Priceless: The Myth of Fair Value (and How to Take Advantage of It) (11 page)

The journalist doubtless got that cold-war spin from Edwards, a RAND Corporation consultant and advisor to governmental agencies. Edwards talked up the Las Vegas game as “one of the few decision-making experiments ever conducted.” Never was it mentioned in the article that this particular game was devised not by Edwards but by two of his former students.

Sarah Lichtenstein had heard that Edwards had an “angel.” This was the attorney and casino backer Charles B. G. Murphy. Over a period of years, Murphy’s Wood Kalb Foundation disbursed several hundred thousand dollars—a fortune at the time—to Edwards. Edwards in turn used it to finance his own research and that of colleagues, much of it done at the Four Queens with the permission of Benny Goffstein and his successor,
Thomas Callahan. The idea was to present decision experiments as casino games. Players would bet money out of their own pockets, and they’d be playing for keeps.

Lichtenstein thought the preference reversal experiment “would be perfect for Vegas.” One criticism of the original study was that the subjects might not have been motivated to make sound decisions. College kids doing repetitive experiments for little or no cash get bored. After a while, they may not even try. Outside the lab, people are motivated to devote more time and attention to a decision when the stakes are high. A Las Vegas trial would be an acid test of whether preference reversal was for real.

The main hitch turned out to be the Nevada Gaming Commission. Any gambling “experiment” conducted in a casino had to be approved. Edwards was called in to meet Wayne Pearson, the commission’s head. Lady Luck was smiling on him: Pearson turned out to be a psychologist, with a Ph.D. from Cornell, who had read Edwards’s work. He quickly green-lighted the project.

For ten weeks in 1969, the Four Queens offered the best odds in Vegas—a fair game with no house advantage. “Stakes and Odds,” as the experiment was called, occupied a balcony space within earshot of a lounge band and the clatter of a casino restaurant. It used a standard roulette wheel, chips, and layout. Pit boss John Ponticello played the role of croupier. Hulking behind him was a PDP-7 minicomputer the size of several tall bookcases. The monitor, looking like an Ed Wood prop, was hexagonal with a circular screen. Any profits were promised to go to a home for unwed mothers.

Lichtenstein and Slovic spent only a few days in Las Vegas. Lichtenstein played the game herself to check up on the dealer. Unlike roulette, Stakes and Odds was strictly solitaire. Because the game was completely unfamiliar and required choosing among and pricing 40 bets, Ponticello had to warn each player that a complete game would take anywhere from one to four hours. For scientific validity, they needed players to go through the whole game. Those who didn’t want to make that time commitment were discouraged from playing.

At the outset, each player was asked to buy 250 chips. He or she got to
name the value of the chips, anywhere from 5 cents to $5 each. There weren’t any high rollers; nobody chose a higher denomination than 25 cents. In the first stage of the game, the player chose between pairs of bets presented on the computer’s monitor, indicating a choice by hitting a set of push-buttons set into the roulette table. The customer then played the chosen bet on the roulette layout. The win probabilities were all divisible by 12, to fit a 36-number roulette layout. Ponticello spun the wheel, tossed the little ball, and called the number. (Any zeros that came up didn’t count. Ponticello ignored them and spun again.) Winnings were paid, losses taken.

In the second stage of the game, players named prices for bets. These prices could be positive or negative, as half the bets were in the house’s favor, the other half in the player’s favor. (The game as a whole had no net advantage for the house.) It can be tricky to get a gambler to name an honest price. We are all so used to bargaining that we instinctively shade asking prices up and offering prices down, figuring we can always come down or up later. This is potentially a serious problem in this kind of experiment. Lichenstein and Slovic needed their off-the-street subjects to name a candid price X, such that they’d be happy to sell the bet for X or anything more, but would truly prefer not to sell for anything less than X.

To ensure candor, they used the Becker-DeGroot-Marschak system, which is a good deal simpler than its name. This is the protocol they used in some of the lab experiments as well. A seller (of a bet, or of anything) is asked to state an honest minimum price. The dealer then spins a roulette wheel to generate a random “bid.” Should the bid be higher than the stated reserve price, the sale goes through at the randomly selected bid price. (The seller is happy because she gets more than her minimum.) If the bid is lower than the seller’s price, there’s no sale. (The seller is happy because she wouldn’t want to sell for less than her honest minimum.) The best strategy here is to name your honest price.

By Las Vegas standards, Stakes and Odds was a flop. According to Slovic, casino patrons like simple, repetitious bets like slot machines. This game was
hard
. Ponticello kept wanting to “improve” the game, over the psychologists’ dogged insistence that he stick to the rules. The game did succeed at drawing the curious. Ponticello noticed that it attracted a diverse crowd: an Air Force pilot, a mathematician, a TV director, college students, a sheep rancher, a computer programmer, a
bus line ticket agent, a real estate broker, and seven fellow Las Vegas dealers.

They managed to get 86 games started. Because some players quit out of boredom or befuddlement, 53 games were completed. That was more than enough.

“The results of this experiment,” Lichtenstein and Slovic reported, “were strikingly similar to the findings of previous experiments based on college students gambling with hypothetical stakes or small amounts of money.” The downtown Las Vegas crowd preferred the P bets when choosing, yet priced the $ bets higher. This time, it was the players’ own money on the line. The highest winning for the complete game was $83.50, and the greatest loss was $82.75. (These figures are around $500 in today’s dollars.) Though the game was fair, the average player lost money to the house. That was the money pump in action.

“There is a natural concern that the results of any experiment may not be replicated outside the confines of the laboratory,” the psychologists wrote. What they learned from Las Vegas was expressed in a masterpiece of understatement: “the widespread belief that decision makers can behave optimally when it is worthwhile for them to do so gains no support from this study.”

From today’s perspective, Lichtenstein and Slovic started a revolution. The preference reversal experiment can with some justice be compared to the Michelson-Morley experiment in physics. That experiment refuted the absolute velocities of nineteenth-century physics, laying the groundwork for Einstein’s relativity. It is tempting to draw a parallel between the physicists’ “ether” and economists’ utility. Both were invisible, impalpable, tasteless somethings that “existed” because everyone assumed they
had
to exist. By showing that there are no invisible valuations dictating all economic decisions, Lichtenstein and Slovic heralded the relativity of prices—a keystone of what would be called behavioral economics.

Lichtenstein and Slovic proposed a simple explanation for preference reversals: anchoring. When asked to price bets, players direct their attention to the prize amounts. The most likely or biggest prize amount becomes a starting point or anchor. The players knew they had to adjust
from the anchor to take into account the probabilities and any other prizes or penalties. This adjustment required tough mental math. Everyone cut corners and guesstimated, with the result that the adjustment was usually inadequate. The final answer was too close to the anchor. It was an acorn that didn’t fall far enough from the oak.

Asking people to
choose
between bets activated a different thought process. Dollar amounts are less relevant, since many gambles are long shots. Obviously, everyone likes winning. There’s a strong tendency to pick the bet most likely to provide that happy outcome. Here, too, players tried to make allowances for dollar amounts and other complicating details. Once again, the adjustments tended to be inadequate.

Amos Tversky and Paul Slovic later generalized this idea into a “compatibility principle.” This rule says that decision makers give the most attention to information that is most compatible with the required answer. Whenever you have to name a price, you will focus on prices or other dollar amounts in the problem. In deciding how much to offer for a used car, Kelley Blue Book value and prices on Craigslist command attention. Everything else that ought to matter (condition, repair history, color, options, whether you
want
the options) gets short shrift. The latter factors are not so easily mapped onto the dollar scale.

Lichtenstein and Slovic used shifting attention to engineer an “impossibility.” The players were convinced their choices and prices had been reasonable throughout, and that they had not been tricked into saying anything they didn’t mean. Yet their values were suddenly revealed to be topsy-turvy. The final twist was the money pump—presto, change-o, your money disappears.

However marvelous the illusions of a conjurer, we know that the woman is not cut in two; the jet plane does not vanish. When perceptions contradict the laws of physics, physics is right and perceptions are wrong. The audience goes home convinced that things are the same as they’ve always been, that the good solid core of reality has not been breached.

No such reassurances are possible with the preference reversal experiment. Nobody is more of an expert than I am on what I want and how much I’m willing to pay for it. In such matters, honest convictions are the only underlying reality there can be. The “illusion” of preference reversal is genuine.

•   •   •

Magic is only one of many metaphors that have been used in coming to terms with Lichtenstein and Slovic’s finding. Another popular trope says that valuations are constructed and not revealed—like architecture, not archaeology. To name a price is to
build
a valuation (rather than to excavate deep into the psyche and uncover one).

A 1990 paper by Amos Tversky and Richard Thaler took its imagery from America’s great wellspring of metaphors, baseball. It involves the old joke about the three umpires:

“I call them as I see them,” said the first. “I call them as they are,” said the second. The third disagreed. “They ain’t nothing till I call them.” Analogously, we can describe three different views about the nature of values. First, values exist—like body temperature—and people perceive and report them as best they can, possibly with bias (I call them as I see them). Second, people know their values and preferences directly—as they know the multiplication table (I call them as they are). Third, values or preferences are commonly constructed in the process of elicitation (they ain’t nothing till I call them). The research reviewed in this article is most compatible with the third view of preference as a constructive, context-dependent process.

What
did
gain support was the relativity of prices. What people want, and how much they’re willing to pay, depends on the granular details of how you phrase the question. “It would be an overstatement to say of preferences, as Gertrude Stein said of Oakland, that ‘there is no there there,’ ” wrote legal scholar Cass Sunstein in this connection. “But frequently what is there is far less fixed, and far more malleable, than conventional theory predicts.”

Values may not be Oakland, but they are something like the elephant in the parable of the blind men. A man who feels the trunk reports that an elephant is like a snake; a man who feels the side says an elephant is like a wall; one who feels a leg compares the elephant to a pillar. “Each of the blind men was partly right,” says a character in an old Walt Kelly
Pogo
cartoon. “Yeah,” his friend adds, “but they were all mostly wrong.”

The Las Vegas experiment threw down the gauntlet. By using real people and real money, Lichtenstein and Slovic had invaded economists’ turf. Their experiment was a challenge to Paul Samuelson’s doctrine of revealed preference, a bulwark of modern economics. In some situations at least, revealed preferences weren’t so revealing at all. Choices failed to predict the prices people would pay. As Lichtenstein put it: “If you can’t talk about a preference, what the hell can you talk about?”

There was a knee-jerk, visceral rejection of preference reversal. “The first time I talked about it to a group of economists, I was astounded,” Lichtenstein recalled. “They were picking at it in trivial ways . . . asking these nit-picking little questions . . . It wasn’t until the economists jumped all over us—and the economists were jumping all over Amos and Danny—that I began taking seriously the incredible hostility.”

The vehemence of that reaction, and those to follow, may puzzle anyone who was not a part of it. It is worth saying a little about economists’ long, complex love-hate relationship with psychology.

Economists live in the same world as everyone else. They have friends who buy overpriced time-shares and brothers-in-law who just don’t
think
. Adam Smith devoted many words to human foibles and their inevitable influence on markets. Psychology was in the lexicon of economics until the Second World War. Then things started to change.

Under the influence of people like Samuelson and Milton Friedman, the field became progressively more mathematical. Much as dogs grow to resemble their owners, the new economics took on the features of
the people now building it. Economists embodied a math-smart, self-controlled stereotype and built theories describing people exactly like themselves.