Priceless: The Myth of Fair Value (and How to Take Advantage of It) (13 page)

The Cornsweet illusion is an open-ended metaphor. People are the same all over; boundaries of various kinds make us think we’re different from “them.” On a more mundane level, it demonstrates the leitmotif of psychophysics. Contrasts matter, and absolute values don’t. It is perhaps not too much of an exaggeration to say that some of Kahneman’s most
celebrated papers applied this general principle to price setting and other types of decision making.

The “most significant intellectual experience” of Kahneman’s grad school years came on a road trip his first summer in America. Kahneman drove to the Austen Riggs Clinic in Stockbridge, Massachusetts, a psychoanalytically oriented asylum for the wealthy. The staff included such famous analysts as Erik Erikson. Each Friday, the doctors met for a group interview of a patient, followed by a freewheeling discussion of the case. Kahneman was permitted to sit in on some of these conferences. The one that lingered in memory was typical except for the fact that the patient didn’t show up. He had committed suicide the night before.

“It was a remarkably honest and open discussion,” Kahneman mordantly observed, “marked by the contradiction between the powerful retrospective sense of the inevitability of the event and the obvious fact that the event had not been foreseen.”

Tversky and Kahneman found they shared a skepticism about the wisdom of experts in psychology or anything else. Tversky mentioned Ward Edwards’s experiment with poker chips. The subjects had failed to appreciate how informative a single poker chip could be. Kahneman countered that the opposite bias was more common. His army psychologists believed that a single data point—performance on the telephone pole test—could predict a future military career, when such things are really not very predictable at all.

Jovial back-and-forth led to the pair’s writing a six-page semi-humorous article for the
Psychological Bulletin
, “Belief in the Law of Small Numbers.” The title is a play on the “law of large numbers” of probability theory. This says that flipping a fair coin a large number of times will give you a percentage of heads close to 50. That is all you can ask of a fair coin. You can’t predict the outcome of a small number of tosses. However, Tversky and Kahneman noted, people want to believe just that. They suppose that flipping a coin ten times will yield five heads and five tails, or something close to it. In reality, lopsided outcomes (like eight heads and two tails) are more common than people believe. Tversky and Kahneman surveyed some mathematical psychologists at a meeting and found that even the experts were subject to this error. The article’s most memorable line displays a playful wit rarely encountered in scientific papers: “People’s intuitions about random sampling appear to satisfy the law of small numbers, which asserts that the law of large numbers applies to small numbers as well.”

This modest paper, published in 1971, inaugurated a decade of intensive collaboration so productive that friends called the pair “the dynamic duo.” Since it was impossible to determine who had contributed more to a given paper, they flipped a coin to determine whose name came first in a publication’s byline.

“There was a lot of irony in our whole research program,” Kahneman told me. “It was not an attack on humanity, it was an amused and ironic look at ourselves.” Tversky had a funny line, or a funny story, for every situation. “In his presence, I became funny as well, and the result was that we could spend hours of solid work in continuous mirth.”

Amos “was the opposite of Danny” in Barbara Tversky’s analysis. “He was a perfectionist in everything, including the words. Amos always wanted to get it right, to do it over and over until it was right. Danny was always moving on to the next idea; he always had a wealth of new ideas.” Amos “couldn’t write a paper without having a title, and the title had to be just the right title.” He rectified any remaining deficiencies in his English by grilling Barbara for just the right word. “ ‘Is it this or this? This or this?’ he’d ask. ‘It’s your language!’ ” Barbara protested. “You’re looking for words that don’t exist!”

Kahneman and Tversky spent the 1971–72 academic year at the Oregon Research Institute. Paul Hoffman was an adept fund-raiser, and ORI then had “a pile of money,” as Kahneman remembered. “We had no schedule, no classes.” He rated the year at ORI “by far the most productive of my life.”

He and Tversky quickly settled in the routine that would define their collaboration. As Kahneman was a morning person and Tversky a night owl, they dovetailed their schedules by meeting for lunch and an afternoon of work. For the most part, “work” meant talking.

“They were so
verbal
,” said Sarah Lichtenstein. “I remember once, with Amos and Danny and Paul, I put my hand up to speak. There were just the four of us—I couldn’t get a word in.” The group at Oregon tossed around ideas such as anchoring, preference reversals, and intuitive conceptions of probability. These discussions grew into Tversky and Kahneman’s now-famous article “Judgment Under Uncertainty: Heuristics and Biases.”

It was a long time in gestation. After the year in Oregon, both Kahneman and Tversky returned to Israel and spent much of the next year hammering out every precious word. They were simultaneously doing research. The paper is essentially a review article citing recent results by the two authors and others. Published in
Science
, “Judgment Under Uncertainty” immediately reached an audience outside the field of psychology. In so doing, it ignited a firestorm of controversy that has only now started to cool.

A “heuristic” is a rule of thumb, something like “No matter how much you’re offered, you can probably get 10 percent more.” The paper discusses three more fundamental examples, named
representativeness
,
availability
, and
anchoring and adjustment
. Anchoring has the most to do with prices. Let me briefly explain the other two.

The best-known example of the representativeness heuristic is “Linda the Feminist Bank Teller” (who makes her first appearance in a later paper, from 1983).

Linda is 31 years old, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.

In a study at the University of British Columbia, 142 undergraduates who read this capsule description were asked which of the following was more likely to be true:

Linda is a bank teller.

Linda is a bank teller and is active in the feminist movement.

Eight-five percent rated the second statement more likely than the first.

That’s
ridiculous
. The only way Linda can be a bank teller
and
a feminist is if she’s also a bank teller. At the risk of beating a dead horse, I’ll draw you a diagram (opposite).

Apparently, in judging how likely it is that Linda is a bank teller, people look at how well the information we have about Linda fits our preconceived notion of bank tellers. The question was written so that Linda fits the stereotype of a feminist and doesn’t fit the stereotype of a bank
teller. Hunches about Linda defied logic. Those hunches were amazingly tenacious, though.

Tversky and Kahneman resorted to “a series of increasingly desperate manipulations” intended to get their subjects to obey simple logic. They tried giving volunteers the Linda problem, followed by two arguments about what the answer
should
be. The subjects didn’t have to commit to an answer, just to say which argument they believed was more convincing.

Argument 1. Linda is more likely to be a bank teller than she is to be a feminist bank teller, because every feminist bank teller is a bank teller, but some women bank tellers are not feminists, and Linda could be one of them.

 

Argument 2. Linda is more likely to be a feminist bank teller than she is likely to be a bank teller, because she resembles an active feminist more than she resembles a bank teller.

Even in this case, 65 percent favored the second argument. After the survey, when the matter was explained at length, many subjects remained unconvinced, uncertain, or unrepentant. Said one in defense of his answer, “I thought you only asked for my opinion.”

•   •   •

Which is more common, words that begin with
r
(like “road”) or words with
r
as the third letter (like “car”)? Most say that words beginning with
r
are more common. It’s easy to rattle off words beginning with
r
; harder and slower to free-associate words with
r
in third place. This is an example of the availability heuristic, and here it leads us astray. Words with
r
in third place happen to be more common. But because words beginning with
r
are more mentally available, we overrate how common they are.

A familiar example of availability is the way we all assume that the tastes, politics, education level, and TV viewing habits of our social set are widely shared. We marvel when such-and-such show is a hit or so-and-so gets elected. “Nobody would vote for that jerk!” Well, they
did
.

Another example: Every year, thousands of kids aspire to become a pro athlete, despite long odds and near-certain disappointment. Why? It’s easy to list names of athletes who beat the odds and became rich and famous. Now try to name some guys who went out for the NBA or NFL and never made it. Can you name any? Hmm, maybe the odds aren’t so bad after all . . .

Anchoring and adjustment had already been proposed as a cause of preference reversal. In Lichtenstein and Slovic’s experiment, the anchors—prize amounts—were at least relevant to the value of the gambles. Tversky and Kahneman supected that anchoring would work even with anchors that were known to be irrelevant. To test this hypothesis, they devised the United Nations experiment. The wheel of fortune was stagecraft, a way of emphasizing that the anchor numbers were completely random and meaningless. The anchors worked nonetheless. Of all of psychology’s challenges to rationality, anchoring is “the easiest to demonstrate,” wrote Fritz Strack and Thomas Mussweiler, but “the hardest to explain.”

The United Nations experiment has become the classic demonstration of anchoring. Yet the 1974
Science
paper is the only article to report it, and it gives scant data. Tversky and Kahneman published other, more detailed papers on representativeness and availability—but not on anchoring. “Amos and I didn’t quite agree on the interpretation of anchoring,” Kahneman explained. “The question was whether it’s adjustment
or (in modern terms) it’s priming. Amos liked the idea of actual adjustment.”

Tversky’s idea was this. When asked to guess the percentage of African U.N. members, people start at an anchor value (the number that came up on the wheel of fortune) and adjust it downward or upward. They would continue adjusting until they reach the outskirts of a broad, fuzzy zone of plausible values. Then they’d stop. The stopping value will be on the anchor’s side of the plausible zone. The greater the uncertainty, the broader the zone and the greater the anchoring effect.

It’s as if I asked you to go get me a hamburger. You would probably stop at the first hamburger place you saw and bring me one of their hamburgers. You wouldn’t search the whole city for the
best
hamburger.

By Tversky’s theory, people adjusting from an anchor stop too soon. Instead of racking their brains for the “best” answer, they settle for the first plausible answer they come to. With a high anchor, that answer will be too high, and with a low anchor, it will be too low.