FORUM on Corrections Research

Risk Markers for Family Violence in a Federally Incarcerated Population

In an attempt to identify risk markers for family violence, the institutional files of close to 600 male offenders from seven federal correctional facilities in Canada were reviewed. Three groups of offenders were identified: non-violent offenders who had no indication of violent behaviour anywhere in their file; stranger-violent offenders who had histories of assault on file but no indication of violence toward their wives or other family members; and family-violent offenders, the majority of whom had also assaulted non family members.

The study examined the offenders' history of experiencing abuse in their family of origin. Significant differences among the three groups were found. While the non-violent group showed the least amount of abuse victimization and the stranger-violent group showed a moderate amount, members of the family-violent group were most likely to have been abused. When more specific types of abuse (physical, sexual and witnessing abuse) were examined, similar differences were found.

There were also differences among the groups in the types of psychiatric disorders they experienced, with the family-violent group being more likely to have non-psychopathic types of personality disorders (e.g., borderline or narcissistic personalities).

The characteristics of perpetrators of family violence seem to be similar to those of incarcerated populations in a number of ways: individuals in both groups often come from abusive family backgrounds, have experienced traumatic separations, abuse alcohol and drugs and suffer from psychiatric and personality disorders.⁽³⁾

The objective of this study was to assess the prevalence of these associated risk markers for family violence among federal inmates, and to estimate incidence of such violence. This information is important in managing risk with federal offenders because inmates with a history of, or propensity for, family violence could continue with such behaviour after their release, particularly if they return to the same relationships they had before incarceration.

Method

The files of 597 randomly selected male offenders from seven federal correctional facilities were reviewed using a file-based risk assessment coding sheet. This coding sheet recorded file reports of childhood victimization (physical abuse, sexual abuse and witnessing interparental abuse), any history of substance dependency or abuse, employment history and any history of physical or sexual assault of family members and others.

Furthermore, the coding sheet highlighted psychiatric diagnoses and personality disorders, such as antisocial personality and borderline, narcissistic, histrionic or mixed personality disorders. These psychiatric disorders were taken from Axis I Diagnoses and Axis II Disorders of the Diagnostic and Statistical Manual of the Mental Disorders (DSM-IIIR).⁽⁴⁾ Psychiatric disorders were recorded only if an inmate's file contained an explicit diagnosis by an institutional psychiatrist or psychologist.

Results Violence Group Classifications

Reports of violence consisted of:

• criminal charges or convictions for offences against persons (such as assault and sexual assault, threatening, use of weapons, robbery, kidnapping or murder) appearing on an offender's criminal record;
• official allegations of violent behaviour (including physical violence, threatening with weapons and serious threats of physical harm) that resulted in the suspension or revocation of conditional release but did not result in criminal charges or convictions; and
• allegations of violent behaviour (as defined above) that were reported or investigated but not officially confirmed.

The reports of violence were classified into one of six categories:physically assaulting a family member;

• sexually assaulting a family member;
• threatening a family member;
• physically assaulting a non-family member;
• sexually assaulting a non-family member; or
• threatening a non-family member.

A family member was defined as a first-degree relative; a spouse, common-law spouse or live-in girlfriend; or a child or stepchild.

For each category of violence, we recorded the types of victims (e.g., adult male, female child, male and female adolescents, unknown). Finally, we noted the weapons used and the severity of harm done to victims (e.g., none, threats only, mild, moderate, severe or death).

Based on these reports, we divided offenders into three groups. Non-violent (NV) offenders were those whose files contained no reports of violence. Stranger-violent (SV) offenders were those whose files contained reports of violence directed toward non-family members only (although it did include assaults on acquaintances). Finally, family-violent (FV) offenders were those whose files contained reports of violence directed toward family members, regardless of any reports of violence toward non-family members.

Of the total sample, 12.4% (74 offenders) fell into the NV group, 58% (346) were in the SV group and 29.6% (177) fell into the FV group.

Abuse in the Family of Origin

The institutional file review of abuse of the inmate in his family of origin revealed some highly significant results. Three forms of victimization were considered: physical, sexual and other. The other category included witnessing physical or sexual abuse of other family members and being severely neglected or abandoned. It excluded emotional or psychological abuse.

---

Figure 1

Overall, about 4 in 10 (41%) of the inmates in the sample had been seriously abused as children or adolescents, according to reports in their files.

As shown in Figure 1, when the groups were compared, 20.3% of inmates in the NV group had been abused, and 38.7% of those in the SV group experienced abuse. Offenders in the FV group were most likely to have been abused, with more than half (54.6%) having reports of childhood abuse in their files. These differences were statistically significant, as were the differences between the FV and the SV groups when they were compared directly.

When specific forms of abuse in the family of origin were examined, again inmates in the FV group were most likely to have been victimized. According to file reports, 41.4% of FV offenders were physically abused, compared with 29.9% of offenders in the SV group and 14.9% of those in the NV group. Of the entire sample, close to one third (31.4%) had been abused physically.

In the FV group, 17.5% of inmates had suffered sexual abuse, compared with 9.8% of SV offenders and 5.4% of NV offenders.

Finally, about 20% of FV offenders had witnessed abuse in their family of origin. This compares with 11% of offenders in the SV group and 5.4% of those in the NV group. All the above findings were statistically significant.

In sum, according to the institutional files, inmates in the FV group were more likely to have reported being abused in their family of origin, regardless of the specific nature of the abuse.

Psychiatric Disorders

Of the entire sample, just over one third (34.4%) showed indications of personality disorders. FV offenders had an incidence rate of 43.5%, compared with 34.1% for SV offenders and 13% for NV offenders.

As can be seen in Figure 2, the types of personality disorders differed significantly within the violence groups. Although SV offenders were as likely to have been diagnosed as antisocial as FV offenders (20.7% and 21.5% respectively), FV offenders were more likely to have had other types of personality disorders (mostly borderline, narcissistic, mixed and other) -22% of FV offenders compared with 13.4% of SV offenders.

---

Figure 2


Discussion

This study indicates, albeit retrospectively, that family-violent offenders are more likely to have been abused during childhood than stranger-violent offenders and, in particular, non-violent offenders.

We also found a high prevalence rate for personality disorders. Although one might expect antisocial personality to be the most common disorder across all groups, the finding that narcissism and borderline personalities are overrepresented, especially among the family-violent group, is noteworthy. One of the diagnostic criteria of borderline personality is having "physical fights," and narcissism is characterized by reacting to criticism with rage, shame and humiliation. Clearly, either reaction would increase the probability of inclusion in a violence group.

What these family-violent offenders seem to require is a transitional form of anger management treatment with a special emphasis on marital and family relations. Such treatment would ideally use the group treatment model developed by Ganley⁽⁵⁾ and others. Such models have been described by Dutton.⁽⁶⁾ The anger management and personal responsibility aspects of this treatment are useful with incarcerated populations, though they have to be modified to encompass anger that develops in a prison setting. However, the high incidence of personality disorders among family-violent offenders bodes poorly for true treatment success.

Future research notwithstanding, the high rate of abuse of federal inmates during childhood and adolescence is of interest. Many profiles of abuse victims have been based on female adults who underwent psychotherapy.⁽⁷⁾ This may give the impression that males either are not abused or do not suffer from long-term consequences of that abuse. The present study indicates that these interpretations are erroneous. Males abused as children are at a higher risk for violent crime.

---

⁽¹⁾Donald. G. Dutton, Department of Psychology, University of British Columbia, 2136 West Mall, Vancouver, British Columbia V6T 124.
⁽²⁾Stephen D. Hart, Mental Health Law and Policy Institute, Simon Fraser University, Burnaby, British Columbia V5A 1S6.
⁽³⁾D.G. Dutton, "Behavioral and Affective Correlates of Borderline Personality Organization in Wife Assaulters," International Journal of Law and Psychiatry, in press. See also K.L. Hamberger and J.E. Hastings, "Personality Characteristics of Spouse Abusers: A Controlled Comparison," Violence and Victims, 3, 1 (1988): 31-48. And see S.£D. Hart, D.G. Dutton and T. Newlove, "The Prevalence of Personality Disorder Among Wife Assaulters," Journal of Personality Disorders, in press.
⁽⁴⁾Diagnostic and Statistical Manual of the Mental Disorders, 3rd ed., rev. (Washington, D.C.: American Psychiatric Association, 1987).
⁽⁵⁾A. Ganley, Participant's Manual: Court-Mandated Therapy for Men Who Batter: A Three Day Workshop for Professionals (Washington, D.C.: Centre for Women Policy Studies, 1981).
⁽⁶⁾D.G. Dutton, The Domestic Assault of Women: Psychological and Criminal Justice Perspectives (Boston: Allyn and Bacon, 1988).
⁽⁷⁾J. Briere, "The Long-Term Clinical Correlates of Childhood Sexual Victimization," Annals of the New York Academy of Sciences, 528(1987): 327-334. See also J.B. Bryer, B.A. Nelson, J.B. Miller and PA. Krol, "Childhood Sexual and Physical Abuse as Factors in Adult Psychiatric Illness," American Journal of Psychiatry, 144, 11(1987): 1426-1430.