Abstract:
When attempting to measure gun ownership in the United States, the problem of missing administrative data arises, making it necessary to find a valid proxy. Several such proxies are employed in economic studies, one of which is the fraction of suicides by firearm of all suicides (FSS). My work validates this proxy from out-of-sample data, namely, Austrian administrative data on firearm licenses. I also reevaluate, with appropriate statistical methods, a result on firearms and suicide from the medical that is often used for public policy advocacy. This result is, unfortunately, heavily biased due to ignoring a well-known fallacy and thus can be only partially confirmed.