Irrational Economist_ Making Decisions in a Dangerous World - Erwann Michel-Kerjan [33]
Kleindorfer, Paul, Howard Kunreuther, and Paul J.H. Schoemaker (1993). Decision Sciences: An Integrative Perspective. Cambridge: Cambridge University Press.
Russo, J. Edward, and Paul J.H. Schoemaker (2002). Winning Decisions: Getting It Right the First Time. New York: Doubleday.
Thaler, Richard H., and Cass R. Sunstein (2007). Nudge. New Haven, CT: Yale University Press.
7
Constructed Preference and the Quest for Rationality
DAVID H. KRANTZ
WHAT CHOICES ARE WISE?
This question has been discussed for millennia. Many proposed answers share a common theme, eloquently expressed by Plato in The Protagoras:
What measure is there of the relations of pleasure to pain other than excess and defect, which means that they become greater and smaller, and more and fewer, and differ in degree? For if any one says: “Yes, Socrates, but immediate pleasure differs widely from future pleasure and pain”—to that I should reply: And do they differ in anything but pleasure and pain? There can be no other measure of them. And do you, like a skillful weigher, put into the balance the pleasures and the pains, and their nearness and distance, and weigh them, and then say which outweighs the other.
Behavioral research has shown that Plato was wrong: The tradeoffs among different pleasures and pains depend heavily on context. “Skillful weighing” has to take account of how weights vary with context. A more flexible definition of rationality must be sought, one that acknowledges a variety of ways of solving decision problems.
Much behavioral research suggests that decision makers are often unwise. One can scarcely disagree! Yet arguments about what is wise must be delicate, because behavioral research also suggests that standards of wisdom need to be set with great care.
CONSTRUCTED CHOICE
Some choices are indeed based on preferences. A person choosing strawberry rather than chocolate ice cream might anticipate the taste of each flavor, and then choose the greater anticipated pleasure. The outcome of weighing strawberry versus chocolate may be transient (“Today I feel like strawberry”) or stable (“I always prefer strawberry over chocolate”), but in either case the choice reveals the preference at that point in time.
In many choice settings, however, a decision maker may not have a “preference.” Disparate goals do not project onto a common scale of pleasure and pain, so the metaphor of “weighing” in a balance does not apply. Rather, the choice settings, including the particular options available, influence the degree to which various goals are activated and/or valued; a setting may sometimes even lead to adoption of goals not previously considered at all. For example, even in a familiar tradeoff, such as saving time versus saving money, the valuation of each dimension varies with context. If a person is asked how much money he would pay to save one hour, the very task highlights and enhances the value of money; if asked how much time would need to be saved to justify the expenditure of $20, the task itself highlights and enhances the value of time. Thus, the tradeoff is not fixed. Moreover, the availability of a new labor-saving device may lead a decision maker to adopt goals not considered at the start of the time/money tradeoff, such as trying out the device, or capitalizing on a low introductory price.
Thus, “preference” among alternative plans is often constructed in the given setting. Often, people go beyond the options given, constructing new feasible plans and then selecting one of them, perhaps on the basis of newly activated goals.
RATIONALITY IN THE CONTEXT OF CONSTRUCTED CHOICE
Activation of additional goals and construction of choice cannot be viewed as irrational. Rationality should not demand that every possible goal be considered in advance, valued, and incorporated into a utility function. Abandoning the common currency of “utility” leads to a key question: Given that choices are constructed, what standards