Correctional Service Canada
Symbol of the Government of Canada

Common menu bar links

The Incidence of Family Violence Perpetrated by Federal Offenders: A File Review Study

Warning This Web page has been archived on the Web.


The Family Violence Incidence Study was designed to provide information about the extent of family violence perpetrated by federal offenders. The study was also intended to investigate correlates of family violence to assist in the identification of offenders who are at high risk of perpetrating violence against their family members. As a major research component of the Family Violence Initiative of the Correctional Service of Canada, the study was viewed as a vehicle for developing basic knowledge about the extent and nature of the problem of family violence within our offender population. This knowledge should prove useful for program planners in the area of family violence treatment.

Federal offenders possess many of the characteristics associated with the profile of men who are violent toward their families. Based on the empirical literature, Appleford (1989) enumerated a number of characteristics of abusers including demographic (e.g., male, under thirty, unemployed, low educational and occupational attainment) psychological (e.g., angry, aggressive, immature, cognitively rigid, and suffering from affective and character disorders), attitudinal (e.g., externalizes blame, rigid definitions of male/female roles), and behavioural factors (e.g., substance abuse, threatens suicide and homicide). Many of the elements in the profile are also well-established predictors of criminal behaviour more generally (Andrews, 1989). Apart from the obvious correspondence between the profiles of offenders and family abusers, the history of generally violent behaviour associated with federal offenders would lead one to suspect that this group is at high risk for the commission of violent acts against their families. Tolman and Bennett (1990) have observed that men involved in criminal subcultures may gain normative support for aggression toward family members.

The current study is the first national study to assess the incidence of family violence among federal offenders. An Ontario Region study by Motiuk and Brown (1993) estimated that 26.8% of released offenders had been physically or sexually abused as a child, 13.6% had been perpetrators of spousal abuse, and 7.9% had been in trouble as a result of abusing children. The estimates were based on a sample of 604 federal offenders who were assessed by community case management officers for risk and need levels. In this study, perpetrators of spousal violence were also found to have an above average suspension rate while under community supervision.

Another study by Dutton and Hart (1992; see also Dutton & Hart, 1993) provided an estimate of the incidence of family violence among federal offenders incarcerated in the Pacific region. They found evidence, including official charges and reports of behaviour that did not result in charges, that 29.6% of a sample of 597 federal offenders had been violent toward a family member (sexual assaults, physical assaults, threats). Dutton and Hart also reported file review data showing that 41% of the offenders had been victimized by family members as children, including physical and/or sexual abuse and/or witnessing abuse of others. Their study also identified positive correlations between childhood victimization and later violence against family members. Diagnosis of a mental disorder, particularly personality disorder, also appeared to be a marker of perpetration of family violence.

The current study places particular emphasis on the relationship between childhood victimization and witnessing of abuse by family members and subsequent perpetration of family violence as an adult. The attention given to childhood experiences follows the current interest demonstrated by family violence researchers in the "cycle of family violence" (Widom, 1989). The link between the experience of violence in the family of origin and wife battering is now well-established in the literature (Tolman & Bennett, 1990). Our focus on childhood victimization is also consistent with the social learning and pro-feminist models which guide the offender programming strategy being developed by the Family Violence Initiative in the Correctional Service of Canada (Correctional Service Canada, 1994).

In planning the Family Violence Incidence Study, the limitations of the file review method were recognised from the outset. The current offender file system lacks a structured format for routinely reporting on history of family violent behaviour. Therefore, the quality of information about family violence perpetration maintained on offender files was to a large extent unknown. At the same time, knowledge gains made in the past using file reviews for studying offenders (see for example the CSC Sex Offender Study, Motiuk & Porporino, 1993), suggested that this methodology was a reasonable strategy for our first attempt to study the phenomena. Alternative methods of estimating the scope of the family violence problem involve soliciting information from offenders or from their family members. Such alternatives were viewed as both costly and difficult to implement. In addition, both approaches possess unique problems of reliability and validity (Saunders, 1991). As Dutton (1988) notes, estimates of the incidence of family violence are difficult to establish because the violent behaviour occurs in private. While the exact extent to which family violence is under reported in research is not known, there is consensus among researchers that a significant amount of victimization remains undetected (Smith, 1987). Therefore, all methods of estimating the incidence of family violence for given populations are likely to be limited. Given these limitations, perhaps the most ideal methodology would be a combination of data sources relating to official charges, reports from victims, and reports from perpetrators. In their study of federal offenders, Dutton and Hart (1992) observed higher estimates of the number of perpetrators of abuse against female partners when they included the reports of offenders and their partners.

The results of the Dutton and Hart study suggest that a very significant proportion of federal offenders are at risk of violence against family members. The study also provided initial validity information on the data in the form of correlates of family violence. This suggests that the file review approach can be a valuable source of research data for the study of family violence in this unique population. While there is no method for determining the accuracy of estimates based on the file review data utilized in this study, it is clear that a large number of family violent offenders can be identified using the file review strategy. Because family violence history data are not systematically recorded in files, it is likely that the incident figures reported for the current study, underestimate the number of offenders who have victimized members of their families. While incomplete, the file review method furnished the best available research tactic for beginning to gain a greater understanding of the scope of the family violence problem in our offender population.