The Undoing Project

by Michael Lewis

  In some strange way, as they approached their moment of greatest public triumph, their collaboration remained a private affair, a gamble with no context. “As long as we stayed in Israel, the whole idea of what the world thought of us didn’t occur to us,” said Danny. “We benefited from our isolation.” That isolation depended on them being together, in the same room, behind a closed door.

  That door was now cracking open. Anne was British. She was also a gentile and the mother of four children, one of whom had Down syndrome. There were about sixteen different reasons she couldn’t, or wouldn’t, move to Israel. And if Anne wasn’t moving to Israel, it followed that Danny would need to leave. Danny and Amos scrambled and found a temporary solution, in 1977, by heading off together from Hebrew University on sabbatical to Stanford University, where Anne might join them. But a few months after their arrival in the United States, Danny announced that he planned to marry Anne and stay. He forced Amos to make a decision about what to do about their relationship.

  It was now Amos’s turn to sit down and write an emotional letter. Danny was messy, in a way that Amos could never be messy even if he wanted to be. Amos had wanted to be a poet when he was a boy. He’d wound up a scientist. Danny was a poet, who somehow happened to have become a scientist. Danny felt some obvious desire to be more like Amos; Amos, too, harbored some less obvious desire to be more like Danny. Amos was a genius. But he needed Danny, and he knew it. The letter Amos wrote was to his close friend Gidon Czapski, the rector of Hebrew University. “Dear Gidi,” it began. “The decision to remain here in the United States is the most difficult decision I have ever made. I cannot ignore my desire to bring to a completion, at least partially, the joint work with Danny. I just cannot accept the idea that the joint work of years could come to naught and that we will not be able to complete the ideas we have.” Amos went on to explain that he planned to accept a chaired professorship offered to him by Stanford University. He knew full well that everyone in Israel would be shocked and angry. “If Danny leaves Israel it is a personal tragedy,” a Hebrew University official had said to him not long before. “If you leave it is a national tragedy.”

  Until Amos actually left, his friends found it unthinkable that he would live anyplace but Israel. Amos was Israel, and Israel was Amos. Even his American wife was upset. Barbara had fallen in love with Israel—its intensity, its sense of community, its disinterest in small talk. She now thought of herself as more Israeli than American. “I had done so much work to become Israeli,” she said. “I didn’t want to stay in the States. I said to Amos, ‘How can I start over?’ He said, ‘You’ll manage.’”

  * * *

  * Here is the simpler version of the paradox. Danny and Amos created it to show how the apparent contradiction might be resolved using their findings about people’s attitudes toward probabilities. And so in a funny way they “solved” the Allais paradox twice—once by explaining it with regret, this time by explaining it with their new theory:

  You are offered a choice between:

  1.$30,000 for sure

  2.A gamble that has a 50 percent chance of winning $70,000 and a 50 percent chance of winning nothing

  Most people took the $30,000. That was interesting in itself. It showed what was meant by “risk aversion.” People choosing between a bet and a certain amount would accept a certain amount that was less than the expected value of the bet (which here is $35,000). That did not violate utility theory. It just meant that the utility of a chance to win 70 grand is less than the utility of a twice as likely chance to win 30 grand—which in this case makes the 30 grand a certainty. But now consider a second choice between bets:

  1.A gamble that gives you a 4 percent chance to win $30,000 and a 96 percent chance to win nothing

  2.A gamble that gives you a 2 percent chance to win $70,000 and a 98 percent chance to win nothing

  Most people here preferred 2, the lower chance to win more. But that implied that the “utility” of a chance to win $70,000 is greater than the utility of a twice as likely chance to win $30,000—or the opposite of the preferences in the first choice. In Danny and Amos’s working theory, the paradox was now resolved differently. It wasn’t that (or at least not only that) people anticipated regret when making a decision in the first situation that they did not anticipate in making the second. It was that they treated 50 percent as more than 50 percent and saw the difference between 4 percent and 2 percent as far less than it was.

  11

  THE RULES OF UNDOING

  In the late 1970s, not long after he’d become superintendent of the Massachusetts Mental Health Center, Miles Shore realized he had a problem. The center was a teaching hospital for Harvard Medical School, where Shore was Bullard Professor of Psychiatry. Newly installed in administration, he found himself faced with a decision: whether to promote a medical researcher named J. Allan Hobson. It shouldn’t have been that hard. In a series of famous papers, Hobson had landed body blows on the Freudian idea that dreams arose from unconscious desires, by showing that they actually came from a part of the brain that had nothing to do with desire. He’d proven that the timing and the length of dreams were regular and predictable, which suggested that dreams had less to say about a person’s psychological state than about his nervous system. Among other things, Hobson’s research suggested that people who paid psychoanalysts to find meaning in their unconscious states were wasting their money.

  Hobson was changing people’s understanding of what happened to the human brain during sleep—but he wasn’t doing it alone. That was Miles Shore’s problem: Hobson hadn’t written his famous papers on dreams by himself, but with a partner named Robert McCarley. “It was very difficult to campaign for promotion for people who did their work collaboratively,” said Shore. “Because the system is based on the individual. It was always: What did this person do to change the field?” Shore wanted to promote Hobson, but he had to argue the case before a skeptical committee. “They basically didn’t want to promote anyone,” said Shore. Resisting the case for Hobson, committee members asked Shore if he could demonstrate exactly how much Hobson had contributed to his partnership with McCarley. “They asked me which one of them did what,” recalled Shore. “And so I went to them [Hobson and McCarley] and asked: ‘Which one of you did what?’ And they said: ‘Which one of us did what? We have no idea. It was a joint product.’” Shore pushed the collaborators a bit until he realized that they really meant it: They had no idea who deserved credit for which idea. “It was really interesting,” said Shore.

  So interesting that Shore decided there might be a book in it. He set out to find fertile pairs—people who had been together for at least five years and produced interesting work. By the time he was done he had interviewed a comedy duo; two concert pianists who had started performing together because one of them had stage fright; two women who wrote mysteries under the name “Emma Lathen”; and a famous pair of British nutritionists, McCance and Widdowson, who were so tightly linked that they’d dropped their first names from the jackets of their books. “They were very huffy about the idea that dark bread was more nutritious than white bread,” recalled Shore. “They had produced the research that it wasn’t so in 1934—so why didn’t people stop fooling around with the idea?” Just about every work couple that Shore called were intrigued enough by their own relationships to want to talk about them. The only exceptions were “a mean pair of physicists” and, after flirting with participating, the British ice dancers Torvill and Dean. Among those who agreed to sit down with Miles Shore were Amos Tversky and Daniel Kahneman.

  Shore found Danny and Amos together in August 1983, in Anaheim, California, where they were attending the American Psychological Association meeting. Danny was now forty-nine and Amos forty-six. They spoke with Shore together for several hours and then, for several hours more, separately. They walked Miles Shore through the history of the collaboration, starting with their early exc
itement. “In the beginning we were able to answer a question that had not been asked,” Amos told him. “We were able to take psychology out of the contrived laboratory and address the topic from the experiences all around us.” Trying to pin them down on the question they thought they were answering, Shore asked if their work fed into the new and growing field of artificial intelligence. “You know, not really,” said Amos. “We study natural stupidity instead of artificial intelligence.”

  The Harvard psychiatrist thought that Danny and Amos had a lot in common with other successful pairs. The way they had created what amounted to an exclusive private club of two, for example. “They were crazy about each other, and not indiscriminate,” said Shore. “They were not generally crazy about other people. They hated editors.” As with some of the other fertile pairs, the partnership had created strains on their other close relationships. “The collaboration has put a lot of pressure on my marriage,” Danny confessed. Like the other pairs, they had lost any sense of individual contribution. “You ask who did what?” said Danny “We didn’t know at that time, not clearly. It was beautiful, not knowing.” Shore thought that both Amos and Danny realized, or seemed to realize, how much they needed each other. “There are geniuses who work on their own,” said Danny. “I am not a genius. Neither is Tversky. Together we are exceptional.”

  What set Amos and Danny apart from the nineteen other couples Shore had interviewed for his book was their willingness to speak about the problems in their relationship. “When I asked about conflicts, most people just ignored it,” said Shore. “A number didn’t want to admit there was any conflict.” Not Amos and Danny. Or, at any rate, not Danny. “It’s been difficult since I got married and since we moved to this continent,” he confessed. Amos remained evasive, and yet great chunks of Shore’s conversation with Danny and Amos wound up being about the many troubles they’d had since leaving Israel six years earlier. With Amos in the room, Danny complained at length about how different the public perception of the collaboration was from its reality. “I am perceived as attending him, which is not the case,” he said, less to Shore than to Amos. “I clearly lose by the collaboration. There is a quality that is clearly contributed by you. Formal analysis is not my strength and it shows up very distinctly in our work. My contributions are less unique.” Amos spoke, at less length, about how the blame for their unequal status fell squarely on other people. “The credit business is very hard,” said Amos. “There is a lot of wear and tear, and the outside world isn’t helpful to collaborations. There is constant poking, and people decide that one person gets the short end of the stick. It’s one of the rules of balance, and joint collaboration is an unbalanced structure. It is just not a stable structure. People aren’t happy with it.”

  Alone with the Harvard psychiatrist, Danny said more. He hinted that he didn’t believe the outside world was entirely responsible for the problems in their relationship. “The spoils of academic success, such as they are—eventually one person gets all of it, or gets a lot of it,” he said. “That’s an unkindness built in. Tversky cannot control this, though I wonder whether he does as much to control it as he should.” Then he came straight out with his own feelings about Amos getting the lion’s share of the glory for work they had done together. “I am very much in his shadow in a way that is not representative of our interaction,” he said. “It induces a certain strain. There is envy! It’s just disturbing. I hate the feeling of envy. . . . I am maybe saying too much now.”

  Shore left the interview feeling that Amos and Danny had just come through a rough patch, but that the worst lay behind them. Their openness about their problems he took as a good sign. They hadn’t exactly been fighting during their interview; their attitude toward conflict was just different than that of the other couples he had spoken with. “They played the Israeli card,” said Shore. “We’re Israeli, so we yell at each other.” Amos, especially, sounded optimistic that he and Danny would continue to work together as much as they had. It helped, Danny and Amos agreed, that the American Psychological Association had just honored both of them with its Distinguished Scientific Contribution Award. “I have lived in some fear that he might get it alone,” Danny confessed to Shore. “That would have been a disaster, and I couldn’t have coped very elegantly.” The award had eased some pain. Or so it seemed to Miles Shore.

  As it happened, Shore never wrote his book about fertile pairs. But years later, he sent Danny an audiotape of their conversation. “I listened to it,” Danny said, “and it is absolutely clear from it that we are finished.”

  * * *

  In late 1977, after Danny had told him that he wasn’t returning to Israel, word spread through academia that Amos Tversky might leave, too. The job market for college professors typically moves slowly and with great reluctance, but in this instance it leapt into action. It was as if an especially deliberate fat man watching TV on his couch suddenly realized that his house was on fire. Harvard University quickly offered Amos tenure, though it took them a few weeks to throw in an assistant professorship for Barbara. The University of Michigan, which had the advantage of sheer size, scrambled to find four tenured professorships—and, by making places for Danny, Anne, and Barbara, also snag Amos. The University of California at Berkeley, which left Danny with the clear impression, when he made overtures, that he was too old to be hired, prepared to offer a job to Amos. But no place moved quite so dramatically as Stanford.

  The psychologist Lee Ross, a rising young star on the Stanford faculty, led the charge. He knew that the big public American universities who wanted Amos might, in the bargain, offer jobs to Barbara and Danny and Anne. Stanford was smaller and didn’t have four jobs to offer. “We figured there were two things we could do that those schools might not,” said Ross. “One was to make the offer early, and the other was to make it fast. We wanted to convince him to come to Stanford, and the best way we can convince him to come is to show him how quickly we can act.”

  What happened next was, Ross believed, unprecedented in the history of the American university. The morning he learned Amos was on the market, he convened Stanford’s Psychology Department. “I was supposed to present the case for Amos,” said Ross. “I said, I’m going to tell you a classic Yiddish story. There’s a guy, an eligible bachelor. A happy bachelor. The matchmaker comes to him and says, ‘Listen, I have for you a match.’ ‘Ah, I’m not so sure,’ says the bachelor. ‘She’s really special,’ says the matchmaker. ‘What, is she beautiful?’ asks the bachelor. ‘Beautiful? She looks like Sophia Loren, only younger.’ ‘What, does she have family money?’ asks the bachelor. ‘Money? She’s an heiress to the Rothschild fortune.’ ‘Then she must be a dope,’ says the bachelor. ‘A dope? She has been nominated for Nobel Prizes in both physics and chemistry.’ ‘I accept!’ says the bachelor. To which the matchmaker replies, ‘Good, we have half a match!’” Ross told the Stanford faculty, “After I tell you about Amos, you will say, ‘I accept!’ and I will say, ‘I’m sorry to tell you we have half a match.’”

  Even to Ross it was unclear that the sales pitch was necessary. “Everyone who came across the work congratulated themselves on their own good judgment and insight in appreciating the work,” said Ross. “But nobody didn’t get it.” That same day, the Stanford Psychology Department went to the Stanford president and said: We have none of the usual paperwork. No recommendations or anything else. Just trust us. Stanford made Amos an offer of lifetime employment that afternoon.

  Amos would later tell people that in choosing between Harvard and Stanford, he imagined the regret he would experience at each. At Harvard he’d regret passing up Palo Alto’s weather and living conditions, and resent the commute; at Stanford he’d regret, and only briefly, not being able to say he was a Harvard professor. If it occurred to him or anyone else that Amos, to be Amos, needed Danny close at hand, he didn’t show it. Stanford showed not the slightest interest in Danny. “There’s a practical issue,” said Ross. “Do you
want two guys doing the same thing? And the cold fact is we got the full benefit of Danny and Amos just by hiring Amos.” Danny would have loved for them all to go to Michigan, but Amos clearly had no interest in anyplace but Harvard or Stanford. After Harvard and Stanford had ignored him, and Berkeley had let him know that he would not be offered a job, Danny accepted a position beside Anne at the University of British Columbia, in Vancouver. He and Amos agreed they would take turns flying to visit each other every other weekend.

  Danny was still floating on air. “We were on such a high from having finished prospect theory and embarked on framing that we must have felt pretty invulnerable,” he said. “There was not a shadow between us at the time.” He watched Amos give the traditional job-application talk at Stanford, after Stanford had made him what was likely the fastest job offer in its history. Amos presented prospect theory. “I noticed that I felt nothing but pride for him,” said Danny. “I noticed it because envy would have been natural.” When Danny left Palo Alto for Vancouver for the start of the 1978–79 academic year, he was even more aware than usual of the serendipity of life. His two children were now on the other side of the world, along with his old lab, a department full of former colleagues, and a society to which he once assumed he belonged. He had left behind in Israel a ghost of himself. “The background to what I was thinking was that I had just changed my life,” he said. “I’d changed my wife. The counterfactuals were with me all the time. I was constantly comparing my life to what it might have been.”