How bad is civil commitment? A study of attitudes toward violence and involuntary hospitalization.

D. Mossman and K. J. Hart,
Bull. Amer. Acad. Psychiatry & the Law 21(2): 181-94, 1993.
Civil commitment statutes throughout the nation authorize involuntary hospitalization for persons who are believed dangerous to others, even though clinicians' ability to predict violence is imperfect. Decision-makers faced with ambiguous evidence about future violence must make either-or decisions about involuntary hospitalization. Such decisions can be characterized as "true positives" (hospitalization of a person who would have acted violently if released), "true negatives" (nonviolent person is not hospitalized), "false positive" (nonviolent person is hospitalized), or "false negative" (person is released and subsequently acts violently). This paper presents two pilot studies of attitudes about false negative and false positive decisions, and explains how Decision Theory can use information gleaned from such studies to establish optimal decision thresholds for initiating involuntary hospitalization. Subjects expressed a broad range of implicit tolerances for false negative and false positive predictions. Though most subjects preferred being hospitalized for three days to being the victim of a knife-wielding attacker, a substantial minority preferred being attacked to being hospitalized. The article briefly explores the practical implications of these findings, which include an implicit endorsement of stringent commitment policies that would release a large fraction of potentially violent persons.