July 20, 1998, pp. 36-37
By Sally Satel and D J Jaffe
A MacArthur Foundation study found that the mentally ill are no more violent than anyone else. Can it possibly be true?
Last May, Raymond Cook went on trial for first-degree murder in the death four years ago of Thomas J. Guinta, a police officer in Fall River, Massachusetts. The same day that Cook shot Guinta, he spared the life of a longtime neighbor because he mistook her for a movie star who had played opposite James Cagney.
The day before Easter, 26-year-old Keith Powell of Red Oak, North Carolina, showed up with a shotgun at his aunt Louise's birthday party and killed his grandfather and two uncles. Months earlier when he walked away from his group home, his mother knew that he had quit taking his medication, and she feared that the voices in his head would soon return. After the shootings, Powell committed suicide.
In the most spectacular recent case, Michael B. Laudor of Hastings-on-Hudson, New York, a brilliant Yale Law School graduate who seemed to have beaten back his mental illness (he was working on a screenplay based on his successful fight) allegedly stabbed his girlfriend to death. He too had stopped taking his medication, and his mother had become so alarmed that she asked police to check on him shortly before his girlfriend's body was found.
Cook, Powell, and Laudor all suffered from schizophrenia. Lacking adequate treatment, they became deranged and violent. Homicide, to be sure, is rare even among schizophrenics, and yet untreated psychosis is so intimately linked to aggression that mental hospitals routinely rely on locked wards and often use physical restraints to control patients until antipsychotic medications take effect.
The connection between certain types of severe mental illness and violence would seem to be a matter of common sense. But until the 1960s, when deinstitutionalization began in earnest, most of the severest cases were out of sight, helping people forget how volatile the psychotic can be. And in the 1970s a movement focused on reducing the stigma of mental illness, and hence downplaying the risk of violence, became better organized and more influential. Thus, the old common-sense view became increasingly controversial.
Now a new study from the MacArthur Foundation says the notion that people with serious mental illnesses may be particularly prone to violence is little more than a stigmatizing myth. The study got wide play in the media-including front-page coverage in the New York Times and some sensational headlines-from the AP, "Mentally Ill Not Especially Violent" and from the Washington Post, "Are Former Mental Patients more Violent? If they Don't Abuse Drugs and Alcohol the Answer is Generally No, Study Finds." It could have a major impact on the running debate over involuntary-treatment laws by providing ammunition to civil libertarians and advocates for the mentally ill who believe that ill people should be left to wallow in their psychosis, haunting the streets of our cities.
The MacArthur findings were published in the May issue of the prestigious Archives of General Psychiatry, under the title "Violence by People Discharged from Acute Psvchiatric Inpatient Facilities and by Others in the Same Neighborhoods." The study does admit that drugs and alcohol increase violence in people with mental illness more than they do in the general population, but it concludes that, otherwise, these people are no more violent than anyone else.
This contradicts the findings of numerous studies over the last thirty years. For example, a study of three hundred patients discharged from California's Napa State Hospital between 1972 and 1975 showed they had an arrest rate for violent crimes ten times higher than the general population. A 1994 British study of schizophrenics in Camberwell found that they were four to five times more likely than their neighbors to be convicted on charges involving serious violence. How did the authors of the MacArthur Violence Risk Assessment Study, all prominent researchers, come to such a radically different conclusion?
First, they excluded many potentially violent people by the way they constructed their sample, such as persons who were currently in or recently released from a prison, jail, forensic hospital, or long- term psychiatric hospital. "We always teach medical students that past violence is the best predictor of future violence," says psychiatrist and schizophrenia researcher E. Fuller Torrey. "While purporting to study violence, the first thing the authors did was omit violent people from the study."
The researchers limited the study to patients in acute-care hospitals. Only one in ten patients stayed longer than thirty days with half of all patients successfully treated and released in under nine days. This largely eliminated anyone too sick to be stabilized acutely, again reducing the chance of including non-psychotic individuals. Among those asked to participate, 29 per cent refused; a disproportionate number were suffering from schizophrenia, the very disorder which, if untreated, is most likely to result in violence.
Second, by choosing a narrow, high-end definition of violence, the study left out many dangerous people. For example, the authors counted as violence only those acts that produced bodily harm. As the father of a woman with schizophrenia and a mean left hook summed it up, "If you're a good ducker, your relative is not considered violent." While most of us would consider setting fires and trashing rooms to be violent acts, the authors of the study did not.
Conversely, the authors chose to study the broad category of mental illness rather than the narrow category of psychosis. Since people suffering from depression rarely commit violence against others, including them in the pool studied skews the results. In fact, 40 per cent of the sample suffered from depression.
Third, the researchers selected a violence-prone control group, residents of poor, chaotic, drug-ridden sections of Pittsburgh (one of the three cities from which the study drew patients) that had higher crime rates than the city as a whole. Since many of the patients were discharged to this same neighborhood, the authors assumed that a higher level of patient violence, if discerned, would be due to their mental illness. But this likely minimized the violence differential between the patients and the control group. That's because many people with psychiatric illnesses-illnesses that can impede economic advancement-end up living in marginal areas with hotheaded neighbors.
Despite all these stratagems, the authors still found that over half of the patients studied engaged in some form of threatening behavior within one year after discharge from the hospital. Specifically, 18 per cent of the patients without a drug or alcohol problem committed at least one act of violence (e.g., throwing objects, kicking, hitting, using a weapon) and an additional 33 per cent engaged in at least one act of aggression (same as above except that no harm resulted). Violence was nearly double (31 per cent) among mentally ill people who also abused drugs and alcohol.
So, in reality, the MacArthur study proves only that nonviolent people tend to be non-violent. But the fact remains that seriously mentally ill people are more prone to violence than the rest of the population. Indeed, one of the MacArthur study authors acknowledges as much.
In a 1992 issue of The American Psychologist, John Monahan of the University of Virginia wrote: "The data that have recently become available, fairly read, suggest the one conclusion I did not want to reach: whether the sample is people who are selected for treatment as inmates or patients in institutions or people randomly chosen from the open community, and no matter how many social or demographic factors are statistically taken into account, there appears to be a relationship between mental disorder and violent behavior."
Advocates will seize upon the MacArthur study as "proof" that all the other research is wrong-and that involuntary-treatment laws are unnecessary. In fact, those laws need to be strengthened, not repealed. Many states still rely solely on the standard of whether a person is "imminently and provably" dangerous to himself or others. What about those who are not yet dangerous but, already prisoners of their delusions, are likely to deteriorate further if not treated soon? A number of states allow the authorities to hospitalize such people against their will, but judges are notoriously reluctant to do so.
And many states do not allow judges to consider past violence when deciding whether someone should be committed. Even a court order for hospitalization does not guarantee medication, since in some states judges may not mandate that a hospitalized psychotic patient take an antipsychotic drug merely because he is dangerous: he must also be judged incompetent to refuse the medication. While over half the states permit judges to order a patient to continue to take his medication once he is discharged from the hospital (if he has a known habit of stopping the medication and becoming dangerous), such laws are actually applied only infrequently.
Obviously, treatment is of profound importance. It can keep people with serious mental disorders out of jails and shelters. It can prevent suicide and help the afflicted rejoin society. And certainty that the mentally ill will be treated might make communities less resistant to supervised housing and other desperately needed community-based programs. Yet, political correctness-an unwillingness to offend or "stigmatize"-prompts efforts to conceal the risk of violence from people suffering from unmedicated psychosis.
After years of denying the association between untreated mental illness and aggression, the National Alliance for the Mentally Ill, the largest and most influential grass-roots organization of family members of mentally ill people, has come full circle. Carla Jacobs, a NAMI board member from California, became an activist for involuntary commitment after her mother-in-law was fatally stabbed and shot by a mentally ill relative.
"We used to think it was stigmatizing to acknowledge violence," she says. "Now we recognize that violence by the minority tars the majority, and makes communities less likely to welcome the community- based housing that can facilitate treatment and reduce violence." Besides, she adds, "too many of our relatives are hurting others, and winding up in jail. The first step to helping them is admitting there's a problem."
Unfortunately, the MacArthur study will make it harder even to take that first, basic step.
Reprinted with permission. Copyright 1998 by National Review, Inc., 215 Lexington Avenue, New York, NY 10016. All rights reserved.