General and victim-specific empathy in violent offenders and non-offenders: development of the violent offender empathy measure

Abstract

The purpose of the present study was to replicate and extend research by Fernandez, Marshall, Lightbody and O'Sullivan (1999) and Fernandez and Marshall (2003) who developed a victim-specific empathy measure for child molesters and rapists respectively. The current research compared the responses of 40 male violent offenders (28 incarcerated, 8 residing in a community-based treatment programme and 4 living in the community) with 40 male non-offenders using the Violent Offender Empathy Measure (VOEM). The Fernandez et al. (1999) and Fernandez and Marshall (2003) studies were not designed to be able to distinguish whether the much less empathic scores of the offenders toward their own victims were caused by a lack of empathy towards the specific individuals they victimised or towards people the offender has harmed himself (rather than people harmed by others). Therefore, the current study aimed to distinguish between these two possibilities by adding this feature to the specific empathy scale designed for this study, the VOEM. The VOEM assesses empathy in 4 contexts: (1) toward an adult individual who had been permanently disfigured in a car accident caused by another (unknown) driver; (2) toward an adult individual who had been permanently disfigured in a car accident that the offender had hypothetically (but accidentally) caused; (3) toward an adult individual who had been violently assaulted by another, unknown offender; and, (4) the offender's own victim. The participants were also required to complete the Interpersonal Reactivity Index (Davis, 1980) purported to be a measure of general empathic capacity and provide demographic information. Violent offenders were found to be significantly higher than non-offenders on the total score of the IRI; however, when the Fantasy subscale was removed the difference between the two groups on the total of the remaining three subscales was no longer significant. Violent offenders and non-offenders were found to give similar ratings for both recognition of the victim's distress (subscale A of VOEM) and their own affective responses regarding the victim's distress (subscale B of VOEM) on each scenario, with 1 exception. For the own hypothetical accident victim scenario, non-offenders' affective responses were significantly more empathic than those of the offender sample. Results suggest that violent offenders do not differ from non-offenders in their capacity to empathise with others, both generally and in specific circumstances. These results have practical implications in terms of treatment and question the validity of current assessment tools used to measure empathy.