p.865

Testing for Neglected Heterogeneity

Andrew Chesher

Abstract

Neglecting across individual heterogeneity can lead to inconsistent or inefficient estimators, and to misleading predictions. This paper develops a specification error test sensitive to neglected heterogeneity, which is viewed as causing parameter variation, by deriving a score test of the hypothesis that parameters have zero variance. The test, whose form is insensitive to the specification of the distribution of varying parameters, turns out to be the Information Matrix test introduced by White [8]. This suggests that the Information Matrix test is a useful diagnostic for researchers using cross-sectional or longitudinal data to estimate models of individual economic agents behavior.