SEMIPARAMETRIC ESTIMATION OF HURDLE REGRESSION-MODELS WITH AN APPLICATION TO MEDICAID UTILIZATION

This paper develops a semi-parametric estimation method for hurdle (tw o-part) count regression models. The approach in each stage is based o n Laguerre series expansion for the unknown density of the unobserved heterogeneity. The semi-parametric hurdle model nests Poisson and nega tive binomial hurdle models, which have been used in recent applied li terature. The empirical part of the paper evaluates the impact of mana ged care programmes for Medicaid eligibles on utilization of health-ca re services using a key utilization variable, the number of doctor and health centre visits. Health status measures and age seem to be more important in determining health-care utilization than other socio-econ omic and enrollment variables. The semi-parametric approach is particu larly useful for the analysis of overdispersed individual level data c haracterized by a large proportion of non-users, and highly skewed dis tribution of counts for users. (C) 1997 by John Wiley & Sons, Ltd.