Priceless: The Myth of Fair Value (and How to Take Advantage of It) (17 page)

They therefore grilled the responders on what offers they would accept. This involved a series of yes-or-no questions. (“If the other player offers you $0.50, will you accept the offer or reject it?”) This approach is known as the strategic method of playing the ultimatum game, now widely used. In effect, it reveals the responder’s reserve price.

The results were similar to Güth’s, half a world away. An even split was the most popular offer, and the average amount offered was about $4.50. Responders were willing to reject offers less than about $2.30.

The vetoing responder is the game’s starkest challenge to economic theory. “It’s the resentment, the willingness to punish at cost, that is the whole thing,” Kahneman explained. The player who vetoes is rejecting logic no less than “free money” and making an economic decision on the basis of emotion. And it wasn’t just an occasional subject who acted contrary to the theory; practically everyone did.

“The thing that’s truly bewildering,” said Kahneman, is that “the theory can stand for hundreds of years, unchallenged, until someone says, ‘look at the emperor, no clothes.’ The counterexample was trivial.”

“Is the Ultimatum Game the Ultimate Experiment?” asked the title of a 2007 paper by Yoram Halevy and Michael Peters. They were referring only half-facetiously to the academic industry the game has become. The ultimatum game is claimed to be one of the most frequently performed of all human experiments today. Psychology and economics grad students are often assigned it as a training exercise in recruiting subjects, getting consent forms signed, and doing chi-square tests. Yet the main reason for its enduring popularity is the belief that the game tells us much about the psychology of prices and bargaining.

What does the game mean, and why should we care? As Güth now sees it, the game has two messages: that “money alone does not rule the world” and that “simple games can be very complex.” Kahneman views the game as a milestone in establishing the importance of psychology in
even simple economic decisions. “Something special had to happen for economists to pay attention,” he explained. “The ultimatum game had that feature.”

One reason economists paid attention is the evident parallels to price setting. The $10 can represent the potential profit (“surplus”) on a sale. The person splitting the money is a “seller,” and the responder is a potential “buyer.” The seller may choose to keep all the profit for himself (set a high price) . . . or surrender all the profit to the seller (“sell at cost”) . . . or share the profit with the buyer. The buyer decides whether to accept the price or reject it as too high.

The game can also be seen as a bare-bones model of negotiation. Lemuel Boulware, General Electric’s labor negotiator in the 1950s, was notorious for putting a wage package on the table and refusing to budge.
It was not GE policy to negotiate
. Boulware’s offers were chosen after much research. They were apparently intended to be the minimum offer that the union leaders would accept, albeit through gritted teeth. Boul-ware (and the many labor negotiators who attempted to emulate him) was acting like a strategic proposer in the ultimatum game.

The more usual back-and-forth kind of bargaining can be thought of as a sequence of ultimatum games. Offers on real estate are structured as ultimatums: This offer must be accepted by 6:00 p.m. Tuesday, or it’s null and void. Unless you accept the latest offer, you run the risk that the other side will walk away.

Bargaining is often a polite, socially sanctioned ritual. I lower my offer and you raise yours in stairstep increments. We meet somewhere in the middle. Sometimes fake “ultimatums” are part of that ritual. “That’s my final offer, take it or leave it. I’m leaving . . . I’m actually walking out the door . . .” Each side may know the other isn’t serious.

The crux of negotiation is how to deal with tough bargainers making lopsided demands. The ultimatum game presents, in concentrated form, the truly
difficult
part of negotiation. Where one or more hard-line bargainers are involved, there must come a moment of truth in which feints, bluffs, and built-in bargaining room are cast aside, leaving only an ultimatum. What do you do then—allow yourself to be exploited, or walk away, leaving money on the table?

New York governor Nelson Rockefeller had the perfect Fifth Avenue penthouse with a panoramic view of Central Park. He also had a problem. There were plans to put up a skyscraper public housing project on the West Side. It would have been a big middle finger blocking Rockefeller’s sunset view. The sponsor of the housing project bill was Meade Esposito, last of the cigar-chomping Democratic party bosses. Rockefeller invited Esposito to his penthouse to discuss the matter as gentlemen. “If you stop construction of that skyscraper,” Rockefeller announced, “I’ll give you that Picasso.”

He pointed to one of the modernist works on the wall. Esposito agreed to do what he could. The skyscraper was never built, and Rockefeller made good on his promise. Esposito got a Picasso, and Rockefeller got a story to tell for the rest of his life. For years afterward, Rockefeller lovingly recounted every detail of the bribery, capping it with the punchline: “It was only a print!”

Negotiation isn’t a pretty picture. Much of the time, the skillful negotiator is the one who best misrepresents value. It is not “fairness” so much as the appearance of fairness that drives the psychology of prices. (In the name of journalistic fairness, I must add that Rockefeller was not the lone scoundrel. Esposito may not have known art, but he knew the art of the deal. In 1987 he was convicted of influence peddling, resulting in a $500,000 fine and a two-year suspended sentence.)

Some of the early commentaries on the ultimatum game experiments mentioned
altruism
. Proposers don’t stiff the responders. They typically
offer a little more than they have to, to get the statistically average responder to okay the deal (through gritted teeth). The game therefore demonstrates an innate and noble generosity.

You still come across this interpretation in some feature stories. Sad to say, this happy notion has mostly been torn to tatters by later research. Bargainers are indeed less concerned with fairness than with what other people will think.

Kahneman, Knetsch, and Thaler took on the altruism question in their first article on the game. They devised what is now called the dictator game. Psychology students at Cornell were given $20 to split with an unknown stranger. The money was divided as the proposer—uh, “dictator”—decreed. That was it; the other player had no say.

In this first experiment, the dictators were allowed only two options. They could be greedy and keep $18 for themselves ($2 for the partner), or be fair and split the money evenly. Seventy-six percent of the subjects opted for the even split.

Kahneman’s group described this result with the noncommittal “resistance to unfairness.” Dictators avoided being unfair. Altruism could be one explanation for that, though not the only one.

The researchers further explored that resistance in another game, “altruistic punishment.” After a round of the dictator game, new subjects were presented with this choice:

(a) They could share $12 evenly with one of the players in the previous dictator game experiment. The player they were sharing with had been a “greedy” dictator (taking $18 for himself, leaving $2 for the other person).

(b) They could share $10 evenly with a different player in the dictator game. This player had been “fair” (chosen the even split).

The majority chose (b). They were willing to penalize themselves a dollar in order to “punish” someone who had done nothing to them personally—but was known to be an “unfair” player.

So far, this sounds encouraging. Dictators were mostly fair, and those who weren’t got their comeuppance. The original dictator game at Cornell was limited, though. It allowed only two options, a fair split and
an
extremely
greedy one (keeping 90 percent of the prize). Not many were comfortable being
that
piggish. Since then, other researchers have done experiments where dictators are allowed a full range of splits. These experiments generally report that dictators are much less generous. When they are free to make any split, they offer around 30 percent on average to the powerless partner. About one in five dictators gives nothing.

Elizabeth Hoffman and colleagues at the University of Arizona performed the definitive dictator game experiment. Hoffman suspected that dictators were being generous only because someone was watching. The experimenter is often the subject’s teacher, someone who would be grading him for months to come. Is it worth a few dollars to have the professor know you’re a greedy bastard?

Hoffman’s group therefore took pains to ensure that no one would know how any specific participant had acted. Each dictator was handed a plain white envelope and directed to the back of the room. There he opened the envelope inside a cardboard box to shield it from prying eyes.

Most of the envelopes contained ten one-dollar bills and ten blank slips of paper cut to dollar-bill size. The dictator was to take as many of the ten bills as he wanted for himself, leaving the rest inside the envelope for his partner. He was then to remove enough blank slips so that the envelope would contain exactly ten pieces of paper (blanks plus bills). This done, he handed the envelope to a “monitor” (who was not the experimenter, and who could not infer anything from the weight or feel of the envelope). The monitor took the envelope into another room to give to the partner.

The scheme’s pièce de résistance was that everyone was informed that a few of the original envelopes contained no bills, just twenty slips of paper. An unlucky dictator who got such an envelope would have to remove ten slips and leave the other ten slips for the partner. The upshot is that even a partner who received nothing could not conclude that he or she had been intentionally stiffed.

Under these conditions, about 60 percent of dictators took
all ten bills
for themselves, leaving only blank slips of paper in the envelopes.

There is no point in being shocked. (No IRS examiner would be.) Concepts like “greedy” and “generous” always depend on a frame of reference.
Right this instant
, you have the opportunity to share the money
in your wallet with a fine charity such as Doctors Without Borders. I’ll give you their address: PO Box 5030, Hagerstown, MD 21741-5030,
www.doctorswithoutborders.org/donate
. You really should donate . . . but no one will be the wiser if you keep practically all of the money for yourself. No one will know if you skip it and don’t send anything at all.

The pessimistic interpretation of Hoffman’s experiment is that it shows how hypocritical people are. When no one was watching,
but only then
, subjects were nearly as selfish as the economists postulated. Colin Camerer and Richard Thaler proposed an alternative interpretation: The outcomes of ultimatum and dictator games have less to do with altruism than with manners. Social norms of fair play are not easily cast off. Even “hypocrisy” is not always a bad thing. Sometimes, just by pretending to be a better person than you are, you end up being that better person, for all intents and purposes.

The ultimatum game has become an ur-experiment, the scientific equivalent of a catchy riff that lends itself to endless sampling and remixing. It has been played with members of the globe’s diverse cultures; with children, the autistic, the high-IQ, and men having exceptionally high levels of testosterone; with players who have been given a hormone that increases trust in strangers; even with chimpanzees splitting a prize of ten raisins. The game’s continuing fascination rests on how behavior changes, or doesn’t change, with context. Like a well-oiled weathervane, the experiment’s archetypic economic choice is sensitive to subtle pressures that affect us all the time yet usually go unnoticed.

Many simple variables dramatically affect behavior. Elizabeth Hoffman’s group at the University of Arizona did a set of trials in which players had to earn the right to be the proposer by winning a trivia quiz. This made proposers
less
generous. They apparently felt they had merited their position of privilege and had a right to the perks. The responders agreed, it seems. They were willing to accept less from someone who had won the proposer role fair and square. Most proposers offered $3 or $4, and these offers were never rejected.

Hoffman’s team also tried presenting the game as a retail transaction. The proposer was called a “seller,” the responder a “buyer” who had to decide whether to buy at the seller’s price or pass. Both parties received a table telling how they would profit from any possible price. The payoffs were identical with the standard ultimatum game.

This shouldn’t make any difference (for a rational actor), but it did. Sellers were greedier, usually allotting $3 or $4 to the buyer. Yet the latter
usually bought. Apparently, participants felt that sellers had a right to set a price. A high price was judged less worthy of punishment than an unequal split in the standard presentation.

One of the most interesting findings of Hoffman’s experiments was that proposers and responders were mostly in sync. When presented with a newly minted variation of the game, proposers instantly sensed how much more or less they should offer, and responders adjusted their expectations too. This happened without communication.