Criminogenic Needs and Recidivism in Male IPV Perpetrators

Journal of Forensic Psychology Research and Practice | p. 1 - 25

Criminogenic Needs Among Men Who Perpetrate Intimate Partner Violence: Association with Risk Management Recommendations and Recidivism 

N. Zoe Hilton; Department of Psychiatry, University of Toronto, Toronto, Canada; Waypoint Centre for Mental Health Care, Waypoint Research Institute, Penetanguishene, Ontario, Canada

Anna T. Pham; Department of Psychology, Carleton University, Ottawa, Ontario, Canada

Kevin L. Nunes; Department of Psychology, Carleton University, Ottawa, Ontario, Canada

Sandy Jung; Department of Psychology, MacEwan University, Edmonton, Alberta, Canada

Liam Ennis; Forensic Behavioural Science Group, Edmonton, Alberta, Canada; Psychiatry Department, University of Alberta, Edmonton, Alberta, Canada

Background: Understanding criminogenic needs is important for risk management of intimate partner violence (IPV). 

Method: We analyzed criminogenic needs and management recommendations in 300 men charged with IPV. 

Results: Case files mentioned five criminogenic needs, from antisocial personality (50%) to family/marital problems (94%). Total needs positively correlated with risk management recommendations substance use positively correlated with IPV recidivism. Needs did not predict recidivism above Ontario Domestic Assault Risk Assessment (ODARA) scores. Risk management recommendations did not reduce the ODARA’s predictive effect. 

Conclusions: We found initial evidence for need principle adherence. Future research should improve criminogenic need measurement and examine risk management implementation.

“Intimate partner violence (IPV) refers to acts of physical aggression against a marital, cohabiting, or dating partner… Scholars have begun examining men’s use of IPV through the lens of the Risk, Need, Responsivity (RNR) principles of effective correctional service (e.g., Bonta & Andrews, 2024; Radatz & Wright, 2016; Richards et al., 2021) and found that RNR adherence is positively associated with treatment effectiveness (Travers et al., 2021). There is substantial progress in applying the risk principle, which posits that treatment and other interventions should be prioritized according to individuals’ risk level, and several well validated tools now exist… However, exploration of RNR in relation to IPV is relatively new, and substantial questions remain (Radatz et al., 2021). Little research has been conducted on the need principle, whereby assessing and addressing modifiable risk factors associated with recidivism is essential to effective risk management. Therefore, the present study examined criminogenic needs and their relation to IPV recidivism among men who were assessed by threat analysts following an incident of IPV against a woman” (p. 2-3). 

“We conducted a secondary analysis of an existing dataset to examine the presence and predictive value of criminogenic needs among men referred to a threat assessment service following a police investigation of IPV against a woman. The dataset included 300 men charged with an assault or other violent index offense against a female intimate partner and referred in 2010 to 2016 to a police threat assessment service that served Alberta, Canada. A portion of this sample (n = 247, 82%) was reported to be high risk compared to routine police samples, based on ODARA and other IPV risk assessment scores” (p. 5).

“In this sample of men who perpetrated IPV and were referred to a threat assessment service, men had, on average, more than three documented criminogenic needs. One in five men had all five of the needs we identified. Criminogenic needs were more common than previously reported in a routine police sample of men who perpetrate IPV (Hilton & Radatz, 2021), adding to evidence that a threat assessment service caseload represents high-risk, high need individuals” (p. 14). 

“Having more criminogenic needs was related to higher actuarial risk of IPV recidivism, as measured by the ODARA. However, contrary to expectations based on previous research with a routine police sample (Hilton & Radatz, 2021), criminogenic needs were only marginally related to subsequent IPV recidivism and did not contribute to a multivariate regression model predicting IPV recidivism in which the ODARA was significant. Of the individual criminogenic needs, only substance use was related to IPV recidivism. Further, only substance use significantly contributed to the multivariate regression model of IPV recidivism in which individual needs were entered” (p. 14). 

“Consistent with the need principle, and as expected, risk management recommendations were related to criminogenic needs. That is, threat analysts made more recommendations for men with more needs… Further, the number of risk management recommendations did not reduce the ODARA’s predictive effect for IPV recidivism, nor did it reduce the (positive but nonsignificant) association of criminogenic needs with IPV recidivism. It appears that, despite some adherence to the need principle, threat analysts’ risk management recommendations were insufficient to mitigate the actuarial risk of IPV perpetration” (p. 15). 

“We found an overall association between criminogenic needs and risk management recommendations; however, beyond substance use, the specific recommendations did not map well onto specific criminogenic needs. Improved assessment and intervention with criminogenic needs in threat assessment service cases may be valuable, including risk/needs assessment tools such as an LS instrument (e.g., Wormith & Bonta, 2021). Incorporating criminogenic needs more intentionally into threat assessment practice could begin with training on the need principle and collaboration with community partners to optimize evidence-based services for criminogenic needs. Threat analysts could then make more specific risk management recommendations. Criminogenic needs are, by definition, dynamic, suggesting that it is important to reassess these factors over time (e.g., Skeem & Monahan, 2011; Stone et al., 2021). In previous research using the current sample (Peters et al., 2023), three distinct classes of men were identified based on levels of IPV-specific risk factors (e.g., recent relationship problems, past IPV, attitudes supporting IPV) and general violence risk factors (e.g., substance abuse, past assault of strangers, employment problems). Subtypes may differ in their criminogenic needs, and these needs may have more predictive value in some subtypes (e.g., generally violent/antisocial) than others (e.g., IPV-specific). As community treatment resources are limited, tailoring resources to subtypes based on their particular needs increases the likelihood of timely and effective intervention.

Our study examined only cases that involved physical violence. However, intimate partner abuse encompasses more than physical violence, and practitioners should be aware of a broader set of behaviors. For example, over half of nonviolent charges against men who perpetrate IPV may involve intimate partner victims (Hilton & Eke, 2016). There may also be responsivity factors such as age, motivation, unemployment, income, or education that promote or impede individuals’ response to criminal justice intervention or treatment (e.g., Hilton & Ennis, 2020). Thus, a broader range of criminal and extra-criminal behaviors should be attended to, although there is currently a limited available evidence base to guide practice specific to IPV. An increasingly recognized area is intersectionality (e.g., Batastini et al., 2022), whereby an individual’s multiple identities, such as gender, sexuality, race, and socioeconomic status, intersect to influence their lived experiences, and potentially their responsivity to intervention. Further, Pham and Jung (2022) found that ODARA items that loaded onto an antisociality factor were most predictive of IPV recidivism in a routine police sample, and a victim vulnerability factor also offered predictive value, suggesting potential avenues for risk management that focus on antisocial and procriminal criminogenic needs as well as victim safety” (p. 17-18).

“Our study contributes to a limited empirical body of literature focused on the criminogenic needs of IPV men. However, it is not without limitations. We analyzed a dataset not originally created to test criminogenic needs. We did not use risk/need assessments such as the Level of Service (LS) instruments (e.g., Wormith & Bonta, 2021) nor interview participants. Instead, we identified variables approximating criminogenic needs that were already coded based on structured assessment tools intended to capture changeable risk/ need factors (e.g., Kropp & Hart, 2015). We focused on criminogenic needs in the Central Eight because existing evidence has established their role in general criminal offending (e.g., Bonta & Andrews, 2024) and we aimed to examine their role in IPV recidivism. Due to the dataset limitations, we could only identify five criminogenic needs and we were missing information on procriminal associates and leisure time, similar to previous studies by Hilton and Radatz (2018, 2021). The needs we did measure were further limited by the archival data; for example, it was difficult for coders to disentangle marital problems from the conflict at the time of the index assault, leading to a possible ceiling effect for this item. Future research should use more comprehensive and validated measures of criminogenic needs. 

Our measures of criminogenic needs were based on archival data captured at a single point in time. Researchers should consider gathering interview and self-report data on criminogenic needs in repeated measures designs to evaluate change in criminogenic needs over time, after recommendations, or following actual intervention for IPV (e.g., Schafers et al., 2021). There is evidence from studies of general offending that repeated and recent measurements may be beneficial for the assessment of short-term or imminent risk (e.g., Simmons et al., 2023; Stone et al., 2021). IPV is often repeated against the same victim and can take place in the context of multiple acts of coercive control (e.g., Hilton et al., 2023). Therefore, examining the potential benefit of repeated assessment of IPV recidivism risk and criminogenic needs is an important next step for research. 

Our matching of risk management recommendations to criminogenic needs was rudimentary, and limited by the broad nature of recommendations that could apply to multiple treatment needs, suggesting limited specificity of risk management strategies used in practice. More sophisticated matching processes could involve identifying whether individuals received treatments that have goals and content designed to address their top ranked assessed needs (e.g., Baglivio et al., 2018); measuring both good matches and bad matches (e.g., intervention given for a need that is not present; Drawbridge et al., 2019); and incorporating information into a weighted or unweighted service fit index (e.g., Gill & Wilson, 2017). 

Our study involved a high risk sample of men who perpetrated IPV against a woman; the findings may not generalize to IPV by women or to persons with diverse gender or sexual identities. Further, our sample included one third non-White men, and used an IPV risk tool that has limited validation research in Indigenous (Hegel et al., 2022) and Black (Hilton & Radatz, 2024) populations; the findings regarding the ODARA may require replication in diverse samples. Also, our sample is not representative of routine police caseloads, although it provided an opportunity to examine criminogenic needs in police-involved cases because threat analysts aimed to identify risks that can be mitigated and they documented sufficiently detailed information to evaluate several criminogenic needs. Storey et al. (2013) reported that police officers’ management recommendations were associated with decreased IPV recidivism in individuals assessed as high risk, but with increased recidivism in lower risk individuals. We were not able to investigate this pattern in our study because the preponderance of cases scored in the highest ODARA risk category. Further research into criminogenic needs in routine police samples of IPV cases would be valuable because the first opportunity to engage in risk management would occur during the police response and pre-trial decisions” (p. 16-17).