BUNMD

Life expectancy in the United States increased by a remarkable 30 years over the course of the 20th century. This impressive progress was driven primarily by advances in the treatment of infectious diseases and delayed mortality for those living with chronic illness (Crimmins and Zhang 2019), but the benefits accrued unevenly. Inequality in mortality between the most advantaged and the least advantaged actually increased over time (Preston and Elo 1995), and by the beginning of the 21st century, the gap in life expectancy between the top and bottom 1% of income earners was over 14.6 years (Chetty et al. 2016).

Despite the longstanding interest in racial and class-based inequalities in health and mortality in the United States (Schwandt et al. 2021; Elo 2009), research is often hampered by data limitations (Card et al. 2010; Song and Coleman 2020). Most research into the general dimensions of mortality disparities using microdata have relied on survey data, with sample sizes that preclude the analysis of smaller population subgroups such as the oldest-old. In the absence of comprehensive population-level registry data such as those found in the Scandinavian countries, researchers are increasingly turning to administrative datasets from agencies such as the Social Security Administration to answer some of the most pressing questions in social science research (Chetty et al. 2016; Card, Dobkin, and Maestas 2008; Card et al. 2010; Meyer and Mittag 2019; Ruggles 2014). While the recent introduction of the United States Mortality Database has created a valuable new resource for studying aggregate mortality trends at the state level (USMDB 2021), public-use administrative microdata for mortality research remains scarce. When available, it is often cumbersome to use.

We introduce the Berkeley Unified Numident Mortality Database (BUNMD), a cleaned and harmonized version of administrative mortality records from the Social Security Administration. The BUNMD represents one of the first publicly available, large-scale administrative microdata resources for studying mortality. We anticipate that the size (N = 49 million) and spatial detail will open up new avenues for high-resolution mortality research. Furthermore, the open-access nature of the dataset will ensure this research is reproducible and extendable. The BUNMD can be downloaded here.