Skip to content | Skip to institutional links

Common menu bar links

Institutional links

Research Reports

Warning This Web page has been archived on the Web.

Factors Related to Unlawful Walkaways from Minimum Security Institutions

1992, No. R-23

by Joseph C. Johnston Laurence L. Motiuk

Research and Statistics Branch
Correctional Service Canada

February, 1992

ACKNOWLEDGEMENTS
EXECUTIVE SUMMARY
I. INTRODUCTION
- Custody Classification
II. PRESENT STUDY
III. METHOD
- Sample Selection
- Data Gathering Process
- 1. Retrieval from Official Automated Records
- 2. Face-to-face Interviews
- 3. Case File Reviews
- 4. Classification Systems
IV. FINDINGS
- CHARACTERISTICS OF WALKAWAYS
- CHARACTERISTICS OF SAMPLE
- CONVICTION HISTORY OF SAMPLE
- SECURITY INCIDENT HISTORY OF SAMPLE
- OFFENDER PERCEPTIONS REGARDING MINIMUM SECURITY PLACEMENT
- OFFENDER PERCEPTIONS REGARDING CIRCUMSTANCES SURROUNDING WALKAWAY
- CASE FILE REVIEW INFORMATION
- OFFENDER CLASSIFICATION SYSTEMS
- Custody Rating Scale
- Statistical Information on Recidivism Scale
- Case Management Strategies
- NARRATIVE ACCOUNTS OF WALKAWAYS
V. DISCUSSION
VI. REFERENCES

LIST OF TABLES

Table 1. Distribution of Walkaways from Minimum Security Institutions
Table 2. Temporal Distribution of Walkaways
Table 3. Days Spent in Custody by Walkaways
Table 4. Percentage Distribution of Offender Characteristics
Table 5. Percentage Distribution of Previous Convictions by Volume
Table 6. Percentage Distribution of Previous Convictions by Type
Table 7. Percentage Distribution of Security Incidents by Type
Table 8. Percentage Distribution of Security Incidents by Volume
Table 9. Percentage Distribution of Responses Relating to the Minimum Security Facility
Table 10. Percentage Distribution of Responses Related to the Walkaway
Table 11. Percentage Distribution of Information from Case File Reviews
Table 12. Percentage Distribution of Custody Rating Scale (CRS) Scores
Table 13. Percentage Distribution of Statistical Information on Recidivism(SIR) Scores

ACKNOWLEDGEMENTS

We would like to express our sincere thanks to all who assisted in the execution of this study. Here at National Headquarters, Robert Dandurand and Rob Sturrock of the Custody and Control Division provided considerable support, especially in regards to the timely supply of security-related data. Ray Belcourt, SIS Head at the Research and Statistics Branch was also helpful in providing data for this study, and deserves considerable thanks. Irving Kulik and Gil Rhodes of the Audit and Investigations Sector provided support from beginning of the project. We would also like to extend our appreciation for the efforts Fred Luciani and those in the Ontario region, namely, the Wardens and case management staff.

EXECUTIVE SUMMARY

This report is the first in a series on the topic of 'walkaways' and provides a descriptive profile of offenders who unlawfully departed from minimum security institutions. A second report is an extension of the initial study and compares the 'walkaway' sample to a matched sample of similarly situated offenders who did not walk away. The final report is a follow-up investigation and provides outcome information (i.e., recidivism) on the 'walkaway' sample.

During the spring of 1990, there was an unusual increase in the number of federally sentenced offenders unlawfully walking away from minimum security institutions in the Ontario region. The Audit and Investigations Sector of the Correctional Service of Canada (CSC) was asked to examine this situation and identify more precisely any factors which could possibly account for those departures. In order to facilitate this endeavour, the Research and Statistics Branch was approached, who in collaboration with the (formerly) Custody and Control Division and the Ontario region, formed a working group to investigate this matter. It was the goal of this group to design and conduct a study that would identify those factors related to, and predictive of 'walkaways'.

The present investigation differed from previous studies on escape or 'walkaway' phenomena in several ways. First, it focused specifically on unlawful departures from minimum security facilities in Canadian federal corrections. Secondly, and perhaps more importantly, it examined both "static" (e.g., criminal history) and "dynamic" or situational factors (e.g., personal crises) that were involved in the 'walkaways'. By generating new information on this issue, it was thought that we could advance our understanding of 'walkaways', and thereby make a contribution towards reducing the likelihood of such events.

In order to conduct the study, data was collected on all offenders who had walked away from federal minimum security institutions in the Ontario region (Bath, Beaver Creek, Frontenac, Pittsburg) over a 15 month period (January 1, 1990 - April 15, 1991). In total, there were 70 offenders who had unlawfully departed during this period.

Data was gathered on the 'walkaway' sample from a variety of sources. These included: automated information systems (i.e., Offender Information System, Security Incident System, Canadian Police Information Centre System), face-to-face interviews with 38 (54.3%) former 'walkaways' who were available and volunteered to participate, reviews of case file documentation and several classification instruments (i.e., Custody Rating Scale [CRS], Statistical Information on Recidivism Scale [SIR], Case Management Strategies [CMS].

The results of the study revealed that unlawful departures from minimum security facilities were most likely to occur at institutions with higher admission rates, during the Spring and Summer months, and within several months of minimum security placement. Moreover, inmates who walked away were more often under 30 years of age (75.4%), Caucasian (97.1%), single (58.6%), having a major offence that was property-related (58.6%) and serving sentences of under four years (71.5%).

Of special interest was the conviction and security incident history of the 'walkaway' sample. As many as 65.4% of the 'walkaways' had more than 20 previous criminal convictions. While there were relatively few offenders who had 10 or fewer convictions on record, 98.6% had property-related convictions, 45.7% had assaults and 41.4% had robbery convictions. Interestingly, almost half (44.3%) of the 'walkaway' sample had previous convictions for escape or Unlawfully at Large.

In exploring further the prison history of the 'walkaway' sample, it was found that the most prevalent incidents while incarcerated were assault on other inmates (25.7%) and possession of contraband (20.0%). It was found overall that three quarters of the 'walkaways' (75.7%) had a history of security incidents while in federal custody.

An important feature of the 'walkaway' study that it sought to capture inmates' perceptions of their minimum security placement and the circumstances surrounding their unlawful departures. It is noteworthy that 76.3% of the 'walkaway' sample reported that they had too much free time on their hands while in minimum security settings. Almost half (47.4%) of the 'walkaway' sample described their free time as "boring". When asked to describe the least liked aspects of their minimum security facility, the things "liked least" included being harassed by other inmates (13.8%), lack of privacy (10.5%), program unavailability (13.2%), staff (10.0%) and a variety of other problems (18.4%). When asked specifically about programming in minimum security facilities, a large percentage (47.7%) noted that they had "no opinion" or it was "not applicable" because of non-participation due to lack of interest or long waiting lists. It is perhaps worthwhile mentioning that the 'walkaways' viewed their job assignments as separate from regular programming. While few 'walkaways' identified themselves as being behaviour problems (15.8%) or as having problems with staff (15.8%), a substantial portion of the sample (76.3%) was preoccupied with their release.

In regards to the circumstances surrounding their unlawful departures, 68.4% of the sample stated that they went alone and 84.2% claimed that their 'walkaways' were unplanned. By far the "main thing" on the minds of 'walkaways' was family/marital relations (42.1%). In fact, the "main reasons" cited for 'walkaways' were family problems (34.4%) and other inmates (15.8%). Another important finding was the 26.3% of the sample indicated that they were intoxicated at the time of their unlawful departure.

A systematic review of case file documentation yielded some further information. It is noteworthy that the entire 'walkaway' sample had a juvenile record, 89.5% had not completed high school, 81.1% had an unstable employment record, 40.0% had a history of parental abuse or neglect, 71.0% had criminal associates, 54.3% showed indications of alcohol problems, 55.9% showed indications of drug problems and 44.1% experienced heavy addictions. It would appear from the foregoing results that 'walkaways' may indeed form a relatively high risk/high need group of offenders.

The scoring of the CRS, SIR and CMS validated previous findings regarding the predictive value of objective classification systems. The 'walkaway' sample was easily differentiated by: 1) the CRS into 'minimum' (37.1%) and 'medium' (62.9%) custody level; 2) the SIR into 'fair' (1.4%), 'fair to poor' (17.1%) and 'poor' (81.4%) risks; and 3) the CMS into Selective Intervention (2.9%), Environmental Structure (14.7%), Casework/Control (38.2%) and Limit Setter (44.1%).

A descriptive profile of the circumstances motivating each 'walkaway' was also provided by reviewing narrative accounts. While it was found that problematic family/marital situations outside the institution and harassment or threats from other inmates were the cause of many 'walkaways', there were also a number of inmates who were simply bored or "sick of" staying in their facility, and who felt that their facility was in some way unresponsive to their needs.

The fact that 'walkaways' appear to be high risk (of re-offending) and multi-need individuals calls for some comment. Aside from the effects of their lengthy criminal records and the need for enhanced case management and programming while in minimum security facilities, a few tentative conclusions can be drawn. Apparently, for some of these inmates there is perceived justification for walking away; a chance to reunite with others, avoiding peer group pressure, alleviation from boredom, and of course, the possibility of freedom.

Finally, the presence of both "static" factors (e.g., criminal history) and "dynamic" or situational factors (e.g., family situation, peer problems, intoxication) points to a need for investigations which encompass a wider range of variables beyond that of documented offender characteristics. Perhaps the utilization of offender risk/needs scales specifically adapted for inmates recently placed in minimum security or nearing release would increase our ability to identify those who are likely to experience adjustment difficulties while in minimum security institutions. Moreover, systematic risk/needs assessments could also provide a useful means of monitoring changes in the offender's behaviour, attitudes and circumstances which are clearly related to the 'walkaway' phenomenon.

I. INTRODUCTION

In accordance with the Correctional Service of Canada's (CSC) Mission Statement (CSC,1989), federal institutions are mandated to provide "reasonable, safe, secure and humane control". For those offenders deemed to be reduced risks (i.e., to staff, other inmates, and the public-at-large), minimum security custody is considered appropriate. Minimum security institutions in this sense are one means of facilitating the safe reintegration of offenders who are approaching their release. These institutions serve as a kind of bridge between higher levels of custody placement (i.e., medium, maximum) and release, and hence, ease the offender into an environment which more closely approximates the community. The effective management of inmates while in minimum security is seen as an important step towards reducing the relative use of incarceration as a correctional intervention.

Although offenders placed in minimum security are considered to be manageable risks, occurrences of offenders who are declared Unlawfully at Large from these facilities, commonly known as "walkaways", is nonetheless a major concern to correctional managers. Given the relative frequency of these "walkaways" (compared to escapes from higher levels of security), it becomes important to investigate the phenomenon, and determine the factors which underlie or mediate it.

Notwithstanding the importance of understanding the phenomena of escape behaviour, there has been little empirical research identifying those factors which are significantly related to unlawful departures from custodial settings. Most of the available research has been conducted in the United States, and there have been only a few Canadian investigations (Basu, 1983; Guenther, 1983; Wharry, 1972). Generally, these studies have examined escapes across a variety of security levels (i.e., minimum, medium, maximum) and focused on particular offender characteristics (i.e., age, sex, family background). For the most part, these studies have examined "static" factors (e.g., offence type) and to a much lesser extent, "dynamic" or situational factors (e.g., family situation, peer problems, alcohol craving). Although several studies have explored environmental conditions, these investigations have examined escape behaviour simply in terms of perimeter security systems such as fences, walls, and alarms (Camp & Camp, 1987).

There have been even fewer studies focusing on unlawful departures from minimum security institutions. Although similar in some respects, the important difference between what is called a 'walkaway' and an 'escape' is that the former takes place in settings without perimeter security systems (e.g., walls, Perimeter Intrusion Detection Systems, etc.). Minimum security inmates typically reside in less secure environments with ample opportunity for involvement in both institutional and community-based programming.

On the basis of a literature review, it was not possible to discern with any degree of specificity, which variables would be most related to 'walkaway' departures (i.e., previous escapes, accessibility of temporary absences, etc.). In one of the few studies which focused on 'walkaways' from minimum security institutions (Murphy, 1984), the only variable reported to have any predictive value was previous escape(s). With respect to the rest of the escape literature, many of the findings were either inconclusive or not generalizable to minimum security settings.

In terms of managing security requirements, one approach that has been taken is to determine the appropriate security level for each offender at the time of initial custody placement. For example, the CSC has a specialized reception centre in Ontario (Millhaven Institution) which performs this function. Upon reception, federally sentenced offenders are evaluated according to security and programming needs.

Custody Classification

Although not yet fully implemented, the CSC has developed an objective classification instrument as an aid to initial penitentiary placement; the Custody Rating Scale (CRS). The CRS is actually composed of two subscales -- Institutional Adjustment and Security Risk -- which together yield a security classification of either minimum, medium, or maximum. The information that goes into the CRS (e.g., the offenders' escape history, past institutional incidents, prior alcohol/drug use, age, sentence length) have been shown to be significantly related to behavioral problems during incarceration. With regard to minimum security placements, it should be noted that in a recent pilot implementation of the CRS (CSC; 1989), roughly a third of warrant committal admissions were rated as being suitable candidates for minimum security. Moreover, in the actual placement decisions that were being made for these offenders, there was a tendency towards over-classification (i.e., medium or maximum versus minimum security).

Increasingly, the values expressed in CSC's Mission Statement are informing the type of custody placement offenders receive. In keeping with the premise that "the least restrictive course of action" be followed in custody placement, an increasing number of offenders are being safely assigned to minimum security facilities. The potential for unlawful departures, however, stands as an issue of focused concern for correctional managers in that it suggests that at some point in an offender's sentence, something could go awry. This does not suggest that placement decisions were necessarily inappropriate but that somewhere along the course of an offender's incarceration, something motivates them to risk the consequences of walking away from custody. Therein lies a gap in our understanding of the factors which underlie or mediate 'walkaway' phenomena.

To date, there have been no systematic attempts to study the factors related to 'walkaways' from federal corrections. The purpose of the present study was to determine those factors, and hopefully provide a better understanding of 'walkaways'. This study should provide useful information to correctional case managers. For instance, there are many situations such as the granting of temporary absences, day parole, reclassification, and accelerated review, that are quite relevant to the issue of 'walkaways'. Viewed from any perspective, this study should be seen as contributing to our understanding of unlawful departures from minimum security facilities.

II. PRESENT STUDY

During the Spring of 1990, there was an unusual increase in the number of federally sentenced offenders unlawfully walking away from minimum security institutions in the Ontario region. The Audit and Investigations Sector of the CSC was asked to examine this situation and identify more precisely any factors which could possibly account for those departures. In order to facilitate this endeavour, the Research and Statistics Branch was approached, who in collaboration with the (formerly) Custody and Control Division and the Ontario region, formed a working group to investigate this matter. It was the goal of this group to design and conduct a study that would identify those factors related to, and predictive of walkaways.

The present investigation differs from previous studies on escape or 'walkaway' phenomena in several ways. First, it focuses specifically on unlawful departures from minimum security institutions in Canadian federal corrections. Secondly, and perhaps more importantly, it examines both "static" factors (e.g., criminal history) as well as "dynamic" factors (e.g., personal crises) that were involved in the 'walkaways'. By generating new information on this issue, it was thought that we could advance our understanding of 'walkaways', and thereby make a contribution towards reducing the likelihood of such events.

III. METHOD

Sample Selection

In order to conduct the study, data was collected on all offenders who walked away from federal minimum security institutions in the Ontario region (Bath, Beaver Creek, Frontenac, Pittsburg) over a 15 month period (January 1, 1990 - April 15, 1991). In total, there were 70 offenders who had been declared unlawfully at large during this period.

Data Gathering Process

Data was gathered on the 'walkaways' from a variety of sources. The range of variables examined were obtained from automated systems, face-to-face interviews, case file reviews and classification instruments.

Retrieval from Official Automated Records. Offender information was retrieved from the following automated sources: CSC's Offender Information System (OIS) and Security Incident System (SIS), as well as the Canadian Police Information Centre (CPIC) system. The OIS contains background information such as: offence type, sentence length, marital status, race, admission data, major offence, and a variety of institutional process variables (e.g., day parole eligibility date, warrant expiry date, admission type, etc.).
The SIS records security incidents (e.g., possession of contraband, inmate assault) that an offender was involved in during his period(s) of federal supervision. This information is broken down into the following categories: violence (e.g., assault on inmate, assault on staff, inmate fight), escape (e.g., prison breach, walkaway, fail to return from temporary absence), contraband (e.g., possession, under the influence, receiving/transporting), and general behaviour problems (e.g., theft, Protective Custody request, vandalism).

CPIC data is basically comprised of each offenders' criminal record. Here, the complete offence history of an offender is recorded, which includes all criminal convictions, the date of each conviction, as well as the sentence imposed for each conviction. Dates of release (e.g., parole, mandatory supervision) are also noted, as are violations of probation, parole or mandatory supervision.
Face-to-face Interviews. Personal interviews were conducted with those inmates who had walked away and subsequently were available to voluntarily participate in this study. This yielded a sub-sample of 38 (54.3% of the cases). In order to conduct the face-to-face interviews, a field researcher was hired on contract. This individual received in-house training to administer both a pre-designed walkaway interview (see Appendix A) as well as a structured case file review instrument (see Appendix B).
The purpose of the face-to-face interviews was to uncover the motivational reason(s) for each walkaway. The interview protocol was structured to gather relevant information on variables that had been identified by a literature review and generated by a series of consultations with CSC staff.

All former walkaway offenders who were subsequently returned to federal institutions in the Ontario region were approached and asked to volunteer in the study Each participant signed a consent form indicating his willingness to participate. It was assured that the information provided would be kept confidential and used only for research purposes (see Appendix C). Given that a sizeable portion of the study sample had been released at the time of the interviews, only 40 cases were available. Of those cases who were approached, there were only 2 refusals, thus giving a total of 38 inmates on whom interview data was collected. Subsequent to completion of the interviews, it was found that the 2 refusals had recent disciplinary actions taken against them which may have, in part, accounted for their non-participation in the study.
Case File Reviews. The structured file reviews which were conducted on the 'walkaway' sample (Appendix B) provided an additional source of information. For example, Case Management Strategies designations (Lerner, Arling, & Baird, 1986) for these offenders were extracted from their case files.
Classification Systems. The various sources that were used to obtain information on the 'walkaways' allowed for the application of several standardized classification instruments. These included the CRS and the Statistical Information on Recidivism (SIR) scale. The CRS (see Appendix D) is an empirically derived scale which classifies offenders into one of the three custody levels (i.e., minimum, medium, maximum), based on their criminal history, age, offence type, institutional incident history, and several personal adjustment variables. The SIR scale (see Appendix E) utilizes offence history, social history, and personal information on offenders to calculate a score which rates the offender in terms of their risk for recidivism. These scores correspond to five levels of risk; "Poor", "Fair to Poor", "Fair", "Good", and "Very Good".
The Case Management Strategies (CMS) classification (Lerner et al, 1986) for the interviewed offenders provides information on differential supervision strategies. These are entitled Casework/Control, Limit Setter, Environmental Structure and Selective Intervention. Casework/Control offenders are characterized by chronic and generalized instability, and having multiple needs. Limit Setters typically have criminal orientations and little interest in behaving in accord with pro-social norms, or sustaining themselves through acceptable, non-criminal means. The principle characteristic of Environmental Structure offenders is a serious lack of social and vocational skills. These individuals' involvement in crime tends to be unsophisticated and often impulsive. Selective Intervention offenders, on the other hand, tend to have acceptable social and job-related skills, as well as pro-social value systems. Their involvement in crime is typically quite limited, and often the result of a relatively isolated life event. The CMS classifications for the interviewed walkaways was retrieved from their case files.

IV. FINDINGS

CHARACTERISTICS OF WALKAWAYS

A distribution of the minimum security institutions from which the inmates "walked away" in the Ontario region is presented in Table 1. Over 75 % of the walkaways were accounted for by Bath (n=26) and Frontenac (n=29) institutions. Interestingly, these institutions also had the highest monthly admission rates (i.e., 16.7 and 9.7 respectively) as well as walkaway rates (i.e., 1.7 and 1.9 respectively).

Table 1

Distribution of Walkaways from Minimum Security Institutions

Institution	Related Capacity	Admission Rate*	Release Rate*	Walkaway Rate*
Bath	102	16.7 (259)	8.8 (137)	1.7 (26)
Beaver Creek	122	5.1 (79)	7.7 (120)	0.4 (6)
Frontenac	161	9.7 (150)	10.6 (164)	1.9 (29)
Pittsburg	105	7.2 (111)	9.4 (146)	0.6 (9)
Total	490	38.6 (599)	36.6 (567)	4.5 (70)

Note: * Average monthly rate; (n)

Table 2 presents the temporal distribution (by quarter) of 'walkaways'. As expected, the majority of 'walkaways' (67%) took place during the Spring and Summer months. This pattern of unlawful departures was found to be consistent across the four institutions.

Table 2

Temporal Distribution of Walkaways

Institution	Jan. 1- Mar. 31, 1990	Apr. 1 - June 30, 1990	July 1 - Sept. 30, 1990	Oct. 1 - Dec. 31, 1990	Jan. 1 - April 15, 1991
Bath	2	13	6	1	4
Beaver Creek	1	1	3	1	0
Frontenac	8	8	10	0	3
Pittsburg	1	0	6	0	2
Total (%)	12 (17.1)	22 (31.4)	25 (35.7)	2 (2.9)	9 (12.9)

Table 3 presents a distribution of the amount of time served in relation to three reference periods for the 'walkaways'. First, the average amount of time spent in custody between admission into federal corrections and subsequent transfer to minimum security was 318 days overall. Second, the average number of days spent between admission and subsequent walkaway was 371 days. Finally, the number of days between transfer to minimum security and 'walkaway' averaged 53 days. For the first two reference periods (i.e., from admission to minimum placement, from admission to walkaway) there were no significant differences found in the number of days the offenders spent across the four facilities. A statistically significant difference between the facilities was found, however, in the number of days between the offenders' minimum placement and their walkaway. Further statistical analyses revealed that walkaways at Bath and Frontenac institutions departed sooner than those at the other two institutions.

Table 3

Days Spent in Custody by Walkaways

Reference Period	Bath	Beaver Creek	Frontenac	Pittsburg	Overall
Admission to minimum placement	M = 233 S = 355 R = 7 - 1336	M = 343 S = 342 R = 54-1082	M = 343 S = 514 R=13-2705	M = 365 S = 463 R=19-1084	M = 319 S = 438 R = 7 -2705

Admission to walkaway	M = 276 S = 369 R = 20 -1428	M = 623 S = 453 R = 198-1447	M = 380 S = 534 R=34-2830	M = 445 S = 453 R=48-1132	M = 371 S = 462 R=20 -2850

Minimum placement to walkaway	M = 43 S = 43 R = 5 - 162	M = 129 S = 144 R = 4 - 365	M = 37 S = 36 R = 0 - 145	M = 80 S = 41 R = 29-152	M = 53 S = 60 R = 0 - 365

Note: M = mean (average)
S = standard deviation
R = range

CHARACTERISTICS OF SAMPLE

General characteristics of the 'walkaway' sample are presented in Table 4. As can be seen, over 75% of these offenders were under 30 years of age, and the majority were Caucasian (97.1%). While 58% of the sample were single, it is noteworthy that at Beaver Creek, all 6 walkaways were married/common-law.

The majority of the 'walkaways' (60%) were currently serving sentences for major admitting offenses (i.e., offence with the longest sentence) which were property-related (i.e., break and enter, theft, possession of stolen property). We note that there was only one offender serving a sentence for homicide (2nd degree murder) and only one for a drug offence. Approximately one third of the sample was serving a term for violent offenses (i.e., Homicide, Attempted Murder, Robbery) and there were no current sex offenders in this group. Further, 71.5% of these 'walkaways' were serving sentences of 4 years or less.

Table 4

Percentage Distribution of Offender Characteristics

Variables	Bath (n=26)	Beaver Creek (n=6)	Frontenac (n=29)	Pittsburg (n=9)	Total (n=70)
Âge:
< 19	3,8	0,0	3,5	0,0	2,9
20-24	42,3	0,0	41,4	0,0	32,9
25-29	34,6	50,0	34,5	66,7	40,0
30-39	19,2	33,3	17,2	33,3	21,4
40-49	0,0	16,7	0,0	0,0	1,4
50+	0,0	0,0	3,5	0,0	1,4
Race:
Caucasian	100.0	100.0	93.1	100.0	97.1
Native	0.0	0.0	6.9	0.0	2.9
Marital Status:
Single	53.8	100.0	62.1	77.8	58.6
Married	46.2	0.0	37.9	22.2	41.4
Major Offence:
Homicide	0.0	0.0	3.5	0.0	1.4
Manslaughter	0.0	0.0	0.0	0.0	0.0
Att. Murder	3.8	0.0	0.0	0.0	1.4
Robbery	19.2	0.0	34.5	33.3	25.7
Sex Offence	0.0	0.0	0.0	0.0	0.0
Assault	0.0	16.7	0.0	0.0	1.4
Property	69.3	50.0	58.6	33.3	58.6
Drugs	0.0	16.7	0.0	0.0	1.4
Other	7.7	16.7	3.5	33.3	10.0
Sentence Length:
<2 yrs	3.8	0.0	17.2	0.0	8.6
2-4 yrs	80.8	50.0	51.7	55.6	62.9
5-9 yrs	15.4	33.3	20.7	44.4	22.9
10+ yrs	0.0	16.7	6.9	0.0	4.3
Life	0.0	0.0	3.5	0.0	1.4

CONVICTION HISTORY OF SAMPLE

The CPIC data provided a more detailed breakdown of the 'walkaway' offenders' criminal history. Table 5 presents the overall volume of previous convictions for the 'walkaway' offenders. Interpretation of Table 4 reveals that for each of the institutions, the majority of offenders (65.4%) had more than 20 previous convictions. This pattern of results was found to be consistent across institutions.

Table 5

Percentage Distribution of Previous Convictions by Volume

Number	Bath (n=26)	Beaver Creek (n=6)	Frontenac (n=29)	Pittsburg (n=9)	Total (n=70)
10 or less	7.7 (2)	16.7 (1)	10.3 (3)	11.1 (1)	10.0 (7)

Between 11 and 20	23.1 (60)	16.7 (1)	34.5 (10)	0.0 (0)	24.3 (17)

Between 21 and 30	46.2 (12)	16.7 (1)	17.5 (50)	33.3 (3)	30.0 (21)

Between 31 and 40	19.2 (95)	16.7 (1)	24.1 (7)	33.3 (3)	22.9 (16)

40 +	3.8 (1)	33.3 (2)	13.8 (4)	22.2 (2)	12.9 (9)

Note: (n)

Types of previous convictions for the sample are presented in Table 6. As was found for their current major offence, property-related offenses (i.e., break and enter, theft, and possession of stolen property) were the most prevalent (98.6%) for these offenders. Also common were previous convictions for assault (45.7%) and robbery (41.4%). There were only 2 offenders (2.9%) who had past convictions for sex offenses. Interestingly, 44.3% of the walkaways had previous convictions for one or more previous escapes or being Unlawfully at Large (UAL).

Table 6

Percentage Distribution of Previous Convictions by Type

Type	Bath (n=26)	Beaver Creek (n=6)	Frontenac (n=29)	Pittsburg (n=9)	Total (n=70)
2nd Degree Murder	0.0	0.0	3.4	0.0	1.4
Manslgtr	0.0	16.7	0.0	0.0	1.4
Assault	42.3	83.3	34.5	66.7	45.7
Escape/UAL	50.0	16.7	48.3	33.3	44.3
Property	96.2	100.0	100.0	100.0	98.6
Impaired Driving	30.8	16.7	31.0	33.3	30.0
Drugs	38.5	50.0	37.9	33.3	38.6
Robbery	30.8	33.3	48.3	55.6	41.4
Sex Offence	0.0	0.0	3.4	11.1	2.9
Weapons	11.5	16.7	48.3	33.3	30.0
Auto or Traffic	34.6	33.6	48.3	44.4	41.4
Other Conviction	96.2	100.0	93.1	100.0	98.4

SECURITY INCIDENT HISTORY OF SAMPLE

Upon examination of security incident data, it was found that 53 (75.7%) of the 'walkaways' had a history of security incidents while in the federal system. Table 7 presents a breakdown of the type of institutional incidents recorded for the sample. Although the types of recorded incidents for these offenders was quite varied, it was found that the most prevalent type of incident was assault on another inmate (25.7%). Further, 20% of the sample had at least one previous walkaway, escape, or fail to return from a temporary absence. While 'walkaways' from Beaver Creek had no history of previous escape recorded, they had the highest rate of assault and possession of contraband relative to the other institutions.

Table 7

Percentage Distribution of Security Incidents by Type

Incident	Bath (n=26)	Beaver Creek (n=6)	Frontenac (n=29)	Pittsburg (n=9)	Total (n=70)
Violence:
Assault on Inmate	26.9	66.7	17.2	22.2	25.7
Assault on Staff	0.0	0.0	6.9	0.0	2.9
Inmate Fight	7.7	16.7	6.9	44.4	12.8
Self-Injury	3.9	33.3	0.0	0.0	4.3
Suicide Attempt	0.0	16.7	3.5	0.0	2.9
Threat	7,7	33.3	10.3	11.1	11.8
Escape:
Walkaway	22.2	0.0	24.1	11.1	14.3
Escape S3+	7.7	0.0	0.0	0.0	2.9
Escort Escape	0.0	0.0	3.5	0.0	1.4
Fail to Return from UTA	0.0	0.0	10.3	0.0	4.3
Contraband:
Posession	3.9	66.7	27.6	11.1	20.0
Receive/Transport	7.7	16.7	10.3	0.0	8.5
Intoxicated	15.4	0.0	6.9	11.1	10.0
Behaviour:
Disciplinary Action	7.7	33.3	10.3	11.1	11.1
Vandalism	0.0	16.7	0.0	0.0	1.4
Theft	3.9	0.0	20.7	0.0	2.9
PC request	7.7	50.0	6.9	11.1	11.4
Intelligence received of involvement	23.1	16.7	20.7	11.1	20.0

Table 8 presents the volume of security incidents recorded while in federal custody for the 'walkaway' sample. While 24.3% of the 'walkaways' had no previous security incident before their unlawful departure, 75.8% had been involved in at least one or more security incidents.

Table 8

Percentage Distribution of Security Incidents by Volume

Number	Bath (n=26)	Beaver Creek (n=6)	Frontenac (n=29)	Pittsburg (n=9)	Total (n=70)
None	26.9 (7)	0.0 (0)	27.6 (8)	22.2 (2)	24.3 (17)
One	26.9 (7)	16.7 (1)	24.1 (7)	22.2 (2)	24.3 (17)
Two	15.4 (4)	16.7 (1)	10.3 (3)	22.2 (2)	14.3 (10)
3 or more	30.8 (8)	66.7 (4)	37.9 (11)	33.3 (3)	37.2 (26)

Note: (n)

OFFENDER PERCEPTIONS REGARDING MINIMUM SECURITY PLACEMENT

We asked each 'walkaway' a number of questions concerning their reflections on the minimum security facility from which they departed (see Table 9). For the most part, the offenders reported that they did not have any significant problems relating to either their institution or its staff. In fact, 78.9% of the 'walkaways' noted that they did not have any problems with the institutions' staff. Interestingly, only a small percentage (15.8%) disclosed that they were considered to be discipline problems. These findings were supported by the frequent comments by the offenders during the interview expressing positive attitudes towards the staff. Negative comments, when noted, tended to be critical of administrative aspects of the correctional process (e.g., availability of passes) rather than any personal conflicts.

In response to questions about the facility walked away from (see Table 9), 34.2% reported that they had no complaints. Among those offenders who reported problems, the nature of their complaints were evenly distributed across poor or non-existent programming (13.2%), harassment from other inmates (13.2%), privacy (10.5%) and staff (10.5%). When asked specifically to comment on available programs, offenders were evenly split between those who expressed positive versus negative attitudes. The 47.4% of "'no opinion'/not applicable" responses was attributed to non-participation in programming due to either lack of interest or the fact that they had walked away from their facility before any programming. It is also worthwhile noting that the 'walkaways' considered job assignments as separate from regular "programming".

As can be seen in Table 9, 76.3% noted that they had "a lot of free time" on their hands, and 47.4% said they felt they had "too much" or that it was "boring".

Table 9

Percentage Distribution of Responses Relating to the Minimum Security Facility

Interview Question	Bath (n=26)	Beaver Creek (n=6)	Frontenac (n=29)	Pittsburg (n=9)	Total (n=70)
Too much free time	85.7	100.0	62.5	80.0	76.3
Opinion of Free time:
Boring	50,0	100,0	25,0	20,0	47,4
OK	35,7	0,0	43,7	80,0	34,2
Nothing	14,3	0,0	31,3	0,0	18,4
Liked Least:
No problems	28,6	66,7	37,5	20,0	34,4
Harassed	14,3	33,3	12,5	0,0	13,2
Privacy	7,1	0,0	12,5	20,0	10,5
Programs	7,1	0,0	25,0	0,0	13,2
Staff	14,3	0,0	6,3	20,0	10,5
Other	28,6	0,0	6,3	40,0	18,4
Programs:
Good	42,9	33,3	6,3	20,0	23,7
Neutral	7,1	0,0	12,5	0,0	7,9
Poor	14,3	0,0	31,3	20,0	21,1
N/A	35,7	66,7	50,0	60,0	47,4
Missed Most:
Freedom	35,7	33,3	6,3	20,0	23,7
Relations	50,0	33,3	12,5	0,0	7,9
Nothing	7,1	33,3	31,3	20,0	21,1
Other	7,1	0,0	50,0	60,0	47,4
Behavior Problems	21.3	0.0	12.5	20.0	15.8
Problems w/ Staff	7.1	33.3	31.3	20.0	21.1
Thought much about release	71.4	81.3	33.3	100.0	76.3

OFFENDER PERCEPTIONS REGARDING CIRCUMSTANCES SURROUNDING WALKAWAY

The interview data in Table 10 revealed that the majority (68.4%) of offenders departed their institutions alone, and that 84.2% reported doing so spontaneously, without any forethought or plans. Further, 56.6% said they were not particular bothered about getting caught. Every offender reported that they knew returning to prison was likely, and it was noted during the interview that many planned on returning voluntarily to their institution. This was largely the case with those offenders who left to take care of family-related problems.

It is noteworthy that the largest portion (42.1%) of the walkaways reported that their family or close friends were the "main thing" on their mind when they departed from their institution. This is supported by the finding that 34.4% cited problems at home as being the principal reason why they walked, and another 10.5% claimed that being unable to obtain a pass was their main reason. In these cases, it was noted that a strong desire to visit a friend or family member was expressed during the interviews.

In keeping with the earlier finding that harassment from other inmates (or another inmate) was problematic for some offenders, 15.8% said other inmates played a threatening role in their walkaway and the same percent reported that other inmates were the 'main reason' they walked. Moreover, fear for their own safety was endorsed by 13.2% of offenders as the main reason they walked, and 21% reported being threatened by another inmate or inmates.

In terms of the mood the 'walkaways' recalled experiencing at the time of their departure, depression (23.7%) and anger/frustration (21.1%) were the most commonly cited emotions. Again, it was noted during the interviews that anger and frustration were often attributed to limitations in programming, or a perceived inability to control events outside the institution.

A somewhat unexpected finding from the interviews was that, without being specifically asked, 26.3% of the walkaways indicated that they were intoxicated at the time of their departure.

Table 10

Percentage Distribution of Responses Related to the Walkaway

Interview Question	Bath (n=26)	Beaver Creek (n=6)	Frontenac (n=29)	Pittsburg (n=9)	Total (n=70)
Went alone	64.3	100.0	68.8	60.0	68.4
Role of others
Partners	28,6	0,0	25,0	40,0	26,3
Threat	14,5	33,3	12,5	20,0	15,8
N/A	57,1	66,7	62,5	40,0	57,9
Unplanned	100.0	100.0	81.3	40.0	84.2
Conflict w/ another inmate	21.4	33.3	12.5	40.0	21.1
‘Main thing’ on their mind:
Relations	71.4	66.7	18.8	20.0	42.1
Capture	7.1	0.0	31.3	20.0	18.4
Safety	7.1	33.3	12.5	20.0	13.2
Summer	0.0	0.0	12.5	20.0	7.9
Nothing	0.0	0.0	18.8	0.0	7.9
Other	14.3	0.0	6.3	20.0	10.4
Mood:
Good	14.3	0.0	12.5	20.0	13.2
Depressed	28.6	66.7	18.8	0.0	23.7
Angry	21.4	33.3	18.8	40.0	21.1
Fearful	21.4	0.0	12.5	0.0	13.2
Other	7.1	0.0	6.1	40.0	13.2
Can’t recall	7.1	0.0	31.3	0.0	15.8
Main Reason:
Family Probs	57.1	33.3	18.8	20.0	34.4
Other inmate	7.1	33.3	12.5	40.0	15.8
No passes	7.1	0.0	12.5	20.0	10.5
Intoxicated	7.1	0.0	18.8	0.0	10.5
Freedom	0.0	0.0	18.8	0.0	7.9
Staff probs.	14.3	0.0	0.0	20.0	7.9
Bored	7.1	33.3	12.5	0.0	10.5
Other	10.2	0.0	6.1	0.0	2.5
Intoxicated	35.7	0.0	18.8	40.0	26.3

CASE FILE REVIEW INFORMATION

The case file review portion of this study, while serving to substantiate information obtained in the face-to-face interviews, also provided additional background on the 'walkaways' (see Table 11). It is noteworthy that the entire sample of offenders had been arrested prior to their 16th birthday, 89.5% had not completed high school, 81.1% had unstable employment prior to admission, and that over 70% had criminal associates. More than half of these individuals were identified as substance abusers and 44.1% were heavily addicted (e.g., $500 per day cocaine habit).

Table 11

Percentage Distribution of Information From Case File Reviews (n=38*)

Variable	Percentage
Arrested at age 16 or before	100.0
Violence used in current offence	18.4
Current offence property-related	100.0
completed: Grade School	89.5
completed: High School	10.5
Unstable employment	81.1
History of parental abuse or neglect	40.0
Other family members with criminal records	31.3
Has criminal associates	71.0
Unstable residence	47.2
Committed current offence while on bail or under supervision	63.2
Indication of alcohol problem	54.3
Indication of drug problem	55.9
Heavy addiction indicated	44.1
Previous suicide attempt(s)	14.3

*Given that case files sometimes did not include, or were ambiguous about the information looked for, the percentages do not always reflect the full sample of 38 interviewed walkaways.

OFFENDER CLASSIFICATION SYSTEMS

The case file review information, along with the automated data, allowed for both the CRS and the SIR classifications to be calculated for each 'walkaway' offender (including those who were not interviewed). It should be noted, however, that since complete information on each offender could not be captured (e.g., in order to score "street stability" or degree of alcohol/drug use), it was decided to give offenders the benefit of the doubt and score in the positive direction (i.e., towards under-classification in terms of security and risk rating) where information was ambiguous or lacking. That is, if there was uncertainty on any of the rating questions, a minimum score was entered. There were few instances where these adjustments had to be made.

Custody Rating Scale: The distribution of CRS classifications is presented in Table 12. Overall, 44 (62.9%) cases were classified as medium security while 26 (37.1%) came out as minimum security. While Beaver Creek and Pittsburg Institutions were roughly equivalent in the distribution of minimum and medium custody level classifications, there were proportionally more medium security offenders in Bath and Frontenac Institutions. This was especially the case for Bath, where nearly three quarters (73.1%) of the 'walkaways' were CRS-rated medium custody offenders.

Table 12

Percentage Distribution of Custody Rating Scale (CRS) Scores*

CRS Rating	Bath	Beaver Creek	Frontenac	Pittsburg	Total
Minimum Security	26.9 (7)	50.0 (3)	37.9 (11)	55.6 (5)	37.1 (26)
Medium Security	73.1 (19)	50.0 (3)	62.1 (18)	44.4 (4)	62.9 (44)

* There were no cases who rated as Maximum security on the CRS.

Statistical Information on Recidivism Scale: With respect to the SIR scores (see Table 13), it was found that more than 80% of the 'walkaway' sample were classified as 'poor' risks (i.e., 2 out of 3 will reoffend). Interestingly, there were no offenders whose SIR classification was in the 'good' or 'very good' range.

Table 13

Percentage Distribution of Statistical Information on Recidivism (SIR) Scores*

SIR Scale	Bath	Beaver Creek	Frontenac	Pittsburg	Total
Fair Risk	3.8 (1)	0.0 (0)	0.0 (0)	0.0 (0)	1.4 (1)
Fair to Poor Risk	11.5 (3)	33.3 (2)	20.7 (6)	11.1 (1)	17.1 (12)
Poor Risk	84.6 (22)	66.6 (4)	79.3 (23)	88.9 (8)	81.4 (57)

* There were no cases whose SIR classification was in the Good or Very Good Risk range.

Case Management Strategies: Of those whose case files that were reviewed, 34 (89.5%) had CMS scores available. While 38% of this sample were classified as Casework/Control, 44% were Limit Setters, 15% were Environmental Structure and 3% were Selective Intervention.

NARRATIVE ACCOUNTS BY WALKAWAYS

Brief narrative accounts were recorded for each offender which describe the circumstances motivating their 'walkaway' (see Appendix F). Although these accounts can only be considered abbreviated summaries, it would appear that problematic family/marital situations outside the institution and harassment or threats from other inmates tend to motivate 'walkaways'. As might be expected, there were also a number of offenders who were simply bored or "sick of" staying in their facility, and who felt that their facility was in some way unresponsive to their needs. Furthermore, drug and alcohol use in the institutions by these offenders was a factor which also undoubtedly underlies the phenomenon of 'walkaways'. In fact, substance use in the institutions seems to have acted as a kind of catalyst in that regardless of the offenders' problems in or out of the facility, the use of intoxicants appeared to disinhibit the offender from remaining in custody.

V. DISCUSSION

The results of the 'walkaway' study revealed that unlawful departures from minimum security facilities were most likely to occur at institutions with higher admission rates, during the Spring and Summer months, and within the first several months of minimum security placement. The 'walkaway' offenders were more often under 30 years of age (75.4%), Caucasian (97.1%), single (58.6%), having a major offence that was property-related (58.6%) and serving sentences of under four years (71.5%).

Of special interest was the criminal conviction and prison history of the 'walkaway' sample. As many as 65.4% of the 'walkaways' had more than 20 previous convictions and only 10% had 10 or fewer convictions on record. Although 98.6% of the 'walkaway' sample had property-related convictions, 45.7% had assaults and 41.4% had robbery convictions. Interestingly, 44.3% of the 'walkaway' sample had previous convictions for escape or being Unlawfully at Large.

The security incident history of the 'walkaway' sample indicated that the most prevalent incidents were assault on other inmates (25.7%) and possession of contraband (20.0%) while in federal custody. Moreover, 75.8% of the 'walkaways' had at least one security incident on record.

An important feature of the 'walkaway' study was the offenders' self-reports regarding their minimum security placement and the circumstances surrounding their unlawful departures. It is noteworthy that 76.3% of the 'walkaway' sample reported that they had too much free time on their hands while in minimum security settings. An Almost half (47.4%) of the 'walkaway' sample described their free time as "boring". When asked to describe the least-liked aspects of their minimum security facility, the things "liked least" included being harassed by other inmates (13.8%), lack of privacy (10.5%), program availability (13.2%), staff (10.0%) and a variety of other problems (18.4%). When asked specifically about programming in minimum security facilities, a large percentage (47.7%) noted that they had "no opinion" or that it was "not applicable" because of non-participation due to lack of interest or waiting lists. It is perhaps worthwhile mentioning that the 'walkaways' viewed their job assignments as separate from regular programming.

While a few 'walkaways' identified themselves as having behaviour problems in their facility (15.8%) or as having problems with staff (15.8%), a substantial portion of the sample (76.3%) said they were preoccupied with the idea of their release.

In regards to the circumstances surrounding their unlawful departures, 68.4% of the sample stated that they went alone and 84.2% claimed that their 'walkaways' were unplanned. By far the "main thing" on the minds of 'walkaways' was family/marital relations (42.1%). In fact, the "main reasons" cited for 'walkaways' were family problems (34.4%) and problems with other inmates (15.8%). Another important finding was that 26.3% of the 'walkaway' sample claimed that they were intoxicated at the time of their unlawful departure.

A systematic review of the interviewed 'walkaways' case file documentation yielded some further information. It is noteworthy that the entire 'walkaway' sample had a juvenile record, 89.5% had not completed high school, 81.1% had an unstable employment record, 40.0% had a history of parental abuse or neglect, 71.0% had criminal associates, 54.3% showed indications of alcohol problems, 55.9% showed indications of drug problems and 44.1% were indicated as having a heavy addiction. It would appear from the foregoing results that 'walkaways' may indeed form a relatively high risk/high need group of offenders.

A descriptive profile of the circumstances motivating each 'walkaway' was also provided by reviewing the brief narrative accounts. While it was found that problematic family/marital situations outside the institution and harassment or threats from other inmates were the cause of many 'walkaways', there were also a number of inmates who were simply bored or "sick of" staying in their facility, and who felt that their facility was in some way unresponsive to their needs.

The fact that 'walkaways' appear to be high risk (of re-offending) and multi-need individuals calls for some comment. Aside from the effects of their lengthy criminal records and the need for enhanced case management and programming while in minimum security facilities, a few tentative conclusions can be drawn. Apparently, for some of these inmates there are many attractions in walking away such as: a chance to reunite with others, avoiding peer group pressure, alleviation from boredom, and of course, the possibility of freedom.

Finally, the success of both "static" factors (e.g., criminal history) and "dynamic" or situational factors (e.g., family situation, peer problems, intoxication) points to a need for investigations which encompass a wider range of variables beyond that of documented offender characteristics. Perhaps the utilization of offender risk/needs scales specifically adapted for inmates recently placed in minimum security or nearing release would increase our ability to identify those who are likely to experience adjustment difficulties while in minimum security institutions. Moreover, systematic risk/needs assessments could also provide a useful means of monitoring changes in the inmate's behaviour, attitudes and circumstances which are clearly related to the 'walkaway' phenomenon.

VI. REFERENCES

Basu, S. (1983). Models of escape behaviour of minimum security inmates in Ontario: The use of routine administrative data for risk assessments. Cornell University: Unpublished doctoral dissertation.

Camp, G. & Camp, C. (1987). Stopping escapes: Perimeter security. Washington, D.C.: National Institute of Justice, Construction Bulletin.

Correctional Service Canada (1989). Mission of the Correctional Service of Canada. Ottawa: CSC Publishing and Editorial Services.

Correctional Service Canada (1989). Pilot implementation of a Custody Rating Scale: Interim report. Ottawa: Research Branch, Communications and Corporate Development.

Fisher, K.L. (1977). Determination of some personality characteristics of women who escape from prison. Unpublished doctoral dissertation: University of Washington.

Green, J.R. and Martin, D.N. (1973). Absconding from approved schools as learned behaviour." Journal of Research in Crime and Delinquency, 10, 73-86.

Guenther, P. (1983). Analysis of escapes from provincial correctional facilities - April 1, 1979 to August 31, 1983. Ottawa: Correctional Service Canada.

Lerner, K, Arling, G., & Baird, C. (1986). Client management classification strategies for case supervision. Madison: National Council on Crime and Delinquency.

Murphy, T. H. (1984). Prediction of minimum security walkaways. Michigan Department of Corrections, research report.

Wharry, J.D. (1972). A study of the nature and frequency of crimes committed by escapees of maximum and medium security institutions and some characteristics of escapes. Ottawa: Canadian Penitentiary Service report.

White, R.B. (1979). Relationship of scores on the escapism scale of the MMPI to escape from minimum security federal custody. Journal of Clinical Psychology, 35, 467-470.

Zamble, E. & Porporino, F.J. (1988). Coping, behaviour, and adaptation in prison inmates. New York: Springer-Verlag.