Limitations
The unit of analysis was the CSA, not the individual. As a result, there is a risk of ecological fallacy in these results and analysis. Rather, redlining at the neighborhood level may not be the same construct [49] as an individual’s exposure to redlining. Small sample size from using CSAs as the unit of analysis is a study limitation as well, and may especially be an issue with relatively rare outcomes such as elevated blood lead levels in any given year. CSAs are also larger than other potential geographic units of analysis like census tracts: a more geographically concentrated unit of analysis may offer better resolution and cleaner categorization of HOLC categories than CSAs could. Future studies could use restricted-access census tract level mortality data to reduce the impact of these limitations. Life expectancy data was calculated by Baltimore’s Health Department using 2013 mortality rates within each age category, and thus may not fully capture “true” life expectancy.
The study scope is limited to Baltimore, which reduces the generalizability of these results. Other redlining maps from across the US are available for use in a future study. However, Baltimore provided a convenience sample that included publicly available mortality and redlining data and has been central to the national conversation around redlining and residential segregation.
The main limitation is the passage of time. While this study adds to the literature that suggests that nearly century-old policies in Baltimore still have measurable and significant impacts on residents’ lives today, the time lag masks a great number of social changes that have taken place in the interim. These include the strong possibility of endogenous and path-dependent [20] changes such as gentrification and concomitant displacement of populations that were living in redlined neighborhoods in the 1930s. Certainly, the demographic makeup of Baltimore has changed drastically even in the last decade [50]. Additionally, the study does not account for the length of time residing in home/residence for Baltimore residents, and potential CSA or city out-migration; if there were mobility across CSAs in Baltimore, this would potentially serve as a mediator of the effect of redlining. These changes to the social and built environment may be either enhancing or blunting the impacts of policies set in the 1930s. We are unable to determine from this analysis whether the measured reduction in life expectancy in redlined CSAs is due to CSAs’ differing resource characteristics—as suggested by several other authors [12,21,28,29,32]—or simply changes in the demographic profile of a CSA through displacement and concentration.
CC BY 4.0. "Association of historic redlining and present-day health in Baltimore" by Huang SJ, Sehgal NJ, PloS one (2022). https://doi.org/10.1371/journal.pone.0261028. Licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Peer-reviewed and open access; full text from PubMed Central (PMC8769359). Headings and layout adapted for reading; figures and tables are in the original.