Analyzing offense patterns as a function of mental illness to test the criminalization hypothesis.

OBJECTIVE Programs for offenders with mental illness seem to be based on a hypothesis that untreated symptoms are the main source of criminal behavior and that linkage with psychiatric services is the solution. This study tested this criminalization hypothesis, which implies that these individuals have unique patterns of offending. METHODS Participants were 220 parolees; 111 had a serious mental illness, and 109 did not. Interview data and records were used to reliably classify offenders into one of five groups, based on their lifetime pattern of offending: psychotic, disadvantaged, reactive, instrumental, or gang- or drug-related affiliation. The distributions of those with and without serious mental illness were compared. RESULTS A small but important minority of offenders with a mental illness (7%, N=8) fit the criminalization hypothesis, in that their criminal behavior was a direct result of psychosis (5%, N=6) or comprised minor "survival" crimes related to poverty (2%, N=2). However, the reactive group contained virtually all offenders with a mental illness (90%, N=100) and the vast majority of offenders without a mental illness (68%, N=74), suggesting that criminal behavior for both groups chiefly was driven by hostility, disinhibition, and emotional reactivity. For most offenders with a mental illness in the reactive group, crime was also driven by substance dependence. CONCLUSIONS Offenders with serious mental illness manifested heterogeneous patterns of offending that may stem from a variety of sources. Although psychiatric service linkage may reduce recidivism for a visible minority, treatment that targets impulsivity and other common criminogenic needs may be needed to prevent recidivism for the larger group.