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Warning This Web page has been archived on the Web.

Results

Violence Perpetrated Against Family Members

There was file review evidence to indicate that slightly more than one third (33.7%) of the offenders in the sample had perpetrated acts of violence against their family members. This includes acts of physical and sexual assault, as well as psychological abuse. Appendix B provides definitions for each of the types of violence we studied and includes examples of acts which constitute family violence. We included female partners, children, step-children, parents, step-parents, siblings, step-siblings, extended family members, and in-laws as family members.

Table 2 displays the weighted percentages of occurrences for the three types of violence included in the study. Physical assault was the most frequent type of violence recorded (26.9%). Sexual assault (10.7%) and psychological abuse (5.2%) were recorded less frequently. There was overlap in the occurrence of abuse types. Of offenders who had psychologically victimized one or more family members, 56.2% had also physically victimized family members. Among those who sexually abused family members, 50.5% also physically abused one or more family members.

Our results are comparable to the earlier results obtained by Dutton and Hart (1992) in their study of offenders from the Pacific region. We found that 32.1% of offender files contained references to physical or sexual assaults on family members compared to 29.4% in the earlier study. However, as presented below, our figures for the Pacific region are almost identical to the Dutton and Hart estimations.

Table 2

Perpetrators of Family Violence (N=935)

Abuse Types
%
n
Sexual Abuse
10.6
99
Physical Abuse
26.9
252
Psychological Abuse
5.2
49
Sexual and/or Physical Abuse
32.1
300
Any Family Violence
33.7
315

Table 3 provides data on the features of family violence reported in the files for acts of physical or sexual abuse against family members. Of the files which contained some evidence of physical or sexual assault of family members, 80.2% of the assaults reported involved official charges and nearly half (47.8%) of the acts resulted in injury (i.e., required medical attention). Moreover, in over 50% of the files (56.3%) which documented physical or sexual assaults against family members, more than one family member had been victimized.

Table 3

Features of Family Violence (n=300)

Physical/Sexual Abuse
%
n
With Official Charges
80.2
241
Causing Injury to Victim
47.8
143
With Two or More Victims
56.3
169

Given the extremely high rate of incidents resulting in official charges it is likely that our estimate of one third falls short of the actual number of offenders who have been violent toward family members. Our assumption is supported by research suggesting that only a small proportion of family violent acts lead to official charges. According to recent Statistics Canada figures from the Violence Against Women Survey, only 26% of wife assaults were reported to police and only 28% of these incidents resulted in official charges (Statistics Canada, 1993). In our file data, however, an opposite trend emerged. The number of violent occurrences that did not result in charges is much lower than the number of occurrences involving charges. Therefore, in all likelihood the current file review method underestimates the level of family violence because it fails to detect a sizeable proportion of men who committed assaults but were never charged.

There is clear evidence that females were the most likely targets of family violence perpetrated by this sample of male offenders. This information is displayed in Table 4. In 91.6.% of the cases involving physical violence toward family members, the victim type was female. The victim was female in 96.8% of cases involving reports of sexual abuse. About one quarter of the files involving physical violence reported male family victims (24.2%) and about one fifth of files involving sexual assault reported male family victims (19.2%).

Table 4

Gender of Victims of Family Violence

Physical (N=252)
%
n
Female
91.6
230
Male
24.2
61
Sexual
%
n
Female
96.8
96
Male
19.2
19

Abuse Perpetrated Against Female Partners

Abuse of a female partner was the most frequently occurring incident of family violence recorded in the files for this sample of federal admissions. In one quarter (24.3%) of the files we reviewed, there was evidence of violence against a female partner. As Table 5 shows, this includes physical, sexual and psychological abuse of a female partner.

Acts of physical violence against female partners were by far the most frequently reported types of abuse. Physical abuse occurred in 92.8% of the files that possessed some evidence of abuse against a female partner. Sexual abuse against a female partner was reported in 9.3% of the cases involving abuse against a female partner. Psychological abuse was also infrequently reported. Only 4.2% of the files contained any references to the psychological abuse of a female partner. This type of abuse occurred in 17.3% of the cases where some form of abuse had been reported. Our estimate of 4.2% of cases of psychological abuse appears low given the wife assault literature which suggests that psychological abuse is a serious problem which frequently occurs either alone or in conjunction with physical violence (MacLeod, 1987). Therefore, we assume that the file review methodology is not a good source of data for estimating the occurrence of this type of abuse.

Our estimate of abuse against female partners increased when we examined only the files of offenders who were involved in a marital relationship at least once in their lives (i.e. marriage or common-law relationship) and excluded offenders who reported no relationships or dating relationships only. There was evidence of abuse of a female partner in 29.0% of offenders who had been involved in at least one marital relationship during their lifetimes (See Table 5).

We also examined the level of abuse against female partners for offenders who reported current involvement in a marital relationship (i.e., while incarcerated). This analysis was meant to provide information relevant to risk of family violence among offenders who have some form of contact with their female partners. About 22.9% of these offenders had files containing evidence of previous abuse against a female partner.

Table 5

Perpetrators of Abuse Against Female Partners

Abuse Types
All Offenders (N=935)
Offenders Involved in Martial Unions
Offenders Involved in Dating Only (N=153)
%
n
%
n
%
n
Sexual Abuse
3
24
3.1
22
1.2
2
Physical Abuse
22.4
209
26.7
193
7.4
11
Psychological Abuse
4.2
39
5.3
39
0.48
1
Sexual or physical Abuse
23.1
216
27.6
199
7.4
11
Any Family Violence
24.3
225
29
209
7.9
12

There was also abuse of female partners (e.g., girl-friend) among those offenders who had never been involved in a marital relationship. Therate of violence against female partners was 7.9% for men who reported only dating relationships. This suggests that some men who have never been involved in the types of marital relationships described above may be at risk of committing violent acts in their dating or casual relationships. In the remaining analyses for this report, we include only men who had been involved in marital or common-law relationships when presenting information on abuse of female partners.

The number of men with official charges for abuse against their female partners was high in this sample. Overall, 16.2% of all files that were reviewed contained information relating to official charges for abuse against a female partner. Moreover, of those men who had been involved in marital relationships, one in five (21%) had an official charge for an assault against their female partners.

In terms of the relative proportion of unofficial and official reports of violence, Table 6 shows that almost three-quarters of the files with records of abuse against female partners involved official charges and approximately half of these cases involved injury to a female partner. Again, we must assume that our file review data underestimates the occurrence of family violence against female partners, at least for violence not resulting in official charges.

Given the number of offenders in this sample who had multiple marital relationships, it was not surprising that some of the files recorded abuse against more than one female partner. Physical or sexual abuse of more than one partner was evident in roughly half of the cases in which there had been evidence of abuse against a female partner.

Table 6

Features of Abuse Against Female Partners

Physical/Sexual Abuse
Offenders Involved in Martial Unions
Offenders Involved in Dating Only (N=153)
%
n
%
n
With Official Charges
76.4
152
81.8
9
Causing Injury to Victim
48.7
97
54.5
6
With Two or More Victims
47.7
95
9.1
1

The files also contained details about the type of violence used by the men who physically assaulted their female partners. Information on type of violence was available for 81.3% of the cases of abuse against female partners. Table 7 displays the frequencies for types of violence. The tabulations indicate that "hitting" was the most frequently reported type of violence used (65.3%), followed by pushing (32.4%) and threats (31.2%). Shooting and stabbing, although occurring at relatively low frequencies (4.1% and 7.1% respectively) were reported for this sample. Given the serious nature of these forms of violence, however, the occurrence rates cannot be regarded as trivial. Mutilation (1.2%) was the most infrequently reported type of violence.

Table 7

Types of Abuse Used against Female Partners (n=170)

Abuse Type
%
n
Hitting
65.3
111
Pushing
32.4
55
Threats
31.2
53
Kicking
12.9
22
Choking
10.0
17
Stabbing
7.1
12
Fighting
4.7
8
Beating
4.1
7
Shooting
4.1
7
Burning
1.8
3
Mutilation
1.2
2

Abuse Perpetrated Against Children

For the purpose of this report "child victims" refer to biological and step-children as well as other child relatives (e.g., nieces and nephews) who were 18 years or younger. Approximately 8% of the files contained information to suggest that abuse of children within the family had occurred.

About 60% of the offenders in the sample reported having children or stepchildren of their own. Table 8 presents the rates of abuse of children for this subset of offenders. According to the file information, 13.3% of these offenders had abused their children. The latter figure includes all forms of abuse including physical, sexual, and psychological abuse. The most frequent type of abuse was sexual abuse (11%), with reports of physical abuse occurring much less frequently (3.1%). Further analysis indicated that the child victims of incest evident in the files were considerably more likely to be female than male. However, males were more likely than females to be victims of the physical abuse and psychological abuse perpetrated by the offenders.

There was evidence of official charges in close to 90% (87.3%) of the files containing reports of child physical or sexual abuse. This information is tabled in Table 9. Again, the high rate of official charges suggests that our figure underestimates the true incidence of child abuse perpetrated by these offenders.

Table 8

Perpetrators of Abuse Against Children Among Offenders with Children (n=553)

Abuse Type
%
n
Sexual Abuse
11.0
63
Male Victims
15.9
10
Female Victims
95.2
60
Physical Abuse
3.1
17
Male Victims
64.7
11
Female Victims
52.9
9
Psychological Abuse
1.3
8
Male Victims
75
6
Female Victims
50
4
Sexual and/or Physical Abuse
12.6
70
Any Family Violence
13.3
74

Table 9

Features of Abuse Against Children (n=71)*

Physical/Sexual Abuse
%
n
With Official Changes
87.3
62
Causing Injury to Victim
36.6
26
With Two or More Victims
14.1
10

The number of occurrences of violence against children increases when we extend our definition of child abuse to take into account child victims who were known to the offenders but were not family members. This group corresponds more closely to the definition of family members employed by the federal family violence initiative (Government of Canada). According to the definition, family members include all individuals with whom relationships of trust or dependency are shared. Twenty-one percent of the files involved abuse of children when the definition included both children within the family and other children who were known to the offenders.

Regional Variations in Perpetration of Abuse

Our analysis of the file review data indicated statistically significant (p<.001) regional differences in the distribution of violence perpetrated against family members. Regional differences in family violence are reported in Figure 2. The reports of perpetration of family violence were highest in the Atlantic and Prairie regions and lowest in the Quebec and Pacific regions. Generally, these regional trends were evident for each of the types of violence that were recorded in the files.

In the Atlantic region, there was evidence that 40% of all offenders who had ever been involved in a marital relationship (marriage or common law relationships) had been abusive toward one or more female partners. This number is quite dramatic when it is considered that the majority of these occurrences involved official charges and that the true rate of perpetration in this sample is being underestimated. The incidence rate was almost as high for this category of violence in the Prairie region.

Figure 2
Rates of Perpetration of Family Violence by Region

The regional variations in rates of family violence are not readily explainable. The recent Violence Against Women Survey conducted by Statistics Canada (1993) found moderate regional variation in rates of violence against women. The survey figures refer to all violence against women over the age of 18 years including violence against partners and other violence. Therefore, the figures are not directly comparable to our file review figures. However, the general population survey is one possible source for exploring the question of regional differences. In the national survey, British Columbia and Ontario had higher rates of violence against women relative to most other provinces, a finding that contradicts the current data. However, the province of Alberta, which is included in our Prairie region, exhibited the second highest provincial rate of violence against women. In addition, P.E.I. and Nova Scotia ranked fourth and fifth highest in rates of violence against women.

In the absence of other compelling evidence that contradicts the current data, it appears that there are regional differences in rates of perpetration of family violence in our offender population. The discrepancy between the general population survey figures for British Columbia and Ontario and file review data may be a function of the differing definitions of violence employed in the two studies. Alternatively, our findings may reflect true differences due to regional differences in the characteristics of offenders.

We also explored the possibility that regional differences may have been caused by variances in the comprehensives or completeness of file information across regions. For example, we hypothesized that regions reporting higher proportions of incidents not involving official charges might maintain more complete or comprehensive file information on family violence. In other words, regional diversity in rates of perpetration of family violence may be an artefact of file quality rather than reflecting actual differences in trends. We tested this hypothesis using the information displayed in Figure 3. The data provide no evidence that this phenomena accounted for the differences occurring in this sample. In fact, the Prairie and Atlantic regions, which exhibit the highest rates of family violence, have the lowest proportion of unofficial charges relative to occurrences resulting in charges.

Figure 3
Proportions of Physical and Sexual Assaults Reported as Occurrences Not Resulting in Official Charges by Region (n = 304)

Aboriginal Versus non-Aboriginal Status

The rates of perpetration of family violence among Aboriginal offenders were almost two-thirds higher than the rate of perpetration reported for non-Aboriginals. The differences observed between the two groups of offenders were statistically significant (p<.001). These comparisons are shown in Figure 4. In particular, the rate of physical violence against family members among Aboriginals was double the rate among non-Aboriginals. The dramatic difference is mainly attributable to higher rates of perpetration against female partners in the Aboriginal sub-sample. We found no statistically significant differences in rates of violence against children when we compared the two groups of offenders. Although the current report was not intended to identify the factors accounting for the high rate of family violence among Aboriginals, our victimization data (to be elaborated below), show that Aboriginal offenders also suffered a higher rate of childhood victimization by family members than non-Aboriginal offenders.

Figure 4
Rates of Perpetration of Family Violence by Aboriginal and Non-Aboriginal Status

The high rate of perpetration of family violence evident in the files of Aboriginal offenders underscores the importance of family violenceprogramming for this group of offenders. The implication of our findings is that at least 50% of Aboriginal offenders require programming oriented toward addressing needs in the area of family violence. The Family Violence Initiative of the Correctional Service of Canada has recognised the importance of specialised programming to meet the unique needs of Aboriginals. The current data support the continuation of this program development strategy.

As mentioned above, regional differences in rates of perpetration reported above may be partially due to differences in the regional compositions of Aboriginal and non-Aboriginal offenders. The Prairie region, which evidenced the highest rates of family violence, also has the highest concentration of Aboriginal offenders.*

* The proportion of Aboriginal offenders by regions for this sample is as follows: Atlantic, 5.5%; Quebec, 1.9%; Ontario, 4.6%; Prairies, 40.8%; Pacific, 20.1%.

Family Violence Treatment

The level of participation in programs designed to address problems of family violence was difficult to establish for this sample of offenders. Only program participation in Correctional Service of Canada programs were recorded. Hence, the history of enrolments in community programs not affiliated with Correctional Service Canada is not included in our figures. Another limitation is that the program records for the files of first-time federal admissions may have been less complete at the time we conducted our file reviews. Given our sampling procedure, some family violent offenders who would receive programming at a later point in their sentence would not be recorded as program participants at the time the files were reviewed. Moreover, since this study was conducted, many offenders have been exposed to new CSC-sponsored family violence programs through the six treatment demonstration projects of the Family Violence Initiative.

Under the category of "family violence" programming we included the core program Living Without Violence along with other less standardized programs geared specifically toward offenders who had committed acts of family violence** . Some programs that focused partially or wholly on family violence issues may have been omitted from our analyses because of our lack of familiarity with program content. Sex offender programming, which may or may not have been used to address incest offending for some men in our sample, is also excluded from our tabulations.

** Parenting Skills Training was not included as a family violence program.

In total, 4.6% (43) of offenders in the sample had received some form of family violence related treatment. Of the offenders who had been identified as having been perpetrators of family violence, 9.2% had received treatment specifically geared toward addressing their problems. Clearly, the number of enrolments in treatment programs falls considerably short of the number of offenders who have been identified as perpetrators of family violence. In terms of program completion, 74.4% of all men who had taken programs had either completed them or their participation was currently in progress. In interpreting the program participation data we must caution that the data do not allow a determination of whether or not family violent behaviour terminated after participation in programming.