Subjective well-being and generic preference-based measures of health: an empirical contribution

Abstract

Background: A growing body of literature proposes alternatives to the current use of generic preference-based measures of health in decisions of resource distribution in health care. One prominent suggestion is the use of subjective well-being. While arguments have developed both for and against this concept there has been limited empirical research carried out into the implications of its use. This study investigates some of the ways in which generic preference based measures and subjective well-being might differ in their sensitivity and responsiveness to differences and changes in health. Methods: Analyses are carried out using ‘satisfaction with life’ and SF-36 values from multiple waves of the British Household Panel Survey. Using ordered probit and tobit regression models this study compares how different factors have an impact upon an individual’s valuation of their current state in terms of life satisfaction, and upon the responses that inform their SF-6D state. Using difference-in-difference analysis, the study investigates the extent to which satisfaction with life and SF-6D values are responsive to objective changes in health. In addition we investigate ‘satisfaction with health’ as a possible measure of health outcome. Results: Results indicate that subjective well-being is not sensitive or responsive to differences and changes in health to the same extent as SF-6D values. Differences in the determinants of each measure fit well with existing theory, and some ∗Corresponding author: C. J. Sampson, Institute of Mental Health, University of Nottingham Innovation Park, Triumph Road, Nottingham, NG7 2TU.chris.sampson@nottingham.ac.uk