Social Networks, Substance Abuse, and Contact with the Criminal Justice System for Persons with Severe Mental Illness

Purpose: Adults with severe mental illness (SMI) tend to have small social networks that are characterized by concentrated disadvantage. Impoverished levels of social and economic capital among adults with SMI have been linked to their overrepresentation in the criminal justice system, with incarceration rates among this population occurring at five times the rate of the general population. In his seminal work on social networks, Granovetter (1973) argues that numerous and diffuse social ties lead to increased pathways of influence, information, and mobility opportunities while more regular contact with a dense network may lead to encapsulated and impoverished social environments. Studies also reveal that substance abuse behaviors contribute to both encapsulated social networks of similar others who are also using as well as higher rates of incarceration. However, no studies to date have examined whether the nature of social networks has an independent effect on the risk of contact with the criminal justice system and whether substance abuse interacts with social network characteristics to increase the likelihood of incarceration among people with SMI. It is hypothesized that larger social networks will decrease the likelihood of arrest while a higher degree of contact with networks (i.e. encapsulation) will increase the likelihood of arrest. Additionally, it is hypothesized that substance abuse behaviors will interact with social network characteristics to increase the likelihood of arrest.

Methods: This study was based on a sample of 119 individuals diagnosed with schizophrenia, schizoaffective disorder, bipolar disorder, major depression with psychotic features, and co-occurring substance abuse from an NIMH funded investigation of consumers living in the community. Three major variables of interest (size of social network, degree of contact with social network, and substance abuse) along with demographic covariates (age, ethnicity, income, gender, and education) were examined as predictors of arrest. Substance abuse was measured by the presence of a substance abuse or dependence diagnosis meeting DSMIV criteria and size of network, degree of contact with network, and the demographic covariates were gathered through self-report surveys. First, a series of bi-variate chi square tests of independence were performed followed by a logistic regression analysis, including substance abuse as a potential moderator of social network characteristics in predicting arrest.

Results: Findings suggest that persons with a higher degree of contact with friends, indicating a more encapsulated network, were 4.8 times more likely to have been arrested (OR=4.766, 95% CI=.1234, 18.410) and those with larger social networks were 28% less likely to be arrested at borderline statistical significance (OR=.277, p=.06). Results also indicate that substance abuse did not moderate network characteristics in influencing the likelihood of arrest but did have a main effect on arrest, increasing the likelihood of being arrested by 23 times (OR=23.230, 95% CI=5.948, 90.722). Lastly, mentally ill men were 11 times more likely to have been arrested than women (OR=10.931, 95% CI=3.019, 39.585).

Implications: Interventions that focus on bridging and linking opportunities to develop connections with broad social networks may be crucial in achieving social integration and reducing arrest and recidivism for this population.