An investigation of health and attrition in the British Household Panel Study (preliminary draft)

Abstract

By 1998 the British Household Panel Study had experienced significant sample loss from cumulative attrition from its initial 1991 membership. We study the effects of this attrition on variations in psychological well-being using the model of Hausman and Wise (1979) which applies to continuous longitudinal data with non-ignorable or informative dropout. The model combines a multivariate linear model for the underlying response with a probit regression model for the drop-out process. Despite the large amounts of attrition, we find no strong evidence that attrition has seriously distorted the determinants of individuals mental health outcomes. The results provide some support for the representativeness of the BHPS through 1998, and some evidence that its cross-sectional representativeness has remained roughly intact.