Why Does State of Residence Matter for the Mortality of Working Age Americans?

Myron Gutmann, University of Colorado Boulder
Sara Hoose, University of Colorado Boulder
Emily Parker, Rutgers University
Stefan Leyk, University of Colorado Boulder
Siqiao Xie, University of Colorado Boulder
Hoeyun Kwon, University of Colorado Boulder
Aleksander Berg, University of Colorado Boulder
L:isa Hollister, University of Colorado Boulder

Since the 1990s, mortality among working age Americans has increased for both men and women, marked by rising deaths from drug and alcohol-related causes, as well as suicide. While individual determinants of age at death are a culmination of a person’s lifetime health experience, recent research suggests that place-based factors also play a role. We can see those processes at various scales, most importantly the neighborhood, which reveals more than analysis at the scale of the county. In this paper we explore the local determinants of midlife mortality for working age Americans who lived in 14 U.S. states between 1990 and 2021. Our raw data are individual death certificates, precisely geolocated at the scale of the census tract, allowing us to analyze the place-based factors influencing U.S. midlife mortality. In this work we raise an important question, asking “why is there mortality variation between states, even when controlling for individuals and the places where they lived?” We ask because in earlier research we see strong state-to-state variation even after we control for other factors. Among our fourteen states, Florida, Michigan, Ohio, and Oklahoma have persistently high working age mortality, while California, Colorado, Georgia and Massachusetts have low mortality. We take two approaches to understanding why states are so significant. We begin with a multilevel analysis in which census tracts are the main units of analysis, but where data about states and counties inform what we know. To that we add a more qualitative perspective, looking at government policy at the state level and at characteristics that might be more embedded in local culture. The result will be a rich contribution to the ways that U.S. lives have changed over the past 30-plus years, with some thoughts about policy approaches that might improve lives in the future.

No extended abstract or paper available

 Presented in Session 30. Mortality Inequalities Across Class, Place, and Borders