Article Text

Download PDFPDF
067 Modeling the association of structural racism with disparities in firearm homicide victimization
  1. Kelsey Conrick1,2,
  2. Avanti Adhia3,4,
  3. Alice Ellyson2,3,5,
  4. Miriam Haviland6,
  5. Vivian Lyons7,
  6. Brianna Mills8,
  7. Ali Rowhani-Rahbar9
  1. 1School of Social Work, University of Washington, Seattle, USA
  2. 2Firearm Injury and Policy Research Program, Seattle, USA
  3. 3School of Medicine, University of Washington, Seattle, USA
  4. 4Harborview Injury Prevention and Research Center, Seattle, USA
  5. 5Center for Child Health, Behavior, and Development Seattle Children’s Research Institute, Seattle, USA
  6. 6Harborview Injury Prevention and Research Center, University of Washington, Seattle, USA
  7. 7Department of Health Behavior and Health Education, University of Michigan, Ann Arbor, USA
  8. 8Harborview Injury Prevention, University of Washington, Seattle, USA
  9. 9Department of Epidemiology, School of Public Health, University of Washington, Seattle, USA


Statement of Purpose Research on the association between individual measures of structural racism (e.g., residential segregation) and racial disparities in firearm homicide often includes the proportion Black population as a covariate. However, including proportion Black as a covariate may imply that race—not racism—is a risk factor for firearm homicide disparities. This study’s goal is to identify a better approach to including measures of structural racism when studying racial disparities in firearm homicide.

Methods/Approach We examined the associations of four indicators for structural racism (poverty, educational attainment, labor force participation (LFP), and arrests for index crimes) with state-level Black-White disparities in age-adjusted firearm homicide victimization rates in the U.S. 2010–2019. Disparities in each indicator were defined as ratios of race-specific rates (Black/White). Ordinary least squares regression coefficients and measures of model fit were obtained to examine the association between indicator variables and racial disparities in firearm homicide. To compare the fit of models including proportion Black to those using an indicator-level disparity, we used four models: 1) absolute indicator rate; 2) absolute indicator rate and proportion Black; 3) absolute indicator rate and Black-White rate disparity; 4) absolute indicator rate, proportion Black, and Black-White rate disparity.

Results For all four indicators, the optimal model included the absolute rate and Black-White rate disparity and did not include proportion Black. Coefficients for the Black-White indicator-level disparity were statistically significant, while proportion Black was not statically significant.

Conclusions The inclusion of proportion Black does not contribute further to the explanation of firearm homicide disparities and should not be included since improved measures of structural racism are available (e.g., Black-White rate disparity).

Significance This study presents a novel approach to including measures of structural racism into research on firearm homicide disparities, consistent with a focus on racism rather than race as a determinant of violence.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.