Firearms were used to commit 16 883 suicides and 12 791 homicides in the USA in 2006.1 One of the most common and possibly least controversial objectives of firearm policies in the USA is to keep firearms from people deemed to be at elevated risk of injuring others or themselves. Firearms laws in the USA commonly disqualify individuals if they are too young, have been convicted of a serious crime, are subject to certain domestic violence restraining orders, have been adjudicated to be “mentally defective,” or because they abuse illegal drugs or alcohol.

Here we discuss the potential of well-crafted firearms prohibitions for persons who abuse alcohol or controlled substances to reduce violence and injuries. There is a large body of scientific evidence indicating that people who abuse alcohol or illicit drugs are at increased risk of committing acts of violence and self-harm. Drug and alcohol abuse has been strongly linked with the perpetration of fatal and non-fatal domestic violence,2 3 4 5 6 youth violence,7 8 incarceration for violent crimes,9 and suicide and suicide attempts.4 10 Although there is some debate about whether these associations are causal or are due to other underlying determinants, there is little doubt that drug and alcohol abusers represent a high-risk group. For example, in one case–control study of risk factors for homicides and suicides within the home, the prevalence of prior alcohol abuse and illicit drug use among homicide perpetrators was 17 and 4.6 times higher, respectively, than among controls. For suicide victims, the prevalence of alcohol abuse was nearly 7 times higher, and the prevalence of illicit drug use 6 times higher, than for controls.4

There has been a long and contentious debate about the risks and benefits associated with gun ownership among the general population. While we believe the evidence points towards gun ownership leading to more violence,11 12 or at least more lethal violence,13 14 15 16 more relevant for this policy discussion is whether guns possessed by high-risk groups such as drug and alcohol abusers would increase or decrease risks. Prior research has not isolated the effects of firearm possession by drug and alcohol abusers. However, a study of another group at high risk of violence—intimate partners in physically violent relationships—found that abusers’ gun ownership increased the risk of lethal outcomes fivefold after other risk factors for domestic homicide, including history of severe violence, had been controlled.2

US federal law prohibits the transfer of firearms to a person who is “an unlawful user or addicted to any controlled substance”17 (illicit drugs), and such people are also proscribed from possessing firearms. Federal regulations specify the evidence needed to infer that someone is an unlawful user of a controlled substance including “a conviction for use or possession of a controlled substance within the past year; multiple arrests for such offenses within the past 5 years if the most recent arrest occurred within the past year; or persons found through a drug test to use a controlled substance unlawfully, provided that the test was administered within the past year.”18 Twenty-eight US states also have statutes that prohibit firearm sales to drug addicts or persons who abuse controlled substances. Two additional states prohibit firearms sales to persons under the influence of controlled substances.19

Despite the evidence that alcohol abusers are at increased risk of violence, US federal law does not bar alcohol abusers from acquiring firearms. Only 16 of 50 US states and the District of Columbia have statutes that include firearms prohibitions for persons who abuse alcohol. Theses statutes generally bar “habitual drunkards” or “alcoholics” from having guns. Carr et al20 in this issue (see page 409) found evidence that alcohol consumption can significantly impair a person’s ability to adeptly handle a gun, and they draw analogies with drunk driving. Consistent with this notion, several additional states prohibit firearm sales to individuals who are intoxicated.

But the gaps in US gun laws that allow drug and alcohol abusers to legally obtain, possess and carry concealed firearms are even broader than what is suggested by simple tallies of state laws. For firearms prohibitions to be useful, statutory law or regulations must provide sufficiently precise definitions of the disqualifying criteria to allow those conducting background checks of prospective firearm purchasers, or those checking the legality of ongoing firearm possession, to determine readily whether a person falls into a prohibited category. With a few exceptions, background checks for firearm purchase applications in the USA are based on searches of administrative databases, primarily of criminal convictions or actions taken by courts (eg, issuance of protective orders for domestic violence, adjudications of mental incompetence). Some state laws prohibit drug and alcohol abusers from purchasing firearms, but do not provide objective disqualifiers that could be used for most background checks. For example, an Alabama statute states “no person who is a drug addict or a habitual drunkard” may possess a handgun.21 However, neither Alabama statutory nor administrative law defines these terms. Similarly, in New Jersey, permits to purchase a handgun may not be issued “to any drug dependent person … or to any person who is presently a habitual drunkard.”22 But the definitions for “drug dependent persons” and “habitual drunkard” in the statute do not provide the sort of objective criteria, such as a history of past convictions, needed by law enforcement to uniformly and efficiently enforce this prohibition.

Abusers of illicit drugs account for 10% of all persons denied applications to purchase firearms through background checks conducted by the US Federal Bureau of Investigation, 13% of the applications processed by local law enforcement agencies, but only 3% processed by state agencies.23 Although alcohol abuse is more strongly associated with lethal violence than is illicit drug abuse,4 24 the US Bureau of Justice Statistics does not specifically report denials for firearm purchase applications due to alcohol abuse or alcohol-related offences.

Pennsylvania is one of the few states with a statute that includes an operational definition of an alcohol abuser that would enable one to readily identify prohibited persons. The statute bars firearms transfers to anyone convicted of three or more alcohol and driving violations within a 5-year span. The District of Columbia changed its law in 2009 to prohibit firearm sales to anyone convicted of two or more alcohol and driving violations within a 5-year period. Using multiple prior drunk driving offences—especially if at least one offence occurred recently—as a criterion for identifying persons with a serious alcohol problem is quite defensible. In addition to having demonstrated a history of reckless behaviour that threatens public safety, repeat DUI/DWI offenders (where DUI is driving under the influence of alcohol and DWI is driving while intoxicated/impaired) have very high rates of drug and alcohol abuse and other psychiatric disorders.25 26 27 Such offenders have less self-control28 and have higher rates of repeated arrests.29

Administrative licence revocation laws (also called “per se” laws) provide an instructive example of an effective injury prevention policy relying on a well-defined standard. Under these laws, persons suspected of drunk driving who fail a lawfully administered sobriety test (eg, by registering 0.08% blood alcohol concentration) automatically lose their driving privileges for a defined period of time. No additional criminal justice process is necessary. Studies have shown the general deterrence value of these laws.30 31

Determining the efficacy of policies to prohibit these high-risk groups from obtaining firearms is fraught with challenges.32 One problem with most prior studies of firearm prohibitions is that they assess population-wide impacts—sometimes immediate population-wide impacts—of laws enacted to keep firearms from one or more high-risk groups, many of whom possessed firearms when the new prohibitions were enacted. An exception is a study in which Wintemute and colleagues33 examined two cohorts of persons: one contained people who had applied to purchase handguns in California just before the state expanded its categories of prohibited purchasers to include those convicted of prior misdemeanours involving violence or firearms, and the other contained people whose handgun purchase applications were denied as a result of the new prohibitions. After age, sex, race and prior arrest record had been controlled for, the cohort whose firearm purchase applications were denied were significantly less likely to subsequently commit violent crimes. Previous research had also shown that handgun purchasers with prior misdemeanour convictions involving violence or firearms were 5 to 10 times more likely to subsequently commit a violent crime than handgun purchasers with no prior convictions.34 Prior research has also shown that state laws prohibiting persons restrained by domestic violence protection orders from possessing firearms were associated with declines in domestic homicides.35 These prior studies demonstrate that policies to prohibit high-risk groups from possessing firearms can reduce the violence committed by these groups. Thus, we are cautiously optimistic that firearm prohibitions for drug and alcohol misusers—based on objective criteria that can be checked against criminal history databases—can reduce violence perpetrated by substance abusers.

Research indicates that alcohol and drug abusers have a substantially increased risk of perpetrating violence, that access to firearms by high-risk groups increases the risk of lethal forms of interpersonal and self-directed violence, and that firearm prohibitions for other high-risk groups can reduce violence. Public polling data from the late 1990s show strong support for laws that keep guns from drug and alcohol abusers.36 (We are not aware of more recent polling information.) Although the gun lobby in the USA is very powerful, politicians may be more willing to support measures to keep guns from drug and alcohol abusers than to support gun restrictions that are not specifically targeted at high-risk groups. Thus, well-crafted policies to prohibit drug and alcohol abusers from having guns seem both prudent and politically feasible. Rigorous evaluations of such policies are needed that can isolate any policy effects on drug and alcohol abusers.