Messy Data Modelling in Health Care Contingent Valuation Studies

(Preliminary draft please don’t cite) This study addresses the complexity in modeling contingent valuation surveys that account for sample selection bias, true zeros and non-ignorable unresolved cases or missing responses including “don’t knows ” and protest responses. Estimation and inference on welfare measure such as mean WTP or health care valuation for demanders, non-demanders as well as counterfactual welfare estimates for these unresolved cases are also provided. Two alternative endogenous switching tobit (EST) models are specified to simultaneously estimate the parameters of the latent willingness to pay (WTP) decision variable and the latent true WTP level. Bayesian techniques are developed using Markov chain Monte Carlo (MCMC) methods data augmentation and Gibbs sampling with Metropolis-hastings for estimating the endogenous switching tobit model. The Bayesian approach presented here is useful even for finite sample size and for models with relatively flat likelihood like sample selection models for which convergence is a problem or even if convergence is achieved correlation of the latent random errors are outside the (-1,1) range. The proposed methodology is applied to