Prospect Theory by Kahneman and Tversky - Book Report/Review Example

Prospect theory as described in Kahneman and Tversky (1979) puts forward an alternative concept of how choosers choose between alternatives under the condition of risk, where the probabilities of outcomes are known to the chooser. Rather than seeing individuals as purely rational agents seeking to raise utility, which is the common view of the standard model of economics, prospect theory proposes that choosers make decisions on the basis of potential value of both gains and losses – as well as by using heuristics, which are simple generalized rules that often serve as the basis of judgments and decisions. Kahneman and Tversky (1979) overview a number of judgment biases (including the certainty effect and the isolation effect) that indicate – in real-life situations – people do not always make the rational, utility-maximizing choice, but instead a choice based on the weighting or un-weighting of probabilities. Prospect theory is a model of decision-making that extends beyond economics and includes knowledge of psychology to form a description of real-life decisions, rather than idealized choices. In the formulation of the model, prospect theory can be expressed graphically. The value function is concave in cases of gains and convex for losses. When losses occur, the value function is generally steeper than it is for gains. The interaction between probability and the value function produces two outcomes: (a) risk-adverse when gains have moderate probability or losses that have small probability and (b) risk-seeking when losses have moderate probability or gains that have small probability. The first outcome is insurance, and the second outcome is gambling. The hypothetical value function that Kahneman and Tversky (1979) present in Figure 3 is s-shaped and asymmetrical with the middle running through the reference point because losses hurt more than gains feel good (we’ll discuss later the “reflection effect”). A protection against losses is insurance, even though insurance only eliminates losses with small probabilities and reduces sure gains – causing an overreaction to small probability outcomes. The alternative that Kahneman and Tversky (1979) spend much of their paper discussing is expected utility theory, which is the dominant view of decision-making (primarily from economics). In utility theory, the rational agent is assumed to be indifferent to the reference point (that is, the starting point where the value of the outcome is evaluated against). In addition, the rational agent only looks at the expected value of an outcome, rather than the gains and losses, irrespective of the reference point. Much of prospect theory’s critique of utility theory has to do with the fact that the latter lacks a reference point. Under utility theory, if two people each have $1 billion dollars in savings, they should both be equally happy; prospect theory, however, would argue with empirical evidence that if one had just gained $500 million to reach the $1 billion dollar mark and if the other had just lost $500 million from $1.5 billion to reach the $1 billion dollar mark, there will be a pronounced difference in their overall happiness level. The key point of prospect theory in contrast to utility theory, then, is how the question is framed – which will inevitably include complicated details about human decision-making into overly simplified economic theories. This will be important in considering the isolation effect later. Prospect theory thus has the potential, according to its authors, to better account for real-life decision making, not necessarily the economic models of rational agents. For example, in the decision to buy insurance, the decision agent has the choice between paying $15,000 for sure – and entering a lottery with two possible events: no cost ($0) at a 99% probability or -$1,000,000 at a 1% probability. For a purely rational agent, the insurance looks unattractive since it equals a certain loss. However, if we look at prospect theory’s suggestion that we set the frame of reference to -$1,000,000, the concavity of the value function (such as the one presented in Figure 3) would bias the decision-maker toward buying the insurance – as both alternatives are set as gains and the unlikely event of losing $1,000,000 is strongly overweighed. Perhaps the most key departure of prospect theory from utility theory is in how they were developed – with prospect theory emerging from a methodology of empirical research involving real decision-makers and utility theory coming about as a philosophically-based concept of the decision-maker as a purely rational agent. Kahneman and Tversky (1979) rely heavily on the response of real study participants to hypothetical questions and choices. Particularly, they state, laboratory experiments where the conditions can be adequately controlled are preferable for obtaining precise measurements of utility and probability from actual choices. However, the extent to which these choices are externally generalizable to real-life is unknown. Regardless, the authors believe that if people are reasonably accurate in predicting their choices (with respect to a situation like, for example, gambling), the regular violation of the rules of utility theory in hypothetical situations would provide a real reason to doubt it. For example, psychological phenomena like the certainty effect, observed by the authors during their empirical testing of decision-making, cannot be accounted for with utility theory. The certainty effect is a negative emotional reaction to a change resulting from the reduction of probability from certain to merely probable. The valence of this reaction is greater when an outcome goes from certain to uncertain as opposed to if the outcome was uncertain to begin with (70% certainty) and became more uncertain (20% certainty). In addition, uncertainty can lead to increased risk-taking. As an example, Kahneman and Tversky (1979) share an example of most people choosing a sure gain of $3,000 relative to a 80% chance of winning 4,000, but most people choosing a 20% chance to win $4,000 relative to a 25% chance to win $3,000. The example helps to illustrate how respondents systematically violate utility theory by virtue of the fact that people consider probabilities as much as they consider outcomes because losses hurt (especially when a “sure thing” is lost). A heuristic that decision-makers use to make a decision that can cause disaccord with utility theory is the isolation effect. It is a psychological phenomenon in which people simply the choice between alternatives by disregarding the components of alternatives that are shared and focus instead on those that distinguish them. Understandably, this may lead to inconsistent preferences and ultimately the weighting or de-weighting of insignificant factors. For example, consider the two people who have $1 billion dollars that were mentioned earlier; the fact of obtaining $1 billion dollars, under utility theory, is expected to be of value to both agents since it is exactly the same outcome. However, agents ignore the issue that under both scenarios, they will obtain $1 billion dollars when it is framed in (1) as a gain from $500 million or (2) as a loss from $1.5 billion dollars. The bias is toward the difference, not the similarities, in the outcome – as well as in the reference point the agent is starting out with. Lastly, the “reflection effect” is a situation in which all positive payoffs of an outcome are replaced by their negatives (or, a reflection around zero). As an example, consider a choice between a 90% chance of getting 3,000 and a 45% chance of getting 6,000; the reflection effect changes the frame by setting the former as a 90% chance of losing 3,000 and a 45% chance of losing 6,000. In this scenario, a risk-averse preference for the high probability safe 3,000 gain is implied and a risky 6,000 gain is relegated to second place. In contrast, utility theory would expect one to choose the same outcome no matter how it was framed, but instead Kahneman and Tversky (1979) found most people chose the safe lottery (90% chance of 3,000) when framed as a gain but only 8 percent chose the safe lottery when it was framed as a loss. Accordingly, it was found that certainty increases averseness to losses as well as the desirability of gains. Going back to the value function (concave for gains, convex for losses), we see this is clearly supported under the model of prospect theory. References Kahneman, Daniel and Amos Tversky. "Prospect Theory: An Analysis of Decision under Risk." Econometrica, 47(2) (1979): 263-291. Read More