FORUM on Corrections Research

An inmate survey: A profile of violent and non-violent offenders

The Correctional Service of Canada conducted its first National Inmate Survey in the fall of 1995.² Nearly 4,300 inmates participated in the survey, providing information about their criminal history, prison experiences and perceptions of a variety of correctional issues. This comprehensive survey included questions concerning the institutional environment, safety of inmates, program experiences, views on staff and health issues, including mental health and risky behaviours associated with HIV. The survey produced a rich research database for exploring a variety of correctional issues which have never been empirically examined by the Correctional Service of Canada.

Inmates supplied information about their current federal incarcerations including both offence and sentence characteristics. The data provided a unique opportunity to compare offenders who had committed different offences on a number of personal characteristics and attitudinal variables. The data were particularly suited to compare violent and non-violent offenders.

Methodology

The survey was based on a random sample of 4,285 inmates who were administered questionnaires, in groups, in their respective institutions. The sample was designed to provide statistical confidence in the results for each institution within the Correctional Service of Canada. About two thirds of inmates who were randomly selected (65.4%) agreed to participate in the survey. Random techniques were also used to select replacements for inmates who refused to participate in the survey. Overall, the sample represented 97% of the quota established for the survey.

The random procedures, the administration of the questionnaires and the data entry of the results of the questionnaires were conducted by an external consulting firm to assure inmates that the information they were providing would be anonymous.

Distribution of offences

A total of 3,972 inmates (93%) provided information about the offences related to their current federal prison sentences. Table 1 shows the distribution of offence types for the sample. In total, 41.8% reported offences from more than one of the offence categories included in the questionnaire. Four out of five inmates (80.6%) reported at least one violent offence including homicide, attempted homicide, manslaughter, assault, weapons offences, robbery and sex offences. This figure is comparable to the violent offender rate obtained using the Offender Management System, which categorizes the types of offences committed by federal offenders.

Robbery offences were reported by 39% of the offenders who fell within the violent offence category, while 27.5% reported homicide, manslaughter or attempted homicide, 23.8% reported weapons offences, 23.5% reported assaults and 22.8% reported sex offences.

Table 1

Distribution of Current Offences
Current Offence Type (n=3,972)	Number	%
All violent (including all homicide/ manslaughter/attempts, assaults, weapons robbery, sex offences)	3,202	80.6
Non-violent offences only (drug, break and enter, fraud/other)	770	20.4
Any non-violent offences only (drug, break and enter, fraud/other)	1,616	40.7
Selected violent offences (homicides, assaults, weapons only - excluding robbery and sex offences)	2,050	51.6
Any robbery	1,248	31.4
Any sex offences	730	18.3
Any homicide/manslaughter/ attempted homicides	1,092	27.5
Any weapons offences	763	19.2
Any assaults	753	18.9
Any drug offences	887	22.3
Any break and enter	976	24.5
Any fraud/other	506	12.7

Among the violent offenders, about one quarter reported at least one non-violent offence (26.4%), including break and enter (18.4%), drug (13.6%) and fraud/other offences (8.4%). In total, two fifths of the entire sample (40.7%) reported at least one non-violent offence on their current sentence (break and enter, drugs, fraud/other). Of the non-violent offences, break and enter was the most frequently reported offence (24.5%), followed by drug offences (22.3%).

Demographic and criminal history variables

Violent offenders differed from non-violent offenders on most of the characteristics examined in the survey. Table 2 compares the demographic and criminal history variables of non-violent and violent offenders. Overall, violent offenders tended to be slightly but (statistically) significantly older, less likely to be married and more likely to be Aboriginal peoples. They were also more likely to have had more than one offence, less likely to have been returned to federal custody on their current sentence and less likely to have served previous federal or provincial incarcerations. Moreover, violent offenders were more likely to have been placed in administrative and disciplinary segregation and more likely to have used illegal drugs inside prison; they were less likely to be past their parole eligibility date and were serving considerably longer sentences.

As Table 2 shows, educational achievement and having children were the only variables on which the two groups were not significantly different. It should be noted, however, that some of the statistically significant differences were not large.

Table 2

Demographic and Criminal History Characterisitics Factors by Offence Type (Means/%)
Characteristics	Any Violent Offences (n=3,202)	Non-violent Offences Only (n=770)
Mean age **	35.2	34.2
Less than grade 8 education	19.1	22.0
Married***	32.5	42.5
Have childern	59.4	60.9
Aboriginal***	16.5	8.9
More than one offence***	43.6	34.1
Return to custody***	21.9	30.8
Previous provincial incarceration***	66.7	73.2
Previous federal incarceration***	49.3	56.3
Administrative segregation***	21.7	13.4
Disciolinary segregation***	25.8	16.3
Used drugs inside prison***	40.2	33.9
Past parole eligibility date***	53.8	66.1
Length of sentence among non-lifters	7.5	5.9
Notes:* age is expressed as mean years; all other figures in Table 2 are pecentages: **p<0.05, ***p<0.001

There was variation in the characteristics of offenders across the subtypes of violent offences. For example, sex offenders tended to be different from other violent offenders on most factors: they were older and more likely to be married and have children. They were also less likely to have been segregated for administrative or disciplinary reasons and less likely to have reported using drugs. Sex offenders were also less likely to have been returned to custody or to have earlier provincial or federal incarceration histories, but more likely to be past their parole eligibility dates.

Robbery offenders, in contrast to sex offenders and other violent offenders (assault, weapons and offences involving the death of a victim), were somewhat younger, more likely to have spent time in segregation, reported higher levels of illegal drug use inside prison, were more often recidivists with previous federal and provincial histories, and had more than one offence type for their current terms.

Staff and programming

Inmates reported their relationships with staff, enrollment in work and education programs, and their views about the effectiveness of programs. Violent and non-violent offenders exhibited few differences on these attitudinal variables (see Figure 1). Both groups reported similar relationship quality with staff, perceptions about the positive aspects of case management and satisfaction with work and programming. Violent offenders, however, showed higher motivation for programs and greater anticrime orientations than non-violent offenders.

Figure 1

The presence of sex offenders in the violent offence category accounted for most of the differences between violent and non-violent offenders for program-related variables. Sex offenders showed the most motivation for programs and were more likely to endorse anticriminal orientations than other inmates. Robbery offenders, on the other hand, had the least motivation for programs and fewer anticriminal attitudes.

Sex offenders also showed higher levels of satisfaction with the quality of their relationships with staff ­ 45% were positive versus 27% of all violent offenders. Robbery offenders reported the lowest levels of satisfaction (18%) with their relationships with staff.

Victimization

The inmate survey contained a large section devoted to information about which offenders were victimized while incarcerated. As indicated in Figure 2, violent offenders reported a significantly greater number of assaults committed against them compared to non-violent offenders. Again, there were interesting differences within the violent offence group. Sex offenders reported the highest level of victimization, reporting, for example, a rate of sexual assault (5.5%) that was nearly double that reported by other violent offenders (2.9%). Interestingly, within the non-violent group, break and enter offenders also reported high rates of physical assaults (25.5%), a rate comparable to that observed for sex offenders (24.9%). In most other comparisons of survey results, break and enter and robbery offenders had similar patterns; however, in this, the victimization category, robbery offenders reported generally lower rates.

Figure 2

Risk of HIV/AIDS

A large section of the survey was devoted to collecting information about inmates' perceptions about HIV/AIDS and the behaviours that place them at risk of contracting the infection. A behavioural risk index used in the survey included injection drug use, unsafe sexual behaviour and the use of unclean tattooing or body piercing equipment. Overall, 26% of inmates engaged in one or more of these risk behaviours. There were no significant differences, however, between violent (26.7%) and non-violent offenders (23.6%). Injection drug use was the most frequently occurring risk behaviour reported by the sample (11%). Violent and non-violent offenders differed on this important variable, with violent offenders reporting the highest levels (11.8%) of injection drug use compared to non-violent offenders (7.9%). Consistent with the patterns reported above, robbery offenders reported the highest levels of injection drug use (18.2%), while sex offenders (4.1%) reported the lowest levels.

Mental health

A number of mental health measures were included in the survey. One measure attempted to assess the level of stress that inmates experience as a result of being incarcerated (e.g., relationships with staff, relationships with other inmates, the inside drug trade, obtaining early release, relationships with family, access to goods and services, prison violence, etc.). About 38% of the sample scored within the "high stress" range of the scale. There were no statistically significant differences by offence type.

Violent offenders reported significantly higher levels (29.1%) of depression than non-violent offenders (22.3%). In keeping with different patterns described above, a higher proportion of sex offenders (32%) reported depression. There was also an indication that violent offenders (45.6%) were more satisfied than non-violent offenders (36.8%) with the mental health services they received while incarcerated. Sex offenders who received mental health services were more satisfied than other groups (49.6%).

Criminal risk

While criminal risk is normally assessed using interview and file review sources, a number of standard actuarial risk indicators were available as self-reported items in the inmate survey (age, return to custody, previous provincial and federal incarcerations). These static items, which normally predict recidivism, were combined to form a "mini" self-reporting risk scale. Three additional items which we believed would be correlated with recidivism were also included: time spent in disciplinary segregation, other disciplinary charges and drug use while in prison. After constructing the risk index, we divided the sample into three levels which were roughly equal in size: low, medium and high risk. Figure 3 shows the proportion of offenders from each offence type who were placed in the high risk category using the available self-reporting indicators.

Figure 3

There were no differences in the proportions classified as high risk when the violent (36%) and non-violent (38%) offence categories were inspected. Sex offenders and offenders who committed offences involving the death of a victim had the lowest proportions falling within the high risk category. Violent offenders who had used weapons or were serving time for assaults, had substantially higher proportions of high risk cases. The robbery (56%) and break and enter (56%) groups had the largest proportions of high risk offenders. Generally, offenders who had committed non-violent offences (break and enter, drug, fraud/other) exhibited higher than average risk using the self-reporting scale.

Heterogeneity within the violent offence category

In terms of public perceptions, many people attribute the same characteristics to all offenders who commit violent offences, regardless of what type of violent offence was committed. This is particularly the case when the notion of criminal risk is discussed. However, the inmate survey results corroborate findings from other research which suggest that there is a high degree of heterogeneity in the characteristics of this large offender group. Sex and robbery offenders, two major offence groups within the violent category, were dissimilar on most of the measures examined in this study. In addition, the generally violent group (assaults, weapons, death of victims) also differed from both the sex and robbery offender groups on many of the survey indicators.

Generally, when robbery offenders were compared with sex offenders, the latter group were more motivated for programming, had more positive relationships with staff, reported higher levels of prison victimization and exhibited higher levels of depression. Most notably, robbery offenders appeared to be at dramatically higher levels of criminal risk. Among the remaining violent offender types, homicide offenders tended to be at lower risk, while those who had committed assaults or weapons offences were more likely to score in the higher risk range.

An important finding concerns the similarity between robbery and break and enter offenders. Generally, these groups displayed the most troublesome behaviour inside prison and had exhibited more extensive criminal histories before they arrived. In short, these two groups of offenders had the most elevated criminal risk profiles. These data confirm earlier findings related to recidivism and responses to correctional programming among property offenders ­ both robbery (violent) and break and enter (non-violent) offenders.³

The simplest conclusion from these inmate survey analyses is that not all violent offenders are the same. Those who generate the greatest fear in members of the public (sex offenders and those who kill their victims) actually show the most evidence of co-operation, treatment motivation and a lower risk of recidivism.

The violent offence category, typically used to define correctional populations, is a very inclusive grouping encompassing more than four out of five federal offenders. Because of the differences observed in offenders across the various violent offence categories, the current findings argue for more detailed breakdowns of offence types when describing federal offender populations. At a minimum, the differences observed across offence types on some of the major inmate survey indicators suggest that researchers should avoid using the violence/non-violence dichotomy as the primary division of offence type. Use of such an offence dichotomy is likely to mask both differences and similarities in the characteristics of different types of offenders.

---

1. 340 Laurier Avenue West, Ottawa, Ontario, K1A 0P9.

2. Correctional Service of Canada, 1995 National Inmate Survey: Final Report, Special Report No. 2 (Ottawa: Research Division, Correctional Service of Canada, 1996). See also D. Robinson and L. Mirabelli, Summary of Findings of the 1995 CSC National Inmate Survey, Report No. B-14 (Ottawa: Research Division, Correctional Service of Canada, 1996).

3. D. Robinson, The Impact of Cognitive Skills Training on Postrelease Recidivism Among Canadian Federal Offenders, Research Report R-41 (Ottawa: Research Division, Correctional Service of Canada, 1995).