Chaos Theory and Microeconomics: An Application to Model Specification and Hedonic Estimation.

This paper is the first to apply the theory of deterministic chaos to a microeconomic problem. Previous applications of chaos theory to time-series data, while successful in uncovering nonlinearities, have not provided guidelines for resolving uncovered misspecification problems. In contrast, the authors show that a modified test statistic from chaos theory is an extremely valuable tool in microeconomic model specification because it shows when excluded information is correlated with included information. This test, applied to hedonic estimation of marginal housing prices, is able to distinguish among alternative regression specifications and assists in discovering a parsimonious specification devoid of nonlinear effects. Copyright 1991 by MIT Press.