Alternatives To Gun Policy?: A Bayesian Analysis Of County-Level Firearm Mortality

Statement of Purpose Policymakers often propose restrictive gun laws to reduce firearm-associated mortality. However, research suggests the relationship between tighter gun policy and preventing gun deaths is complex. We examined whether several other factors were associated with firearm-related mortality as they may be less contentious routes to improving health in this domain. Methods/Approach We used CDC data to calculate firearm-related mortality in US counties in 2012. The INLA package in R was used to run Bayesian Poisson mixed effects models for two county-level outcomes: counts of gun suicides and gun homicides. Predictors included urban development, healthcare access, gun policy leniency, and several demographic covariates. A random effect was placed on state to account for clustering; a non-informative prior was used. Results Stricter gun policy was associated with fewer gun suicides, but not homicides. In addition, living in a rural county was protective against both gun suicide and homicide. With factor analysis, the healthcare access variables condensed into factors that describe each county’s prevalence of 1) large healthcare system infrastructure (e.g., tertiary care hospitals) and 2) smaller-scale healthcare infrastructure (e.g., Conclusions Stricter gun policy was not associated with lower gun mortality, and a complex urban/rural dynamic exists. Planned future analyses will include a spatial framework, longitudinal data, and examination of urban/rural subpopulations to further evaluate these relationships. Significance and Contributions to Injury and Violence Prevention Science There may be several channels by which policymakers can decrease rates of firearm-associated mortality at the county level. Importantly, the best policy approach may differ between urban and rural settings.