Barriers and bounds to Rationality

This review article takes Peter Albin’s book Barriers and Bounds of Rationality : Essays on Economic Complexity and Dynamics in Interacti6e Systems as the starting point of a general assessment of complexity analysis from the point of view of the scope and limits of human rationality. In particular, this article examines the constraints upon human rationality. In particular, this article examines the constraints upon human rationality deriving from Gödel’s theorem, computational complexity, chaotic behavior and interconnections among system components. Gödel’s theorem is especially relevant in cases of self-reference (as in game-like situations); computational complexity makes search for optimal solutions often to exceed the capacity of the existing computers even in the case of relatively ‘simple’ problems; chaotic behavior makes long-term prediction impossible; cellular automa suggest that entirely different outcomes may result from small differences in patterns of interdependence. Special attention is devoted to the use of cellular automa in identifying the nature of non-linearities compatible with aggregation; in assessing the role of a central actor in a in a decentralized market economy; in discovering ‘simple-minded principles’ by which market equilibria can be brought about. In conclusion, the article suggests that a new theoretical paradigm is emerging, in which the search for invariant laws would be substituted by attention for historical conditions and localized patterns of interaction. © 2000 Elsevier Science B.V. All rights reserved. JEL classification: D80; C90; E37; O41