Can self-reported health measures be more useful to health economists?
How would you rate your health today? This question is commonly included in health surveys both nationally and internationally, thus regularly used by health economists in analyses of the determinants of health as a true reflection of an individual’s actual or latent health. However, the literature suggests that individuals may mis-report their self-assessed health (SAH) levels. It has been argued that systematic forms of mis-reporting, if ignored, will invalidate any econometric modelling strategy. If we can appropriately account for such fundamental mis-reporting, SAH measures, the research that utilises these, will become more reliable and useful in a practical policy related manner. This session reports on a program of work to refine the modelling of such measures to explicitly take into account the possibility of mis-reporting. We do this using novel data and statistical modelling techniques, including collecting our own data, utilising large secondary data sets, employing a range of SAH measures, and proposing new econometric models to validate the use of such SAH measures. We employ anchoring vignettes, monte carlo simulation and test for over-inflation and middle response bias. Such approaches allow us to test for the characteristics of who is mis-reporting their SAH, and consequently to point to the groups in society who really do require resources to alleviate health problems. In so-doing we will make simple and easily collected SAH measures less uncertain, thus more usable for informing policy.