022 Do handgun purchase waiting periods save lives? Evidence from a synthetic control approach

Abstract

<h3>Statement of Purpose</h3> Firearm suicide is a leading cause of death for nearly every age group 10–14 years and older in the U.S. In 2019, a firearm was used in half of all suicides. Despite research demonstrating that impulsive suicide attempts are far from uncommon, few studies have examined the effectiveness of firearm purchase waiting periods as a tool to restrict access to lethal means and prevent suicide. The present study estimates the effect of a waiting period repeal on handgun suicide rates. <h3>Methods/Approach</h3> I use a quasi-experimental design and the synthetic control method to estimate the effect of Wisconsin’s handgun purchase waiting period repeal on the handgun suicide rate. Annual state-level suicide and sociodemographic data from 1999–2019 were included. Placebo tests and difference-in-difference regression were used to supplement results from the synthetic control analysis. <h3>Results</h3> The proportion of suicides involving handguns increased from 25.3% to 32.4% following the repeal; post-intervention suicides were more likely to involve handguns than those in the pre-intervention period [χ² (1, N = 8269) = 49.25, p &lt; .001]. A ‘synthetic Wisconsin’ closely approximated the pre-intervention handgun suicide rate in Wisconsin (mean squared prediction error = 0.04). The estimated treatment effect of the waiting period repeal is 1.16 deaths per 100,000, which translates to a 31% increase in the handgun suicide rate. Estimates from the difference-in-difference regressions indicated a similar treatment effect. <h3>Conclusions</h3> Allowing for the immediate acquisition of a handgun at the point-of-sale in Wisconsin resulted in an estimated mean annual increase of 67 handgun suicides. <h3>Significance</h3> The findings add to the limited research on firearm purchase waiting periods by using a data-driven method to estimate the causal effect of repeal. The synthetic control’s ability to closely approximate pre-intervention trends despite a limited donor pool of control states has implications for future policy analyses.