Who's 'Dangerous'?

In the wake of last month's tragedy at Newtown, some now assert that the law should make it easier to hospitalize dangerous people against their will. Their reasoning is intuitively seductive: If Adam Lanza (the Newtown shooter) and James Holmes (the Colorado "Batman" shooter) had been identified and forcibly confined as "dangerous" before they reached the point of committing murder, many innocent lives would have been saved. Therefore we must restructure the legal standard for involuntary commitment so that dangerous people like Holmes and Lanza can be institutionalized until such time as they no longer pose a risk of harm to innocent others.

The emotions which motivate this view are understandable. But the argument is not rationally defensible.

The case for locking up the dangerous usually proceeds from three assumptions: (1) that any person who commits atrocities such as the Newtown or Colorado shootings must be mentally ill (and thus, a proper subject of psychiatric treatment); (2) that "dangerousness" is a quantifiable entity which can be accurately identified and measured, like flour or sugar; and, (3) that involuntary civil commitment on grounds of dangerousness poses no threat to individual rights because it does not risk a criminal conviction or imprisonment.

All three assumptions are demonstrably wrong.

It is not true that all murderers -- even all mass murderers -- suffer from diagnosable mental illness. In an interesting Wall Street Journal piece last month, David Kopel reported two recent pieces of such evidence: "A 2000 New York Times study of 100 rampage murderers found that 47 were mentally ill. In the Journal of the American Academy of Psychiatry Law (2008), Jason C. Matejkowski and his co-authors reported that 16% of state prisoners who had perpetrated murders were mentally ill." In those numbers a mental health professional might see a serious need for treatment. The law, on the other hand -- whose precious charge is to protect individual rights -- sees the danger of over-relying on mental illness as a basis for involuntary commitment. If 47% of the Times's 100 rampage killers were mentally ill, then 53 % -- more than half -- were not. If 16 % of state-imprisoned murderers were mentally ill, then 84 % -- more than four-fifths -- were not. The presence or absence of mental illness is not a reliable gauge of a person's potential for homicidal violence.

Of course the legal standard for involuntary civil commitment does not rely solely on a finding of mental illness. To justify involuntary commitment, a mentally ill person must also be adjudged dangerous to self or others (or gravely disabled and unable to care for self -- a standard that is not relevant to the argument for dangerousness-based involuntary commitment). The details and the wording vary, but the two-pronged standard -- requiring findings of both mental illness and dangerousness before the law may override a person's desire to refuse treatment -- has been the norm in the U.S. since the 1970s, when civil libertarians successfully fought to add the dangerousness component to a commitment standard which had previously required only psychiatric findings of mental illness and "need for treatment" to forcibly confine a patient.

Which brings us to the second, and perhaps most central, faulty assumption of the three named above -- that "dangerousness" is an easily identifiable and readily measurable entity on which a court can reliably ground its decision about whether or not to override a respondent's refusal of psychiatric treatment. Again, that is simply not true. Psychological research into the phenomenon of "dangerousness" suggests that the rate of "false positives" -- cases in which mental health professionals have determined that someone is dangerous when, in fact, they are not -- ranges from 25% to 50% of cases. Thus, in as many as one out of every two cases in which a person is adjudged "dangerous", that finding may be false.

In an informative article published last month, Deborah Brauser interviewed Dr. Stephen Dubrovsky, M.D., chair of psychiatry at the University of Buffalo in New York and professor of psychiatry at the University of Colorado, Boulder. Dr. Dubrovsky told Brauser: "There are a lot of angry people in the world who act on their anger -- and the vast majority do not have anything wrong with them psychiatrically." Thus, "the assumption that most mass murderers are psychiatrically ill and have been in a psychiatrist's or therapist's office is incorrect." According to Dr. Dubrovsky, who has studied the 1999 Columbine shootings and interviewed survivors of that tragic event, "the perception that people can rely on the mental health profession to catch these things and prevent disastrous outcomes is really a vain hope." In terms of assessing a patient's level of dangerousness, Dr. Dubrovsky's conclusion echoes the findings of many research studies: "There are a variety of assessment methods available. But I don't know that any of them have ever been shown to actually predict dangerousness to any degree, and that's because a study like that is almost impossible to do."

Child and adolescent psychiatrist Michael Houston, M.D., highlights the problems inherent in distinguishing troubled youths from those who might turn violent. Dr. Houston told Brauser: "Signs such as social withdrawal, irritability, and a change in habits are seen in both groups.... We've found that the best predictor of future behavior is past behaviors. A history of violence towards family members, toward others, towards animals is a warning sign of sorts. But it's not universal." As far as anyone knows, neither James Holmes nor Adam Lanza had a history of violence or lawbreaking before they committed the horrific murders which brought them both to national attention. Given an absence of violent history, and the apparent inability of either law or psychiatry to correctly distinguish the dangerous from the nondangerous, on what should the law rely in deciding whether to label a troubled young person as a threat to society?

Our criminal law embraces the axiom that it is better for ten guilty defendants to walk free than for one innocent person to be unjustly convicted. Yet we are now told that the civil courts should move more aggressively to incarcerate mentally ill people on the flimsy grounds of future "dangerousness" -- though they have committed no crime at all! Protecting the public from danger is an important problem -- but this cannot be the right answer.
Some would argue that it is permissible for civil courts to adopt a less rights-protective standard because involuntary commitment does not raise the prospect of criminal conviction or imprisonment. From the perspective of a respondent who has been falsely judged "dangerous" and forcibly committed on that basis, that argument is specious in the extreme. A criminal conviction may add an extra measure of social stigma to civil incarceration. But if individual freedom is an important issue -- as it must be for a legal system which seeks to protect the rights of innocent individuals from unjustified intrusions by the majority - then the prospect of being forced by the law into a psychiatric hospital until such time as the mental health profession decides you no longer pose a threat which you never posed in the first place, raises strong and important concerns which the law should -- which it must -- act upon.

Finally, the issue of involuntary civil commitment should not be conflated with the problem of "deinstitutionalization" -- a collection of public policies, dating from the 1960s, which resulted in the defunding and emptying of psychiatric hospitals across the country and a resulting dearth of resources available to treat mental illness, including serious mental illness which might require long-term hospitalization. Deinstitutionalization has been a failure, and as a society we should seek and support ways of expanding the availability of psychiatric treatment for those suffering from mental illness who want and would benefit from such help. But making treatment more widely available for those who desire it is entirely different from forcing treatment on those who would otherwise refuse it.

On the question of involuntary civil commitment, psychiatry and law have importantly different perspectives. Where psychiatrists may look at the mentally ill and see a need for treatment, the law sees persons who, like all other persons, have the strong presumptive right to choose, or refuse, medical and psychiatric treatment. It is the law's particular obligation to protect that right unless there is powerfully persuasive proof of incapacity or the propensity for violence. In guarding that right, the law should ask its own questions, establish its own standards, and stand its ground against unfounded (although often well-intentioned) efforts to calm the public's fear of danger by depriving innocent persons of basic freedoms.

Cynthia Ward is Professor of Law at the College of William and Mary. 

In the wake of last month's tragedy at Newtown, some now assert that the law should make it easier to hospitalize dangerous people against their will. Their reasoning is intuitively seductive: If Adam Lanza (the Newtown shooter) and James Holmes (the Colorado "Batman" shooter) had been identified and forcibly confined as "dangerous" before they reached the point of committing murder, many innocent lives would have been saved. Therefore we must restructure the legal standard for involuntary commitment so that dangerous people like Holmes and Lanza can be institutionalized until such time as they no longer pose a risk of harm to innocent others.

The emotions which motivate this view are understandable. But the argument is not rationally defensible.

The case for locking up the dangerous usually proceeds from three assumptions: (1) that any person who commits atrocities such as the Newtown or Colorado shootings must be mentally ill (and thus, a proper subject of psychiatric treatment); (2) that "dangerousness" is a quantifiable entity which can be accurately identified and measured, like flour or sugar; and, (3) that involuntary civil commitment on grounds of dangerousness poses no threat to individual rights because it does not risk a criminal conviction or imprisonment.

All three assumptions are demonstrably wrong.

It is not true that all murderers -- even all mass murderers -- suffer from diagnosable mental illness. In an interesting Wall Street Journal piece last month, David Kopel reported two recent pieces of such evidence: "A 2000 New York Times study of 100 rampage murderers found that 47 were mentally ill. In the Journal of the American Academy of Psychiatry Law (2008), Jason C. Matejkowski and his co-authors reported that 16% of state prisoners who had perpetrated murders were mentally ill." In those numbers a mental health professional might see a serious need for treatment. The law, on the other hand -- whose precious charge is to protect individual rights -- sees the danger of over-relying on mental illness as a basis for involuntary commitment. If 47% of the Times's 100 rampage killers were mentally ill, then 53 % -- more than half -- were not. If 16 % of state-imprisoned murderers were mentally ill, then 84 % -- more than four-fifths -- were not. The presence or absence of mental illness is not a reliable gauge of a person's potential for homicidal violence.

Of course the legal standard for involuntary civil commitment does not rely solely on a finding of mental illness. To justify involuntary commitment, a mentally ill person must also be adjudged dangerous to self or others (or gravely disabled and unable to care for self -- a standard that is not relevant to the argument for dangerousness-based involuntary commitment). The details and the wording vary, but the two-pronged standard -- requiring findings of both mental illness and dangerousness before the law may override a person's desire to refuse treatment -- has been the norm in the U.S. since the 1970s, when civil libertarians successfully fought to add the dangerousness component to a commitment standard which had previously required only psychiatric findings of mental illness and "need for treatment" to forcibly confine a patient.

Which brings us to the second, and perhaps most central, faulty assumption of the three named above -- that "dangerousness" is an easily identifiable and readily measurable entity on which a court can reliably ground its decision about whether or not to override a respondent's refusal of psychiatric treatment. Again, that is simply not true. Psychological research into the phenomenon of "dangerousness" suggests that the rate of "false positives" -- cases in which mental health professionals have determined that someone is dangerous when, in fact, they are not -- ranges from 25% to 50% of cases. Thus, in as many as one out of every two cases in which a person is adjudged "dangerous", that finding may be false.

In an informative article published last month, Deborah Brauser interviewed Dr. Stephen Dubrovsky, M.D., chair of psychiatry at the University of Buffalo in New York and professor of psychiatry at the University of Colorado, Boulder. Dr. Dubrovsky told Brauser: "There are a lot of angry people in the world who act on their anger -- and the vast majority do not have anything wrong with them psychiatrically." Thus, "the assumption that most mass murderers are psychiatrically ill and have been in a psychiatrist's or therapist's office is incorrect." According to Dr. Dubrovsky, who has studied the 1999 Columbine shootings and interviewed survivors of that tragic event, "the perception that people can rely on the mental health profession to catch these things and prevent disastrous outcomes is really a vain hope." In terms of assessing a patient's level of dangerousness, Dr. Dubrovsky's conclusion echoes the findings of many research studies: "There are a variety of assessment methods available. But I don't know that any of them have ever been shown to actually predict dangerousness to any degree, and that's because a study like that is almost impossible to do."

Child and adolescent psychiatrist Michael Houston, M.D., highlights the problems inherent in distinguishing troubled youths from those who might turn violent. Dr. Houston told Brauser: "Signs such as social withdrawal, irritability, and a change in habits are seen in both groups.... We've found that the best predictor of future behavior is past behaviors. A history of violence towards family members, toward others, towards animals is a warning sign of sorts. But it's not universal." As far as anyone knows, neither James Holmes nor Adam Lanza had a history of violence or lawbreaking before they committed the horrific murders which brought them both to national attention. Given an absence of violent history, and the apparent inability of either law or psychiatry to correctly distinguish the dangerous from the nondangerous, on what should the law rely in deciding whether to label a troubled young person as a threat to society?

Our criminal law embraces the axiom that it is better for ten guilty defendants to walk free than for one innocent person to be unjustly convicted. Yet we are now told that the civil courts should move more aggressively to incarcerate mentally ill people on the flimsy grounds of future "dangerousness" -- though they have committed no crime at all! Protecting the public from danger is an important problem -- but this cannot be the right answer.
Some would argue that it is permissible for civil courts to adopt a less rights-protective standard because involuntary commitment does not raise the prospect of criminal conviction or imprisonment. From the perspective of a respondent who has been falsely judged "dangerous" and forcibly committed on that basis, that argument is specious in the extreme. A criminal conviction may add an extra measure of social stigma to civil incarceration. But if individual freedom is an important issue -- as it must be for a legal system which seeks to protect the rights of innocent individuals from unjustified intrusions by the majority - then the prospect of being forced by the law into a psychiatric hospital until such time as the mental health profession decides you no longer pose a threat which you never posed in the first place, raises strong and important concerns which the law should -- which it must -- act upon.

Finally, the issue of involuntary civil commitment should not be conflated with the problem of "deinstitutionalization" -- a collection of public policies, dating from the 1960s, which resulted in the defunding and emptying of psychiatric hospitals across the country and a resulting dearth of resources available to treat mental illness, including serious mental illness which might require long-term hospitalization. Deinstitutionalization has been a failure, and as a society we should seek and support ways of expanding the availability of psychiatric treatment for those suffering from mental illness who want and would benefit from such help. But making treatment more widely available for those who desire it is entirely different from forcing treatment on those who would otherwise refuse it.

On the question of involuntary civil commitment, psychiatry and law have importantly different perspectives. Where psychiatrists may look at the mentally ill and see a need for treatment, the law sees persons who, like all other persons, have the strong presumptive right to choose, or refuse, medical and psychiatric treatment. It is the law's particular obligation to protect that right unless there is powerfully persuasive proof of incapacity or the propensity for violence. In guarding that right, the law should ask its own questions, establish its own standards, and stand its ground against unfounded (although often well-intentioned) efforts to calm the public's fear of danger by depriving innocent persons of basic freedoms.

Cynthia Ward is Professor of Law at the College of William and Mary.