ABSTRACT
Background
Criminal careers have been extensively studied in general population sample, but less is known about such patterns among people with major mental illness, and where so, criminal justice expenditure has not been taken into account.
Aims
Our aim was to examine criminal justice system expenditure over time in one Florida county. Our main research question was whether treatment for mental disorders was related to a change in criminal offending and expenditure trajectory.
Methods
We used the Pinellas County (Florida) Criminal Justice Information System to identify individuals under age 65 arrested between July 2003 and June 2004. Archival medical service, social and homeless services data were used to identify individuals with a serious mental illness. A two‐step analysis was used to examine the data: first, we identified groups of people with similar patterns of criminal justice expenditures over 4 years (July 2002 to June 2006); second, we evaluated their demographic characteristics, diagnosis and treatment as potential predictors of group membership.
Results
Three thousand seven hundred sixty‐nine people with serious mental illness were identified in the Pinellas County jail population. Their average length of stay in jail was 151 days and in prison was 48 days. The trajectory analysis identified three groups of individuals with distinct trajectories of criminal justice expenditures: those with low stable, those with initially high but decreasing and those with initially high and sustained or increasing. Mental health treatment, whether acute or sustained, voluntary or mandatory, was associated with membership of the low stable group.
Conclusion
Review of criminal justice expenditure over time on individuals with major mental disorder may provide important indicators of unmet need for mental health services. Furthermore, it seems probable that improved provision of such services for them could reduce recidivism as well as improving health. Interventions may also be better focused if criminal justice expenditure trajectories are examined; programmes targeting re‐offending as well as specific mental health problems may be most effective. Copyright © 2011 John Wiley & Sons, Ltd.