Avoiding Retransformation in Nonnnegative Health Economic Models with Sample Selection
*Joseph Vincent Terza, University of Florida 

Keywords: marginal effects, incremental effects, endogeneity, Heckman's method, instrumental variables

This paper focuses on econometric estimation of incremental and marginal effects using parametric regression results from a model whose outcome is non-negative valued and observable only if a nonrandom selection criterion is satisfied.We offer a parametric,but robust, control function method that does not require the typical joint normality assumption.This allows, for example,the analysis of count data models.Transformation is not required to facilitate estimation so marginal and incremental effects can be directly estimated based on the conditional mean regression specification without retransformation or smearing.Along with the explicit functional forms of these effects,we derive corresponding consistent estimators complete with correct asymptotic standard errors.The performance of the method is evaluated relative to the conventional approach using simulated and real world data.

Print Friendly
For more information send email: ichps@amstat.orgHPSS 2005