Predicting Pathways From Victimization to Perpetration of Violence by Women

Background: Extant research illustrates the association between victimization and perpetration of violence. Even though most women involved in criminal justice system have histories of victimization, women comprise only 5% of those sentenced for a violent offense. Theory about women’s use of violence postulates three reasons: 1) self defense, 2) prevention of perceived threat, or 3) aggression. Other factors such as mental illness and substance use disorders have also been associated with perpetration. Most research on women and violence has focused on those that use force against intimate partners, with relatively few studies examining a broader spectrum of perpetration. Some studies indicate that more serious violence (e.g. killing abuser) may be the result of repressed rage or expressive anger, however little is known about differences among women and what factors influence the use, or target, of violence. Our research questions: Do different forms of victimization lead to variation in the target of violence? Do factors (i.e., anger, substance abuse and/or mental health) mediate the relationship between victimization and perpetration?

Method: An anonymous survey was collected from a random sample of women within one state’s prison system (N=574). This analysis focuses on variables depicting childhood adversity; victimization and perpetration of abuse by/toward partners and non-partners; serious mental illness (SMI), alcohol and drug use (SUD); and inventories of expressive and instrumental anger. Structural equation modeling (SEM) was conducted to assess a theoretical framework linking childhood adversity, victimization by partner/non-partner, and perpetration of violence against partners/non-partners. Mediation by SMI, SUD and type of anger was also tested.

Results: The hypothesized model was an excellent fit to the data (CFI=.99), revealing significant indirect paths from childhood adversity through victimization by partners and non-partners to perpetration of violence against partners (β=.20) and non-partners (β=.19). However, there were differences: women’s perpetration of violence against partners was influenced by victimization by both partners (β=.15) and others (β=.13), but their perpetration of violence against non-partners was influenced only by victimization by non-partners (β=.33). Perpetration against partners was mediated by both instrumental anger (β=.24) and expressive anger (β=.11), but perpetration against non-partners was mediated only by instrumental anger (β=.29). SUD and SMI had significant direct associations with childhood adversity and victimization but only indirect associations with perpetration, through anger.

Implications: Among this sample of incarcerated women, childhood adversity and victimization across the lifecourse were associated with perpetration of violence against both partners and non-partners, with important mediators of perpetration -- anger, SUD, and SMI. Clearly, trauma-informed social work prevention and treatment interventions are needed, especially in conjunction with mental health and substance abuse treatment. Associations differed by the target of violence, with women’s perpetration against non-partners associated only with victimization by non-partners and with instrumental anger. This suggests that interventions may benefit from tailoring to address specific forms of victimization and anger presented by women with different experiences. SUD and SMI were associated with violence perpetration only indirectly, through anger, suggesting that anger, in addition to SUD and SMI, should be addressed directly in interventions with women who perpetrate violence.