Abstract
Purpose
This study examined whether state-level racial disproportionality in homelessness is associated with racial disproportionality in overdose mortality.
Methods
Counts of individuals experiencing homelessness (2015–2017; by state and racial/ethnic group) were obtained from the US Department of Housing and Urban Development; population estimates and counts of drug overdose deaths (2018–2021; by state and racial/ethnic group) were obtained from the National Center for Health Statistics. Homelessness and overdose mortality disproportionality scores were calculated to indicate the extent to which each racial group was over- or under- represented among those experiencing homelessness, or among overdose deaths, respectively (relative to each racial group’s proportional share in the general population). For each racial group examined, ordinary least squares regression models with robust standard errors (SEs) examined associations between state-level disproportionality in homelessness and disproportionality in overdose mortality, adjusting for percent aged 18–64 and US Census Region, as well as disproportionality in educational attainment and unemployment.
Results
State-level racial disproportionality in homelessness was significantly and positively associated with racial disproportionality in overdose mortality for Black (b = 0.16 [SE = 0.05]; p < .01), American Indian/Alaska Native (b = 0.71 [SE = 0.23]; p < .01), and Hispanic populations (b = 0.17 [SE = 0.05]; p < .01), in models adjusting for region and percent aged 18–64. The significant positive associations in these three populations persisted after adjusting for educational attainment disproportionality, yet the association was no longer significant in the Black population after adjusting for unemployment disproportionality.
Conclusion
States with the highest levels of racial/ethnic minority overrepresentation in homelessness generally also had relatively higher levels of racial/ethnic minority overrepresentation in overdose deaths.