The Effects of Community-Based Treatment and Neighborhood Disadvantage on Arrest Among Mental Health Court Participants

Background: Due to the over-representation of people with mental illness in criminal justice system, mental health courts (MHCs) was established to provide community-based treatment instead of incarceration. Previous research found that MHC participants have successful outcomes (e.g., more treatment and fewer arrest). While most MHC studies investigate the factors associated with recidivism, little is known about effect of treatment and neighborhood characteristics on arrest and how the effect of those factors changes after MHC court enrollment as compared to traditional court. The purpose of the study is to examine whether offenders with mental illness in MHCs (N=357) have more positive effects of treatment and neighborhood factors on recidivism compared to traditional court participants (N=384).

Method: Data were from the MacArthur MHC project, which include the survey from four MHCs and comparison traditional courts at each site. Participants were interviewed at baseline and six months follow-up. The objective arrest data were collected from FBI. Logistic regression and propensity score weighting was used to investigate the effects of treatment and neighborhood disadvantage on crime after controlling for selection bias. Treatment variables include treatment adherence (i.e., medication and appointment), treatment perception (i.e., treatment motivation and perceived voluntariness), and treatment usage (mental health and substance abuse service). Neighborhood disadvantage data include 5 indexes (i.e., rate of poverty, unemployment, vacant/abandoned housing, public benefits recipients, and single-parent families) as a census tract level from 2005-2009. Neighborhood disadvantage data was obtained from the American Community Survey, and linked with residential information. All statistical analyses were conducted using STATA version 13 and ArcGIS.

Result: The regression model indicated that MHC participants had significant more positive effect of treatment adherence, treatment perception, treatment usage, and neighborhood disadvantage on arrest compared to the traditional court participants both at baseline and follow-up. In terms of the changes of effects after the MHC enrollment, MHC participants had positive changes for appointment adherence, medication adherence, treatment motivation, and perceived voluntariness, but those changes were not statistically significant (p>.05). Counter to what was hypothesized, MHC participants had negative changes of effects for mental health service, substance abuse service, and neighborhood disadvantage on arrest over time but not statistically significant (p>.05).

Conclusions and Implications: Regression analyses found that there were significant positive effects in the MHC group which led to fewer arrests than in the TAU group at baseline and follow-up. Consistent with the MHC goals, findings suggest the efficacy of the MHC program. Recognizing the effects of treatment factors and neighborhood disadvantage can help offenders with mental illness to have successful community rehabilitation. Given the fact that quality of treatment may also affect the future crime among offenders with mental illness, further research is recommended to examine the links between capacity of the agencies (e.g., credentials of clinical staff, community resources, and programs) and justice outcomes in the MHC system. In addition, future research could take advantage of variation in neighborhood disadvantage by type of housing (e.g., residential facility, rented apt., or own house).