Does Gender Influence Subjective Health Reports? Evidence Using Anchoring Vignettes

Megan Todd, Princeton University
Jennifer Dowd, Baruch College, City University of New York (CUNY)

Gender differences in health are well established in many contexts, with women often paradoxically reporting worse health but with lower rates of mortality. Self-rated health (SRH) measures are commonly used to quantify and explain differences in health status across groups, but population groups may differ in their understanding and interpretation of SRH survey questions. Anchoring vignettes have been introduced as one potential solution to this problem, whereby respondents’ reports of their own health are “anchored” by their ratings of the health of a fictitious individual. This paper uses anchoring vignettes across several domains of health to test hypotheses about how gender may influence reports of health, taking advantage of randomization of the gender of the vignette characters in the Health and Retirement Study. The results will shed light on whether measurement of sex differences based on self-reports are likely to be biased.