Introduction

Making predictability problems visible

Epistemology teaches: he who wants to see everything will see nothing. Understanding why it is so difficult to predict behaviour is no exception to this rule. Like a magnifying glass, theories make visible what is in their purview. But this greater visibility comes at a price. The fringe of the field of observation is blurred. Rational choice theory is a strong magnifier. It makes one class of predictability problems crystal clear. The theory allows for random preferences. It demonstrates how interaction is nonetheless possible, with or without institutions. One can even interpret recent economic theory as concentrating more and more on the predictability problem. For information economics has assumed centre stage. Mechanism design demonstrates the conditions under which rational actors can overcome information asymmetries. But when employing this magnifying glass, a large environment remains invisible. It assumes full rationality, of both the informed and the uninformed actor. Both possess well-behaved utility functions. They optimise anew from scratch, whenever new information comes up. Their computational capacities are unlimited.

This part of the book uses a different magnifying glass. Many proponents of the approach on which this book draws tell a story of fact versus fiction. They criticise rational choice analysis for being purely speculative. They contrast it with what they claim is a realistic approach. The rhetoric of this criticism is epistemologically naïve. Perfect realism would imply a one-to-one description of reality.