ABSTRACT
Perhaps more than anybody else in economic theory, Herbert A. Simon stressed that individual decision makers have no choice but to make decisions under the constraints of limited cognitive resources (e.g., Simon 1978). On the basis of this indisputable truth about the human cognitive system, he challenged classical economic theory, which in his view projected an omniscient rationality assuming unbounded knowledge, computational capacities, and time. He also targeted Milton Friedman’s (1953) famous defense of classic economic theory, “The Methodology of Positive Economics.” Responding to the criticism that economic theory rests on unrealistic assumptions, Friedman argued:
Complete “realism” is clearly unattainable, and the question whether a theory is realistic “enough” can be settled only by seeing whether it yields predictions that are good enough for the purpose in hand. (Friedman 1953, 41)
In Friedman’s view, the purpose in hand is to account for aggregate behavior, that is, the behavior of firms, institutions, or, more generally, the market. Therefore, unrealistic assumptions and possible discrepancies between the predictions of the theories and individual choice behavior need not be detrimental to the fate of economic theory. Not without some smugness, Simon pointed out that “economists who are zealous in insisting that economic actors maximize [subsequently] turn around and become satisficers”—people satisfied with workable, if not optimal, solutions-“when the evaluation of their own theories is concerned,” as in Friedman’s good-enough criterion (Simon 1979, 495). Moreover, he argued that psychologically plausible theories of decision making, which assume realistic limits on the knowledge and computational abilities of the human agent, also lead to conclusions at the level of aggregate phenomena. Importantly, these conclusions are not always the same as those suggested by neoclassical theory, thus rendering possible crucial tests (Simon 1979). Simon’s vision of a different rationality of economic behavior, bounded rationality (Simon 1956, 1990), has not only posed a challenge to economic theory but has also suggested a new research agenda revolving around the following key question: how rational are people, given their limited computational capabilities and their incomplete knowledge?
