Email | Print

Questions are often asked whether the mentally ill as a group are more dangerous or violent than the general population. The concern here pertains essentially to the goal of social defense -- public safety -- the protection of the community.

Although these questions can be simply framed and are often answered in similarly simple fashion, brief consideration will indicate some of the inherent ambiguities and difficulties in providing accurate and meaningful responses. Moreover, the issue tends quite often to be further confounded by various attitudinal and ideological considerations.

To address these questions it is necessary first to clarify the key terms involved. How, for example, are the terms "mentally ill" and "violent" being used? The absence of clear and specific definitions leads to various uses. For example, in referring to the "mentally ill" does one have in mind some specific estimate of prevalence -- e.g., the number of persons in the general population suffering from at least one psychiatric disorder during the course of a six month period (about 19% of American adults)? Or, is one referring not only to those either suffering from serious mental disorders (e.g., psychoses) and/or with recent histories of psychiatric hospitalization? Or, is the reference to persons who are currently receiving some form of mental healthcare and treatment -- including the vast majority who receive such services in various private and public outpatient settings?

Similarly, the terms "dangerous" and "violence" can have several meanings. For example, although "dangerous" is broader in meaning and can be used to include a wide range of antisocial and criminal acts, "violence" generally refers to acts that are intended to inflict physical harm to other persons or to oneself, or to damage property.

Among the major mental disorders there are many which, by their very nature and pattern of symptoms, will generally be associated with reduced likelihood of criminal and violent behaviors toward others (e.g., anxiety and depressive disorders); other disorders have characteristics that will generally be associated with increased likelihood of such behaviors -- e.g., alcohol dependence and abuse, drug dependence and abuse, impulse control and sexual disorders, and some of the personality disorders (all of which will be more common among criminal and prison populations).
Among persons suffering from serious mental disorders (e.g., schizophrenia and mood disorders), those with co-occurring substance abuse and especially those with prior histories of assaultive and violent behaviors will typically present increased risks of violence.

Considering the larger empirical studies that have been conducted over the past several decades, the vast majority has focused on mentally ill persons who had been hospitalized in state or county mental facilities. Obviously, these samples constitute a highly selective subgroup of the class of "mentally ill" persons. Such samples are biased in terms of the nature and seriousness of the mental disorders and in terms of economic and social class factors. One must therefore be very careful not to generalize from such studies to other categories of mentally ill persons. Similarly, studies involving seriously mentally ill persons with histories of assaultive, disruptive and violent behaviors will typically reveal high rates of subsequent violence -- most often towards persons with whom they have frequent interactions (e.g., family members). Again, one must be very careful about generalizing to other groups.

Rather consistently, studies prior to the Sixties found that mentally ill persons who had been hospitalized in public sector facilities tended to have post-release arrest rates that were lower than those for the general population. Understandably, arrest rates for alleged criminal acts do not provide a very good or precise measure of "dangerous" or "violent" behaviors towards others, but they constitute an easily available and relevant, albeit crude measure. One must also remember that prior to the Sixties both voluntary and involuntary psychiatric hospitalizations were much easier to accomplish and patients typically had rather long -- and for those involuntarily hospitalized, even indeterminate -- periods of confinement. Hence, the exposure to risk for arrests in the community was markedly reduced.

In contrast, almost every large study since the Sixties has found that persons with histories of mental hospitalization (typically in public sector facilities) tend to have subsequent arrest rates higher than those for the general population. For hospitalized patients with histories of prior arrests, post-release arrest rates were markedly higher.

The above changes in research findings are not surprising if one considers the nature of the major reforms of mental health laws and other policy changes that have occurred since the late Sixties -- e.g., more demanding criteria for voluntary admissions, more restrictive and dangerousness-based legal criteria for involuntary hospitalization, durational limits on the length of confinement for committed patients and considerably shorter average lengths of stay for voluntary admissions, major shifts away from hospitals to the community as the primary focus of treatment, limited resources for intensive community-based care, treatment, and other social supports for the seriously mentally ill, etc.

It is more important to recognize that the answers obtained to questions about violence among the mentally ill depend on how questions are framed. Thus, in terms of public policy that is ostensibly focused on social defense concerns, the fundamental question would be: Which groups in American society pose the greatest threat to the community in terms of inflicting death and serious injury on others?

Framed this way, the response -- based on the evidence -- would be that the seriously mentally ill (despite the shocking and headline-grabbing nature of some of their violent acts) would clearly not be in top contention. In terms of the sheer numbers and lethality involved, top place would most likely go to persons with repeated arrests for driving under the influence of alcohol.