From neighborhood to household: connections between neighborhood vacant and abandoned property and family violence

Res Sq [Preprint]. 2024 Mar 11:rs.3.rs-4022003. doi: 10.21203/rs.3.rs-4022003/v1.

Abstract

Rates of family violence, including intimate partner violence (IPV) and child maltreatment, remain high in the U.S. and contribute to substantial health and economic costs. How neighborhood environment may influence family violence remains poorly understood. We examine the association between neighborhood vacant and abandoned properties and family violence, and the role collective efficacy may play in that relationship. Data were used from a longitudinal cohort of 218 maternal-child dyads in a southern U.S. city known for elevated rates of violence. Women were matched on their propensity score, for living in a neighborhood with elevated vacant and cited properties. Analyses accounting for clustering in neighborhood and matched groups were conducted to examine the association between neighborhood vacant and abandoned property and family violence, and the potential mediating relationship of collective efficacy. The likelihood of experiencing child maltreatment at 12-months of age was more than twice as high for children living in neighborhoods with a high vacant and cited property rates compared to women living in neighborhoods with fewer vacant and cited properties (OR=2.11, 95% CI=1.03, 4.31). Women living in neighborhoods characterized by high levels of vacant and cited properties were also more than twice as likely to report IPV (OR=2.52, 95% CI=1.21, 5.25). Associations remained mostly stable after controlling for key covariates. Collective efficacy did not act as a mediator in the relationship between vacant and cited properties and family violence. Reducing neighborhood vacant and cited properties may be an important target for interventions focused on reducing family violence.

Keywords: Family violence; child maltreatment; intimate partner violence; neighbourhood disorder; vacant property.

Publication types

  • Preprint