• Sessions
• Presentations
• Posters
• Rooms

Presentation

Session: Posters
Room: TBA
Time: Fri 13:00-14:30

Presenter: Silvana Robone (University of York. )

Abstract

This paper investigates heterogeneity in the reporting of health and health system responsiveness using data for nine European countries from the World Health Survey (WHS). Previous studies have shown evidence of heterogeneity in the reporting of health (i.e. Kapteyn et al. 2007, Bago d’Uva et al. 2008) and responsiveness (Rice et al., 2008, 2009), across and within countries. The main contribution of our paper is to compare reporting behaviour for these two constructs, in order to assess the extent to which they are driven by common factors. This has to our knowledge not been done previously. We further assess the magnitude of the respective effects and the percentage of the total variability in the dependent variable that these factors are able to explain.

Reporting behaviour can be identified by exploiting variation in the ratings of anchoring vignettes. Anchoring vignettes describe the health status or health care encounters of hypothetical individuals. Since these descriptions are fixed and pre-determined, any systematic variation in the rating of vignettes across individuals can be attributed to differences in reporting behaviour. Reporting behaviour equations are estimated through generalized ordered probit models, which model the thresholds between categorical responses as functions of age, gender, income, education and country dummies. We estimate such models for the health domains of “pain” and “mobility”, and responsiveness domains of “waiting times” and “clarity of communication”. We allow for reporting effects of observable and unobservable factors (i.e. personality traits). To account for unobservable individual specific reporting effects, we follow two strategies. First, we develop a latent class model of heterogeneous reporting. Secondly, we consider a bivariate normal distribution for unobserved reporting heterogeneity in health and responsiveness.

Our preliminary results suggest the existence of reporting heterogeneity in both self-reported health and responsiveness and that this is largely driven by age and education. Moreover, the way in which these factors affect reporting behaviour appears to be consistent across the domains analysed. Compared to their younger counterparts, older individuals appear to have lower expectations and hence down-report a given level of health or responsiveness. Further, the more educated appear to have higher expectations than the less educated at low levels of health and responsiveness, but lower expectations at high levels. In contrast income appears to influence only the reporting of health.

Key Terms

Anchoring Vignettes, Reporting Heterogeneity, Latent Class Model, Bivariate Model, Health, Health System Responsiveness

Authors:

Teresa Bago d’Uva (Erasmus University Rotterdam. Department of Applied Economics) , Andrew M. Jones (University of York. Department of Economics and Related Studies) , Nigel Rice (University of York. Centre for Health Economics) , Silvana Robone (Centre for Health Economics. University of York) and Peter C. Smith (Imperial College)

Search with Google