Investigating the Difference in Mortality Estimates between the Social Security Administration Trustees’ Report and the Human Mortality Database

Published: 2018

Abstract

This study’s goal was to determine whether differences in data or differences in methods explain the divergence between the mortality estimates at ages 65 and older of the Social Security Administration (SSA) and the Human Mortality Database (HMD). These differences, increasing since 1968, are an issue of significant value considering the importance of SSA estimates and projections to determine the long term solvency of the Social Security Trust Funds, as well as of other government programs such as Medicare and Medicaid. The two organizations use different data and different methods to construct their estimates. In particular, the HMD relies on national statistics from the vital registration system and the Census Bureau, while the SSA uses Medicare program enrollment data. Applying the SSA methods to the HMD data showed that differences in the data, rather than in the methods, explain the entire gap in life expectancy at age 65, with the HMD indicator 0.4 years higher for 2014 than the SSA. The study also determined that the gap resulted mostly from lower mortality rates at ages 65 to 84 years (rather than at 85 and older) up to about 2005 to 2006, but that the growing divergence since then is nearly entirely due to increasingly lower mortality at ages above 85. The pattern was found to be similar for men and for women, though the gap is slightly larger for the latter. Additional investigations, with more detailed data, will be necessary to assess whether data reliability or issues of representativeness explain the difference.

Key Findings

    • There is a gap in the life expectancy at birth as estimated by the Social Security Administration (SSA) and by the Human Mortality Database (HMD)
    • The gap is attributable to differences in mortality risks at ages 65 and above only.
    • It is not due to differences in methods but to differences in the data (national statistics for the HMD, Medicare enrollment data for the SSA).
    • Further investigation is necessary to determine whether the gap results from mortality risks which are genuinely different in the two datasets or whether it results from reliability issues in the sources of data.

Authors

Magali Barbieri

Citation

Barbieri, Magali. 2018. “Investigating the Difference in Mortality Estimates between the Social Security Administration Trustees’ Report and the Human Mortality Database,” University of Michigan Retirement Research Center (MRRC) Working Paper, WP 2018-394. Ann Arbor, MI. https://mrdrc.isr.umich.edu/publications/papers/pdf/wp394.pdf