What Do Self-Reported, Objective, Measures of Health Measure?

Survey reports of the incidence of chronic conditions are considered by many researchers to be more objective, and thus preferable, measures of unobserved health status than self-assessed measures of global well being. The former are 1) responses to specific questions about different ailments, which may constrain the likelihood that respondents rationalize their own behavior through their answers, and 2) more comparable across respondents. In this paper we evaluate this hypothesis by exploring measurement error in these 'objective, self-reported' measures of health. Our analysis makes use of a unique data set that matches a variety of self-reports of health with respondents' medical records. Our findings are striking. For example, the ratio of the error variance to the total variance ranges from just over 30 percent for the incidence of diabetes to over 80 percent for the incidence of arthritis. Furthermore, for many conditions the error is significantly related to individuals' labor market activity, as hypothesized in the literature. In the final section of the paper we compare estimates of the effect of these different measures of health on labor market activity