Priceless: The Myth of Fair Value (and How to Take Advantage of It) (14 page)

Originally, Tversky instructed the participants to start at the wheel of fortune number and mentally adjust it upward or downward. This instruction embodied what Tversky thought was actually happening. It’s now clear that this instruction is unnecessary. The important thing is that there be some kind of mental comparison between the anchor and the quantity to be estimated. This happened naturally in the preference reversal experiments. When the anchor is blatantly uninformative—a random number or an obviously wrong one—a comparison can be prompted by asking a preliminary question of the form “Is the percentage of African U.N. members more or less than 65?”

Tversky’s adjustment theory can’t readily account for the anchoring seen with plausible values. Was the year of Einstein’s first visit to America earlier or later than 1939? German students who heard this question gave a later estimate of the year than those asked whether Einstein’s first visit was before or after 1905. Both dates are reasonably believable (the real date was 1921). It should not be necessary to adjust a figure that is already plausible.

Many other explanations for anchoring have been proposed. It has been argued that anchoring is logical, that subjects clutching at straws reasonably take a “conversational hint” from the experimenter. The experimenter would not be asking whether Einstein visited America
before or after 1939 unless that was a reasonable answer. Therefore, you can’t look too stupid by naming a year close to 1939.

Kahneman had a distinct theory. “I didn’t know about priming,” he said, explaining that “the term didn’t exist. But I was pushing an idea that was very close to priming, that it’s a suggestibility thing.”

“Priming” is a fairly new term for phenomena that have long been part of the world’s store of knowledge, not necessarily of the scientific kind. Have you ever bought a car and suddenly noticed that “everyone” on the freeway, practically, is driving that model? Have you ever learned a new word (or heard of an obscure sea mammal or an ethnic dance) and then encountered it several times in the space of a few days? You come across it in the news, you overhear it mentioned on the bus and on public radio, and the old issue of
National Geographic
you’re thumbing through falls open to an article on it . . .

This is priming (fortified with a few low-grade coincidences). When you skim the newspaper, half-listen to TV, or drive on the freeway, you ignore most of what’s going on around you. Only a few things command attention. Paradoxically, it is unconscious processes that choose which stimuli to pass on to full consciousness. Prior exposure to something (priming) lowers the threshold of attention, so that that something is more likely to be noticed. The upshot is that you have probably encountered your “new” word or car many times before. It’s just that now you’re noticing.

Priming affects not only what you notice but what you do. In this case, priming can be identified with the power of suggestion. A yawn becomes contagious in a boring meeting, a cough in a concert hall. Visit Scotland, or Alabama, and you start talking that way.

“When it comes to our behavior from moment to moment, the big question is, ‘What to do next?’ ” said the Yale psychologist John Bargh,
author of a number of papers on priming. That question is unlikely to have a clear-cut, logical answer. Instead, says Bargh, “we’re finding that we have these unconscious behavioral guidance systems that are continually furnishing suggestions through the day about what to do next, and the brain is considering and often acting on those, all before conscious awareness.”

In the current understanding of priming, words and other stimuli activate relevant mental processes. Once “switched on,” this cognitive machinery remains accessible for a while, influencing subsequent thoughts and actions. When priming affects the estimation of number values, psychologists call it anchoring.

“Anchoring effects are (mostly) caused by the fact that when I ask you if the tallest redwood is more than 800 feet tall, I have primed you to think of very tall trees,” explained Kahneman. “The sample of trees you recover from memory is biased upward.” You think of redwoods and giant sequoias and eucalyptus, and everything you ever read or half-remember seeing on the Discovery Channel about extremely tall trees. You think of things that are 800 feet tall, and reasons why a tree might or might not be that tall. All these trains of thought remain active and help to legitimize high estimates for the height of the tallest redwood. Meanwhile, other ideas must compete with the anchor for attention. The final answer is to some extent a weighted compromise between the values considered—reflecting any biases in the set of considered values. Even if you conclude (correctly) that no tree on this planet is anywhere near 800 feet tall, you cannot completely ignore what you’ve just been thinking. “What I tell you three times is true,” said the Bellman in Lewis Carroll’s “The Hunting of the Snark.” Thinking is believing—to a degree, anyway.

Some decisions are the result of logical deliberation. These decisions come with an inner dialogue you “hear”—a dialogue that is often voiced aloud to spouses, accountants, agents, and corporate boards. The conscious mind flatters itself that
all
decisions are like that. Other decisions are completely unconscious, though, like whether to cough. Most important decisions fall somewhere in between these two extremes.

Despite their numerical nature, price decisions usually have a strong intuitive component. A price is not an answer to a math problem; it’s an expression of desire or a guess about what other human beings will do (accept your offer or turn it down). You name a price that “feels” about
right. As the rest of this book will show, price numbers are influenced by factors the conscious mind would reject as irrelevant, irrational, or politically incorrect.

In the film
The Usual Suspects
, Kevin Spacey plays a con man who fabricates a confession to a crime. His deception is exposed when his police interrogator swivels in his chair to gaze at the bulletin board behind his back. He realizes that every name or detail in Spacey’s story has been lifted from one or another of the posted memos. The detective is so shocked that he drops his coffee mug. Picking up the shards, he notices that the cup’s manufacturer, Kobayashi Porcelain, is identical with the name of an attorney Spacey has just mentioned.

The ability to confabulate—to tell stories smoothing over the rough edges of experience—is part of being human. The mind generates an ongoing fiction in which it knows more and acts more logically and nobly than it does in reality. We
believe
this fiction. Anchoring is one small part of it. We feign accuracy in mapping our feelings onto numbers or dollars. In truth, we are always grabbing handy numbers in the environment and transforming them into guesstimates and prices.

This somewhat unsettling idea raises an extreme possibility that University of Virginia psychologist Timothy Wilson called the
basic anchoring effect
. “There are many, many arbitrary numbers in our minds throughout a given day, such as the temperature that was just announced on the radio, the numbers on a computer keyboard that we just pressed, the numbers on the dial of the clock we just consulted, or the page numbers of a book or questionaire we read,” Wilson and colleagues wrote in a 1996 paper. “It seems unlikely that numbers considered as briefly as these would be used to make an unrelated judgment.”

Wilson and company tried to find out how subtle a “background” anchor could be. In one experiment, volunteers were given questionaires with adhesive notes attached. Written on each sticker was a four-digit “ID number” between 1928 and 1935. One group of participants was required simply to copy this number onto the questionnaire. They were then asked to estimate the number of physicians in the local phone book. The average estimate was 221 doctors.

The important thing here is that the ID code was just a number that
happened to be there, not a meaningful part of the problem. Other groups were given slightly different instructions that caused them to pay a little more attention to the ID number. Some were told to note whether the ID number was written in red or blue ink (on the pretext that this would determine which page of the questionnaire to fill out). For these people the average answer was 343 doctors. A split second’s extra attention to the number had raised the estimate 55 percent. (All the ID numbers were big. As anchors they would have pulled the estimates up.)

Another group was asked to note whether the ID number was in the range of 1920 through 1940 (as they all were). Unlike the question about ink color, this forced participants to consider the number as a number. This group put the number of doctors in the phone book at 527.

One group was asked a two-part question. They first had to guess whether the number of physicians in the phone book was greater or less than their ID number, and then give their estimate of the number of physicians. This group’s average was 755.

The anchoring effect was by far the strongest when people had to make an explicit comparison between their anchor and their estimate. Still, the anchor numbers “leaked through” and affected answers even when they were fairly peripheral.

The researchers later asked some participants whether they thought their judgments might have been influenced by the ID number on the adhesive sticker. The answer, overwhelmingly, was no. As Kevin Spacey says in
The Usual Suspects
(quoting Baudelaire): “The greatest trick the devil ever pulled was convincing man he didn’t exist.”

For part of the 1970s, Barbara Tversky’s husband drove her and the kids up the wall. “I would go batty because he’d be asking questions all the time. ‘Yeah, we had that question yesterday,’ ” she’d protest. “ ‘No, no, it’s different,’ ” Amos would insist.

Maurice Allais was to blame—his paradox, anyway. Early in his collaboration with Tversky, Kahneman decided that Allais’ paradox was the foremost unsolved problem in the psychology of decision making. The great prize would be to give a proper accounting of it. In an attempt to do that, he and Tversky began devising “interesting choices.” When a choice seemed interesting enough, the families became guinea pigs. One example:

Would you rather have $3,000 (a sure thing)
or
an 80 percent chance of winning $4,000 (and a 20 percent chance of ending up empty-handed)?

Just about everyone preferred the $3,000 sure thing to the gamble. That’s not surprising. Tversky had a clever idea about how to make things more interesting. “Reflect” a question by putting a negative sign in front of each dollar amount. To make it realistic, imagine you’re being sued for $4,000. Would you rather settle for $3,000 right now (a sure loss, –$3,000) or go to trial, knowing there’s an 80 percent chance you’ll lose and owe the full amount (–$4,000), and a 20 percent chance you’ll win and owe nothing at all? (Ignore legal fees.)

Were attitudes toward risk consistent, the answer to this second question seemingly ought to be the same as to the first. But that’s not the way most people think. A majority rejects the sure loss and prefers to take its chances with the trial. The train of thought is, “I don’t want to lose thousands of dollars, and the gamble is my only shot at avoiding it. The difference between a $3,000 loss and a $4,000 loss isn’t all that great.”

By posing many other questions of this kind, Kahneman and Tversky were able to probe attitudes toward gains, losses, and risk from all directions. They stopped annoying family members and started doing careful surveys of student volunteers. This led to their 1979 article, “Prospect Theory: An Analysis of Decision Under Risk.” The newly coined title was intentional, Kahneman said. “We reasoned that if the theory ever became well known, having a distinctive label would be an advantage.”

Prospect theory is founded on several simple, powerful ideas. One is the relativistic nature of money (or gains and losses generally). Kahneman and Tversky invoked the parallel to psychophysics:

Our perceptual apparatus is attuned to the evaluation of changes rather than to the evaluation of absolute magnitudes . . . an object at a given temperature may be experienced as hot or cold to the touch depending on the temperature to which one had adapted.

Likewise, people get used to a certain level of wealth or income and react mainly to changes. For instance: You expect your rich aunt to give you a $1,000 check for a wedding present because that’s what she’s given all your siblings. Instead she gives you a lousy $25 gift card! You are apt to feel you’ve “lost” $975 rather than gained $25.

In Kahneman and Tversky’s terminology, the anticipated $1,000 is a
reference point
. This is much like the “adaptation level” of psychophysics. The reference point determines whether something is entered as a gain or a loss on the mental ledger. That can make a huge difference in behavior.

A second key idea of prospect theory is
loss aversion
. Losing money (anything of value) hurts more than gaining that same thing delights. You can demonstrate loss aversion by offering a bet on a coin toss. Tails
you lose $100, and heads you win X. How big does X have to be for you to take the bet?

Surveys show that few want to accept a “fair” bet with X = $100. Few accept X = $110, which offers a nice expected profit. (Those who do accept at this price tend to be gamblers, arbitrageurs, or economists.) The average person requires roughly a $200 prize to balance the prospect of an equally probable $100 loss.