The Changing Contours of Racialized Inequities in the United States: Geographic Disparities in Mortality Over the Past 65 Years

Abstract

This dissertation examines how racial and geographic inequities in premature mortality in the United States have evolved from 1959 to 2023. While Black-White disparities and geographic variations in mortality have each been widely studied, less is known about how these dimensions have changed jointly over time. Using national vital statistics data, I combine decomposition methods, Bayesian spatio-temporal modeling, and causal inference techniques to analyze long-run race-specific mortality patterns across both time and space. I find that Black-White inequities in premature mortality declined over this period, but in a non-linear fashion. Distinct periods of convergence and divergence were driven by changing cause-specific and race-specific mortality patterns. Reductions in racial inequity stalled and reversed prior to the COVID-19 pandemic, reflecting broader deterioration in population health in the past decade. At the same time, geographic disparities widened within both Black and White populations, with mortality risk becoming increasingly heterogeneous across counties. The specific spatial patterns of relative mortality advantage and disadvantage changed substantially over the 65 years, although counties have become increasingly entrenched in their relative positions over time. Finally, I show that the Second Great Migration contributed to long-run changes in the geography of Black mortality. Areas that experienced larger increases in Black populations between 1940 and 1970 had persistently higher Black premature mortality rates in subsequent decades, with no corresponding effect on White mortality. Together, these findings document reconfigurations of racialized and spatial mortality inequities in the United States. They emphasize the importance of studying structural inequities as dynamic and contingent processes that unfold over long periods of time in a geographically heterogeneous manner.