The author is a J.D. Candidate at the University of Pennsylvania Carey Law School. The author remains anonymous out of concern that even a careful, fact-based discussion of firearms law might invite professional or social backlash. That the author has such fears within a law school environment, where open inquiry should be protected, reflects growing unease about the shrinking space for candid debate on important questions.
This post is part of a series written by students in the University of Pennsylvania Carey Law School’s The Second Amendment – A Modern and Historical Perspective class. The posts were written by students as an assignment for the class and edited by their professor, Jonathan Goldstein, in preparation for publication.
With the proliferation of militaristic government agents wearing face masks and operating without identification, the right to bear arms in defense of a free state has never been less urgent. These paramilitary squads act without impunity: emerging from unmarked vehicles with weapons drawn, shooting into vehicles, grabbing people off the street and putting them in unmarked vans. While these are certainly not Second Amendment issues, freedom from unreasonable search and seizure is essential to the security of the “free state” protected by the 1787 constitution. There’s no disincentive to unidentified goons kicking down doors quite like widespread firearm ownership.
But there remain weaknesses in the pro-firearm position. We explore here some social scientific data connecting firearm ownership to violence against the self. The ease and finality of suicide for firearms owners intuitively increases the rate of suicide. This conclusion, that firearm possession is correlated with suicide, is born out in the data.
A central question is whether firearm possession can be demonstrably linked to an increase in overall suicide rates or only to an increase in the proportion of suicides by gun. A landmark 2004 study of firearm possession in the Northeast found that states with higher household gun prevalence had higher total suicide rates (Miller, Hemenway & Azrael). This relationship was driven by higher firearm suicides, i.e., non-firearm suicide rates did not rise with gun prevalence. The implication is that as fire-arm prevalence increases, suicide increases. In other words, suicide by firearm is not merely a substitution effect, with suicidals substituting one more lethal method (suicide by firearm) for a less lethal method (such as suicide by hanging or by bleeding).
Detractors of this study have criticized the geographic selection process for 1) choosing to look at states rather than smaller geographic areas like cities and 2) choosing states by decision rather than through a process of random selection. The study justifies using states over smaller areas based on availability of data and the considerable variation in firearm ownership across states. The specific states provided for in the study were chosen for the completeness of electronic hospital discharge data (again record availability) and the cohesiveness of telling a story about the specific Northeastern geographic area. Intuitively, these results have applicability outside of the Northeast. However, the study itself does not make claims about other geographical regions.
The major countervailing thought is well represented by a similarly macro-level analysis by Gary Kleck. Kleck uses a WLS (weighted least squares) design and adds controls of hypothesized co-variate factors to achieve a null result. In other words, by changing the methodology and adding controls, he achieves a result that indicates the opposite of the Miller study, that volumes of total suicides are not correlated with firearm possession. Critics of Kleck’s approach have pointed out the fragility of the cross-sectional WLS design and the power of small changes to covariate controls. One major problem with Kleck’s study is reproducibility. When other researchers have performed similar analyses, they have found, and continue to find, a relationship between firearm prevalence and total suicide rates.
Firearm possession has another problematic link to suicide: the rapidity of escalation from suicidal ideation to action. A person having one bad day or one bad night is at higher risk of death or lifechanging injury when readily supplied with the mechanism of death. The data bears out this conclusion. While substance abuse or depression are highly predictive of suicide, the acute period of heightened risk of suicide can be only minutes or hours long.
In a Houston study, 153 survivors of suicide attempts were asked how much time passed between the decision and the suicide attempt. 24% said less than 5 minutes, 24% said 5-19 minutes, 23% said 20 minutes to 1 hour, 16% said 2-8 hours, 13% said 1 or more days (Simon). An Alabama study on suicidal impulsivity reported that for nearly half (47.6%) of patient inmates at a psychiatric hospital there was less than 10 minutes before their first current thought of suicide and the attempt (Deisenhammer). The N value (or total population) is admittedly low for both studies, but the intuitive conclusion warrants more research on the proportion of impulsive suicides and the link between these and firearm possession.
Is there a compromise position between firearm possession and impulsive suicide prevention? Possibly. A recent 2021 study linked handgun possession, not shotgun possession, to self-inflicted gunshot wounds. The danger of self-harm produced by handguns taken alongside the statistics associating the majority of firearm-involved crimes with handguns rather than long guns, may point to a policy turn which restricts side-arms rather than shotguns and rifles (long guns having legitimate uses in hunting and deterring brownshirts). Besides, when a trio of masked gunmen appear at your front door, unannounced, wearing full riot gear, wouldn’t you rather have a bigger, uglier machine?