ABSTRACTTitle of Dissertation: THE LONG TAIL OF HISTORY: COMBINING THE 1940 CENSUS, REDLINING MAPS, AND HRS: METHODS FOR ANALYZING THE IMPACT OF REDLINING ON HEALTH, ECONOMIC, AND HEALTHCARE OUTCOMES IN OLDER ADULTS TODAY

	Shuo Huang, Doctor of Philosophy, 2023	

Dissertation directed by: Neil Jay Sehgal, PhD, MPH Department of Health Policy Management

BackgroundAs part of the New Deal in the 1930s, the Federal government used the Home Owners’ Loan Corporation (HOLC) to draw real estate security maps that were color coded or redlined to discourage lending in majority Black neighborhoods. Redlined areas still have worse health and economic outcomes in the present day. Current literature is focused on present-day residents of redlined areas. Tracking exposures to redlining and conditions of redlining close to the 1930s with present-day health is an unexplored area. Methods We utilize geo-referenced 1930s HOLC maps to locate individuals and map demographic considerations. We use novel algorithmic solutions to geolocate unknown 1940 enumeration districts. Using a 1940 census-linked sample of the Health and Retirement Study to locate individuals in HOLC areas at the time, we conduct survival analysis on HOLC categories’ effect on age at death as well as other analysis on health, economic, and healthcare utilization in the near present. We test for a potential mediator. Results Population density is not associated with either HOLC category or present day life expectancy, and is unlikely to be a mediator. In uncontrolled models, for HRS individuals in the 1940 census HOLC category is associated with greater hazards, worse odds of self-rated health, and worse economic outcomes. With controls, HOLC category is only associated with worse odds of self-rated health. HOLC category is not associated with health insurance or healthcare utilization in this sample. Conclusion Redlining is associated with health and economic outcomes which are attenuated when controlling for likely pathways between redlining and health. Future research should focus on whether individuals stay in redlined areas, and on identifying policy and initial state matrix that can describe what redlining may be a proxy for.