Targeting the Optimal Domestic Violence Offender: Shifting from clinical risk to actuarial harm

Mr Andrew Hurst1,2

1NSW Police Force, Sydney, Australia
2University of Cambridge, Cambridge, United Kingdom


Following an Intimate Partner Domestic Violence Incident (IPV), what relevant factors can predict the amount of harm caused by the same offender over the next two years?

This study concerns 48,293 incidents of ‘Domestic Violence’ with a classification of ‘intimate partner’, recorded by the New South Wales Police Force in the Computerised Operational Policing System (COPS) over a two-year period between 1st of January 2015 through to 31st December 2016. The independent variables were 97 predictors, obtained from the prior offending history of each offender involved in each presenting IPV incident.

The 48,293 incident-offenders committed 72,292 offences and generated 2.4 million days of crime harm. The offences committed during the presenting IPV incident included Violence Offences (60.5%), Property Offences (16.9%), and Serious Offences (1.9%).

The findings revealed the average offender was engaging in non-crime contact with police (23.18 years), serious offending (27.43 years), and violent offending (28.47 years), all prior to the average age of the first domestic related offence (30.46 years), pointing towards generalist offending.

83.5% incident-offenders had a prior history of violence, 82.6% had a prior history of domestic violence, and 33.7% had a prior history of property offences.

58.0% of offenders had committed a prior serious offence, 65.0% had a prior drug offence, and 83.5% had a prior violence offence.

80% of the post-incident harm was attributed to just 11.80% of the incidents, generating a minimum of 147.65 days of crime harm, committed by the same offender, over the following two years from the presenting IPV incident.


Bio to come.