The Undoing Project

by Michael Lewis

  Three days later Amos called Danny. He’d just received some news. A growth that doctors had discovered in his eye had just been diagnosed as malignant melanoma. The doctors had scanned his body and found it riddled with cancer. They were now giving him, at best, six months to live. Danny was the second person he’d called with the news. Hearing that, something inside Danny gave. “He was saying, ‘We’re friends, whatever you think we are.’”

  * * *

  * After the article appeared, in the October 1983 issue of Psychological Review, the best-selling author and computer scientist Douglas Hofstadter sent Amos his own vignettes. Example: Fido barks and chases cars. Which is Fido more likely to be: (1) a cocker spaniel or (2) an entity in the universe?

  Coda

  BORA-BORA

  Consider the following scenario.

  Jason K. is a fourteen-year-old homeless boy who lives in a large American city. He is shy and withdrawn but extremely resourceful. His father was murdered when he was young; his mother is an addict. Jason takes care of himself, sleeping sometimes on the sofas at friends’ apartments but mostly on the streets. He manages to stay in school until the ninth grade. He often goes hungry. One day in 2010 he accepts an offer from a local gang to sell drugs, and drops out of school. A few weeks later, the night before his fifteenth birthday, he is shot and killed. He was unarmed when he died.

  We are seeking ways to “undo” Jason K.’s death. Rank the following in order of their likelihood.

  1)Jason’s father was not murdered.

  2)Jason carried a gun and was able to protect himself.

  3)The U.S. federal government made it easier for homeless kids to obtain the free breakfast and lunch to which they are entitled. Jason never went hungry, and remained in school.

  4)A lawyer steeped in the writings of Amos Tversky and Daniel Kahneman took a federal government job in 2009. Drawing upon Kahneman and Tversky’s work, he pushed for changes in the rules, so that homeless kids no longer needed to enroll in the school meal program. Instead they automatically received free breakfast and lunch. Jason never went hungry, and remained in school.

  If you found #4 more probable than #3, you violated perhaps the simplest and most fundamental law of probability. But you’re also onto something. The lawyer’s name is Cass Sunstein.

  Among its other consequences, the work that Amos and Danny did together awakened economists and policy makers to the importance of psychology. “I became a believer,” said Nobel Prize–winning economist Peter Diamond of Danny and Amos’s work. “It’s all true. This stuff is not just lab stuff. It’s capturing reality, and it’s important to economists. And I spent years thinking of how to use it—and failing.” By the early 1990s a lot of people thought it was a good idea to bring together psychologists and economists, to allow them to get to know each other better. But as it turned out, they didn’t particularly want to know each other better. Economists were brash and self-assured. Psychologists were nuanced and doubtful. “Psychologists as a rule will only interrupt a presentation for clarification,” says psychologist Dan Gilbert. “Economists will interrupt to show how smart they are.” “In economics it is completely normal to be rude,” says economist George Loewenstein. “We tried to create a psychology and economics seminar at Yale. We had our first meeting. The psychologists came out completely bruised. We never had a second meeting.” In the early 1990s, Amos’s former student Steven Sloman invited an equal number of economists and psychologists to a conference in France. “And I swear to God I spent three-quarters of my time telling the economists to shut up,” said Sloman. “The problem,” says Harvard social psychologist Amy Cuddy, “is that psychologists think economists are immoral and economists think psychologists are stupid.”

  In the academic culture war triggered by Danny and Amos’s work, Amos served as a strategic advisor. At least some of his sympathies were with the economists. Amos’s mind had always clashed with most of psychology. He didn’t like emotion, as a subject. His interest in the unconscious mind was limited to a desire to prove it didn’t exist. He was like a man in stripes wandering a land settled by people dressed in plaids and polka dots. Like the economists, he preferred neat formal models to mixed-chocolate boxes of psychological phenomena. Like them, he found it completely normal to be rude. And, like them, he had worldly ambitions for his ideas. Economists sought influence in the arenas of finance and business and public policy. Psychologists hardly ever entered those arenas. That was about to change.

  Danny and Amos both saw that there was no point trying to infiltrate economics from psychology. The economists would just ignore intruders. What were needed were young economists with an interest in psychology. Almost magically, after Amos and Danny arrived in North America, they began to appear. George Loewenstein was a good example. A trained economist disillusioned by the psychological sterility of economic models, Loewenstein read Amos and Danny’s work and thought: Wait, maybe I want to be a psychologist! As he happened to be Sigmund Freud’s great-grandson, this was an even more complicated than usual thought. “I had tried to escape the family’s past,” said Loewenstein. “I realized I had never taken a single class in what really interested me.” He approached Amos and asked him for advice: Should he move from economics to psychology? “Amos said, ‘You should stay in economics—we need you there.’ He already knew in 1982 that he was starting a movement. And he needed people inside economics.”

  The argument that Danny and Amos started would spill over into law and public policy. Psychology would use economics to enter these places and others. Richard Thaler—the first frustrated economist to stumble onto Danny and Amos’s work and pursue its consequences for economics single-mindedly—would help to create a new field, and give it the name “behavioral economics.” “Prospect Theory,” scarcely cited in the first decade after its publication, would become, by 2010, the second most cited paper in all of economics. “People tried to ignore it,” said Thaler. “Old economists never change their minds.” By 2016 every tenth paper published in economics would have a behavioral angle to it, which is to say it had at least a whisper of the work of Danny and Amos. And Richard Thaler would have just stepped down from his tenure as president of the American Economic Association.

  Cass Sunstein had been a young law professor at the University of Chicago when he came upon Thaler’s first war cry on psychology’s behalf. A paper Thaler had titled in his mind “Stupid Shit That People Do” he’d finally published as “Toward a Positive Theory of Consumer Choice.” Thaler’s bibliography led Sunstein directly to the article written by Danny and Amos in Science about judgment, and to “Prospect Theory.” “For a lawyer both of these were difficult,” said Sunstein. “I had to read them more than once. But I remember the feeling: It was an explosion of lightbulbs. You have thoughts in your mind and you read something that immediately puts them in order and it’s electrifying.” In 2009, at the invitation of President Obama, Sunstein went to work at the White House. There he oversaw the Office of Information and Regulatory Affairs and made scores of small changes that had big effects on the daily lives of all Americans.

  The changes Sunstein made had a unifying theme: They sprang directly or indirectly from the work of Danny and Amos. You couldn’t say that Danny and Amos’s work led President Obama to ban federal employees from texting while driving, but it wasn’t hard to draw a line from their work to that act. The federal government now became sensitive to both loss aversion and framing effects: People didn’t choose between things, they chose between descriptions of things. The fuel labels on new automobiles went from listing only miles per gallon to including the number of gallons a car consumed every hundred miles. What used to be called the food pyramid became MyPlate, a graphic of a dinner plate with divisions for each of the five food groups, and it was suddenly easier for Americans to see what made for a healthy diet. And so on. Sunstein argued that the government needed, alongside its Council of Economic Adviser
s, a Council of Psychological Advisers. He wasn’t alone. By the time Sunstein left the White House, in 2015, calls for a greater role for psychologists, or at any rate for psychological insight, were coming from inside governments across the world.

  Sunstein was particularly interested in what was now being called “choice architecture.” The decisions people made were driven by the way they were presented. People didn’t simply know what they wanted; they took cues from their environment. They constructed their preferences. And they followed paths of least resistance, even when they paid a heavy price for it. Millions of U.S. corporate and government employees had woken up one day during the 2000s and found they no longer needed to enroll themselves in retirement plans but instead were automatically enrolled. They probably never noticed the change. But that alone caused the participation in retirement plans to rise by roughly 30 percentage points. Such was the power of choice architecture. One tweak to the society’s choice architecture made by Sunstein, once he’d gone to work in the U.S. government, was to smooth the path between homeless children and free school meals. In the school year after he left the White House, about 40 percent more poor kids ate free school lunches than had done so before, back when they or some adult acting on their behalf had to take action and make choices to get them.

  * * *

  Even in Canada, Don Redelmeier still heard the sound of Amos in his head. He’d been back from Stanford for several years, but Amos’s voice was so clear and overpowering that it made it hard for Redelmeier to hear his own. Redelmeier could not pinpoint the precise moment that he sensed that his work with Amos was not all Amos’s doing—that it had some Redelmeier in it, too. His sense of his own distinct value began with a simple question—about homeless people. The homeless were a notorious drag on the local health care system. They turned up in emergency rooms more often than they needed to. They were a drain on resources. Every nurse in Toronto knew this: If you see a homeless person wander in, hustle him out the door as fast as you can. Redelmeier wondered about the wisdom of that strategy.

  And so, in 1991, he created an experiment. He arranged for large numbers of college students who wanted to become doctors to be given hospital greens and a place to sleep near the emergency room. Their job was to serve as concierges to the homeless. When a homeless person entered the emergency room, they were to tend to his every need. Fetch him juice and a sandwich, sit down and talk to him, help arrange for his medical care. The college students worked for free. They loved it: They got to pretend to be doctors. But they serviced only half of the homeless people who entered the hospital. The other half received the usual curt and dismissive service from the nursing staff. Redelmeier then tracked the subsequent use of the Toronto health care system by all the homeless people who had visited his hospital. Unsurprisingly, the group that received the gold-plated concierge service tended to return slightly more often to the hospital where they had received it than the unlucky group. The surprise was that their use of the greater Toronto health care system declined. When homeless people felt taken care of by a hospital, they didn’t look for other hospitals that might take care of them. The homeless said, “That was the best that can be done for me.” The entire Toronto health care system had been paying a price for its attitude to the homeless.

  A part of good science is to see what everyone else can see but think what no one else has ever said. Amos had said that to him, and it had stuck in Redelmeier’s mind. By the mid-1990s, in startling ways, Redelmeier was taking what everyone could see and thinking to say what no one had ever said. For instance, one day he had a phone call from an AIDS patient who was suffering side effects of medication. In the middle of the conversation, the patient cut him off and said, “I’m sorry, Dr. Redelmeier, I have to go. I just had an accident.” The guy had been talking to him on a cell phone while driving. Redelmeier wondered: Did talking on a cell phone while driving increase the risk of accident?

  In 1993, he and Cornell statistician Robert Tibshirani created a complicated study to answer the question. The paper they wrote, in 1997, proved that talking on a cell phone while driving was as dangerous as driving with a blood alcohol level at the legal limit. A driver talking on a cell phone was four times as likely as a driver who wasn’t to be involved in a crash, whether or not he held the phone in his hands. That paper—the first to establish, rigorously, the connection between cell phones and car accidents—spurred calls for regulation around the world. It would take another, even more complicated study to determine just how many thousands of lives it saved.

  The study also piqued Redelmeier’s interest in what happened inside the mind of a person behind the wheel of a car. The doctors in the Sunnybrook trauma center assumed that their jobs began when the human beings mangled on nearby Highway 401 arrived in the emergency room. Redelmeier thought it was insane for medicine not to attack the problem at its source. One point two million people on the planet died every year in car accidents, and multiples of that were maimed for life. “One point two million deaths a year worldwide,” said Redelmeier. “One Japanese tsunami every day. Pretty impressive for a cause of death that was unheard-of one hundred years ago.” When exercised behind the wheel of a car, human judgment had irreparable consequences: That idea now fascinated Redelmeier. The brain is limited. There are gaps in our attention. The mind contrives to make those gaps invisible to us. We think we know things we don’t. We think we are safe when we are not. “For Amos it was one of the core lessons,” said Redelmeier. “It’s not that people think they are perfect. No, no: They can make mistakes. It’s that they don’t appreciate the extent to which they are fallible. ‘I’ve had three or four drinks. I might be 5 percent off my game.’ No! You are actually 30 percent off your game. This is the mismatch that leads to ten thousand fatal accidents in the United States every year.”

  It is sometimes easier to make the world a better place than to prove you have made the world a better place.

  Amos had said that, too. “Amos gave everyone permission to accept human error,” said Redelmeier. That was how Amos made the world a better place, though it was impossible to prove. The spirit of Amos was now present in everything Redelmeier did. It was present in his article about the dangers of driving while speaking on a cell phone—which Amos had read and commented upon. That was the article Redelmeier was working on when the call came with the news that Amos had died.

  * * *

  Amos told very few people that he was dying, and, to those he did tell, he gave instructions not to spend a lot of time talking to him about it. He received the news in February 1996. From then on he spoke of his life in the past tense. “He called me when the doctor told him that it was the end of it,” said Avishai Margalit. “I came to see him. And he fetched me from the airport. And we were on our way to Palo Alto. And we stopped somewhere on the road, with a view, and talked, about death and about life. It was important to him that he had his death under control. And the feeling was that he was talking not about himself. Not about his death. There was a kind of stoic distance that was astonishing. He said, ‘Life is a book. The fact that it was a short book doesn’t mean it wasn’t a good book. It was a very good book.’” Amos seemed to understand that an early death was the price of being a Spartan.

  In May Amos gave his final lecture at Stanford, about the many statistical fallacies in professional basketball. His former graduate student and collaborator Craig Fox asked Amos if he would like for it to be videotaped. “He thought about it and said, ‘No, I don’t think so,’” recalled Fox. With one exception, Amos didn’t change his routine, or even his interactions with those around him, in any way. The exception was that, for the first time, he spoke of his experience of war. For instance, he told Varda Liberman the story of how he had saved the life of the soldier who had fainted on top of the bangalore mine. “He said this one event in a way kind of shaped his entire life,” said Liberman. “He said, ‘Once I did that, I felt obliged to keep this image of hero. I did tha
t, now I have to live up to it.’”

  Most people with whom Amos interacted never even suspected he was ill. To a graduate student who asked if he would supervise his dissertation, Amos simply said, “I’m going to be very busy the next few years,” and sent him on his way. A few weeks before his death, he called his old friend Yeshu Kolodny in Israel. “He was very impatient, which never happened,” recalled Kolodny. “He said, ‘Listen, Yeshu, I’m dying. I take it not tragically. But I don’t want to talk to anyone. I need you to call our friends and tell them—and tell them not to call or visit.’” To his rule against visitors Amos made an exception for Varda Liberman, with whom he was finishing a textbook. He made another for Stanford president Gerhard Casper—but only because he’d gotten wind of Stanford’s plan to commemorate him, with a lecture series or a conference in his name. “Amos told Casper, ‘You can do anything you want,’” recalled Liberman. “ ‘But I beg you, don’t have a conference in my name with mediocre people who will talk about their work and how it is “related” to mine. Just put my name on a building. Or a room. Or a bench. You can put it on anything that is not moving.’”

  He accepted very few phone calls. One he took came from the economist Peter Diamond. “I learned he was dying,” said Diamond. “And I learned he wasn’t taking phone calls. But I had just finished my report to the Nobel Committee.” Diamond wanted to let Amos know that he was on a very short list for the Nobel Prize in economics, to be awarded in the fall. But the Nobel Prize was awarded only to the living. He didn’t recall what Amos said to that, but Varda Liberman was in the room when Amos took the call. “I thank you very much for letting me know,” she heard Amos say. “I can assure you that the Nobel Prize is not on the list of things I’m going to miss.”