Thinking, Fast and Slow.

• Author: Stein, Alex
• Position: Book review

THINKING, FAST AND SLOW. By Daniel Kahneman. New York: Farrar, Strauss and Giroux. 2011. Pp. 3, 418. Cloth, $30; paper, $17.

INTRODUCTION

Daniel Kahneman's (1) recent book, Thinking, Fast and Slow, is a mustread for any scholar or policymaker interested in behavioral economics. Behavioral economics is a young, but already well-established, discipline that pervasively affects law and legal theory. (2)Kahneman, a 2002 Nobel Laureate, is the discipline's founding father. His pioneering work with Amos Tversky and others challenges the core economic concept of expected utility, which serves to determine the value of people's prospects. (3) Under mainstream economic theory, the value of a person's prospect equals the prospect's utility upon materialization (U) multiplied by the probability of the prospect materializing (p). (4) When the prospect is advantageous, its utility is a positive sum that augments the person's well-being. When the prospect is disadvantageous, its utility is a negative sum (a disutility) that decreases the person's well-being. Under both scenarios, the full amount of the person's utility or disutility is discounted by the prospect's probability of not materializing. Economic theory holds that the expected-utility formula, P x U, ought to determine a rational person's choice among available courses of action. The action yielding the highest expected utility is the one that the person ought rationally to prefer over the alternatives.

This normative assumption underlies all economic models that predict human behavior. For example, when a legal system apprehends 10 percent (0.1) of all drivers who run a red light and forces each violator to pay a $300 fine, the expected fine for each prospective violator equals $30. Economic theory consequently predicts that a risk-neutral driver will run a red light when her expected gain from doing so exceeds $30. Hence, when the average social loss from a red-light violation (including the marginal increase in society's cost of enforcing the law) exceeds $30, economically minded scholars and policymakers recommend upping the fine to a sum that will eradicate the drivers' antisocial incentive to violate. (5)

Kahneman and his collaborators have carried out numerous experiments that examined people's determinations of probability and utility. These experiments purport to identify a systematic mismatch between the P-U formula and the ways in which people typically make decisions in the real world. Specifically, the experiments have been interpreted as demonstrating that people systematically err in calculating probability and appraising utility. According to Kahneman and other behavioral economists, these errors manifest people's bounded rationality. (6)

As far as probability is concerned, people often seem to allow familiar and stereotypical scenarios to override statistical information (pp. 112-13, 119-23, Chapter Twelve). They ignore general statistical information-specifically, they ignore "base rates" (pp. 146-69)--which causes them to underestimate the probability of unfamiliar events (pp. 149-53) and overestimate the probability of scenarios that fall within their experience or easily come to mind, such as natural disasters or startling events that have occurred recently (Chapter Thirteen). In the domain of utility, people fare no better. They irrationally allow their choices to be influenced by the framing of future prospects as either "gains" or "losses" (Chapter Twenty-Six). Specifically, the average person strongly prefers the prospect of not losing a certain amount of money to an equally probable prospect of gaining the same amount of money. (7) This preference explains people's loss aversion (8) and their unwillingness to sell property that they already own at what should be an economically attractive price (the "endowment effect").(9) Additionally, people's actions are often driven by sunk costs--expenses that have already been incurred, and that an economically rational individual ought to ignore. (10)

These findings regarding bounded rationality have a far-reaching implication: they prompt policymakers and scholars to abandon the neoclassical homo economicus assumption used to predict people's choices and evaluate the effects of policy interventions. (11) Kahneman argues that scholars and policymakers would do well to shift their attention to the ways in which typical real-world people appraise probability and utility. These ways, according to Kahneman, exhibit misjudgments that are potentially harmful to the person making a decision and anyone else who depends on her decision.

These findings have won many adherents among legal scholars. (12) Their broad acceptance in the legal academy has led to the establishment of the discipline of Behavioral Law and Economics. (13) Scholars working in this discipline use these findings as a basis for recommending a variety of legal reforms: the mandatory supply of information to error-prone individuals, (14) "soft" choice architecture, (15) and regulatory intervention

to prevent and correct people's mistakes. (16) Areas that these reform proposals try to nfluence include accidents and risk regulation, consumer agreements, business contracts, credit and lending, employment, insurance, prenuptial agreements, and adjudicative factfinding. (17) Kahneman's book outlines and approves of some of these proposed reforms (pp. 412-15).

Two features--theory and accessibility--make Thinking, Fast and Slow an indispensable book. Behavioral economics has existed for nearly four decades and is unlikely to fade away. Yet, prior to the publication of Thinking, behavioral economics had developed no integrated theory of bounded rationality's causes and characteristics.(18) Instead of developing such a theory, behavioral economists engaged in largely experimental work that uncovered discrete manifestations of people's bounded rationality. These manifestations include "representativeness," (19) "availability," (20) "anchoring,,, (21) "overoptimism," (22) "base-rate neglect," (23) "hindsight bias," (24) loss aversion (pp. 283-88), and other misevaluations of probability and utility (pp. 222-29, 363-74). Behavioral economists had developed no causal explanations for these misevaluations.(25) All they had done was identify, describe, and categorize these misevaluations one by one.

Thinking does not content itself with an untheorized catalog of descriptions; it goes considerably further, and seeks to develop a unified explanation of causes and effects. Remarkably, it is the first book to provide a comprehensive and theorized, as well as fully updated, account of bounded rationality. Kahneman's account of bounded rationality combines experimental and empirical findings with a causal theory. This theory builds on two reasoning mechanisms that, according to Kahneman and other psychologists, define people's mental makeup. (26) One of those mechanisms, identified as System 1, relies on intuition (pp. 19-30, 89-96, 105). Another mechanism, identified as System 2, relies on deliberation (pp. 19-30).

A person using System 1 thinks fast. Her thinking invokes instincts and hunches, of which some are experience based and others are biologically hardwired (pp. 24-25). These instincts and hunches enable her to form quick and effortless responses to tasks, questions, and challenges (p. 25). By and large, she bases her responses on familiar causal associations, while relying on the "What you see is all there is" assumption (pp. 85-88). This assumption conveniently establishes the person's informational base as a mix of her individual observations and experiences (p. 87). This mix is dependable for the most part, but it also results in wrong decisions being made in a nonnegligible number of cases (pp. 87-88). These wrong decisions result from an individual's failure to account for information that lies beyond her cognitive horizon (pp. 87-88). This cognitively remote and consequently unaccounted-for information includes statistical data that are often crucial for a person's decision (pp. 109-18).

System 2 relies on deliberation but still fails to protect people against probabilistic errors, as people use it selectively (pp. 39-49). Paradoxically, it is System 1 that decides whether a person will use System 2 (p. 44). As a result, people resort to System 2 only in a state of uncertainty and disbelief (p. 81)--that is, when they encounter a difficult task that calls for a disciplined analytical solution (e.g., finding the square root of 2,226,064). Since mental energy is a scarce resource, people use System 2 sparingly and slowly. As a result, System 2 becomes busy, lazy, and depleted (pp. 39-42), while System 1 dominates people's decisionmaking27 and drives them into probabilistic misconceptions that substitute heuristics for data (pp. 109-84).

Thinking is beautifully written: its insights are rich, profound, and at the same time lucid and exceptionally well presented. These virtues make the book accessible not only to a specialized readership, including economists, decision psychologists, and economically minded legal scholars, but also to readers possessing a basic familiarity with social science. As a grandmaster of psychology, Kahneman knows a lot about human intuition. In Thinking, he utilizes this knowledge to develop intuitive explanations for complex economic and psychological phenomena. Kahneman accompanies these explanations with brilliantly selected examples from real life and controlled experiments--a presentation that immensely benefits the book's reader. For nonspecialized readers, the book reproduces two famous articles in which Kahneman and Tversky developed the insights that defined behavioral economics as a field. (28)

The book's virtues, however, do not make it uncontroversial. In what follows, I examine Kahneman's account of people's probabilistic irrationality, an account that occupies the majority of the book. (29)...