Draft: On Concepts of Rationality in Games

Models in the classical theory of games and in neoclassical economics normally assume rationality in the sense that agents have compete and transitive preferences. The paper labels this fundamental rationality and distinguishs two other sorts of rationality pertinent to the study of strategic interaction: individual economic rationality (IER) and effective rationality. IER is, we observe, characteristic of the classical theory of games and neoclassical economics. Philosophers and others have discussed fundamental rationality (and its concomitant concept, the Nash equilibrium), which has a number of problems, limitations, and anomalies. This paper focuses instead on criticisms of IER and develops a critique of it. Our main complaint is that in general, and in very many cases of interest, there is no realistic, effective procedure by which agents can realize IER. As an alternative concept for explanation and prediction in strategic contexts we sketch a “type B” game setup for which effective procedures of play are available. We call this effective rationality. The paper offers suggestions for assessing effective rationality in type B games, with the aim of indicating something of how game theory (and economics) might be pursued eschewing IER as other than a benchmark or mere idealization. $Id: two-concepts-rationality.tex,v 1.9 2006/01/05 23:57:23 sok Exp $