The populations most likely to offend and most likely to be victimised are substantially the same people, which complicates any clean division into offenders and victims.
Across surveys and cohort studies, prior offending predicts victimisation and prior victimisation predicts offending, with shared risk factors including age, routine activities, and the settings in which both occur. The finding sits awkwardly with policy and with public narrative, both of which are built on two separate groups.
In practice
- A programme designed for victims and a programme designed for offenders are competing for much of the same caseload.
- A neighbourhood's most frequently assaulted residents appear in the same records as its most frequently arrested.