Revealed preference analysis of boundedly rational choice (Herbert A. Simon)

This dissertation proposes a theory of choice behavior that addresses Herbert Simon's long-standing critique of the economic paradigm of full rationality. Chapter 1 begins by reinterpreting Simon's argument in terms of four categories of problems that can interfere with optimal decision making, proceeds to argue that a facile “reductionist defense” against his critique is unsatisfactory, and concludes by introducing an alternative response to Simon centered on a distinction between notions of “perceived” and “founded” preference. Chapters 2 and 3 develop in parallel; the former reviewing in detail the concepts, methods, and conclusions of classical (full rationality) choice theory, and the latter constructing a less demanding theory that allows for three of the four aforementioned categories of problems. Finally, Chapter 4 demonstrates how the classical theory can be further modified to allow for the remaining category, albeit with only partial success in obtaining the desired results.; As in revealed preference analysis more generally, the objective in Chapters 2–4 is to axiomatize plausible hypotheses about the determinants of choice behavior by imposing restrictions on this behavior in the form of conditions on a choice function. While classical hypotheses of preference maximization have been shown by Amartya Sen to correspond in this sense to conjunctions of contraction and expansion consistency conditions, weaker hypotheses of perceived preference maximization are found in Chapter 3 to correspond to conditions of expansion consistency alone. Similarly, the still weaker compound hypotheses put forward in Chapter 4 are shown to correspond to conditions of relational congruence, and these too have their analogues in the classical theory. Throughout this progressive dilution of the axiomatic basis for full rationality, numerical representations of the postulated behavior are obtained wherever possible.