The demand for health: some new empirical evidence.

This paper presents some new estimates of the 'pure investment' and 'pure consumption' models of the demand for health. In contrast to previous studies, the paper presents parameter estimates of both reduced-form and structural demand for health care equations. The latter--which are obtained by treating health capital as a latent variable in a recursive equation system--provide a safer way of checking the consistency of the model's predictions with the data. The present analysis of the pure consumption model differs from previous analyses in its recognition of the dependency of the shadow price of initial assets on the entire lifetime profiles of the model's exogenous variables. The paper also explores the implications in terms of reliability of parameter estimates of the failure on the part of previous studies to recognize this.