The Criminal Docket


July 2007

Volume II, Number 2


Predicting Violence in the Mentally Ill
Hon. Cynthia Loo
Los Angeles County Superior Court


Many have drawn parallels between the April 16, 2007 shooting at Virginia Tech, referred to as the deadliest school shooting in United States history, and the 1999 Columbine High School massacre.  In trying to piece together why theses tragedies occurred, great significance has been made to the unsubstantiated reports that the twenty-three-year old shooter at Virginia Tech, Seung-Hui Cho had been hospitalized in 2005 for suicidal ideation and depression, and possibly had paranoid schizophrenia; and that one of the shooters at Columbine, Eric Harris, was said to be taking an antidepressant commonly prescribed to treat obsessive-compulsive behavior. These cases do nothing to dispel the enduring perception that people with mental disorders are predisposed to violence.  Is there a link between mental illness and violence?  And if so, can violence be predicted accurately so that incidents like Virginia Tech and Columbine be prevented?

Nor can the justice system ignore a potential link, given that psychological disorders are much more prevalent in prison populations than in the general population.  On average 16 percent of incarcerated adults (Bartol, 2002), and an astounding two-thirds of incarcerated youth have serious mental health and/or substance abuse problems (Grisso, 2005).  Not surprising, the Los Angeles Times (November 21, 2005) reported that due to inadequate mental health services, the Los Angeles County Jail is often referred to as the largest de facto mental health facility in the nation (Torrey, 1995).

Given the number of mentally ill that are incarcerated, as well as the need to understand what lead to such tragedies as Virginia Tech and Columbine, efforts must be made to further understand the connection between mental illness and violence, and whether future violence in the mentally ill can be accurately predicted.

The Link Between Mental Illness and Violence.  Is it Valid?

Recent studies suggest a complicated link between mental disorder and violence.  In general there is agreement of at least a small relationship between severe mental disorders and violent behavior (Monahan, 2001). Approximately 90 percent of people with mental disorders are in no way violent or dangerous (Swanson et. al., 1990). In a report which followed patients for one year after discharge from hospitalization, patients who did not have a co-occurring substance abuse disorder were no more likely to have a violent incident than others living in the same neighborhoods (Steadman et al., 1998).  After reviewing a number of studies, John Monahan Professor of Law,  Psychology and Legal Medicine, University of Virginia (2001, 1993, 1992),  concluded the rate of violent behavior among persons with mental disorders, is at least somewhat higher than of people without such disorders:
1. Around 25 percent in mental hospitals assault another person during hospitalization.

2. Approximately 12 percent of all people with schizophrenia, major depression, or bipolar have assaulted other people, compared with 2 percent of persons without a mental disorder.

3. Approximately 4 percent who report having been violent during the past year suffer from schizophrenia, whereas 1 percent of nonviolent persons suffer from schizophrenia.

The mentally ill may in fact be more likely to withdraw or harm themselves than to act aggressively toward others (Hillbrand, 1994).  If they do act aggressively, it tends to occur within 20 weeks after discharge from institutionalization.   Further, most incidents occur in the normal flow of daily life, family members are likely to be involved, most incidents occur in the home, and alcohol plays a large role. In only 13% of the violent incidents did the mentally ill seek out the other person with the intent of harming them. Strikingly, one-fourth of the violent incidents involved a situation where prescribed medication was not being taken (Monahan, 2001).

The relationship between diagnosis and violence has long been confusing and contested.  Some research indicates those with schizophrenia have higher rates of violence than those with other Axis I diagnoses (Baxter, 1997). One recent study showed a significant association between schizophrenia and a higher rate of criminal convictions for violent offenses (Wallace, 2004).  Other research suggests those with schizophrenia present less risk of violence than persons with other serious mental illnesses such as bipolar disorder or depression (Monahan, 2001).  Clinical depression and feelings of hopelessness as well as anger and impulsiveness is tied with half of suicides (King, 2003).  However, some researchers have found that the highest rates of violence are not among Axis I diagnoses at all, but rather among Axis II diagnoses (Monahan 2001). 

Psychopathy and Violence

Psychopathy in adulthood is more powerfully related to future illegal aggression than any other single characteristic investigated by violence-prediction researchers (Monahan, 2001). As many as 80 percent of the most dangerous predatory criminals exhibit sociopathic behavior patterns.  Although they comprise about 4 percent of the total male population they are responsible for half of all serious felony offenses (Cohen, 1996). Experts note the most sadistic serial murderers are not mentally ill, instead, they typically exhibit a sociopathic personality that deprives them of pangs of conscience or guilt to guide their behavior.  They are typically driven to extreme acts, experience long-term frustration, blame others for their problems, and then are set off by some catastrophic loss they are unable to cope with or get help for (Fox, 1994).

Other Explanations for Violence:  Substance Abuse and Childhood Risk Factors

Some experts believe a DSM-IV diagnoses plays little role in the adult prediction of violence, and instead conclude that traits associated with some mental disorders such as impulsivity and anger offer better potential for prediction for violence.  Some argue that people suffering from disorders other than schizophrenia, such as substance abuse and neuroticism, are most at risk for chronic criminal behavior (Berenbaum, 1994). 

A co-occurring diagnosis of substance abuse is strongly predictive of violence. When substance abuse and psychiatric disorder are both present, the risk of violence increases substantially (Monahan, 2001).  Monahan also noted childhood violence together with a history of mental disorder was significant in predicting violence. For example, the seriousness and frequency of having been physically abused as a child predicted subsequent violent behavior in mentally ill patients.  An injury from an adult received before a child is 15 years old has been found to be predictive of subsequent violence in schizophrenic male patients (Klassen & O’Conner, 1988).  Severe paternal discipline has been found to be predictive of violence among male schizophrenic patients (Yesavage, 1984).  As with those without mental illness, there is a great falloff in rates of violence as people age (Swanson et al. 1990). 

There are substantial gender differences.  Men are likely to have been drinking or using drugs, less likely to have been taking prescribed psychotropic medications before committing violence, and women likely to target family and to be violent in the home (Monahan, 2001).

Contrary to popular wisdom and the results of several other studies, the presence of delusions does not predict higher rates of violence among recently discharged psychiatric patients (Monahan, 2001).  On the other hand, nondelusional suspiciousness – involving the tendency toward misperception of others’ behavior as indicating hostile intent – does appear to be linked with subsequent violence (Arseneault et al., 2000). 

Consistent with the literature that ranks prior history of violence as the most important factor for predicting risk of future violent behavior, overall the best predictor of violence in psychiatric patients is past behavior (Bartol,  2002).  Presence of a juvenile record, the number of prior arrests, prior incarcerations and seriousness of prior offenses have been found to be predictive of adult violence among male psychiatric patients (Steadman & Cocozza, 1974; Monahan 2001).  Monahan’s research indicates that the propensity for violence is the result of the accumulation of risk factors. Not one of which is either necessary or sufficient for a person to behave violently (Monahan, 2001).   

Prediction of Dangerousness

Today many researchers consider the prediction of dangerousness to be one of the most important issues in both criminal and civil matters nationwide.  Despite the pervasiveness of violence risk assessment in the law it has been known for a very long time that clinicians are not very good at predicting violence.  Early research on the accuracy of predicting violent behavior was reviewed by Monahan (1981). In one study, researchers were unable to predict nearly two-thirds of the violent crime that ultimately occurred and nearly two- thirds of the persons whom they predicted would be violent were not.  Recent research continues to indicate that the unaided abilities of mental health professionals to predict violence are modest at best (Monahan, 1981, 2001). When psychologists and psychiatrists make long-term predictions of violence, most often they overestimate the likelihood that patient will be violent (Eccleston & Ward, 2004).  On the other hand, studies suggest that short-term predictions of imminent violence are more accurate (Binder, 1999).  Recently the science has improved considerably and shows considerable promise in accurately predicting dangerousness as researchers have developed new statistical approaches that are more objective than the subjective judgments of clinicians (Monahan, 2005).


Understandably, there will be little comfort to the parents of the victims of the Virginia Tech and Columbine shootings that most diagnosed with mental illness are not violent, and that age simply matures most youth out of crime by their late teens or early twenties.  Nor will it be a comfort to parents who have children with mental illness who are incarcerated due to fears of possibly being violent and suicidal due to lack of adequate mental health services; locked up as a knee-jerk reaction to the horrors of Virginia Tech, under the premise it is better to be safe than sorry.  Fortunately the science of predicting dangerousness is said to have improved. 
Judges are in a profound place of responsibility; they are in a powerful position to act, in a way that the parents of Cho and Harris were not.  Judges are in a position to demand that quality mental health treatment is available.  Judges can not only mandate appropriate treatment, especially in the months immediately after discharge from incarceration, but to ensure that at risk mentally ill continue in the longer-term therapy that is needed to reduce the risk of violence and suicide.  Hopefully the painful, yet important lessons of Virginia Tech and Columbine will not be overlooked.


Arseneault, L., Moffit, T., Caspi, A., Taylor, P., & Silva, P. (2000). Mental disorders and violence in a total birth cohort: Results from the Dunedin Study. Archives of General Psychiatry, 57, 979-986.

Bartol, C. (2002). Criminal behavior, a psychosocial approach. Upper Saddle River, NJ: Prentice Hall.

Berenbaum, H., &  Fujita, F. (1994). Schizophrenia and personality: exploring the boundaries and connections between vulnerability and outcome. Journal of Abnormal Psychology, 103, 148-58.

Binder, R.L. (1999). Are the mentally ill dangerous? J. Amer.Acad. Psychiat.Law, 27(2), 189-201.

Cohen, L., &  Villa, B. (1996). Self-control and social control: An exposition of the Gottfredson-Hirschi/Sampson-Laub debate. Studies on Crime and Crime Prevention 5, 31-58.

Eccleston, L., & Ward, T. (2004). Assessment of dangerous and criminal responsibility. In W.T. O’Donohue & E.R. Levensky (Eds.),  Handbook of forensic psychology: Resource for mental health and legal professionals (pp.85-101). New York: Elsevier Science.

Fox, J., & Levin, J. (1994). Overkill: Mass murder and serial killing exposed. New York: Plenum Press.

Hillbrand, M., Krystal, J., Sharpe, K., & Foster, H. (1994).  Clinical predictors of self-mutilation  in hospitalized patients.  Journal of Nervous and Mental Disease, 182, 9-13.

King, R., & Apter, A. (2003). Suicide in children and adolescents. New York: Cambridge University Press.

Klassen, D., & O’Connor, W. (1988).  Crime, inpatient admissions, and violence among male mental patients. International Journal of Law and Psychiatry, 11, 305-312.

Monahan, J. (1981). The clinical prediction of violent behavior. Washington, DC:Government Printing Office.

Monahan, J. (1992). Major mental disorder and violent behavior: Perceptions and evidence. Amer. Psychologist, 47(4),511-512.

Monahan, J. (1993). Mental disorder and violence: Another look. In S. Hodgins (Ed.), Mental disorder and crime. Newbury Park, CA: Sage.

Monahan, J., Steadman, H., Silver, E., Appelbaum, P., Robbins, P.C., Mulvey, E.M., Roth, L.H., Grisso, T., Banks, S. (2001).
Rethinking risk assessment, the MacArthur Study of Mental Disorder and Violence. New York: Oxford University Press.

Monahan, J., Redlich, A.D., Swanson, J., Robbins, P.C., Appelbaum, P.S., Petrila, J. (2005).  Use of leverage to improve adherence to psychiatric treatment in the community. Psychiat. Serv., 56(1), 37-44.

Steadman, H., & Cocozza, J. (1974). Careers of the criminally insane. Lexington, MA: Lexington Books.

Steadman, H., Mulvey, E., Monahan, J., Robbins, P, Appelbaum, P., Grisso, T, Roth, L., &  Silver, E. (1998). Violence by people discharged from acute psychiatric inpatient facilities and by others in the same neighborhoods. Archives of General Psychiatry, 55, 393-401.

Swanson, J., Holzer, C., Ganju, V., & Jono, R. (1990). Violence and psychiatric disorder in the community: Evidence from the epidemiological catchment area surveys. Hospital and Community Psychiatry, 41, 761-770.

Torrey, E. (1995). Jails and prisons - America’s new mental hospitals. Am J Public Health, 85, 1611-1613.

Wallace, C., Mullen, P., Burgess, P. (2004). Criminal offending in schizophrenia over a 25-year period marked by deinstitutionalization and increasing prevalence of comorbid substance use disorders. Am J Psychiatry, 161, 716-727.

Yesavage, J. (1984). Correlates of dangerous behavior by schizophrenics in hospitals. Journal of Psychiatric Research, 8, 271-280.

Back to Beginning of This Issue