Can Educating Adult Offenders Counteract Recidivism?

Prepared for

The ACA Winter Conference
Portland, Oregon

Research and Statistics Branch
Communications and Corporate Development
Correctional Service of Canada
340 Laurier Avenue West
Ottawa, Ontario

Executive Summary

Participation in Adult Basic Education grew very quickly in the Correctional Service of Canada since the Government of Canada announced special measures to combat literacy in 1986/87. At any one time, about 1,600 offenders, or 15% of the federal inmate population, is engaged in ABE programming. Almost half of those who participate are enroled on a full-time basis. Although, ABE programming requires a large share of our programming resources (about $5.6 million spent in 1990-1991), until recently there was only scant evidence regarding the "correctional" impact of ABE . The Correctional Service of Canada has now completed research on the effects of Adult Basic Education on recidivism and other indicators of post-release adjustment in the community. The current results have furnished grounds for optimism about the positive benefits of our efforts in providing literacy and numeracy training for federal offenders.

CSC researchers approached the question of the effectiveness of ABE from two perspectives. A large follow-up study was done on ABE participants who were conditionally released to determine whether or not program completion was related to subsequent readmissions. Secondly, in order to determine offender perceptions about the benefits of ABE , interviews were conducted with a small sample of offenders in two regions. In both studies, the evidence was very positive.

In the large study, readmissions were monitored for an average of 1.1 years for 1,736 offenders who had been participants in ABE in 1988 and were subsequently released on full parole or mandatory supervision. The released offenders fell into three categories of ABE participation: Completion (grade 8 equivalent), Release before Completion, and ABE Withdrawals. The latter group was composed of offenders who had voluntarily decided to withdraw from the program before reaching the grade 8 achievement target. The poorest outcomes were associated with this group. Only 30.1% of the ABE completers had readmissions during the follow-up period compared to 35.5% of those who were released before completing, and 41.6% of the offenders who had withdrawn from ABE.

A possible explanation for the findings was that ABE completers were more successful in the community, not because of ABE , but because they were lower risk offenders. However, a series of statistical analysis that we used to test this argument indicated that even high risk offenders benefit from ABE . In fact, those offenders who possess "high risk" characteristics appeared to benefit from ABE completion more than lower risk offenders. For example, while the positive effects of ABE on recidivism for mandatory supervision cases was apparent, ABE completion did not influence post-release outcomes for full parole cases. In other words, ABE seemed to help reduce recidivism for mandatory supervision cases, but had no impact for full parole cases. Other evidence showed benefits of ABE when alternative "high risk" indicators were examined. The positive influence of ABE held for offenders serving longer sentences, those who had served previous federal terms, violent offenders and younger offenders.

The second study surveyed, using in-depth face-to-face interviews, 38 offenders in the Pacific and Ontario regions who had participated in ABE and were released on full parole or mandatory supervision. Since their release, the majority of these offenders (79%) were employed full-time and most had gained employment within the first month after release. About a third of the respondents felt that skills gained in ABE were very helpful in their job search. About half of the offenders said that ABE skills continued to be "very useful" with respect to matters they encountered "on-the-job". Close to 90% said they read newspapers and a large proportion also read books or magazines. Thirty percent felt that their ABE experience helped them in family matters and about three quarters of the respondents said they felt "more in control" of their lives. The survey data also suggested that the acquisition of numeracy skills may be particularly important in giving offenders a sense of self-mastery.

Taken together, the results of the both studies lend new support to the notion that gaining literacy and numeracy skills may be important factors in successful community reintegration.

Can Educating Adult Offenders Counteract Recidivism?

"It is axiomatic that no one in prison is reprehensible and antisocial all of the time. Another way of putting this dictum is that no inmate is impervious to intervention all of the time, and this means that if the "right" overture happens to intersect with the "right" intermission or crisis in even a discouraging chronology, the inmate is likely to resonate and respond." (Toch 1987, p.66).

Correctional programs should have one fundamental goal that transcends variation in content, mode of delivery, duration or theoretical orientation --- to engage the offender in developing a new perspective on the possibility, and the possibly rewarding outcomes, of a non-criminal lifestyle. The acquisition of necessary skills for pro-social living can easily follow this basic change in attitude and outlook.

In this light, the question of whether education for offenders can have an effect on their reintegration into the community is, in a very real sense, a tautology. The process of reintegration, when it catches, is essentially regenerative and educative. A more appropriate question, and one that correctional agencies need to address to more effectively target their program resources, is when and with what types of offenders is educational programming more or less successful?

This paper describes some of the efforts that have been made within the Correctional Service of Canada to study the benefits of our Adult Basic Education programs, both for the general process of reintegration and in reducing the likelihood of recidivism for different categories of offenders.

Research Linking Basic Education With Reduced Recidivism

"It cost the Government half a million bucks to keep me in jail and $450 to teach me to read and write." (Ex-con and former illiterate)

"In a meritocratic society those whose powers of oral and written expression are extremely limited are left at the starting line in the race for social status. It is perhaps difficult for the educated person to appreciate the sense of isolation and inferiority which the illiterate must feel. These people are already cut off in a host of ways from society. Those amongst them who are shut away in prisons are doubly punished." (Palfrey, 1974, p. 84)

School failure in childhood and adolescence is widely accepted by researchers as one of the most consistent precursors of later adult criminality (Loeber, 1991). Whether the opposite is the case, that some success in schooling can contribute to desistance of adult criminality, has certainly not been as convincingly documented.

There have been a number of excellent reviews of the research literature examining evidence for the effectiveness of correctional education programs (Linden & Perry, 1983; Ross, 1978; Ryan, 1991). One author has asserted that this literature is :

"…a mixture of exaggeration, unsubstantiated affirmations, inadequate definition and specification of terms, conflicting statements re incidence, inadequate consideration of confounding factors (e.g., socio-economic class variables) and persisting myth." (Ross, 1978, p. 207)

Without doubt, the measurement of program impact in corrections is fraught with difficulties. Reduced recidivism, as one end-point index that some change has occurred, is clearly inadequate in revealing either the nature or the source of that change. Undoubtedly, the pathways to change for offenders, as with individuals more generally, are multiply determined; change typically arises out of a chain of events, reinforced and supported by improvement in both personal and social circumstances.

Acknowledging that recidivism is only one indirect measure of effectiveness, and that other factors may confound the effects of educational programming, dismissing recidivism altogether as a measure of outcome is clearly inappropriate. Indeed, our own reading of the literature on educational programming in corrections is quite optimistic. It suggests that there is more support than is commonly recognized for the beneficial effects of educational upgrading on recidivism.

From approximately 30 studies on the relationship between educational programming in prisons and recidivism among offenders, our analysis examined seven, which dealt specifically with Basic Education among adult male offenders.

Table 1 presents a brief description of the types of studies completed and their results. It should be noted that all seven studies used relatively sound methodologies. Sample size ranged from 75 offenders to more than 3,000, and the follow-up period for examining recidivism was also substantial (the shortest period being approximately two years). More importantly, we observed that the studies with the soundest methodologies yielded the most positive findings.

At first glance, there appears to be only weak support for the conclusion that participation in Basic Education programs has a positive impact on recidivism - only four of the seven studies found that the recidivism rates of participants were significantly lower than those of the comparison groups. Upon closer examination, however, the results become more convincing: Only three of the seven studies used a random or matched comparison group design, that is, an appropriate and convincing method of selecting their samples. Notably, all three of these studies found that the recidivism rate of participants in the basic education programs was significantly lower than that of the comparison group. We consider this an important finding.

In summary, although our review of the literature was by no means exhaustive, we did find research to support the contention that participation in Basic Education programs by adult male offenders has a positive effect on their recidivism rates.

The Correctional Service of Canada's Approach to Basic Education

We have found a similar effect of reduced likelihood of recidivism in our analysis of participation in Adult Basic Education programming offered by the Correctional Service of Canada.

Education represents a major programming endeavour for the Correctional Service of Canada. In the 1990-91 fiscal year, approximately 44% of inmates who were available for education or work -- a total of 4,364 offenders -- participated in some type of educational program. Educational programs available to inmates include Adult Basic Education, and secondary, vocational, college, and university studies. The total cost of these programs in 1990-91 was approximately $26,300,000 for an annual average full-time equivalent enrolment of 3,231 students.

A large share -- close to one third -- of educational enrolments are in Adult Basic Education (ABE). Another 8% of correspondence registrations are for ABE . Although often referred to as "literacy" training, ABE leads to certification in both literacy and numeracy skills at the grade-eight level. More recently, we have set out to encourage grade-ten completion since this is a prerequisite for many Vocational Training courses.

On admission, offenders are asked to take achievement tests to assess language (reading and writing) and mathematical skills (we have adopted the Canadian Adult Achievement Test). All offenders who score below the grade-eight level are offered the opportunity to enrol in ABE programs. We complement this grade-level testing with formative evaluations assessing functional skills. Moreover, progress is determined through the formative evaluation of teachers and final certification at a given grade level is attained only when local community standards are met (i.e., testing to determine achievement of provincial curriculum standards).

The average inmate who enrols in ABE has grade-six-level skills and spends an average of three to six months achieving grade eight.

Our approach to Basic Education over the past number of years has increasingly adopted a functional view of literacy wherein offenders are regarded as voluntary adult learners who study to acquire literacy and numeracy skills appropriate to their self-defined needs and goals. We attempt to provide, for example, a teacher to student ratio that will allow for individualized attention (one teacher for six students for up to grade five, and one to eight students for grades six to eight). We rely extensively on peer tutors, emphasize variety in currucula and resource materials, provide private learning spaces as much as possible, and support computer-assisted learning for those offenders who are so inclined.

Participation in ABE is voluntary, although we clearly endorse an active motivational approach as our overriding strategy for intervening with offenders. There are constraints in the kinds of work placements, and consequently in the level of pay, for those offenders who refuse basic upgrading. We believe, however, that there are more significant intrinsic incentives that we are tapping into. We "contextualize" ABE , and encourage participation, as part of a much broader programming strategy focussing on the personal development of offenders. In addition to our specialized programming (e.g., for substance abuse or sexual offending), in all of our major institutions there are now a range of Living Skills programs targetting particular cognitive, attitudinal, and interpersonal problems that we see as propelling offenders towards criminal behavior (Fabiano, Porporino, & Robinson, 1991; Fabiano, Robinson, & Porporino, 1990). We believe that this purposeful and mission-driven approach to programming has been successful in "signaling" to offenders that we wish to make a genuine investment in their reintegration. Participation in ABE for many offenders becomes a first step, often serving to eliminate antipathy towards programming more generally.

ABE participation has grown very quickly, beginning in 1986-87 when the Government of Canada announced special measures to combat illiteracy. The Correctional Service of Canada made ABE its priority in educational programming. Now, at any one time, approximately 1,600 offenders, or over 15% of the federal inmate population, are enroled in some type of ABE program. Almost half of ABE participants are full-time students. The remainder combine their enrolment with work or other types of programming. Participation rates are fairly uniform across our five regions.

In the 1990-91 fiscal year, 903 inmates achieved grade-eight levels in language and mathematics through participation in the Correctional Service of Canada's ABE programs. The enrolment to completion ratio for grade eight was approximately four to one (considerably higher at about 2.6 to 1 for grade five), and the cost per completion was $6,151, an increase of about two thousand per completion from 1989-90. Relative to overall participation in educational programming over the last five years, Figure 1 shows how enrolment in ABE programming rose sharply in 1987-88, and has been sustained at a level of about one-third of all educational enrolments.

A special effort has been made to address the educational needs of our native offenders (e.g., by relying on culturally relevant educational materials and activities and incorporating cultural life skills components). With native offenders making up about 10% of the offender population, about 12% of the participation in ABE programs is by natives.

With the infusion of resources into ABE (about $5.6 million spent by the Correctional Service of Canada in the 1990-1991 fiscal year), program planners and correctional administrators are anxious to learn if there is measurable impact on recidivism. Obviously, offenders who participate in ABE increase their language and mathematical skills -- an advantage that Canadians believe every member of society should possess. However, an important question remains: What contribution does ABE make in enhancing the quality and the likelihood of successful reintegration in the community. Despite the programming resources necessary to sustain a high level of participation in ABE , the Correctional Service of Canada had little evidence that ABE was having the desired "correctional" impact.

Figure 1 - Participation Rates for Educational Programming and ABE Full-Time Equivalent Enrolments

ABE Effects on Recidivism

We approached this issue from two perspectives. From a more qualitative angle, we were interested in whether those offenders who participated in ABE , and were now released, would "perceive" and "attribute" benefits to their participation in the program. The methodology we adopted was to conduct semi-structured interviews with a random sample of ABE graduates, who were under conditional release in the community, in two regions of the country. The second more quantitative study we conducted involved examining the postrelease outcomes of a large sample of offenders who participated in ABE programs in 1988. The outcomes of those offenders who had achieved a grade-eight level were compared with those of ABE participants who, for whatever reason, had not completed the program prior to release.

The findings from both our qualitative and more quantitative analysis generally provide encouraging evidence that involvement in Adult Basic Education during incarceration can improve adjustment after release.

Quantitative Analysis of ABE Effects

The sample for our quantitative analysis included a total of 1,736 offenders who participated in ABE during 1988 and were subsequently released to the community. We selected participants from 1988 in order to allow sufficient time to observe whether they would recidivate upon release. All the offenders were monitored from the time they were released until October 1990. The average postrelease follow-up period was 1.1 years. The majority of the offenders in the study (57%) were followed up for at least one year, and only 16% were followed up for less than six months.

A number of inmates who enrol in ABE never complete their programs. Some consciously decide to withdraw from their program of studies; others terminate their participation because they are transferred to another institution; and still others are granted conditional release before they have an opportunity to achieve the grade-eight completion target. In 1987, a total of 3,278 inmates participated in ABE programs run by the Correctional Service of Canada. However, by the end of 1989, only about half of these inmates had completed ABE .

The sample we examined included 899 offenders who successfully completed the grade-eight level, 462 who were released before completing ABE , and 375 who chose to withdraw from the program.

When we compared the characteristics of these released ABE participants with our general population of released offenders (see Table 2), two differences were striking in illustrating the kinds of offenders that ABE appears to involve (either because they are attracted to ABE or because our staff make special efforts to target these individuals). It was clear that ABE participants are disproportionately younger (i.e., 25 or under) and more likely to be on their first federal term. There were no differences in sentence length, nature of offence, native status, or type of release when ABE releases were compared with the general population of released offenders.

Younger offenders on their initial term are typically at higher risk for recidivism. Our first finding, then, supports the fact that our ABE programs are appropriately capturing a higher-risk category of offender.

The readmission patterns that we documented were clearly most favourable for the ABE program graduates. As Figure 2 shows, those who completed the targeted grade-eight level had the best community outcomes: only 30.1% of the graduates were readmitted during the follow-up period, compared with 35.7% of those released before completion, and 41.6% of those who withdrew from ABE . Hence, there was a difference of almost 12% in readmission rates for the group of offenders who had successfully completed ABE and the group who had withdrawn.

Of course, one danger of comparing ABE graduates with offenders who did not complete the program is that the different outcomes observed may be related to differences in group characteristics that have nothing to do with ABE participation. For this reason, the three ABE groups were compared on a number of key characteristics.

The three groups had very similar average sentence lengths (3.5 years) and proportions of inmates who were serving sentences for a violent offence (44%). They were also similar in proportion of native offenders (about 14%). However, the graduates differed from the other two groups in their incarceration history: Only 28% had previous federal sentences, compared with 38% of those who withdrew from ABE and 33% of those who were released before completing the program. As well, the three groups differed in age with about 50% of the graduates and only 40% of those who withdrew being twenty-five years of age or older. Finally, the groups were granted different types of releases. The graduates were much more likely to receive full parole (33%) than were the offenders who withdrew (19%) and those who were released before completion (30%).

Figure 2 - Readmission Rates of ABE Participants^{Footnote *}

These initial differences between the groups suggested that ABE participants who fail to complete the program may be higher-risk offenders than those who do complete it. Normally, a history of previous incarceration is associated with higher chances of reoffending, and cases released under mandatory-supervision generally exhibit poorer community outcomes than those granted full parole. For this reason, it might be argued that the graduates, being at lower-risk to begin with, would be more successful after release regardless of whether or not they had completed ABE . In order to rule out this hypothesis, we adjusted for the initial differences observed between the three ABE participation groups before examining the effects of ABE on postrelease outcomes.

The supplementary analyses indicated that the initial group differences related to risk of recidivism did not account for the link between ABE completion and lower readmission rates. In fact, ABE appeared to have the greatest impact on offenders who could be initially defined as higher risks for reoffending.

These findings are illustrated in Figure 3, which compares the effects of completing ABE for different categories of offenders. We note a consistent effect where completion of ABE is reflected in lower rates of recidivism for offenders serving sentences of three years or more, for those with a prior federal term, for younger offenders, for natives, and for those offenders convicted of a violent offence. The differential effect on recidivism, contrasting completing and withdrawing from ABE , ranges from a high of 18% for violent offenders, 15% for longer sentenced offenders, 12% for younger offenders, 10% for recidivists, and 7% for natives.

The conclusion that involvement in ABE may have a particularly beneficial effect for higher-risk individuals was further supported when we compared cases granted parole with those released under mandatory-supervision (see Figure 4). Among the full parole cases, achieving grade-eight level skills was found to be only marginally associated with lower readmission rates. Whether or not a parolee had completed ABE , had been released before completing or had withdrawn seemed to have little consequence for success in the community. Among mandatory-supervision cases, on the other hand, there appears to be a clear link between ABE completion and later success in the community. Although this group did more poorly overall than the full-parole group, mandatory-supervision cases who had completed ABE showed readmission rates 10% lower than those offenders who had withdrawn from their studies prematurely.

Figure 3 - Readmission Rates of ABE Participants By Selected Characteristics

Figure 4 - Readmission Rates and ABE Participation by Release Type

This finding is consistent with what has been referred to as the "risk" principle, which contends that correctional programming shows the most benefit among cases that are at the highest risk (Andrews, 1989). The finding of positive ABE effects on mandatory-supervision cases gives additional credence to the belief that when we increase the levels of numeracy and literacy in our higher-risk offenders, we may also be successful in motivating expectations of success and reducing recidivism.

Qualitative Analysis of ABE Effects

Another approach to studying the impact of ABE is to survey offenders directly to obtain their views on the benefits of educational programs in institutions. In a recent study conducted in collaboration with Dr. Stephen Duguid of Simon Fraser University and Dr. Joel Tax, we interviewed 38 released offenders who had participated in ABE programs while serving sentences in federal institutions in two regions of the country (Ontario and Pacific). The study was considerably smaller in scale than the readmission study. However, the findings provide some insight into how offenders perceive the usefulness of ABE skills and how they link these skills to various facets of their community adjustment.

The majority of offenders we spoke to noted that they had voluntarily chosen to take ABE (74%). Only 21% indicated they took ABE because they were required to and another 5% more to pass the time. Most offenders were enroled full-time (76%) and many simultaneously enroled in other courses (71%); mostly lifeskills or vocational courses. As well, when asked whether involvement in ABE made them want to take other courses, 42% reported "very much" and another 37% "some". Almost all of the offenders perceived their teachers as very supportive (84%).

In terms of general community adjustment, 79% of the offenders reported being employed full-time, most of them having found a job within one month of release (78%). Since their release, the majority displayed employment stability, having held only one or two jobs.

Table 3 shows some of the sorts of perceived benefits that were noted by the offenders we interviewed.

Approximately 47% of the offenders indicated that ABE helped them "a great deal" to do things they had not been able to do before. About 30% felt that the skills acquired through ABE had helped "very much" in their job search, and another 30% felt that the new skills had helped at least "somewhat." ABE skills were rated as "very useful" in on-the-job matters by about half the offenders. Reading, writing, spelling, and mathematics were all viewed as valuable job skills, although reading and numeracy skills were seen as especially valuable.

Approximately one third of the offenders claimed that what they learned in ABE helped them "very much" with their families and children, and the majority (over 70%) regarded reading as an important skill for entertainment, to acquire news, and for hobby and job information. More concrete evidence that literacy skills were actually being applied is the fact that a substantial number of the offenders reported that they regularly read newspapers, magazines and books (see Table 3).

When the offenders were asked to assess how much positive change had occurred in their lives over the last five years, a significant number (21 %) cited their participation in ABE as "very important" in helping them achieve positive changes (see Table 4). Specifically, offenders strongly agreed that they had made significant personal changes such as:

• when I get mad at someone I am better able to work it out (58%)
• other people think better of me now (58%)
• feel more in control of myself now (74%); and,
• am more concerned about other people's feelings (68%).

An interesting finding of the study -- and one that warrants further inquiry -is the link between the acquisition of mathematical skills and other changes in the lives of the offenders. Those who reported feeling "more in control" of their lives rated the usefulness of mathematical skills very highly (correlation of .65). The ability to manipulate numbers may be even more important than reading and writing skills for the development of a feeling of self-mastery among offenders. Numeracy skills are also probably crucial to survival in today's labour market.

Our research has still not clearly identified which aspects of ABE programs are most helpful to the postrelease community adjustment of offenders. The specific intellectual skills that are gained through ABE training may equip offenders to deal more effectively with daily problems encountered in the community. Moreover, the sense of achievement and confidence that results from successfully completing a program may encourage some offenders to make further positive changes in their lives.

The research conducted to date is only a first step by the Correctional Service of Canada to answer some of the questions about the presumed benefits and effectiveness of ABE programs. Nevertheless, the existing evidence suggests that our current commitment to teaching basic language and mathematical skills should not be diminished.

References

Andrews, D.A. 1989. Recidivism is Predictable and Can be Influenced: Using Risk Assessments to Reduce Recidivism.Forum on Corrections Research, Vol. 1, No.2.

Fabiano, E., Porporino, F., and Robinson, D. (1991) Canada's Cognitive Skills Program Corrects Offenders' Faulty Thinking.Corrections Today, August 1991.

Fabiano, E., Robinson, D., Porporino, F. (1990) A Preliminary Assessment of the Cognitive Skills Training Program: A Component of Living Skills Programming.Correctional Service of Canada.

Glaser, D. (1964) The Effectiveness of a Prison and Parole System.New York: Bobbs-Merrill.

Ingalls, G.R. (1978) The Relationship Between Educational Programs and the Rate of Recidivism among Medium Security Prison Parolees and Mandatory Supervision Cases from Drumhellar Institution in the Province of Alberta.Dissertation, Washingon State University, Department of Education, Ann Arbor, Michigan: University Microfilms International.

Linden, R.; Perry, L. (1983) The effectiveness of prison education programs.Journal of Offender Counselling, Services, & Rehabilitation, 6, 43-57.

Loeber, R. (1991) Risk Factors and the Development of Disruptive and Antisocial Behaviour in Children.Forum on Corrections Research, Vol. 3, No. 3.

Mace, J. L. (1978) Effect of correctional institutions' Education Programs on Inmates' Societal Adjustment as Measured by Post-Release Recidivism.Doctoral Dissertation: West Virginia University.

Palfrey, C. (1974). Remedial education and the adult offender.Howard Journal of Penology and Crime Prevention, 14, (1), 78-84.

Ross, R. R. (1978). Reading and rehabilitation.Crime and Justice, 6, 4, 207-218.

Ryan, T.A. (1991) Literacy Training and Reintegration of Offenders.Forum on Corrections Research, Vol. 3, No. 1

Stevens, R. D. (1986) The effect on recidivism of attaining the general education development diploma.Journal of Offender Counselling. 7, 1, 3-9.

Toch, H. (1987). Regenerating prisoners through education.Federal Probation, September 1987, 61-66.

Walsh, A. (1985) An evaluation of the effects of adult basic education on rearrest rates among probationers.Journal of Offender Counselling Services and Rehabillitation, 9, 4, 69-76.

Zink, T. M. (1970) A study of the effect of prison education on societal adjustment.,22, 2, 18-20.