This post is based on a paper that was presented at the 2025 Firearms Law Works-In-Progress Conference. The Conference is held each year on a home-and-away basis with the Duke University Center for Firearms Law. This post also appears on the CFL’s Second Thoughts blog.
In the realm of firearms research, the debate often revolves around the theoretical mechanisms through which gun control laws influence different outcomes. While one can construct statistical models which can produce all sorts of outcomes—and, as a quantitative researcher, I believe in their merit—any statistically significant finding must be supported by a plausible underlying theory. Without a sound theory, the result should be ignored as one of the many potential statistical artifacts: omitted variable bias, model mis-specification, miscalculated standard errors, et cetera. This is especially true when a model produces a result that many theories would not predict, such as prior research claiming that permitless carry laws increase suicide rates: a tenuous link, given that such laws expand firearm access outside the home when death by suicide is a problem that primarily occurs inside the home.
Unfortunately, in the realm of quantitative firearms research, when theories are proposed to explain underlying results, the theorizing does not go much further than telling a “just-so” story. These just-so explanations are often untestable, or go untested even if they are testable. My future paper on firearms theft aims to, in a small way, fill that gap. For years opponents of liberalized concealed carry laws have claimed one mechanism through which these laws increase crime is through increased gun thefts. John Donohue, an esteemed scholar at Stanford University, has published multiple studies whose quantitative models suggest liberalized concealed carry laws increase crime (my research fails to find this). In a 2023 working paper, he claims liberalized right to carry laws increase rates of gun theft. In another paper, he states that carrying a gun outside the home tends to increase gun thefts. This may be counterintuitive, as one might hypothesize that allowing lawful owners to carry their firearms would reduce the number of guns left unattended in homes or vehicles, potentially decreasing theft.
Other gun laws have been tied to firearms theft as well. The Center for American Progress, for example, claims safe storage laws, which require gun owners to lock up their firearms, deter gun theft: “the people most susceptible to gun theft are those who own a large number of guns, frequently carry guns outside of their homes, or do not practice safe storage methods.” The theoretical effect of safe storage on theft is ambiguous, however. In fact, research on child-resistant packaging has noted that there exists a “lulling” effect which actually led to an increase in childhood medical ingestion because it caused people to stop taking other precautions. While I am not claiming this effect applies to safe storage laws, the hypothesis is at least as compelling as the one offered by the Center for American Progress and warrants investigation.
Motivated by these competing “just-so” stories, my recent paper, presented at the Firearms Law Works-in-Progress Conference, directly tests the theory: do gun laws influence firearms theft rates? Are reductions in firearms theft a mechanism through which gun control laws save lives? Using the value of stolen firearms collected by the FBI as a proxy for gun theft, the answer for most laws was “no.” The analysis focused on eight key policies: permitless carry, child access prevention (CAP) laws, mandatory reporting of lost or stolen firearms (MRLF) laws, permit-to-purchase (for all firearms and for handguns) laws, firearms training mandates, and universal background checks (for all firearms and for handguns). With one exception, the results were statistically insignificant. The exception was MRLF laws, which require gun owners to report to the police when their firearms are stolen. Preliminary gun trace studies suggest states with stricter laws mandating the reporting of lost or stolen firearms are less likely to be sources of crime guns to other states, meaning these laws may have positive impacts on efforts to fight firearms trafficking. These efforts may deter gun thefts by making it more difficult to get away with the crime without punishment. Likewise, they may change gun owner behavior since stolen guns will be immediately linked to the owner, a potentially undesirable legal situation, causing owners to store guns more safely to avoid unwanted police contact.
The preliminary results of my paper show a substantial effect: MRLF laws were significantly associated with lower firearms theft rates, with states implementing these laws experiencing a 28.9% to 30.1% decrease in theft rates in the log-linear panel model and count model, respectively. The question is, should we buy it?
Despite this paper’s goal of testing others’ “just-so” stories, explaining the mechanism behind the MRLS result requires venturing into similar speculative territory. As noted, these laws could change owner behavior, perhaps by increasing the rate at which they report lost or stolen firearms. But do gun owners really not report their stolen firearms in the absence of these laws? For those of us who own firearms, we know how expensive they are: of course we would report them stolen, even if there were no law requiring us to do so. Only criminals, who would not report gun thefts anyway and would not be encouraged to report even under MRLF laws, would fail to report. Perhaps there are some law-abiding gun owners who would fail to report a stolen firearm in the absence of such a law but would report it if there were a MRLF law think, for example, a reluctance to report a family member who stole the firearm, who distrust the police, or fear of reprisal (which is why regular crimes are also underreported). Perhaps in these situations MRLF laws could help. But, once again, we are simply just-so story telling.
Furthermore, these results raise the question of whether MRLS laws are enforced vigorously enough to impact citizen behavior. As noted in another paper at the conference by Professor Jonathan Abel, rates at which firearms laws are enforced vary widely across jurisdictions. If MRLF laws are underenforced, it is unlikely that they would change gun owner behavior, and these results would have to be chalked up to one of the statistical artifacts I mention above. To summarize briefly, I would say this: Most gun control laws seem to have no impact on rates of gun thefts, but laws that require civilians to report their stolen firearms to the police might have some impact. Future research is needed to determine if this correlation is causal and whether or not the “just-so” stories told above hold any water.