Personal Social Networks of Women with Legal Involvement in Substance Abuse Treatment

Background & Purpose: Substance abuse is a key common correlate of criminal offending among women. Few studies have examined personal network characteristics (composition, support availability, and structure), limiting our understanding of the social context in which justice-involved women with substance use disorder (SUD) are embedded. Women, especially women in the criminal legal system, are reported to have less social capital and social support in comparison to men yet indicate the importance of relationships as motivational drivers for sobriety and behavior change.

Methods: Women (N = 376) were interviewed in three inner-city county-funded women-only substance abuse treatment programs at intake (T1) and follow-ups 1 (T2), 6 (T3), and 12 (T4) months later (80% retention at T4). A network software program, EgoNet, elicited 25 network members (alters) with whom they had had any type of contact in the past six months. Respondents were asked about their relationship with each alter (composition; e.g., family, substance user, very close alter); concrete, emotional, and sobriety support availability; and connections between each unique pair of relationship (structure; e.g., density). Density, a score between 0 and 1, quantifies cohesiveness as calculated by the proportion of ties in a network relative to the total of all possible ties. Mixed model repeated measures analyses with maximum likelihood estimation and unstructured covariance matrix were used to account for correlated responses within a subject, controlling for relevant confounders including trauma symptoms, co-occurring mental disorder, and self-efficacy.

Results: Participants were 60% African American with a mean age of 36.5 (SD = 10.4). Slightly less than half (45%, n = 168) were legally involved at the time of intake, with 23% (n = 86) reporting a history of a juvenile offense. The majority (73%, n = 275) had a co-occurring mental disorder, with more than half (57%, n = 216) being diagnosed with Major depression. The primary SUD was cocaine (56%), followed by alcohol (47%), marijuana (40%), and opiate (23%). Women with current legal involvement had fewer family members (9.7 vs. 10.7) and less number of alters who are “very close” (11.5 vs. 12.6), and their networks were less densely connected (0.30 vs. 0.35) than women with no current legal involvement over the 12-month study period. However, current legal involvement was not related to the number of alters reported as providing emotional, concrete, or sobriety support and the number of substance-using alters. Association of current legal involvement with the network characteristics was consistent over the follow-up period. Greater trauma symptoms were related to fewer family members and “very close” alters in the network.

Conclusions: Few differences were found in personal network characteristics between women with current legal involvement vs. those without current legal involvement, suggesting the social networks of these two populations are more similar than they are distinct. However, the study did not measure the severity of women’s legal involvement (e.g., misdemeanor and/or felony offending history). The null effect of negative social capital (i.e., number of substance-using alters) across groups has implications for both treatment intervention and criminological theory.