomes like years of education and high-paying jobs. Test scores also determine how much money a school district receives from city and state governments. Unfortunately, at the present time, minority students just cannot seem to catch their white counterparts when it comes to test scores. The racial achievement test gap remains both considerable and stubborn, and many urban schools fail at their mission to close it.
To answer Amadio’s challenge, we decided to run another set of field experiments that involved over 7,000 students in a variety of elementary and high school settings in Chicago and Chicago Heights. These tests took place in the schools’ computer labs, where students took a standardized test three times a year.7
As an introduction to our experimental premise, perhaps you remember the images of two young girls who were gymnasts at the 2008 Summer Olympics. Both were winners. As the girls stood on the podium, each of them was overcome with intense emotion. And no wonder: they had both trained for years for this moment, sacrificing normal lives as teenagers to reach the very apex of gymnastic performance. The photos were taken after they received their respective medals. One was decorated with the silver medal, the other with the bronze. When their photos were published in the press, one was beaming, the other appeared to be holding back tears.
Which do you think won the silver and which the bronze?
We all know that silver is better than bronze, but context is everything. The silver medalist who missed out on the gold was devastated, and her face looked as if she were sucking on a sour lemon. But the bronze medalist who just barely made the podium was clearly ecstatic.8
Over the past forty years two psychologists—Daniel Kahneman and Amos Tversky—have revolutionized our understanding of the importance of human emotions like sensitivity to context in everyday choices we make. One of the things these two “fathers” of behavioral economics have shown is that the way humans understand the world has to do with the way we interpret (or “frame”) phenomena. Depending on how you frame something when you speak, you influence someone’s behavior in various ways. A parent might say to a child, “If you don’t eat those peas, you won’t grow up big and strong.” (That’s what behaviorists call “loss framing”—it frames a statement as a loss or punishment.) Alternatively, the parent could phrase the same thing in a more positive light and say, “If you eat your peas, you will grow up big and strong.” (That’s called “gain framing”—it frames the statement as a benefit or reward.)
Imagine you are a thirteen-year-old boy coming into the computer lab to take a standardized test. It’s a nice fall day, and you are restless, a little hungry, and all you can think about is that last round of your favorite video game and the pretty girl sitting at the desk behind you. You wish you were anywhere but stuck in this stupid lab to take another stupid test.
In walks the school’s assessment coordinator, Mr. Belville, who asks for everyone’s attention. (Mr. Belville also happens to be the school’s reading coordinator and head of the school’s technology department; he is the sort of overqualified and overdedicated administrator that single-handedly makes a school run.) The process of just getting the students to stop talking takes a minute, but finally they quiet down.
“Today,” Mr. Belville announces, “you’re going to take the next level of the standardized tests that you took back in the spring. But this time we’re going to be doing something different. If you do better on today’s test than you did the last time you took it, you’ll receive a reward of $20.”
Your eyebrows shoot up. So do everyone else’s. “Awesome!” someone yells. Suddenly everyone starts chattering at once. Mr. Belville immediately quiets the room.
“Now, before we begin the test, I’m going to be handing each of you a $20 bill,” he continues. “I want each of you to fill out this receipt confirming you’ve received the money. On the receipt sheets, I want you to write a little bit about what you plan to do with the cash. You will keep the money in front of you on your desk while you take the test. Remember, you will get to keep the $20 if you improve on the test. But if you don’t improve you will lose the $20.” He passes around the receipt forms and the $20.
You dutifully fill out the form and think about what you want to do with the $20, which you would like to put toward a new skateboard. You write down your dream on the form, and then place the $20 to the right of the keyboard, just above the mouse. You smile as you look at it. “My wheels,” you think. You imagine walking into the skateboard store and plunking down your money.
Mr. Belville
returns to the front of the room, interrupting your daydream. “We will begin the test in two minutes. Please sign in on the computer.”
You sign in, and the clock advances. You watch the second hand. You can’t wait to start.
“Ready? Begin!”
Now, when you’ve taken these tests in the past, you usually whiz through them because you really don’t care about them—you think they are pretty pointless and you leave many questions blank. But this time, with your $20 sitting in front of you, you take your time. Some questions stump you initially, but instead of guessing and moving on, you start to really think about what the best answer might be.
At the end of an hour, Mr. Belville announces that the test session is over. You’re the last student still working on the exam. You answer the last question and hit “submit.” Almost immediately, your score shows up on the teacher’s computer screen. Once the whole class finishes up you can see how well you did compared to last spring’s test.
So how did you do?
In this field experiment, we actually divided the schoolkids into one of five groups. As described above, kids in one group received a $20 bill and were told if they did not improve on their previous test score, we would take the money away. This is what we described above as the “loss” group: the kids had the $20, and stood to lose it by not achieving on the test.
Students in the comparison, “gain-framing” group were told that if they improved on their previous test scores we would give them $20 immediately after the test, but they did not receive the $20 beforehand. Because they didn’t have the $20 directly in front of them, they stood to gain.
