The accuracy of economic measurement in the health and retirement study

We assess the quality of the HRS’s measures of out-of-pocket medical spending and health insurance premia, both in the “core interviews” and in the “exit interview” data. We provide detailed evidence on the quality of the HRS insurance premia data, and we compare the HRS exit data to exit data in the MCBS. We document how changes in survey questions, including the introduction of “unfolding brackets,” affect the HRS measures. We document what we believe are errors in the HRS imputations and provide some suggestions for improving the accuracy of some imputed variables. Overall, we find the HRS data to be of high quality. However, we believe that many interesting variables in the HRS are under-utilized because users must perform imputations themselves.