Herbert Simon advocates that economists should study procedural rationality instead of substantive rationality. One approach for studying procedural rationality is to consider algorithmic representations of procedures, which can then be studied using the concepts of computability and complexity. For some time, game theorists have considered the issue of computability and have employed automata to study bounded rationality. Outside game theory, very little research has been performed. Very simple examples of the traditional economic optimization models can require transfinite computations. The impact of procedural rationality on economics depends on the computational resources available to economic agents. Citation Copyright 1994 by Kluwer Academic Publishers.
