Firearm Availability, Suicide, and the Role of Public Inpatient Mental Health Capacity

Background: The rate of firearm-related deaths in the U.S. is the highest among all industrialized nations, and firearm suicides represent the vast majority of all male suicides. Although previous work supports a positive relationship between firearm availability and suicide, it does not address the availability of inpatient mental health services for persons at high risk of suicide as a potentially important modifier of the relationship. Aims of the Study: We examine whether a relationship between firearm availability and suicide rates varies according to the capacity of the inpatient mental health system. Methods: We analyze state-year panel data for 50 U.S. states for years 1980-2010 (N=1,550). Outcomes include firearm and nonfirearm suicide rates, separately for males and females. The per capita number of hunting license holders and total hospital psychiatric beds are proxies for state-level firearm availability and inpatient mental health system capacity, respectively. We estimate a first-order moving-average dynamic panel data model of suicide that explicitly models persistence in the data-generating process. The natural logarithm of suicide rates is modeled as a linear combination of logged per capita hunting license holders and logged total hospital psychiatric beds as well as a comprehensive set of covariates. To test the modifying role of inpatient mental health system capacity we examine the interaction term of per capita hunting license holders and hospital psychiatric beds. We obtain a heteroskedasticity-based instrumental variables estimator to address potential endogeneity of main regressors. Results: The number of per capita hunting license holders is positively associated with firearm suicide rates for both males and females. The relationship is statistically significant for male firearm suicide rates: a 1% increase in per capita hunting license holders is associated with a 0.014% increase in the male firearm suicide rate. For the non-firearm suicide rates, the coefficients on per capita hunting license holders were negative but not statistically significant. The coefficient on the interaction term of hunting license holders and hospital psychiatric beds is consistently negative and statistically significant for both firearm and non-firearm suicides. Discussion: Our findings suggest a positive effect of state firearm availability on male firearm suicide rates. We also find that greater inpatient mental health capacity offers a statistically detectable protective effect in that an increase in inpatient mental health corresponds with fewer suicides by firearm among males and nonfirearm suicides among both genders. Implications for Health Policies: Identifying and maintaining appropriate levels of inpatient mental health system psychiatric capacity may attenuate the adverse effect of firearm availability on suicides. Implications for Further Research: Further investigation appears warranted regarding the extent to which alternative community treatment options may affect suicide by firearm, and whether they could provide an adequate substitute for the dramatic reductions in inpatient psychiatric care.