Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
June 2013

Public Safety, Mental Disorders, and Guns

Author Affiliations

Author Affiliation: Department of Psychiatry, Columbia University, and New York State Psychiatric Institute, New York.

JAMA Psychiatry. 2013;70(6):565-566. doi:10.1001/jamapsychiatry.2013.315

In the aftermath of the shooting of 20 schoolchildren and 6 teachers in Newtown, Connecticut, on December 14, 2012, attention quickly focused on the presumed link between mental disorder and violence. With no more than rumors to rely on, the media speculated wildly on the gunman's diagnosis and drew parallels to earlier shootings involving persons with mental illness. Wayne LaPierre, executive vice president of the National Rifle Association, announced at a press conference that the problem of violence was largely due to people with mental illness, “genuine monsters . . . that are so deranged, so evil, so possessed by voices and driven by demons, that no sane person can even possibly comprehend them.”1

The validity of this widely accepted causal connection between mental disorder and violence is of critical importance for the formulation of appropriate policy responses to the risk of violence. Moreover, inaccurate representations of the relationship between mental disorder and violence have the potential to further stigmatize people with mental disorders and impede their treatment and integration into the broader community.


Conflicting conclusions regarding the influence of mental disorder on violence have complicated our professional understanding and the public's understanding of the issue. For several decades, many advocates and researchers argued that mental disorder conferred no increased risk of violence. However, the weight of the evidence since the early 1990s suggests that violent behavior is modestly increased among people with mental disorders. Much of the research has focused on schizophrenia and related disorders, with meta-analyses suggesting a roughly 2-fold to 4-fold increase in the risk of violence, albeit with considerable heterogeneity among studies.2 However, epidemiologic data show that other mental disorders are also associated with an increased risk of violence, and often to a greater extent than schizophrenia, including depression, bipolar disorder, anxiety disorders, and personality disorders.3,4

Two caveats must be kept in mind regarding these data. First, much of the increased risk seen in people with mental disorder is attributable to other variables rather than to the disorders themselves. Substance abuse, for example, accounts for a large proportion of the incremental risk.5 Indeed, the effect of substance abuse on violence is significantly greater than the effect of mental disorder.4 Premorbid personality traits, including psychopathy, may similarly increase the risk of violence, and insofar as they also lead to substance abuse, their effect on violence in mental illness may be magnified.6

Second, the vast majority of people with mental disorders do not engage in violence, and the proportion of overall risk of violence attributable to mental disorders is small. A meta-analysis of the murders of strangers by people with psychotic disorders, for example, found a rate of 1 murder per 140 000 persons with schizophrenia.7 The best US data put the population attributable risk percentage for violence due to mental disorder between 3% and 5%.8 Studies from other countries with lower overall rates of violence find somewhat higher population attributable risk percentages, but none exceed 10%.2 Thus, even if the proportion of violence accounted for by mental disorders could be eliminated entirely, which is unlikely under any circumstances, 90% to 97% of violent behavior would remain.

There are, of course, several types of violence for which the connection to mental disorder is unquestioned and substantial. People with mental disorders are much more likely to be victims of violent crime than the general population; data from a sample of people with severe mental illness in Chicago, Illinois, for example, indicate an 11-fold increase in victimization.9 Suicide is also disproportionately associated with mental disorder. In 2009, according to data from the Centers for Disease Control and Prevention, the age-adjusted suicide mortality rate for the total population (11.8 suicides per 100 000 persons) was approximately twice the age-adjusted rate for homicide (5.5 homicides per 100 000 persons); most people who commit suicide have diagnosable mental disorders at the time of their deaths.

Finally, given the attention that episodes of mass violence attract, it is worth noting the absence of hard data linking serious mental disorders to such tragedies. Compilations of incidents of mass shootings suggest that people with severe mental disorders may be overrepresented among the perpetrators, but given the possibility of bias in the nonsystematic collection of such data, firm conclusions are impossible at this point. Moreover, such incidents are rare and account for only a very small proportion of homicides. Indeed, gun violence of any sort is not commonly committed by people with mental disorders. The MacArthur Study of Mental Disorder and Violence,10 which observed hospitalized psychiatric patients for 1 year after discharge, found that only 2% to 3% of the cases of violence in that sample involved the use of a gun.


What are the implications of these data for our response to horrific acts of violence such as the Newtown murders? Given the very small share of violence toward others attributable to mental disorder, policies aimed exclusively at people who experience mental disorders are unlikely to lead to significant increases in public safety. Although there may be some benefit from improved reporting to the National Instant Background Check System of patients who are judicially committed for treatment, the gain is likely to be marginal. If we are really serious about substantially reducing the risk of gun violence, the answer almost certainly lies elsewhere.

In addition, ill-thought-out policies adopted in haste can wreak havoc to the mental health system and can lead to counterproductive consequences. New York's new statute requiring mental health professionals to report the names of people whom they conclude are “likely to engage in conduct that would result in serious harm to self or others” is an example. Patients' names will be checked against a database of people with gun licenses to determine whether their permits should be revoked. The law, which was pushed through the legislature in 24 hours, threatens to deter people with thoughts of harming themselves or others from coming to therapy or speaking openly about their ideation, lest they find their privacy abrogated and the state in possession of their identities. If treatment can prevent violence in some cases, this statute makes that less likely to occur.

A third implication of taking the data seriously is that mental health professionals and other advocates for improved mental health services must exercise caution in their endorsement of proposals for increased mental health funding. Such offers are often premised on the proposition that the problem of violence is largely a problem of untreated mental illness, and its corollary that better treatment will preclude a repetition of mass shootings such as those in Tucson and Newtown. The steady deterioration of funding for treatment of mental disorders, including the loss of $4 billion from state mental health budgets since 2008, makes it tempting to embrace any policy that promises to make more services available. However, tying the need for increased funding to public safety will lead to further demonization of people with mental disorders, as well as an inevitable backlash when it becomes clear that more mental health clinics or inpatient beds have not had a major effect on the prevalence of violence. This is not an argument for rejecting needed funding, but for honesty and clarity in making the case for better services. An adequately funded mental health system should be a national priority—but for the right reasons.

Violence is a complex, multicausal phenomenon, and its prevention requires attention to the means used to perpetrate violence; in the United States of the 21st century, that means guns. Pointing the finger at people with mental illness as the cause of the problem of violence in this country is misleading, counterproductive, and just plain mean.

Back to top
Article Information

Correspondence: Dr Appelbaum, New York State Psychiatric Institute, 1051 Riverside Dr, Unit 122, New York, NY 10032 (

Published Online: April 3, 2013. doi:10.1001/jamapsychiatry.2013.315

Conflict of Interest Disclosures: None reported.

 Remarks from the NRA press conference on Sandy Hook school shooting, delivered on Dec. 21, 2012 (Transcript). Washington Post. December 21, 2012. Accessed February 1, 2013 
Fazel S, Gulati G, Linsell L, Geddes JR, Grann M. Schizophrenia and violence: systematic review and meta-analysis.  PLoS Med. 2009;6(8):e1000120PubMedArticle
Swanson JW, Holzer CE III, Ganju VK, Jono RT. Violence and psychiatric disorder in the community: evidence from the Epidemiologic Catchment Area surveys.  Hosp Community Psychiatry. 1990;41(7):761-770PubMed
Pulay AJ, Dawson DA, Hasin DS,  et al.  Violent behavior and DSM-IV psychiatric disorders: results from the national epidemiologic survey on alcohol and related conditions.  J Clin Psychiatry. 2008;69(1):12-22PubMedArticle
Fazel S, Långström N, Hjern A, Grann M, Lichtenstein P. Schizophrenia, substance abuse, and violent crime.  JAMA. 2009;301(19):2016-2023PubMedArticle
Volavka J, Swanson J. Violent behavior in mental illness: the role of substance abuse.  JAMA. 2010;304(5):563-564PubMedArticle
Nielssen O, Bourget D, Laajasalo T,  et al.  Homicide of strangers by people with a psychotic illness.  Schizophr Bull. 2011;37(3):572-579PubMedArticle
Swanson JW. Mental disorder, substance abuse, and community violence: an epidemiological approach. In: Monahan J, Steadman HJ, eds. Violence and Mental Disorder. Chicago, IL: University of Chicago Press; 1994:101-136
Teplin LA, McClelland GM, Abram KM, Weiner DA. Crime victimization in adults with severe mental illness: comparison with the National Crime Victimization Survey.  Arch Gen Psychiatry. 2005;62(8):911-921PubMedArticle
Monahan J, Steadman H, Silver E,  et al.  Rethinking Risk Assessment: The MacArthur Study of Mental Disorder and Violence. New York, NY: Oxford University Press; 2001