Estimating the effect of state-level gun purchasing policy on county-level firearm suicide mortality

Abstract

Although gun violence is one of the most widely discussed issues in American politics, research funding freezes have hindered the development of public health strategies for its prevention. A substantial majority (approximately two-thirds) of annual gun deaths in the United States are suicides, but most political efforts to prevent gun violence focus on homicide. In this study, I evaluated the effectiveness of handgun permit-to- purchase (PTP) laws at preventing firearm suicide. PTP requirements mandate that all potential handgun purchasers at certified sellers must present a permit or license issued by the state. I used a county-level Poisson model controlling for several relevant covariates to estimate the relative risk of firearm suicide one through 35 years after the passage of a state-level PTP requirement. I also tested the popular theory that prevented firearm suicides will manifest as suicides by other means by running the same model with non-firearm suicide as the dependent variable. I found that time since PTP passage was significantly associated with lower firearm suicide rates and that the relative risk by year-since-passage exhibited a pattern that was consistent with my theoretical expectations. Counterintuitively, I also found that PTP requirements had a significant protective effect against non-firearm suicide, a result that suggests that my analysis was subject to unaccounted-for confounding. I ran several sensitivity analyses, which were able substantially clarify the effect of PTP requirements on firearm suicide from the theoretically implausible effect on suicide by all other means I had identified. This study contributes relatively robust evidence to the growing body of literature supporting gun control as a means of suicide prevention, but it also underscores the need for the development of more rigorous data sources and methodologies in the field of gun violence prevention research.