Do Reporting Differences Bias the Measurement of Health Inequalities in U.S. Adults? Evidence Using Anchoring Vignettes from the Health and Retirement Survey

Jennifer Dowd, Baruch College, City University of New York (CUNY)
Megan Todd, CUNY Institute for Demographic Research

The accurate measurement of population health is crucial for monitoring health levels and trends across different groups. Measures of health inequalities based on self-reports may be biased if individuals from different groups have systematically different expectations or reporting standards for health. This paper tests for reporting heterogeneity in self-reported disability measures by level of education and race/ethnicity in the U.S. using anchoring vignettes from the Health and Retirement Survey in the domains of mobility, sleep, pain, memory, shortness of breath, and depression. We then estimate inequalities in each health domain by education and race/ethnicity before and after adjustment for reporting heterogeneity to quantify the magnitude of potential biases in health reporting using self-reported measures. Preliminary results suggest strong evidence of reporting heterogeneity in self-reported health measures by level of education and race/ethnicity, with models that do not account for reporting differences strongly underestimating health inequalities by education and race/ethnicity.