by Gneezy, Uri
“This is the smoking gun,” he said proudly. “It shows a clear positive relationship between ads and sales. When we placed 1,000 ads, sales were roughly $35 million. But see how sales dipped to roughly $20 million when we placed only 100 ads?”
To see why the relationship between ad placements and sales might not be as clear-cut as the executive believed, take a look at a similar figure that we have produced:
This second chart shows two very different phenomena: the number of drowning incidents from 1999 to 2005, and the number of retail ice cream cone sales (in millions) from one of the biggest ice cream companies in the United States over the same time period. Of course, it’s shocking to see such a relationship between these two variables.
Parents persuaded by charts like this might believe that the correlation is causal, and never let their kids eat ice cream near open water. But, of course, there is a hidden third variable lurking in the background. In the summertime, people eat more ice cream and swim more. More swimming leads to more drownings. Even though people do eat more ice cream cones in the summer, eating ice cream doesn’t cause people to drown. Swimming does.
So what was the hidden variable lurking in the background of the chart the marketing executive showed us? We learned later that the retailer placed a lot of ads during the November and December holiday shopping season when, not surprisingly, the company sold a lot of products. This gave the illusion that ads and sales were related causally. But when we dug deeper into the data and took account of the fact of when the ads were placed, we found no causality in the data—just correlation. Consumers bought more products because of the holidays, not necessarily because of the retail ads.
Our world is beleaguered by mistakes like this. In cases where we think a causal relationship could exist, it’s easy to mistake simple correlations with causality. In so doing, we stand to waste a lot of money and effort for nothing. The problem is that the world is filled with complicated relationships, and it’s difficult to find true causal relationships.
Then there’s the current trend “big data.” By gathering mounds and mounds of data and observing the patterns, people using big data can draw interesting conclusions. Big data is important, but it also suffers from big problems. The underlying approach relies heavily on correlations, not causality. As David Brooks has noted, “A zillion things can correlate with each other, depending on how you structure the data and what you compare. To discern meaningful correlations from meaningless ones, you often have to rely on some causal hypothesis about what is leading to what. You wind up back in the land of human theorizing.”2
The other problem with big data is that it is so big that it’s hard to find your way in it. Companies have so much data that they don’t know what to look at. They collect everything, and then become overwhelmed because they have so many possible permutations of variables of interest that they really don’t even know where to start. Because our work focuses on using field experiments to infer causal relationships, and because we think hard about those causal relationships of interest before generating the data, we go well beyond what “big data” can ever deliver.
Fortunately, field experiments can provide the kind of hard data that citizens, educators, philanthropists, policy makers, and CEOs need in order to not only avoid making big mistakes but also to develop a better understanding of the people they are supposed to serve: What really motivates people and why?
What kinds of incentives cause people to do the “right” things? When do incentives in the form of punishments and sanctions steer people away from undesirable behaviors? And when do incentives just plain not work?
As economists, we clearly believe that there’s more to motivation than meets the eye, and that when one does find a causal relationship between variables, the implication can be profound. In fact, incentives are not simple blunt instruments. Hidden motives are actually very complex, and they don’t always operate the way we think they should. Until one fully understands what incentives motivate people, it is impossible to predict how new policies, or changes, will actually work.
In this book, we show the many ways that incentives can work to change ourselves, our businesses, our schools, and the world for the better; but before we try to apply them, we need to understand how these incentives change our hidden motives.
We3 are also fueled by our personal interests and passions. For example, consider how we got interested in the question: Why do people discriminate against each other? It wasn’t just because discrimination hurts society in general, or because it is a murky issue that has vexed researchers for years. We chose to study it because we, and our loved ones, have been on the receiving end.
Uri will never forget the nightmarish stories his father, Jacob, a Holocaust survivor from Budapest, told him about what happened to his tight-knit neighborhood. When the Nazis took over Hungary and the Holocaust swept into Budapest in 1944, Jacob was no longer allowed to work. His mother, Magda, managed to move the family into one of three safe houses run by the Swedish diplomat Raoul Wallenberg outside the Jewish ghetto. But the houses turned out to be not so safe after all.
One night, members of the pro-Nazi Arrowcross party flushed their Jewish neighbors from their homes, marched them to the Danube river, and shot every man, woman, and child. The next night, the same thing happened to the people in the second building. On the next night, Uri’s father and his family were expecting to go to a similar destination. But instead the Nazi sympathizers forced them at gun-point into the ghetto, where Magda fended off the family’s starvation by fighting over the decidedly un-Kosher meat of dead horses. They escaped death by sheer luck. Many years later, not far from the sites of those roundups, Uri lectured at Budapest University—the same institution from which his grandfather was summarily ejected on the basis of his religion. Uri could not help shuddering as he stood at the lectern.
When we think of discrimination, these are the types of ugly, virulent prejudice that we think about. But John faced a much different kind of discrimination when he entered the job market as a newly minted Ph.D. in 1995. Although he applied to more than 150 academic jobs and had completed several field experiments, he was given only one interview. He later learned that other nearly identical applicants received thirty interviews from just forty or so applications. The main difference between John and these other applicants was that John received his Ph.D. from the University of Wyoming, whereas they had received theirs from “brand-name” schools like Harvard and Princeton. Employers were using that bit of information to screen their applicants—effectively discriminating between the “haves” and the “have-nots.”
You, too, have likely experienced this type of discrimination—maybe without even knowing it. And like most people, you may think that human beings treat each other unfairly because we’re simply wired that way. It’s easy to understand why most of us assume the worst of each other. All around us, every day, accusations of racism fly. President Obama’s supporters accuse his detractors of racism and vice versa; bloggers, news organizations, politicians, and other public officials routinely jump to conclusions about people’s motivations before the facts are out.
What does all this have to do with economics? The answer is this: rather than accept that humans are hardwired to be racists or bigots, we wanted to learn more about the underlying motivations for why people actually discriminate. Clearly discrimination has serious, long-term effects on people’s lives, and we wanted to understand how discrimination works in real markets, where people function every day. What causes it? Is it driven by deep-seated prejudice alone, or is there another explanation?
Using various field experiments in real markets, we have learned that the kind of discrimination that John faced is today much more common than the kind that Uri’s family faced. Unabashed hatred and pure animus are not as pervasive as most of us believe. As a result, if you really want to end discrimination, don’t just focus on the ugly, racist side of things—that’s the wrong culprit. Instead,
consider the economic incentive for the discrimination, and then look through the microscope. As it turns out, most cases of modern-day discrimination are caused by people or companies trying to increase their profits.
But that does not mean that outright hatred is dead. As it turns out, people often discriminate in a bigoted way when they perceive others as having a choice in the matter. As Archie Bunker, the racist protagonist of the old television sitcom All in the Family, asked Sammy Davis Jr. in one famous episode: “Your bein’ colored, now, I know you had no choice in that. But whatever made you turn Jew?”4
These insights turn out to be important not only for society, but also for you. Furthermore, policy makers cannot begin to battle something that they don’t understand. If you are someone who designs laws, understanding how not to be discriminated against is invaluably important.
Another issue that really bothered us was the gender gap in labor markets. Women still earn less than equally skilled men do, and are still too scarce in boardrooms and the C-level offices of companies.
Between us, we have four smart daughters (and four beautiful sons). Like you, we want all of our kids to get a fair shake as they grow up, go to college, and compete for jobs. But from their earliest years, we noticed that fair shakes weren’t always happening for our daughters. Why did one of our girl’s teachers seem to be telling her that she wasn’t as good at math as the boys, even though it was clear that she had mathematical talent? Why did the sports coaches at her school berate the boys in her class to “stop playing kickball like a girl”? And why were Uri’s two daughters—one competitive, one not so much so—so different?
We both wondered whether our daughters would be able to compete for great schools and great jobs, or whether they would be discouraged and sidelined along the way. After combining our observations from their early days in school with the facts about the great differences between men’s and women’s ability to command high salaries, climb the corporate ladder, and hold prominent public positions, we wondered if differences in competitiveness could help to explain the gender gap. So we asked a simple question: Are women different than men in terms of competitiveness? After finding important differences, we asked the age-old question: Is this difference in competitiveness because of nature or nurture?
To find answers, we boarded planes, helicopters, trains, and automobiles and went to the far corners of the earth to investigate gender competitiveness among the most and least patriarchal societies on the planet (that’s how we met Minott). The results of our research come down strongly on the side of nurture. In the right environment—one in which women are not deterred from competitive situations and are accepted by their society as powerful individuals—women grow up to be just as competitive as men, and sometimes even more so. This has important implications for our daughters and yours, and for policy makers who want to reduce the gender gap in labor markets. If you set the incentives correctly, the gender gap can be reduced drastically.
Another question we’ve explored is: How can we get people to donate more money to charity? Beyond our desire to be good citizens, we each had selfish reasons for our curiosity.
For his part, John has been interested in the economics of charity since he was a wet-behind-the-ears professor at the University of Central Florida, where he discovered that an integral part of our economy—the charitable sector—was largely driven by anecdotes and outdated rules of thumb devoid of scientific validation. Along the way, he came to know Brian Mullaney, the founder and CEO of Smile Train and WonderWork.org—whose ubiquitous magazine ads and direct-mail envelopes appeal for donations that can correct cleft lips and palates (and, through WonderWork.org, other maladies) with a simple surgery.
A large-scale field experiment touching roughly 800,000 direct-mail recipients revealed something about giving that no one would have guessed: allowing people to check a box saying “never contact me again” actually lead to higher levels of gifts, not lower. Many fundraising experts thought the idea was crazy; why on earth would any charity invite people to stop contributing? But as it turned out, people loved it. We raised much more money using the opt-out rather than the standard treatment, and only 39 percent of recipients opted out. Smile Train and WonderWork.org ended up saving money on postage, because they only needed to re-mail to those who were interested in giving in the future. It was a true win-win.
For his part, Uri became intrigued by the idea of getting people to give more to charity while experimenting with a new pricing mechanism at various companies—“pay what you want.” Under pay-what-you-want pricing, a company tells its customers that they can have the goods or services they need for any price they set (including $0). We were able to convince Disney to test this new and unusual pricing mechanism in one of its large theme parks. We found that when a charitable donation is combined with pay-what-you-want pricing, people pay a lot—much more, in fact, than they do according to the traditional pricing models.
And, as we discovered, human beings have more complicated and, yes, more complex reasons to give than simple altruism. When we looked at all kinds of techniques—door-to-door campaigns, direct-mail solicitations, matching grants, and so on—we found out what works best in setting the right incentives and convincing people to open their hearts and wallets. As you will see, a running theme throughout the book is this: once we discover what people value, then we can design useful policies that influence their behavior and induce change.
Here’s another dilemma that has grabbed us: How can you use incentives to keep kids in school and curtail youth gun violence?
This question is anything but abstract. Public schools in some areas of Chicago have horrendous attrition rates, in some cases as high as 50 percent, and one out of every thousand public school student gets shot. When the mayor of Chicago Heights asked John for some help, John responded as any good citizen would—and he brought the tool kit of an economist to the job. The large-scale experiments we describe in this book—the first of their kind anywhere in the country—are demonstrating that certain kinds of incentives, offered in the right way, can go a long way toward improving student performance. They can save lives, too.
In investigating student performance, we had to delve deep into motivation. What really happens when you use money as an incentive? When do incentives work and when don’t they? These questions first started to bother us years ago, when our children were in day care. The principal of the preschool, frustrated by parents who failed to pick up their kids at the appointed time, decided to impose a small fine for late pickups. The fine actually acted as a counterincentive, because it put a price—and a fairly low one at that—on inconveniencing the teachers and staff. Parents might have felt guilty about being late before, but once the fine was instituted, they decided it was downright silly to show up on time. Why rush through traffic like a crazy person for the sake of saving a few bucks? We did more research, and concluded that if you want someone to do something, you had better be pretty careful about the details—the who, what, when, where, why, and how much you motivate. Money works, but only at the right levels.
As you might have gathered by now, we’re not like most economists. While we use important insights from economic theories, we didn’t develop our thinking in intellectual hothouses.
For example, John, as we mention above, took his first forays into the business world as a hungry college student, when he learned to buy, sell, and trade sports memorabilia. He received an unforgettable lesson about cutthroat competition and capitalism when he traded a valuable collection of his own sports cards for a set of worthless counterfeits. But through the process he learned how to bargain more effectively, and even how to price his goods correctly. To his surprise, he later observed that most firms—even international corporations—haven’t the foggiest idea about how to set prices for their goods and services.
Uri loves good California wine. Often, when visiting wineries, he wondered how owners priced their wines—a particularly tricky task, since
quality is hard to judge objectively. When a vintner asked him to help with just that, Uri told him that he had no clue how much the wine should cost—but he did have a tool that could do the job of finding out simply and cheaply. We conducted a small field experiment in the winery, and few weeks later we were able to find the best price—one that raised the winery’s profits considerably. Our field experiments in companies have shown how to raise both productivity and profits in a way that increases everyone’s share of the pie.
Often, businesspeople think that running experiments is a costly undertaking, but we believe it’s prohibitively costly not to experiment. How many product and pricing failures can be laid at the feet of insufficient investigations and tests? Just ask the people at Netflix, who blundered badly in 2011 when they introduced new pricing that substantially damaged both their brand and their stock value.
Every transaction is an opportunity to learn something about customers. Companies that learn to run field experiments, and run them well, will lead in their markets. In the past, skilled managers could rely on intuition and received wisdom from their predecessors. But tomorrow’s successful manager will generate her own data via field experiments and use those insights to drive the bottom line.
So there you have it. By the time you finish this book, we hope you will come away with a much better idea of what works—and what doesn’t. We also hope that you will see economics as a passionate science, not a “dismal science,” the name conferred on it by Victorian historian Thomas Carlyle.5
To us, economics is a discipline fully engaged with the entire spectrum of human emotions, with a laboratory as big as the whole world, and with the capacity to produce results that can change society for the better. We believe you’ll find that our field experiments are not just eye-opening, but fun and full of surprises. We hope you’ll discover that economics is not boring or dismal at all. We think you’ll come away with new understanding of the hidden motives that drive people to behave the way they do and of how we can all achieve better outcomes for ourselves, our companies, our customers, and society in general.