Sample Selection Bias in Health Econometric Models

Abstract

This article discusses nonrandom sample selection (SS), its causes, and consequent biases for causal inference and prediction. SS bias in the conventional linear framework is detailed, and the classical Heckman-type two-stage SS estimator is reviewed. Linear models and methods are not, however, compatible with most empirical contexts in health economics and health services research which often involve outcomes that are qualitative or otherwise limited in range. For this reason, SS in a general nonlinear framework is reviewed, and a recently developed two-stage estimation approach for such nonlinear models is outlined.