Thinking, Fast and Slow

Kahneman was unfamiliar with decision theory when Tversky brought him on board the project to develop a new theory. Perhaps for that reason, he immediately saw the problem described in Chapter 25 and expected Tversky to educate him. Instead, it became clear that the question of what to do about the glaring error in existing theory—how to develop a new, more accurate and effective theory—was precisely the task that Tversky had in mind.

Unlike most theorists in the field, Tversky also intended to measure changes in wealth rather than absolute states. The approach was not entirely new, but it went against the overwhelming orthodoxy of decision theorists.

Further, Tversky and Kahneman realized that while most people would prefer a sure $900 over a 90% chance for $1,000 (as Bernoulli predicted), things changed when the matter was framed in terms of losses. That is, most people would prefer a 90% chance of losing $1,000 to a sure loss of $900. They were not the first to notice that a slate of negative options creates risk-seeking behavior, but the orthodoxy had yet to recognize the implications of this observation.

Although this idea was not present in their original work on prospect theory, here Kahneman explains how System 1 characteristics explain the phenomena he and Tversky observed. The first is that evaluation occurs relative to a neutral reference point (e.g., water is hot compared to something else, not in an absolute sense). Second, such evaluation follows a principle of diminishing sensitivity (i.e., water that is hot and water that is a few degrees hotter will not be perceived as very different from each other but as very different from cold water). Finally, System 1 displays loss aversion (that is, losses loom larger than gains, and the negative is perceived more directly and intensely), which results from the evolutionary reality that organisms that prioritize threat-response are more likely to survive.

The impact of these three cognitive tendencies shapes the insights of prospect theory. Kahneman provides a graph that illustrates a fairly simple point: the potential loss of some amount is perceived as far more impactful (or valuable) than a potential gain of the same amount (283).

The fundamental claims of prospect theory are that (1) people are risk averse where there is a mixed gamble (meaning one may gain or lose); and (2) people become risk-seeking when all options are bad (only losses are possible). The second point directly reflects the principle of diminishing sensitivity—that is why people will take a 90% chance of losing $1,000 over a certainty of losing $900.

Kahneman identifies two significant aspects of prospect theory that make it inaccurate in some situations. These include the use of a neutral reference point valued at zero (because no status quo results are actually felt in the same way) and the inability to account for regret (or other emotions) that actually influence many decisions. He also suggests, however, that more detailed theories accounting for emotions have generally lost significant utility to researchers in doing so.

The “endowment effect” relates to the fact that all gains or losses are evaluated on the basis of one’s history and current position. In explaining this concept Kahneman details how behavioral economics was born, introducing its founder Richard Thaler, the graduate economics student who used his professors’ irrational conduct to debunk their theories of rationality.

One of Thaler’s professors enjoyed collecting wine, and his behavior suggested that merely owning a bottle increased its value to him (which was irrational from the traditional perspective). Once the bottle was his, the professor would not sell it even for several times the price he paid for it. This is one example of the endowment effect that established Thaler as the founder of behavioral economics.

Thaler met one of Kahneman’s students by chance and obtained an advance copy of the article introducing prospect theory. Loss aversion as explained in the article solved the key problem that the endowment effect represents to traditional theory.

Thaler arranged to visit Stanford for a year when he knew Tversky and Kahneman would be there, and the three developed a very productive relationship. Thus, the economist Thaler’s interest in the endowment effect led him to the psychologists Tversky and Kahneman, which produced the movement toward behavioral economics that majorly influenced numerous fields, including law and several social sciences.

The picture of eyes on Page 301 may be invaluable to understanding the point of this chapter. One knows, instantly, that one set of eyes expresses terror, and the other does not. Why? The answer involves ancient evolutionary brain circuitry, which includes a direct line from the eyes to the threat-processing center of the brain that bypasses conscious recognition. This connection explains just how hard-wired and intense loss aversion is for the human animal.

The primacy of the negative is aptly expressed in an observation Kahneman borrows from psychologist Paul Rozin: A single cockroach destroys the appeal of a bowl of cherries, but a single cherry has no effect on a bowl of cockroaches.

The reference point for a prospect theory analysis need not be the status quo. Depending on how a person thinks about a situation, the reference point may be the achievement of a goal. Drawing on golf’s use of “par” as a reference point, Kahneman describes research that has demonstrated professional golfers do better when putting to avoid going over par than when putting to stay below it. Likewise, research shows that existing entitlements (such as to a current wage) are viewed as a reference point so that any reduction is perceived as a loss.

Working on prospect theory led to a clear pattern now known as the fourfold pattern, which is illustrated in a table on Page 317. In brief, it shows the following: With regard to high-probability events, people are (1) risk averse regarding gains but (2) risk-seeking regarding losses, and with regard to low-probability events, people are (3) risk-seeking regarding gains and (4) risk averse regarding losses. The two outcomes for high-probability events reflect what is known as the certainty effect, whereas the remaining two reflect the possibility effect. The pattern results when people assign value to changes rather than wealth itself and when people attach decision weights to various outcomes that are different from probabilities.

Chapter 30 discusses two general observations: that people overestimate the possibility of rare events and, when required to weight possible outcomes in a decision (as with a bet, for example), they tend to overweight the unlikely events. This is especially true where the outcome or event is easily imagined and contrasted with a more abstract alternative. Cognitive ease contributes, as does the confirmatory bias of memory where there has been a focus on the possibility of the event. These patterns are not hard rules but reflect a tendency applicable in most situations.

Kahneman explains that it is costly to be risk averse for gains and risk-seeking for losses, and that framing decisions broadly or narrowly both illustrates the point and can also be used to significantly affect most people’s choices. Combing multiple decisions into one broad decision tends to reduce risk aversion. This is reflected in the culture of stock traders, who insulate themselves from the emotional pain of losses by creating broad framings of their financial decisions. Deconstructing a decision into several narrow decisions causes greater risk aversion because that pain is experienced more frequently.

Traders make their decisions broader by limiting the number of times they check the status of investments, thus reducing their tendency to act on small losses due to risk aversion. This is an example of the type of risk policy that Kahneman suggests can play an important role for decision-makers who may run into problems caused by excessive risk aversion.

Chapter 32 discusses the “mental accounts” that people keep, which essentially use money as a proxy for emotional gains and losses. The effects of these underlying drivers of many choices are predictable and observable.

The disposition effect, known especially to financial researchers, leads people to sell a stock that is up rather than an alternate stock that is down the same amount, presumably because gaining feels better than solidifying a loss. Yet, for tax reasons, people would actually gain more by selling the loser. The tax-based reasoning does affect people’s decisions in one out of 12 months. The other 11 months reflect the disposition effect even though the tax benefit is available all year. The disposition effect results from narrow framing—in this instance, framing is reflected in the arrangement that requires a decision on each stock rather than treating multiple stocks as one broad account that requires only one decision.

Mental accounting is also apparent in the sunk cost fallacy that causes people to make decisions driven by the extent of their previous investment rather than by rational analysis of probable future outcomes. Factors such as regret and blame operate similarly, as do the distinctions that people draw between commission and omission of actions.

This brief chapter illustrates the role of framing. Kahneman shows that preferences can be reversed depending on how questions are presented. Again, it is clear that broader and more comprehensive decisions are more likely to be guided by rationality. Thus, narrow framing can and does produce gross injustices.

Framing relates to how the presentation of information affects the way it is perceived. To illustrate, Kahneman uses two logically equivalent sentences that evoke different associative responses from System 1.

Regarding the 2006 World Cup soccer match, one could say “Italy won” or “France lost.” Both phrases are logically correct and have the same meaning in terms of the game’s outcome. Yet they have very different “meanings” in terms of the System 1 associations that are generated.

Kahneman explains what has become known as the “Asian disease” problem, which he and Tversky used to test framing effects. It turns out that even the public health officials who actually make decisions about such questions are affected by framing—a fact Tversky learned when he was invited to speak to them and arranged an experiment.

In the problem, the subject is told that a new variant of a disease has been identified in Asia and will arrive soon. The death toll is expected to be 600 if no action is taken. Two actions are possible. From there, two different presentations are given.

One group of subjects, Group A, is told that if the first program is adopted, 200 people will be saved, and if the second program is adopted, there is a one-third probability that all 600 people will be saved. Group B, the other group of subjects, is told that if the first program is adopted, 400 people will die, and if the second program is adopted, there is a two-thirds probability that 600 will die.

A majority of subjects in Group A chose the sure thing (the first program), whereas a majority of subjects in Group B chose to gamble (the second program).

As Kahneman emphasizes, framing effects affect very real and significant decisions in our lives and in society, such as public health choices. Similar examples have been demonstrated in numerous policy areas, such as the tax code.

Because of these differences, which can be enormous in the aggregate, Kahneman argues that there some frames which are better to adopt than others. At the very least, framing choices are important to advancing a particular purpose or policy goal.