Neighborhood Predictors of Intimate Partner Violence: A Theory-Informed Analysis Using Hierarchical Linear Modeling

Background and Purpose: Due to high prevalence rates and deleterious effects on individuals, families, and communities, intimate partner violence (IPV) is a significant public health problem. Because IPV occurs in the context of neighborhoods, research must examine the broader environment in addition to individual-level factors to successfully facilitate behavior change. Previous research has provided useful insights; however, this body of research has notable limitations, including statistical analyses that do not account for multilevel structure of the data, a tenuous theoretical framework resulting in limited predictors under study, and little research examining neighborhood risk factors for men-as-perpetrators. Consequently, the following research question was addressed using Hierarchical Linear Modeling (HLM): Are neighborhood-level factors identified in the Determinants of Health framework, (e.g., social environment and social capital), and by Social Disorganization Theory, (e.g., social disorganization and collective efficacy), associated with the occurrence of male-to-female IPV, after controlling for important individual-level risk-factors for men, e.g., alcohol and drug use, level of education, employment, age, female-to-male IPV, and impulsivity?

Methods: The Fragile Families and Child Wellbeing Study, a national sample of families at risk for breaking up and living in poverty, spans nine waves of data collection. A cross-sectional analysis was run using HLM and a random intercept model on wave 4 data because data on key variables were collected during this wave. Data were from a subsample of fathers in in-tact intimate relationships with the mother of their child (n = 474) who resided in the same census tracts (n = 186), a proxy for neighborhoods. Variables were constructed from 2000 U.S. Census data, father-report of neighborhood- and individual-level variables, and mother-reports of IPV.  IPV items drawn from the Conflict Tactic Scale were coded such that low numbers reflected less violence exposure and high numbers indicted more violence exposure. Robustness checks were completed with a sample of mothers and fathers who were not romantically involved at wave 4, but who were in a relationship with each other previously (n = 843).

Results: Women’s IPV exposure was compared to national incidence rates. On average, men in the sample were 33 years old, had at least a high school degree, were employed in the last week, reported low to moderate drinking, very little drug use, low to moderate levels of dysfunctional impulsivity, and similar levels of female to male IPV as compared to women’s report of IPV. Results indicated that neighborhood-level, i.e., concentrated disadvantage, and individual-level factors, i.e., female-to-male IPV, were robust predictors of women’s IPV victimization in this sample.

Conclusion & Implications: Results from this study suggests that neighborhood-level factors have an effect over and above individual-level factors on women’s IPV victimization. Future research should improve upon measurement and design to examine theoretically-informed neighborhood-level predictors related to IPV. Policy-makers and community stakeholders should consider investing in community economic development strategies to revitalize the economic, physical, and social infrastructures in low-income neighborhoods. Additionally, these results indicate that IPV programming directed at violent or aggressive behaviors for both partners may lend itself towards a more successful model.