DEP

ARTMENT OF JUSTICE

OF

FIC

E

OF JUSTICE PRO

GR

AM

S

BJA

N

IJOJJ DP BJS

OV

C

U.S. Department of Justice

Office of Justice Programs

Office of Juvenile Justice and Delinquency Prevention

Shay Bilchik, Administrator July 1999

Reintegration, SupervisedRelease, and IntensiveAftercareDavid M. Altschuler, Troy L. Armstrong, andDoris Layton MacKenzie

Over the past decade, interest in theissue of aftercare for juvenile offendershas grown tremendously. Jurisdictionshave sought new ways to reintegrateyouth being released from confinementinto their communities while also en-suring public safety, and juvenile jus-tice policymakers and professionalshave begun experimenting with after-care and other reintegration models.In the late 1980’s, the Office of JuvenileJustice and Delinquency Prevention(OJJDP) began supporting a long-term,multistage research and developmentinitiative to design an intensive juve-nile aftercare model. The final stages(implementation and testing) of theinitiative, an experimental evaluationof the Intensive Aftercare ProgramModel (IAP) using random assignmentconducted by the National Council onCrime and Delinquency (NCCD), arewell under way.

In “Reintegrative Confinement andIntensive Aftercare,” Dr. David M.Altschuler and Dr. Troy L. Armstrongdescribe the IAP model, distinguish itfrom other models and programs that

have been implemented and assessedwith varying degrees of success, andanalyze individual intensive aftercareprograms. While other aftercare evalu-ations have not all been experimentalin design, the IAP evaluation uses ex-perimental methodology to gauge thesuccess of the four OJJDP-supportedprojects currently implementing theIAP model. Following Drs. Altschulerand Armstrong’s comparative analysis,Dr. Doris Layton MacKenzie provides acommentary that reviews existing af-tercare programs and studies of theseprograms and reacts to Drs. Altschulerand Armstrong’s conclusions. Dr.MacKenzie examines aftercare initia-tives, including OJJDP’s IAP model, inlight of the findings and recommenda-tions of the University of Maryland’sreport entitled Preventing Crime: WhatWorks, What Doesn’t, What’s Promis-ing, which she coauthored. Together,these analyses present a cutting-edgeexamination of what has worked inreintegrating juvenile offenders, whathas not worked, and why.

From the Administrator

If we are to succeed in our effortsto combat juvenile delinquency andrecidivism, it is not sufficient to knowwhat works, or even to implementprograms based on that knowledge.We need to ensure that the juvenilejustice system conducts comprehen-sive front-end assessments ofcourt-involved youth, encompassesa system of immediate and interme-diate sanctions, and provides bothnonsecure and secure community-based programs and facilities.

We must not stop there, however,because the juvenile offenderscurrently placed in secure confine-ment will one day return to thecommunity. Hence, aftercare isessential for youth released fromresidential programs.

This Bulletin describes an intensivejuvenile aftercare model developedfrom a long-term OJJDP researchinitiative and compares it with otherapproaches. An analysis of intensiveaftercare programs is also offered inlight of the publication of the Univer-sity of Maryland report, PreventingCrime: What Works, What Doesn’t,What’s Promising.

This analysis helps us to understandwhat works—and what does not—inreintegrating juvenile offenders intotheir communities. More needs to bedetermined, but this Bulletin is afirst step toward accomplishing thatcrucial goal.

Shay BilchikAdministrator

2

Reintegrative Confinement andIntensive Aftercare

David M. Altschuler and Troy L. Armstrong

As the trend toward confining greaternumbers of juveniles in corrections facili-ties continues (see table 1), increasing at-tention is being paid to what happens oncethey are released back into the community.The “what happens” question frequently isasked in reference to two closely relatedissues. The first is whether released offend-ers will commit additional crimes, particu-larly person offenses, and thereby threatenpublic safety. In fact, one of several motiva-tions for prolonging incarceration is thatconfinement is regarded by some as theprimary way to prevent offenders fromcommitting additional crimes. Implicit inthis view is the belief that incarceration isinsufficient to prevent or deter offendersfrom committing crimes when released. Asecond, and very closely connected, issuecenters on what is being done to ensurethat released juvenile offenders will notcontinue to offend. Because there is somuch uncertainty surrounding the commu-nity adjustment of juvenile offenders afterrelease, some believe the best policy is topostpone release as long as possible.Prolonged incarceration is problematic,however, for several reasons. First, it is ex-ceedingly expensive; second, many juvenileinstitutions are already dangerously over-crowded (see table 2) and space is scarce;and third, its increased use has not demon-strated measurable reductions in juvenilearrests following the release of incarceratedoffenders.

In short, there is a growing interest andneed to learn more about what steps totake to promote law-abiding behavior inthe community by juvenile offenders re-turning from institutions. What can belearned from prior and ongoing researchon corrections sanctioning, supportiveprogramming, and the imposition of socialcontrol techniques when emphasis isplaced during the confinement phaseon linkage with aftercare? What type ofapproach is likely to generate the mostpositive outcome, and how can it be imple-mented? Fortunately, considerable re-search has been conducted on programsthat, to varying degrees and in distinctlydifferent ways, pursue a “reintegrative”form of confinement. Much can be gleanedfrom these programming initiatives andtheir evaluations that can help not onlyto shape the design and development of

future efforts and initiatives, but also toguide their implementation and opera-tions. Reintegrative confinement is definedas an incarceration experience that in-cludes a major focus on structured transi-tion and a followup period of aftercarecharacterized by both surveillance andservice provision in the community.

Transition and postinstitutional cor-rections programming and supervisionhave attracted considerable attentionacross the country, in part because re-search findings tend to indicate thatgains made by juvenile offenders in cor-rections facilities quickly evaporate fol-lowing release.1 Other research findingssuggest that either better outcomes areapparent or the potential for positiveimpact is increased when a highly struc-tured and enhanced transition from cor-rections facilities into the community isimplemented in accordance with certain

specifications.2 An important implicationof these findings is the growing realiza-tion that incomplete, flawed, or highlyuneven implementation cannot producebetter outcomes for participating offend-ers. Stated simply, when requirementsfor implementing the basic program de-sign are not met, success is unlikely.

What kind of requirements are in-volved? In general terms, reintegrativeconfinement emphasizes:

◆ Preparing confined offenders for reen-try into the specific communities towhich they will return.

◆ Making the necessary arrangementsand linkages with agencies and indi-viduals in the community that relateto known risk and protective factors.

◆ Ensuring the delivery of required ser-vices and supervision.

To the extent that these general speci-fications are not met, there is little reasonto expect that reoffending behavior willdiminish or that the overall performanceof youth returning to the community will

1 See, for example, Altschuler, 1984; Altschuler andArmstrong, 1991; Baird, Storrs, and Connelly, 1984;Catalano et al., 1988; Coates, Miller, and Ohlin,1978; Whittaker, 1979.

Table 1: The 1-Day Count of Juveniles Held in Public Facilities Rose 47%From 1983 to 1995

Public Facility 1-Day Count Percentage of Change, 1983–1995

Law violation 48%Delinquency 52

Person 109Violent Index 99

Property –17Drug 95Public Order 87

Status Offense –21Total 47

◆ The increase was not evenly distributed across all offense categories, however. Thenumber of juveniles held for Violent Crime Index offenses doubled. The broader categoryof person offenses (that includes such offenses as simple assault and kidnaping and theViolent Crime Index offenses) more than doubled.

◆ The categories of drug and public order offenses also saw large increases.

◆ In contrast, there was a drop in the number of juveniles held for property crimes and statusoffenses.

Source: Sickmund, M. (1997).Note: Analysis of data from OJJDP’s Children in Custody Census 1982/83 and 1994/95[machine-readable data files].

2 See, for example, Altschuler, 1998; Deschenes,Greenwood, and Marshall, 1996; Fagan, 1990;Greenwood, Deschenes, and Adams, 1993;Goodstein and Sontheimer, 1997; MacKenzie, 1997;Sealock, Gottfredson, and Gallagher, 1995, 1997;Sontheimer and Goodstein, 1993.

3

improve. Accordingly, reintegrative con-finement initiatives must be carefully as-sessed to determine the extent to whichimplementation adheres to a prescribedmodel. In addition, different reintegrationinitiatives must be examined in terms ofthe specific required program elements,components of the elements, and proce-dures. Adherence to a theoreticallyflawed model is no more likely to producea positive result than is low-quality imple-mentation of a sound model.

This Bulletin provides an overview ofwhat has been learned from research andpractice about designing, developing, andimplementing aftercare initiatives thatplace a high priority on reintegrativeconfinement, structured transition, andfollowup in the community. Correctionsapproaches incorporating reintegrativeconfinement are not widespread. Fewof these efforts have been rigorouslyevaluated. However, a small numberof such initiatives have been well docu-mented and analyzed in considerabledetail. There is also related researchand program development work on in-tensive aftercare. Collectively, this knowl-edge base offers important insight andguidance.

It is critical to note that much of the re-cent experimentation with innovative juve-nile aftercare programming has focused onways to develop more effective “intensive”approaches. However, the approaches dif-fer in terms of what “intensive” means andwhat specialized modalities and practicesmust be incorporated programmatically.These differences emphasize a variety ofissues, including anticipated caseload sizeand frequency of contact, classification andassessment procedures, criteria for target-ing youth appropriate for participation inthis kind of intervention framework, andthe respective roles of surveillance andtreatment/service provision activities tomaximize long-term, prosocial communityadjustment and normalization. Questionsinclude:

◆ What constitutes a measurable thresh-old of intensity in terms of supervisionand services?

◆ What range of offender profiles (withregard to delinquent histories and/or special problems/needs) definesthe parameters for referral to theseprograms?

◆ What technology can be brought tobear to better identify and matchclients to effective intervention?

The programs and developmentalwork discussed in this Bulletin representattempts to answer such questions.

In the following pages, the small bodyof research and developmental work onintensive aftercare is briefly reviewed,highlighting both the lessons learned andthe pitfalls experienced. First, however, itis useful to explore the underlying ratio-nale, both theoretical and empirical, thathas led selected jurisdictions across thecountry to focus on intensive juvenile af-tercare. This discussion is followed by adescription and brief analysis of the indi-vidual intensive aftercare initiatives.

The IAP ModelOne model or conceptual yardstick

against which transition and aftercareprograms can be measured is the IAPmodel, developed with OJJDP funding byDrs. Altschuler and Armstrong (1994a,1994b). Its usefulness as a guide for exam-ining program design and implementationrests in its identification of specific pro-gram elements, components of the ele-ments, and services that address what arecommonly regarded as essential aspectsof reintegrative corrections programming(see figure 1). One of IAP’s components,the requirement that both surveillanceand treatment services be provided, hasbeen found relevant to success in both

intensive supervision programs (ISP’s) forprobationers (Petersilia and Turner, 1993;Byrne and Pattavina, 1992) and bootcamps (MacKenzie and Souryal, 1994).Many researchers believe that the suc-cess is related to active, direct interven-tion in the home community and socialnetwork within which the offending origi-nated. This is also where various prob-lems and needs related to family, school,employment, peer group, and drugs sur-face. However, when the response is pre-dominantly, or exclusively, a matter ofoffender surveillance and social control(e.g., drug and alcohol testing, electronicmonitoring, frequent curfew checks, strictrevocation policies) and the treatmentand service-related components are lack-ing or inadequate, the indication is thatneither a reduction in recidivism nor animprovement in social, cognitive, andbehavioral functioning is likely to occur.

Attention is thereby drawn to the ex-tent and nature of both the surveillanceand service components as reflected inthe implementation and day-to-day opera-tion of the aftercare program. Regardingservices in particular, the question iswhether institution-based treatment fo-cusing specifically on “criminogenic” (i.e.,predictive of future criminal activities)needs (see Andrews and Bonta, 1994) iscompatible and consistent with treatmentin the community. Specialized treatment

Table 2: On February 15, 1995, 69% of Public Facility Residents WereHeld in Facilities Operating Above Their Design Capacity

All Public Facilities ResidentsPercentage Held

Percentage in FacilitiesOperating Operating

Above AboveDesign Capacity Total Design Capacity Total Design Capacity

All public facilities 1,080 40% 69, 929 69%Fewer than 31 residents 595 21 8,543 2931–110 residents 324 58 18,506 59111–200 residents 90 63 13,141 66201–350 residents 39 82 10,075 82More than 350 residents 32 88 19,664 91

◆ 40% of public facilities housed more residents than they were constructed tohold—a greater proportion than in 1991 (36%).

◆ The larger a facility’s design capacity, the more likely it was to be operating over capacity.

◆ Small facilities (designed for fewer than 31 residents) accounted for the largest number ofover-capacity facilities.

Source: Sickmund, M., Snyder, H.N., and Poe-Yamagata, E. (1997).

4

in the institution is likely of little long-lasting value if it is not relevant to pressingconcerns in the daily lives of offendersin the community and not carefully andconsistently reinforced in this setting.The lack of such services in either theinstitution or the community is equallydetrimental, because the former offersthe potential for establishing a powerfulfoundation on which to build and the lat-ter offers the potential for transferringnewly learned skills and competencies tothe very community in which the offenderwill reside (see, for example, Altschuler,1984; Altschuler and Armstrong, 1995b;Whittaker, 1979).

Strategies to develop a service struc-ture that spans institution and communityinvolve several major challenges. Allocat-ing sufficient numbers of qualified staffand funds to support service provision atthe level required in both the institutionand community is critically importantand challenging. Cost sharing, leveragingfunds, in-kind contributions, contracting,public-private partnerships, and reallocat-ing portions of existing budgets are someof the approaches that are being used.Developing the organizational capacityand wherewithal to facilitate the consis-tency and compatibility of service deliv-ery between the institution and com-munity is another critical challenge.Strategies designed to foster such com-patibility include bringing into the institu-tion specialized service providers and

agency staff based in the community,providing joint staff training, establishinginteragency case management teams,adopting and tailoring for institutionaluse those practices and approaches thatclosely resemble promising treatment andservice modalities found in the commu-nity programs, and conversely, applyingpromising techniques initially developedfor institutional use (such as anger man-agement or aggression replacement) tocommunity-based programs. In short, theintent is to have community-based after-care services parallel those that are firstinitiated in the institution and institu-tional services geared to achieve essen-tially the same purposes as those that willbe achieved in the community. The keyservice areas around which both the insti-tution and community-based providersneed to organize their respective effortsin tandem are family, peers, schooling,work, and drug involvement (i.e., druguse and drug selling). Program develop-ments in these areas need to be encour-aged by funding support, reflected in or-ganizational policies and procedures, andpromoted through carefully designatedstaff roles and responsibilities, training,and career advancement.

Regarding supervision and control inthe community, a critical question relatesto how various practices can work in con-cert with the required services. Drug andalcohol testing, attendance and curfewchecks, electronic monitoring, and track-

ing are all valuable supervision practicesthat can be used to encourage participa-tion in required services and adherenceto rules and conditions. In fact, close su-pervision and tracking that increase theprobability of detecting noncompliancewith, and nonparticipation in, requiredservices may well discourage lack of co-operation, especially when coupled withgraduated responses. The key is having astrategy to heighten surveillance in a waythat promotes participation in treatment.Such a strategy is essential because re-search suggests that recidivism declinesonly when offenders are simultaneouslyreceiving both supervision and treatment-related services.3 Again, adequate re-sources and organizational ability areclearly necessary to promote the imple-mentation of programs that truly incorpo-rate sufficient levels of supervision andservices.

Research findings repeatedly haveshown that providing high levels ofsupervision to lower risk offenders re-sults in poorer performance, not better.4

One reason frequently cited to explainthis pattern is that intensive supervisiontends to be accompanied by an increasein detected technical violations that, bydefinition in many studies, is one measure

Underlying PrinciplesProgressively Increased

Responsibility & Freedom

Facilitating Client-CommunityInteraction & Involvement

Working With Both Offenders & Targeted Community Support Systems

Developing New Resources,Supports, & Opportunities

Monitoring & Testing Asse

ssm

ent &

Cla

ssific

atio

n

Indi

vidua

l Cas

e Pl

anni

ng

Surv

eilla

nce/

Serv

ice M

ix

Ince

ntive

s &

Cons

eque

nces

Brok

erag

e &

Link

ages

Special Needs & Special Populations

Education & School

Vocational Training, Job Readiness, & Placement

Living Arrangements

Social SkillsLeisure & Recreation

Client-Centered Counseling (individual & group)

Family Work & Intervention

HealthSpecial TechnologyCase ManagementOrganizational

& Structural Characteristics

Management Information

& Program EvaluationProgram Elements

Service Areas

Strain Theory Social Learning Theory Social Control Theory

An Integration of:

Figure 1: Intervention Model for Juvenile Intensive Aftercare

3 See, for example, Byrne and Pattavina, 1992;Gendreau, 1996; Petersilia and Turner, 1993.

4 See, for example, Andrews, 1987; Baird, 1983; Erwinand Bennett, 1987; Markley and Eisenberg, 1986.

5

of program failure. Moreover, when in-creases in technical violations becomethe basis for more revocations andreincarcerations, intensive supervisionactually becomes a contributor to insti-tutional crowding. Another problem isrelated to the lack of evidence indicatingthat technical violations, per se, are pre-dictive of future criminality (see, forexample, Petersilia and Turner, 1991;Turner and Petersilia, 1992). This raisestwo fundamental questions. First, whatis accomplished from the perspectiveof crime prevention and control byreincarcerating technical violators?Second, what is accomplished by impos-ing intensive supervision on offenderswho are already at low risk for reoffend-ing? Another reason why lower risk of-fenders tend to perform poorly whensubjected to intensive supervision is thetendency of some individuals, particu-larly adolescents, to react negatively tothe pressures created by highly intrusivesupervision. Given the negative reac-tions, it appears that intrusive supervi-sion techniques are counterproductiveto the intended goal of supervision. In-sight into the dynamics between level ofsupervision and offender performancehas prompted numerous observers tosuggest that the level of community su-pervision provided be commensuratewith the actual level of risk posed by theoffender in the community.5

Recent JuvenileAftercare Initiatives

As discussed above, there has beenincreasing interest in initiatives that canprovide:

◆ Institutional services that subse-quently lend themselves to applicationand reinforcement in the community.

◆ Highly structured, smooth transitionalexperiences at the point of communityreentry.

◆ Intensive multimodal and multiphasedprogramming during a period ofcommunity-based aftercare.

Unfortunately, few evaluations haveexamined the design, implementation,and impact of these efforts. However,these evaluations are enormously valu-able in identifying strengths and weak-nesses of the programs and in pinpoint-ing how future efforts can build on what

has been learned. The programs arequite diverse, not only in design, but inthe extent to which they have been suc-cessfully implemented and in the natureof their impacts. A critical examinationof these evaluated programs must focuson at least three dimensions:

◆ Are there an identifiable programmodel and a philosophy that specifyhow the program design and strategyare expected to induce change, anddo the model and philosophy makesense?

◆ Did the program actually implement itsmodel and philosophy in accordancewith the requirements?

◆ What kind of impact did the programhave?

The following sections examine fivejuvenile aftercare initiatives. Each pro-gram is briefly described and analyzed,and the basic findings and implicationsare highlighted in the context of the threekey dimensions.

The Philadelphia IntensiveProbation AftercareProgram

The Philadelphia Juvenile ProbationDepartment’s IAP was developed to testthe efficacy of an intensive reintegrationapproach for adjudicated youth in transi-tion from State juvenile corrections facili-ties back into the community. The experi-ment was prompted by the realization thatthe most serious, violent, and habitualsegment of the State’s delinquent popula-tion was being inadequately served. Theinadequacy was due, in part, to the ex-tremely large caseloads of the supervisingprobation officers, the resulting low levelof contact with offenders in both the insti-tution and the community, and a lack ofspecialized resources and services in thecommunity. Using an experimental design,Sontheimer and Goodstein (1993) evalu-ated the program. Eligible incarceratedjuveniles had to have at least one prioradjudication for aggravated assault, rape,“involuntary deviate sexual intercourse,”arson, robbery, or a felony-level narcoticsoffense or at least two prior adjudicationsfor burglary. These juveniles were ran-domly assigned to the program or a

control group that received the customaryaftercare supervision. The participatingjuveniles were predominantly African-American (81 percent). Their average ageat placement was 17.2 years, and they hadan average of five prior arrests—more than90 percent had at least one prior arrest fora felony-level offense—and had spentan average of 10.8 months in confinement.

Additional aftercare probation officerswere hired to work exclusively with IAP of-fenders. Each officer was given a caseloadof no more than 12 youth in the communityand also was responsible for making regularcontact with incarcerated IAP offendersassigned to their caseload. IAP officerswere expected to meet monthly with theconfined offenders, institutional staff,and the parents or guardian in the homeand to prepare a postrelease plan. Afteroffenders were released, officers were givenoperational guidelines on:

◆ Minimum number of contacts with thejuvenile per week, which was to de-cline gradually with satisfactoryperformance by the juvenile.

◆ Minimum number of contacts withparents and collaterals (e.g., school,work).

◆ Contacts during nonbusiness hoursincluding evenings and weekends.

Despite these requirements, implemen-tation fell considerably short in severalkey areas (Sontheimer and Goodstein,1993; Goodstein and Sontheimer, 1997).These shortcomings are not surprising,because program planners did not addresssome fundamental issues related to pro-gram design and philosophy. In fact, after-care staff reportedly received few guide-lines about the philosophy or mission ofthe program (Sontheimer and Goodstein,1993, p. 204):

The program was not defined, forexample, as emphasizing a socialcontrol or rehabilitative perspec-tive. No effort was made to articu-late whether the emphasis of theprogram would be on enhancingfamily ties and prosocial relation-ships, on facilitating educationalor vocational growth, on increasingprobationers’ perceptions of ac-countability through surveillance,or on some other combination ofprinciples assumed to reducecriminality.

Supervising officers were simply giventhe contact requirements and then fol-lowed a relatively traditional casework

5 See, for example, Andrews, 1987; Baird, 1983; Erwinand Bennett, 1987; Markley and Eisenberg, 1986.

6

approach to supervision. These officersalso maintained a reactive, nonindividual-ized approach to noncompliance. As aresult, the quality of probation service de-livery remained unchanged for some timeand contact between officers and juvenileswas problematic during nontraditionalhours (Goodstein and Sontheimer, 1997).Another operational component not incor-porated into the program design and,therefore, absent in program implementa-tion for a substantial period of time was agraduated response capability in the formof incentives and consequences. In termsof staffing and leadership, following anenthusiastic startup period, the six-person intensive aftercare team experi-enced complete turnover. Consequently,many of the participating line staff had nosupervising officers for extended periods.This severe turnover problem—and thedifficulties that both preceded and fol-lowed it—most likely created enormousprogrammatic turmoil and confusion. In-deed, it is hard to imagine how the pro-gram could not have faltered somewhatunder such circumstances. Over the fullcourse of implementation, however, theprogram found its footing and evolvedinto a model that began to incorporatemany of the social control and servicedelivery elements necessary for an effec-tive reintegrative model of incarceration,transition, and aftercare (Goodstein andSontheimer, 1997).

The outcome evaluation of the Philadel-phia IAP employed a classic experimentaldesign with random assignment of cases.The evaluation was based on the perfor-mance of 44 experimental and 46 controlcases. The juvenile offenders in this samplewere released from a single youth correc-tions facility between December 1988 andJanuary 1990 and were tracked until May1990. Thus, the followup period that wasdefined as time following completion ofaftercare ranged from 3 to 17 months, aver-aging 11 months. The study found that theintensive aftercare group exhibited a sig-nificantly lower average number of rear-rests than the control group (1.65 versus2.79) and a significantly lower number offelony arrests (0.41 versus 0.76), but thepercentage of subjects rearrested was thesame (Sontheimer and Goodstein, 1993).In short, the findings indicate that whenroutine aftercare is compared with thereintegrative intensive aftercare imple-mented in Philadelphia, the latter pre-vented participating juvenile offendersfrom incurring multiple arrests and did noworse than the former in the percentageof offenders who were rearrested.

Juvenile Aftercare in aMaryland Drug TreatmentProgram

The Maryland Department of JuvenileJustice received a grant from the Centerfor Substance Abuse Treatment (CSAT) todevelop an aftercare program that wouldcomplement short-term residential treat-ment for chemically addicted or drug-abusing juvenile offenders. Sealock,Gottfredson, and Gallagher (1997, 1995)evaluated this program, which includedonly youth who resided in Baltimore City.Drug-abusing youth from a number ofother Maryland counties were also com-mitted to residential treatment facilitiesby judges, but no aftercare was provided.The residential treatment was to includeAlcoholics Anonymous group sessionsand offer academic courses, recreationalopportunities, vocational education, workassignments, and social activities. After-care for Baltimore City juvenile offend-ers was to consist of three phases. Dur-ing prerelease, the services of a familytherapist were to include assessment, de-velopment of a treatment plan, and fam-ily work. During the intensive phase (thefirst 2 months in the community), staffwere to have daily contact with the par-ticipants and hold youth support groupmeetings and family support sessions.Additionally, an addiction counselor wasto provide individual counseling and in-home family therapy. The final transi-tional aftercare phase was to include atleast two meetings per week with thecase manager, two meetings per monthwith the addiction counselor, and thecontinuation of family support groups. Onan as-needed basis, other community-based services and family therapy wereto be provided.

The evaluation examined both the resi-dential and aftercare experience. Youthin aftercare treatment, all of whom werefrom Baltimore City, varied substantiallyin several characteristics from those whoreceived residential treatment and no af-tercare, all of whom were from outsideBaltimore. The group from Baltimore Cityhad a higher percentage of nonwhites, ex-hibited a greater number of prior offenses,were younger at first referral to the justicesystem, and had offense histories that re-flected greater involvement with drugs and

more property offenses than comparisonyouth. The residential treatment servicesprovided were found to be highly unevenin scope and quality. Although assess-ments for drug problems and drug educa-tion took place, much less happened inrelation to encouraging family participa-tion, providing family therapy, conductingpsychological assessments, and holdingindividual counseling sessions. In fact, theevaluators found that most of the residen-tial program’s intermediate goals (e.g.,increasing coping skills, internal control,family communication) were not realized.Further, it was noted that a 2-month resi-dential treatment program might not havebeen sufficient for a youthful, drug-involved population (Sealock, Gottfredson,and Gallagher, 1997). Finally, althoughsome positive effects were observed dur-ing residential treatment, the evaluatorsfound that this component required addi-tional strengthening (Sealock, Gottfredson,and Gallagher, 1995).

Curiously, the evaluators also foundthat youth who received aftercare treat-ment spent less time in the residentialdrug treatment program (by 12 days) thanyouth in the comparison group, who alsoexperienced some additional residentialplacements (Sealock, Gottfredson, andGallagher, 1997). According to the re-searchers, this finding suggests that theaftercare services actually may havereplaced more expensive, and possiblymore effective, residential treatmentservices. The problem related to familyparticipation is again noted in relation toaftercare, but at this point, of course, theyouth were back in the home directly in-teracting with family. Additionally, al-though the average number of weeks spentin the intensive phase of aftercare was 33,not the 8 weeks planned, the average juve-nile had only 29 contacts with staff, lessthan one per week, nowhere close to thedaily contact envisioned by the model.Finally, attrition became a significant prob-lem. Of the 162 juveniles who began theaftercare prerelease phase, only 54 enteredthe transitional phase, and of those, only36 entirely completed the aftercare phases.In short, few of the aftercare clients re-ceived much aftercare, and for those whodid, the quality and nature of the servicesprovided were highly suspect.

The evaluation found that aftercare ser-vices of the quality and intensity deliveredin the Maryland program were not benefi-cial (Sealock, Gottfredson, and Gallagher,1997). Specifically, aftercare clients had nofewer alleged or adjudicated offenses

7

offenders were confined in one of theState of Michigan’s training schools foran average of 17.1 months. Early releaseplayed no role in the effort. The averageage at first arrest was 14.4, and the par-ticipants averaged 2.5 prior arrests.More than half of the Detroit participantswere known to be drug dealers, nearlyhalf had drug use problems, and the cur-rent offense of slightly more than halfwas a crime against persons. In Pitts-burgh, a privately run wilderness pro-gram with an average length of stay of10.2 months was used for this experi-ment. The average age at first arrest was14, and the participants averaged 4.6prior arrests and 3.7 adjudications. Theircurrent offenses were mostly propertycrimes. The study found no differencebetween experimental and controlgroups in the proportion of youth ar-rested, self-reporting of offenses, or druguse during a 12-month followup period.

Equally important, youth in the experi-mental programs did not participate anymore frequently in educational or workactivities than did control group youth.Also, most of the families viewed delin-quency as the youth’s personal problemand were not interested in making majorchanges in their own behavior or activi-ties. Further, in neither of the two sitesdid the aftercare program have a signifi-cant effect on the youth’s associationswith delinquent peers. In the Detroit pro-gram, which was characterized by longerlengths of stay and no possibility of earlyrelease, no savings were apparent in resi-dential placement costs. Consequently,the aftercare program simply producedan overall increase in cost per placement.In Pittsburgh, where reduced time in resi-dential placement was an explicit part ofthe program, total placement costs wereslightly reduced.

Given the absence of any impact onthe participation of the experimentalgroup in school and work, family involve-ment, and delinquent peer associations,there is little reason to expect loweredrecidivism. Greenwood and colleagues(1993) took the position that a number offactors explain the results, including:

◆ Aftercare workers provided only gen-eral support and assistance, ratherthan targeting specific problems thatwere contributing to risk.

◆ Aftercare workers did not devotesufficient attention to programmingthat addressed risk factors relatedto delinquent behavior, for example,

overall than youth in the comparisongroup, meaning that there was no differ-ence in the level of reoffending in general.Aftercare clients were, however, adjudi-cated delinquent for more drug offensesthan those in the comparison group, andthere was no evidence indicating a pro-gram effect related to increasing familysupervision or communication, reducingfamily violence, decreasing health prob-lems, or increasing problem-solving skills.On the positive side, aftercare clients com-mitted significantly fewer new crimesagainst persons than their counterparts inthe comparison group.

The Skillman IntensiveAftercare Project

Two experimental intensive aftercareprograms for chronic delinquents in De-troit and Pittsburgh were evaluated byGreenwood, Deschenes, and Adams(1993). Over 2 years, approximately 100juveniles completing residential place-ments in each city were randomly as-signed to either intensive aftercare orregular supervision. The programs weredeveloped and operated by two separateprivate providers committed to theSkillman program model that emphasizedfive components:

◆ Prerelease contacts and planning in-volving the assigned aftercare case-worker, the youth, and the family,beginning at 3 months before release.

◆ Intensive supervision contacts in thecommunity, starting at several per dayand gradually diminishing.

◆ Assistance in family stabilization.

◆ Mobilization of supportive communityresources, particularly in relation toeducation and jobs.

◆ Enlistment of role-modeling, motivatedcaseworkers.

In terms of actual implementation, thetwo programs differed in the timing ofthe youth’s release from placement, theintensity of implementation, and thesanctions that could be imposed. Theprofiles of participating offenders alsovaried. In the Detroit program, juvenile

substance abuse treatment and angermanagement.

◆ Deployment of a surveillance/caseworkapproach was inappropriate, particu-larly given the kind of problems andhigh level of temptations encounteredby these youth after they returned totheir home communities.

◆ More formal methods of assessingongoing needs and progress wereneeded, including drug testing, reportsby third parties, or tests of specificskills.

The Michigan NokomisChallenge Program

The Nokomis Challenge Program wasstarted in 1989 by the Michigan Depart-ment of Social Services (DSS) as an alter-native placement to traditional custodialsettings for medium- and low-risk juve-niles. Instead of placement in a long-termresidential facility, the program offered3 months in a remote, 40-bed wildernesschallenge facility, followed by 9 monthsof aftercare surveillance and treatment.The model called for a three-stage wilder-ness challenge experience that includedorientation and assessment, challenge,and community survival. The model em-phasized cognitive/behavior training, so-cial and survival skills, and family work.During the residential stage, the youthand their families were to be seen every2 weeks at the facility by a communitytreatment worker (CTW), who was alsoexpected to meet with the family once aweek in the community. CTW’s were pro-vided under contract by seven differentprivate agencies located across the State.

The aftercare component included anumber of distinct phases, each withseparate tasks and goals for the youthand family. Reentry into the communitywas marked by virtual house arrest for30 days. During the initial 3 months, theminimum level of contact required ofthe CTW was three contacts per weekwith the youth, including one with thefamily. The program placed great em-phasis on family participation in thetreatment process, with the CTW acting

8

as the primary family worker. The pri-vate agencies also provided communitycontact workers (CCW’s) to conductsurveillance, initially at least threetimes per day with the youth and onceper day with the family.

Using a quasi-experimental design,Deschenes, Greenwood, and Marshall(1996) evaluated this programming effort.Participation in the Nokomis Challenge Pro-gram was limited to adjudicated youth whowere 14 years of age or older. The majority(64 percent) of the juveniles targeted forNokomis were African-American youth whowere approximately 14 years old at the timeof their first arrest, with an average of threeprior arrests plus two prior adjudications.Their average age at the time of placementin the program was 16.5 years. Roughly 29percent of the offenses committed by theyouth entering Nokomis were crimesagainst persons. Thirty-seven percent wereproperty crimes, 16.5 percent were drug-related offenses, and 17.5 percent wereother types of crime. Approximately 20 per-cent of youth participating in the experi-ment were known gang members; 42 per-cent were drug dealers; and 55 percentwere drug users.

The outcome evaluation was based onan analysis of 97 youth in the NokomisProgram and a comparison group of 95youth in a traditional residential pro-gram. The evaluation focused on basicsocial adjustment and familial function-ing and also on criminality and drug use.Although both the Nokomis and compari-son group participants showed somepositive changes in coping mechanismsduring the residential period, bothgroups had experienced setbacks by the24-month followup. In the area of familyfunctioning, the evaluation revealed a gen-eral decline in both groups at 24 months.Arrest records indicated no difference be-tween the groups at 24 months in the over-all proportion with a new felony arrest;however, the Nokomis youth self-reportedless involvement in drug sales than didthose in the comparison group. Overall,the self-reported frequency of substanceuse declined slightly from intake to 24-month followup, but there was no differ-ence between the two groups.

What might explain the overall similar-ity in impact? One possibility is suggestedby the finding that Nokomis participantsonly received formal substance abusetreatment during the residential phaseand that, compared with traditional resi-dential programs, the alternative programapparently offered less family counseling.

Even so, the families of youth in the ex-perimental program were no worse offthan the families of youth in traditionalresidential care.

Nokomis also encountered a substan-tial problem in successfully retainingparticipants during the first 12 months(including residential and communityphases). A staggering 60 percent of youthin Nokomis were either transferred to orplaced in another custodial program dur-ing the first year, and an additional 10 per-cent were rearrested during the second 12months of the 24-month study period. Incontrast, only 16 percent of the partici-pants in the traditional residential pro-gram (where length of stay averaged 15.5months) did not successfully completethe program; 14 percent were rearrestedduring the remaining months in the 24-month study period. Deschenes, Green-wood, and Marshall (1996) conclude thatthe main weakness in Nokomis was re-lated to the community phase, which isthe ultimate test of any sanction or dispo-sition. During that phase, youth in thetraditional residential program were rear-rested at about the same rate. It shouldalso be noted that the initial 3 months ofresidential placement in Nokomis couldwell be regarded as relatively short-term,particularly since that was the only timespent by offenders in drug treatment.Deschenes, Greenwood, and Marshall(1996) conclude that, regardless of theintervention, youth who were releasedback into the same environment faced thesame difficulties in readjusting to thecommunity setting without relapse. Theresearchers recommend strengtheningthe community phase, particularly withreference to treating substance abuse,improving family functioning, and target-ing younger juveniles.

OJJDP’s IntensiveAftercare Program

Since 1987, OJJDP has been fundingresearch and development activities inthe area of intensive juvenile aftercare.A decade ago, growing concerns aboutcrowding in juvenile corrections facilities,high rates of recidivism, and escalatingcosts of confinement prompted OJJDP toexamine the juvenile aftercare philosophyand practice and to explore options forreform. As originally formulated, theprogram had four stages:

◆ Assessing programs currently in op-eration or under development andreviewing the relevant research andtheoretical literature.

◆ Developing a program prototype(model) and related policies andprocedures.

◆ Transferring the prototype design toa training and technical assistancepackage.

◆ Implementing and testing the proto-type in selected jurisdictions.

Initiated as a research and developmentproject conducted by the Johns HopkinsUniversity Institute for Policy Studies incollaboration with the Division of CriminalJustice at California State University atSacramento, the IAP project culminatedin a four-State national demonstration de-signed to test a model of intensive after-care developed by this Bulletin’s authors.The four pilot programs are as follows:

◆ Colorado. The IAP project in Colorado isoperated by the State Division of YouthCorrections (DYC), Department of Insti-tutions and serves parts of Arapahoe,Denver (including greater metropolitanDenver), and Jefferson Counties. Thesite benefits from its proximity to thejuvenile offenders’ home communities.Only 18 miles from downtown Denver,Lookout Mountain Youth Services Cen-ter (LMYSC) is a secure facility whoseresidents include the most serious andviolent delinquent youth in the DYC sys-tem. LMYSC houses IAP participants in asingle cottage.

◆ Nevada. The Division of Nevada YouthCorrections Services’ Parole Bureauoperates the State’s IAP project. ClarkCounty, which has the greatest concen-tration of serious juvenile offenderscommitted to State confinement, wasselected as the pilot site. The 150 milesbetween the offenders’ home commu-nity of Las Vegas and the CalienteYouth Center, the participating youthcorrections facility, presented a signifi-cant challenge to implementing the IAPmodel.

◆ New Jersey. New Jersey’s IAP projectfocuses on high-risk youth from Camdenand Essex (Newark) Counties. Theseyouth are incarcerated in a single cot-tage at the New Jersey Training Schoolfor Boys (NJTSB) in Jamesburg. FromNJTSB, IAP participants are moved intoaffiliated residential centers in the twocounties that provide a stepdown transi-tion for community reintegration.

◆ Virginia. The Intensive Parole Program(IPP), Virginia’s IAP project, is designedfor chronic offenders who have beencommitted to the Beaumont JuvenileCorrectional Center by the Norfolk

9

Juvenile and Domestic Relations Court.A Norfolk Youth Network CommunityAssessment Team (CAT) handles all IPPcases. CAT works with parole officers,offenders, and offenders’ families toidentify treatment, service needs, andagencies that can address problems.

For a more detailed description anddiscussion of these sites, see Altschulerand Armstrong (1995b, 1996, 1997). De-tails of program eligibility and selectionare found in table 3.

These demonstration projects followed7 years of research, development, and train-ing activity and are presently in the midstof their third year of operation, with theexception of the New Jersey site, which hasbeen discontinued due to implementationdifficulties related to restructuring andsystem reform.

The IAP model currently being testedis theory-driven, risk and needs assess-ment based, and empirically grounded

(Altschuler and Armstrong, 1995a, 1994a,1994b, 1994c, 1991). The model empha-sizes the identification, preparation, transi-tion, and reentry of “high-risk” juvenileoffenders from secure confinement backinto the community in a gradual, highlystructured, and closely monitored fashion.Consequently, it can be viewed as a formof reintegrative confinement. A multifac-eted and integrated approach to commu-nity reentry, the IAP model requires anoverarching case management process

Table 3: IAP Eligibility and Selection

Eligibility Criteria Colorado Nevada Virginia

Legal status Committed Committed Committed

County of residence Denver, Arapahoe, Clark (Las Vegas) City of NorfolkJefferson

Facility placement Lookout Mountain Caliente BeaumontHanover (since 3/97)

Risk of reoffending High risk High risk High risk

Gender Males Males Males

Age 12–18 12–18 13–18 (16–18prior to 3/97)

Excluded offenses None Sex offenders Murder, rape, arson(with determinantcommitment to age 21)

Excluded conditions Severe mental health Severe mental health Pending charges orproblems; developmental or medical problems. sentence in adult court;disabilities. potential rescinded

commitment; severemental health or sub-stance abuse problems;prior IAP.

Location and timing At separate diagnostic While in local At separate diagnosticof selection facility; after completion of detention; prior to facility; after 60-day

30-day assessment and assessment and classifi- assessment/classificationclassification process and cation process. (IAP process and facilityfacility placement decision. selection determines placement decision.

facility placement.)

Number of youth,randomized to 11/30/98

IAP 82 104 76Control 68 108 45

Total 150 212 121

Source: Weibush, McNulty, and Le, 1998.

IAP Site

10

that guarantees substantial control overreleased juvenile offenders and enhancedservice delivery focusing on recognizedrisk and protective factors. To reduce thelevel of recidivism and relapse, the IAPmodel also requires that working collabo-rations be forged across diverse profes-sional and agency boundaries.

A number of previous research and pro-gram development efforts have developedframeworks for intervening with seriousand chronic juvenile offenders (Elliott andVoss, 1974; Elliott, Huizinga, and Ageton,1985; Weis and Hawkins, 1981; Fagan andJones, 1984), but these projects have gener-ally not directed much attention to the spe-cial structural and systemic problems thatmust be confronted in devising strategiesthat will enable high-risk offenders to makea successful transition back into the com-munity. Distinctive to the IAP model is thefocus on the numerous issues and concernsarising from the mostly disconnected andfragmented movement of offenders fromcourt disposition to juvenile authority and/or institution, to aftercare supervision anddischarge. Consistent with this approach, anumber of principles for programmatic ac-tion have been identified and incorporatedas a foundation for the IAP model:

◆ Preparing juveniles for progressivelyincreased responsibility and freedomin the community.

◆ Facilitating interaction and involve-ment between juveniles and thecommunity.

◆ Working with offenders and targetedcommunity support systems (families,peers, schools, employers) on thosequalities needed for constructive inter-actions that advance the juveniles’reintegration into the community.

◆ Developing new resources and supportservices as needed.

◆ Monitoring and testing the capacity ofjuvenile offenders to receive—and thecommunity to provide—services andsupport.

The demonstration programs have beengiven flexibility to structure and apply theIAP model within local contexts, as longas the program meets certain specifica-tions. Many of these requirements revolvearound the IAP design for overarchingcase management. It is this dimension ofthe model that defines how clients areidentified for particular levels and typesof supervision, how clients can be trackedthrough the system without fallingthrough the cracks, and how specific

techniques can aid in the provision ofsupportive activities and sanctioningmeasures necessary for client supervisionin the community. The requisite compo-nents of case management are:

◆ Risk assessment and classification forestablishing eligibility.

◆ Individual case planning that incor-porates a family and communityperspective.

◆ A mix of intensive surveillance andenhanced service delivery.

◆ A balance of incentives and gradu-ated consequences coupled with theimposition of realistic, enforceableconditions.

◆ Service brokerage with communityresources and linkage with socialnetworks.

To date, the demonstration sites havebeen engaged in selectively fine-tuningand elaborating certain components andfeatures in their particular program ap-plications. The major challenge has beenthe need to adapt the generic IAP modelto the specific problems, needs, and cir-cumstances of the individual jurisdic-tions. As a group, all have identified andacted on the following programmingstrategies vital to following the basicframework of the model:

◆ Defining the overall aftercare functionin a fashion that guarantees the inclu-sion of staff and program componentsacross the entire continuum, from thepoint of judicial commitment and resi-dential placement to the terminationof community supervision (see table 4).

◆ Designing the network of community-based services in a way that respondscomprehensively to the problems andneeds of serious and chronic juvenileoffenders.

◆ Devising a framework for case man-agement that ensures continuity ofsupervision and service delivery,matches clients with appropriateinterventions, and brings the mostobjective procedures to informdecisionmaking in the areas of riskand need.

◆ Focusing on collaborative, interagencyapproaches to supervision and serviceprovision.

The IAP initiative has been funded toinclude an independent evaluation that in-corporates random assignment using anexperimental design. The evaluation, whichis being conducted by the National Council

on Crime and Delinquency (NCCD), in-cludes both process and outcome dimen-sions. Because IAP participants have onlyrecently begun to be discharged from after-care, outcome results involving substantialnumbers of participating youth are not yetavailable.

Comparative Analysisof the Five AftercareProjects

Heavily focused on design and imple-mentation issues, the initial IAP researchand development project sponsored byOJJDP was planned to build on the existingknowledge base in the field about inten-sive aftercare. From this starting point, theOJJDP project was to propose a testablemodel that would include clear guidelinescovering program principles, components,and features that appear most promisingfor reducing rates of recidivism among tar-geted youth making the transition frominstitutional confinement back into thecommunity. The factfinding, model devel-opment, and implementation work thathas been part of the IAP initiative has asits goal identifying and incorporating pre-cisely those factors vital to success. Fromits inception, the IAP project was con-ducted with the idea of building on theexisting knowledge base in the youth cor-rections field about juvenile aftercare. Ofcourse, the final word on the effectivenessand suitability of the IAP model awaits theresults of the experimental design,multisite outcome evaluation.

The five projects presented in thisBulletin are being widely discussed inthe field. This review has highlighted theirbasic design, the status of their imple-mentation, and, if known, the outcomes.Table 5 takes the IAP model and uses it tocapture critical design and implementationfeatures of these five projects.6 The tableenumerates the essential components andprocedures that characterize IAP opera-tions. They are organized within thephases of the aftercare continuum (i.e.,institution, transition, community). In eachphase are listed the specific characteris-tics being used as criteria for comparisonacross projects. In addition, there areseparate headings regarding implementa-tion and evaluation issues. The fourprojects are listed side-by-side on the topof the table. IAP is used as the baseline(represented by the enumeration of

6 For a detailed description of the IAP model, seeAltschuler and Armstrong, 1994b.

11

characteristics) against which compari-sons are being made.

In light of the discussion of the IAPmodel, the five aftercare programs dis-cussed here, and the information in table5, a number of key issues and challengesfor program policy, design, implementa-tion, and evaluation become evident.

The implementation of juvenile aftercareprogramming is still in its infancy. Therehave been notable omissions in both pro-gram design and operation, yet there havealso been some positive results. Althoughthe overall picture is mixed, the evaluationsand the IAP research and developmentwork clearly point to reforms and changesthat are needed. Recommended reformsand changes are highlighted below.

First, community-based aftercareis one part of a reintegrative corrections

continuum that must be preceded byparallel services in the correctionsfacility and must include careful prepa-ration for the aftercare to follow. Institu-tional services need to be geared to theservices, opportunities, and challengesthat exist in the community to which thejuvenile will return. The institution orresidential corrections facility cannot op-erate in isolation from aftercare and thecommunity. Institutional services that areinadequate, inconsistent, incompatible, ordisconnected in relation to what will beencountered in the aftercare communityare likely of little long-term value. Addi-tionally, high-quality institutional servicesare likely of little value if they are notcarefully reinforced and followed up inthe aftercare community. Accordingly,aftercare is only one phase of the correc-tions process. The development, imple-

mentation, and evaluation of aftercarerequire equal attention to what occursduring the institutional and transitionalstages of corrections jurisdiction. Thechallenge is that institutional correctionsis often highly resistant to change andopposed to interference from the “out-side.” Institutional programming has be-come reintegrative in numerous in-stances, but it typically requires strongleadership from the top and a commit-ment to developing a working partnershipbetween the institution, community cor-rections, and the judiciary.

Second, aftercare is frequentlyfunded and staffed at levels far belowwhat is required to provide truly inten-sive supervision and enhanced servicedelivery. The community aftercare

Table 4: IAP Management and Staffing*

Component Colorado Nevada Virginia

Administrative agency Colorado Division of Nevada Youth Parole Virginia Department ofYouth Corrections Bureau Juvenile Justice

Program coordinator DYC Community Services Clark County Parole Parole Services ManagerCoordinator Unit Manager (central office)(central office) (local office)

Primary IAP staff

Institution IAP Institutional/ • 2 IAP Case Managers†Community Liaison†

3 IAP Client ManagersCommunity • 2 IAP Case Managers • 3 IAP Parole Officers

• 2 Field Agents • Parole Aide†• Parole Unit Manager• Education Liaison

Other key staff • Cedar Cottage Treatment • “B” Cottage Manager • Reception/DiagnosticTeam Coordinator • IAP Data Coordinator Facility IAP Case

• 4 Group Leaders Manager• 1–3 MSW Interns • Data Coordinator• IAP Researcher†

IAP staff/client ratio‡

Institution Liaison: 1/22 (in) Case Manager: 1/15 (in)Client Managers: 1/18(18 = in + out)

Community Parole Officer + Agent: Parole Officer 1/152/20 (out) (in + out)

Source: Weibush, McNulty, and Le, 1998.Note: “in” = in the institution; “out” = in the community.* Data current as of 5/31/98.† The position is funded by OJJDP through the IAP grant.‡ Staff/client ratios shown are based on program design.

IAP Site

(continued on page 15)

12

Table 5: Juvenile Aftercare Matrix

PROGRAM DESIGN

Institutional Phase

Prerelease Community sources Family therapist Aftercare caseworker Planning forplanning offer input via assesses, diagnoses, commences contacts community reentry

probation officers. develops family with youth and family is initiated 30 daysPreparation of contract, and begins 3 months prior to after placement inpostrelease plan. weekly family release. residential phase.

group sessions.(Highly unevenimplementation.)

Involvement of Probation officers Family visits facility Not indicated. Parents meet withoutside agencies meet with at least once; therapist confined children,and individuals institutional staff involves youth in institutional staff,in institution and juveniles. family assessment and a community

session. (Less than worker once everyhalf of youth involved 2 weeks.in family assessmentsession.)

Targeted Probation officers Family attends weekly Not indicated. Communitycommunity meet with parents on group sessions with workers seeactivities during regular basis in the therapist and support parents onceconfinement community. groups. (Low family per weekperiod involvement.) at their home.

Transitional Phase

Testing and Not indicated. Not indicated. Not indicated. Not indicated.probing of reentryprior to placementin community

Structured First 6 weeks with Initial period of intense First few weeks after Initial month ofstepdown very high level of contact, followed by release from facility virtual house arrest.process using probation officer/ lesser contact with case with carefully pre- Level of communityresidential client contact. No manager, addiction scribed program. worker/clientplacement or use of intensive day counselor, and family Average number of contact alsointensive day treatment or short- therapist. No use of monthly contacts was high during firsttreatment term residential intensive day treatment. 10 over 6 months in 3 months. No use

treatment. Detroit; 60 over 6 of intensive daymonths in Pittsburgh. treatment or short-Contacts tapered off term residentialafter the first 2 months. treatment.Pittsburgh uses atransition group home.

Philadelphia MichiganProgram/Study Intensive Maryland Skillman NokomisCharacteristics Probation Aftercare Intensive Challenge

(IAP Model) Aftercare Program Aftercare Project Program

13

Philadelphia MichiganProgram/Study Intensive Maryland Skillman NokomisCharacteristics Probation Aftercare Intensive Challenge

(IAP Model) Aftercare Program Aftercare Project Program

Table 5: Juvenile Aftercare Matrix (continued)

Community Followup

Provision of Few prescribed Wide spectrum of Efforts to improve A variety of re-multimodal activities but some services offered with family functioning quired program-treatment emphasis on education links made to com- through counseling ming activities.services and vocational munity resources. and to link clients Some major

activities. with education questions aboutprogram. Jobs fell far quality of delivery.short of expectations.

Discrete case Required procedures Three articulated levels Not highly developed. Not emphasized.management neither highly de- of intervention: pre-services veloped nor clearly release, initial inten-

articulated. sive aftercare, andtransitional aftercare.

Use of graduated Not indicated. Not indicated. Not indicated for Not indicated.sanctions and incentives. Pittsburghpositive sanctions permittedincentives return to group or

wilderness program.

Provision of Thirty percent of Not indicated. Not indicated. Supplemental sur-supervision and contacts by probation veillance activitiessurveillance officers required to provided by spe-beyond ordinary occur outside normal cialized commu-working hours office hours. nity workers.

Reduced caseload Aftercare caseload Caseload size unknown. Caseload size of 6. Aftercare caseloadsize/increased of 12 youth under Clients had 3.2 average Experimental group of 10 youth. Higherfrequency of community super- monthly contacts received far more level of contactclient contact vision versus standard during aftercare or contacts than control for supervision,

70–120. Far higher 32.4 contacts over group. treatment, andlevel of contact than approximately 10 surveillance.usual. months.

Multistage Procedures for Intensive stage of after- Contacts tapered Framework anddecompression gradual, phased care was 33 weeks long, off over time in procedures for aprocess reduction in not 8 weeks as planned; aftercare. diminishing level

level of imposed youth had less than 1 of supervision andcontrol during contact per week on control during6 months of after- average. During aftercare.care supervision. transitional phase of

aftercare, clients metwith case managersless than once every3 weeks on average.

Table continues on next page.

14

Philadelphia MichiganProgram/Study Intensive Maryland Skillman NokomisCharacteristics Probation Aftercare Intensive Challenge

(IAP Model) Aftercare Program Aftercare Project Program

Table 5: Juvenile Aftercare Matrix (continued)

IMPLEMENTATION

Designated No special procedures None indicated. None indicated. No specialprocedures to or activities. procedures orfacilitate full activities.implementation

Documentation Research team Researchers studied Implementation Evaluator observedand tracking of assessed quality and implementation through studied through youth program activities,implementation extent of implemen- client interviews, and staff interviews, administeredprocess tation through selec- official records, staff program records, and questionnaires, and

tive interviews of staff, interviews, and official record data. interviewed clientsclients, and parents. tracking forms. and parents.

Extent of Evaluators determined All three phases of Mixed results. Mixed results in allintended that program ran aftercare suffered from program sectors.implementation smoothly only in later serious implementation See programachieved months. deficiencies, and most summary for more

objectives of the short- details.term residentialprogram were not met.

OUTCOME EVALUATION

Research design Experimental. Nonexperimental with Experimental. Quasi-experimentaluse of a comparison with effectivegroup. matching

procedure.

Target population Male delinquents Drug-involved juveniles Chronic offenders. Chronic seriouscommitted to State committed to residen- male delinquentsyouth corrections and tial facilities with after- committed toexhibiting chronic care compared with State youthhistories of severe drug-involved youth corrections.criminality. committed to facilities

without aftercare.

Sample size 90 cases: 44 in 162 youth entered pre- 99 cases in Detroit: 192 cases: 97 inexperimental group release aftercare; of 50 in experimental experimentaland 46 in control these, 54 entered group and 49 in control group and 95 ingroup. transitional aftercare; group. control group.

of these 36 completedaftercare. Recidivism: 87 cases in Pittsburgh:120 in aftercare and 46 in experimental132 in comparison group and 41 in controlgroup. group.

Significant Yes. Generally no, though No. No.findings slightly mixed.favoring inten-sive aftercare1

1See program summary for details.

15

portion of reintegrative confinementcannot be accomplished “on the cheap.”Employing sufficient staff who are well-trained, closely supervised, profession-ally qualified, and personally committedis an absolute requirement for effectiveaftercare. Resolving issues related tofamily, peers, education, employment,and substance abuse requires knowl-edgeable individuals who have the requi-site competency in these areas and arewilling to go “that extra mile” in problemsolving. Whether through the mecha-nisms of partnering with other publicagencies or through contracts with pri-vate organizations, there can be nodoubt that a sustained response in theareas of family, education, employment,and substance abuse must be a funda-mental part of the corrections responsein juvenile aftercare.

Third, intensive aftercare, in contrastto “standard” aftercare, requires closeattention via formal assessment proce-dures to determine which offenders arein need of a level of intervention thatincludes both highly intrusivesupervision and enhanced treatment-related services. Identifying which seg-ment of the incarcerated juvenile offenderpopulation is most likely to recidivate is akey to successful intensive aftercare pro-gramming. This approach ensures alloca-tion of limited resources to those juvenileoffenders who are most at risk of recidi-vating and who frequently fail unlesshighly structured, intensive community-based interventions are deployed whenthey are released from confinement. Tar-geting the appropriate group also pre-cludes the possibility of applying thesekinds of stringent and highly intrusivetechniques to lower risk offenders, whohave been shown not to benefit from theimposition of such corrections strategies.

Fourth, it is clear that a reduction incaseload size and an intensification inlevel of contacts are widely acceptedoperational principles for intensive af-tercare programming. Yet, “more” con-tact with staff is not necessarily a mea-sure of more productive interaction,since a higher level of contact in itselfreveals virtually nothing about what ishappening during these important peri-ods of contact. Further, specific guide-lines and policies about the nature andpurpose of increased contact are criticalif these interactions are to have positive,longer term impact. Tied to intensificationof supervision is a need to incorporate a

graduated response capability, in termsof both administering sanctions and pro-viding incentives during the communityphase of these programs. Given the high-risk potential of this identified offenderpopulation, it is inevitable that the re-quired increased level of contact will re-sult in the detection of technical viola-tions and program infractions. Theavailability of graduated incentives tominimize the frequency of violations andgraduated consequences to respond pro-portionately and appropriately to mis-conduct is critical.

Finally, it would be a misreading ofthe research discussed above to con-clude that it is not possible to craft a

workable model of reintegrative con-finement, which necessarily includesaftercare as its final phase. Movementtoward reintegrative confinement withinthe youth corrections system is occur-ring, but much remains to be accom-plished. Examples of success—in termsboth of implementation and of outcome—can readily be identified. However, falsestarts also abound, characterized by inad-equate and poorly articulated frameworksthat seem to lend themselves to unevenimplementation. Nationwide support foreffective juvenile aftercare programmingis growing. Policymakers must seek inno-vative reforms in the juvenile justice sys-tem to promote effective aftercare. ■

Commentary: The Effectiveness ofAftercare Programs—Examining theEvidenceDoris Layton MacKenzie

The first essay in this OJJDP Bulletin, byAltschuler and Armstrong, critically evalu-ates recent juvenile aftercare initiatives andpresents a proposed model for an effectiveaftercare program. The authors review theaftercare initiatives by asking whether theprogram had an identifiable philosophy,whether it was implemented in line withthis philosophy, and what impact the pro-gram had on the participants.

This commentary assesses what isknown about juvenile aftercare programsbased on a report entitled PreventingCrime: What Works, What Doesn’t, What’sPromising (Sherman et al., 1997). Thereport added an additional dimensionto the examination of new initiatives—namely, whether there is sufficient evi-dence to conclude that such initiativesare effective in preventing crime. Thereport weighed both the scientific meritand the outcomes of the research todraw conclusions about the effectivenessof the programs in reducing recidivism.Juvenile aftercare programs must beevaluated on the basis of the scientificevidence. The question addressed inthis commentary is whether there isevidence that aftercare programs ofthe type proposed by Altschuler andArmstrong are effective in reducing therecidivism of juveniles.

Crime Prevention andWhat Works

The 104th Congress directed the Attor-ney General to provide a “comprehensiveevaluation of the effectiveness” of themoney given in grants from the U.S. De-partment of Justice to State and localcommunities. In 1997, a research team atthe University of Maryland prepared theabove-cited report. The research teaminvestigated the effectiveness of crimeprevention programs in seven differentinstitutional settings: communities, fami-lies, schools, labor markets, places (spe-cific premises), police, and criminal jus-tice. The report, referred to as “TheMaryland Report,” assessed effectivenessby weighing the strength of the scientificevidence.

While traditional crime prevention ef-forts are directed toward people who arenot yet involved in crime, the broaderdefinition adopted in The Maryland Re-port includes any setting that reducescrime in the community. By definition,therefore, programs in the courts andcorrections that focus on reducing thecriminal activities of adult and juvenileoffenders were considered crime preven-tion efforts. The chapter on criminal jus-tice settings examined interventions thatfocus on six different potential methodsfor reducing crime in the community:

16

◆ Be implemented in a way that is appro-priate for the participating offendersand that uses therapeutic techniquesknown to work (for example, the pro-gram must be delivered as designed,and treatment must be provided byappropriately educated and experi-enced staff).

◆ Require offenders to spend a reason-able length of time in the program con-sidering the changes desired (deliversufficient dosage).

◆ Give the most intensive programs tooffenders who are at the highest riskfor recidivism.

◆ Use cognitive and behavioral treat-ment methods based on theoreticalmodels such as behaviorism, sociallearning, or cognitive behavioral theo-ries of change that emphasize positivereinforcement contingencies for pro-social behavior and are individualizedas much as possible.

The question is: How closely do theseprinciples of effective treatment apply tothe model of juvenile aftercare proposedby Altschuler and Armstrong? Most nota-bly, none of these principles refer directlyto the reintegration focus of their model.The principles of rehabilitation summa-rized above give little guidance on whetheran emphasis on reintegration will be moreeffective than other types of programs. Ithas not been shown that recidivism will bereduced by the emphasis on “preparingconfined offenders for reentry into the spe-cific communities to which they will re-turn” (p. 2), and by “making the necessaryarrangements and linkages with agenciesand individuals in the community thatrelate to known risk and protective fac-tors” (p. 2). On the other hand, thecomponents of the treatment process ad-vocated by Altschuler and Armstrong aresupported by the meta-analyses research.As Altschuler and Armstrong argue, pro-grams must be implemented in a mannerthat is consistent with the design of theprogram (i.e., have therapeutic integrity)and provide sufficient time in treatmentto permit change to occur.

Treatment Programsfor JuvenileDelinquents

Juvenile crime is often serious andmay represent a significant proportion ofthe total criminal activity in a community.It is usually assumed that adolescentsdeserve and require special handling

7 Andrews and Bonta, 1994; Andrews, Bonta, and Hoge,1990; Andrews et al., 1990; Palmer, 1975; Gendreau andRoss, 1979, 1987.

incapacitation; deterrence; rehabilitation;community control; structure, discipline,or challenge programs; and combinationsof rehabilitation and control. The assess-ment of the model of aftercare proposedby Altschuler and Armstrong that followsdraws on the findings of The MarylandReport on the effectiveness of juvenileprograms in reducing the recidivism ofdelinquents.

Judging the Scientific MeritThere is an enormous body of criminal

justice literature on crime preventionefforts. However, little of this literatureexamines the impact of crime preventionstrategies. Instead, much of the researchdescribes different types of programsand the manner in which they are imple-mented. The research that does exist of-ten is of such poor quality that it does notpermit one to draw conclusions regardingthe effectiveness of the program studied.

The scientific standards for inferringcausation have been clearly establishedand can be used to evaluate the strengthof evidence included in each programevaluation. The Maryland Report used ascale of 1 to 5 to summarize the scientificrigor of the studies examined. The scoresgenerally reflect the level of confidencethat can be placed in an evaluation’s con-clusions about cause and effect, with ascore of 5 indicating the strongest evi-dence and a score of 1 considered so lowin scientific rigor that the results wereexcluded from conclusions about a topic.Studies were evaluated by determiningtheir scientific merit and the outcomes.The scientific method scores reflect thestrength of the evidence about the effectof the programs on recidivism. The out-comes (direction and size of the effect)were evaluated based on differences be-tween the treatment group, which re-ceived the intervention, and the controlor comparison group, which did not re-ceive the intervention.

A large body of research on correctionsprogramming for juveniles is in agreementwith Altschuler and Armstrong. However,the quality of much of this research is dis-appointingly poor. Many of the studiesonly describe the program being evaluatedand give recidivism rates for the partici-pants without providing any informationon the rates for a comparable group of ju-veniles who did not participate. Therefore,it is impossible to draw conclusions aboutthe impact of the program. Other researchattempts to make comparisons betweendifferent groups of participants and

nonparticipants. However, the research isso poorly designed (a score of 1 or 2 onthe Maryland scale) that it is impossible torule out alternative explanations for theoutcome results.

CorrectionsRehabilitationand Treatment

While there is still some debate aboutthe effectiveness of rehabilitation (e.g.,Lab and Whitehead, 1988; Whitehead andLab, 1989), recent literature reviews andmeta-analyses provide strong evidencethat rehabilitation programs can effec-tively change offenders.7 This body of lit-erature can guide the examination of whatworks in corrections programming forjuveniles. In general, reviews of the litera-ture show positive evidence of treatmenteffectiveness (Andrews et al., 1990). Forexample, in a series of literature reviews,the proportion of studies reporting posi-tive evidence of treatment effectivenessvaried from near 50 percent to 86 percent.In reviewing these studies, Andrews andcolleagues conclude, “This pattern of re-sults strongly supports exploration of theidea that some service programs areworking with at least some offenders un-der some circumstances” (1990:372).From this perspective, the importantissue is not whether something works,but what works for whom.

Some approaches to treatment are bet-ter than others. Psychological research-ers emphasize that effective treatmentprograms must follow some basic prin-ciples (Gendreau and Ross, 1979, 1987;Cullen and Gendreau, 1989). Recent meta-analyses found that effective correctionstreatment programs follow these basicprinciples (Lipton and Pearson, 1996;Andrews et al., 1990). It appears that tobe effective in reducing recidivism, treat-ment programs must:

◆ Be carefully designed to target thespecific characteristics and problemsof offenders that can be changed intreatment (dynamic characteristics)and that are predictive of future crimi-nal activities (criminogenic character-istics), such as antisocial attitudesand behavior, drug use, and angerresponses.

17

because they are in a formative periodand criminal behavior at this stage of lifewill not necessarily be continued intoadulthood. Therefore, rehabilitation hasparticular appeal for use with juveniles.Theoretically, rehabilitation is the focusof corrections programs for juveniles. Inpractice, however, as occurs with adultprograms, juvenile rehabilitation pro-grams may be poorly implemented.Strengthening implementation of existingrehabilitation and delinquency preventionprograms could substantially reduce fu-ture criminality.

Broad assessments of the effectivenessof delinquency treatments have greatlybenefited from the rise of meta-analysis,in which researchers aggregate the con-tinuously growing research literature toexamine and compare the effect sizes(magnitude of differences betweengroups) for comparisons of treatment andcontrol groups. The most extensive meta-analysis examining the effectiveness ofjuvenile delinquency programs was con-ducted by Lipsey (1992), who examined443 different research studies.8 Lipsey’sanalysis focused on interventions ortreatments designed to reduce, prevent,or treat delinquency or antisocial behav-ior problems similar to delinquency. In64.3 percent of the studies he examined,the treatment group did better (in mostcases this finding refers to a reduction inrecidivism) than the control group. Con-sidering all treatment program studiescombined, 45 percent of those who re-ceived treatment were expected to recidi-vate, in comparison with 50 percent of thenontreated control group. In more de-tailed analyses, Lipsey worked to identifythe characteristics that were most impor-tant in determining differences betweentreatment and control groups. The moreeffective programs were predicted to re-duce recidivism substantially (once themethodology effects were controlled for).For instance, as compared with a 50-percent recidivism rate for the controlgroup, only 32 to 38 percent of the juve-niles who were given employment and

multimodal or behavioral programs wereestimated to recidivate.

Overall, the results of Lipsey’s meta-analysis indicated that more effectiveprograms:

◆ Provided larger amounts of meaningfulcontact (treatment integrity) and werelonger in duration (more dosage).

◆ Were designed by a researcher or hadresearch as an influential componentof the treatment setting.

◆ Offered behavioral, skill-oriented, andmultimodal treatment.

There was also evidence that moreeffective programs targeted higher riskjuveniles, but this difference was smalland nonsignificant. On the other hand,treatment in public facilities, custodialinstitutions, and the juvenile justice sys-tem was less effective than other alter-natives, suggesting that treatment pro-vided in community settings may bemore effective. If this effectiveness is theresult of increased linkages with agen-cies and individuals in the community,then Lipsey’s work supports the pro-posed emphasis on reintegration in theAltschuler and Armstrong model. How-ever, it is also possible that other factorsmay be important. Lipsey himself cau-tions that the conclusion that treatmentin community settings is more effectivecannot be separated from the differencesin the intensity (number of meetings,length of time in treatment) and needs amore refined breakdown before definiteconclusions can be drawn.

The programs that were effective werethose that were either provided by the re-searcher or implemented in treatment set-tings where the researcher was influential.This may indicate that treatment deliveredor administered by the researcher was bet-ter implemented than typical programs,supporting Altschuler and Armstrong’spoint that poor implementation of a soundtheoretical model is unlikely to produce apositive outcome.

Examining theResearch on JuvenilePrograms

Although the literature reviews and themeta-analyses provide strong evidence ofthe effectiveness of rehabilitation pro-grams, they give little information aboutthe specific characteristics of the effec-tive programs. The Maryland Report re-viewed two types of juvenile programs:

8 This was a more extensive analysis than previousmeta-analyses, which had focused on delinquents inresidential programs (Garrett, 1985) and treatment ofadjudicated delinquents (Gottschalk et al., 1987; White-head and Lab, 1989). Although the conclusions fromthese analyses differed, all yielded a positive mean ef-fect of about the same order of magnitude (one-fourthto one-third of a standard deviation superiority for thetreatment group outcome compared with the controlgroup outcome). See also the early discussion of theAndrews et al. (1990) meta-analysis in this Bulletin.

wilderness/challenge-type programs andcommunity supervision.

The wilderness or Outward Bound-type programs were particularly popularfor juveniles during the late 1970’s andearly 1980’s. These programs empha-sized physical challenge and requiredparticipants to do more than what theybelieved they could do. Assessment ofthese programs is relevant to conclu-sions about the effectiveness of aftercarebecause most of the programs includedsome type of aftercare. Outcome evalua-tions of these programs have been ex-tremely rare (Gendreau and Ross, 1987).The Maryland Report identified fourprogram evaluations that receivedscores of 2 or higher on the Marylandscale: the Greenwood and Turner (1987)study of VisionQuest; the Deschenes,Greenwood, and Marshall (1996) studyof the Nokomis Challenge Program in theMichigan Department of Social Services;the RAND research examining the effec-tiveness of the Paint Creek Youth Centerin southern Ohio (Greenwood and Turner,1993); and the Castellano and Soderstrom(1992) study of the Spectrum program inIllinois.

Overall, these studies of wildernessand challenge programs produced mixedresults. The VisionQuest participantshad significantly fewer arrests (39 per-cent) than the control group (71 percent)(Greenwood and Turner, 1987). TheNokomis participants had significantlymore arrests (48 percent) than the con-trol group (23 percent) (Deschenes,Greenwood, and Marshall, 1996). PaintCreek youth had fewer official arrests(51 percent) than control group youth(61 percent), but they self-reported moreserious offenses (75 percent) than thecontrol group (62 percent), although nei-ther of the comparisons was statisticallysignificant (Greenwood and Turner,1993). Spectrum youth did not differfrom control group youth in recidivism(Castellano and Soderstrom, 1992).

Although several of the studies werewell designed, problems that arose in theresearch with the small number of sub-jects, attrition, and study implementationlimit the conclusions that can be drawnabout the effectiveness of the programsin preventing crime. The studies ofVisionQuest and Spectrum were evaluatedas 2’s on the Maryland scale, making ithard to draw any conclusions from theresults. The remaining two programs wereevaluated as 3’s on the Maryland scaleand, thus, of reasonable scientific merit.

18

The one program that included both astrong research design and a reductionin recidivism was Paint Creek (althoughthe reduction was not statistically signifi-cant). Interestingly, this program followedmany of the principles proposed byAndrews and colleagues (1990). High-riskyouth were targeted for participation inthe intensive program, which used a cogni-tive/behavioral mode of treatment. How-ever, problems with the research designseverely limited the study’s potential fordetecting differences, even if the PaintCreek program had been effective. Theother programs targeted individuals atlower risk for recidivism (Nokomis, Spec-trum), were of short duration (Spectrum),were less behavioral in treatment philoso-phy, or focused on noncriminogenic fac-tors such as physical challenge (Spec-trum). Thus, from the perspective of TheMaryland Report, studies of the wildernessand challenge programs do not provideevidence that they are effective in reduc-ing future criminal behavior.

These programs attempted to providereintegration services to the participants.As a result, the mixed aftercare findingswere disappointing. For example, Nokomiswas designed to focus on relapse preven-tion. The youth were expected to spendless time in the residential facility but alonger time in community treatment thanthe comparison youth in the trainingschools. However, the study of the programimplementation revealed that the aftercarephase of Nokomis failed to provide manyof the expected treatment programs. Theyouth received limited substance abusetreatment, and the control group youth hadmore family counseling than the treatmentgroup.

The Paint Creek Youth Center alsosought to provide reintegration services.The center’s small size, problem-orientedfocus, cognitive/behavioral methods, fam-ily group therapy, and intensive commu-nity reintegration and aftercare werepromising features. However, many of thePaint Creek youth were dismissed fromthe program and sent to the trainingschool. Thus, it is difficult to draw con-clusions about the impact of the reinte-gration and aftercare provided, becausemany of the youth did not receive the fullPaint Creek program.

Community Supervisionand Aftercare for Juveniles

A majority (53 percent) of adjudicatedjuvenile delinquents are given probationwhile just 28 percent are placed outside

the home. Those knowledgeable aboutjuvenile corrections increasingly arguefor aftercare and transitional services forjuveniles who are incarcerated. In sup-port of this position, two of the recentmeta-analyses (i.e., Andrews et al., 1990;Lipsey, 1992) suggest there will be greaterreductions in recidivism if treatment isprovided in community settings instead ofin institutions. However, when Lipsey andWilson (1998) studied serious juvenileoffenders, they found no difference in re-cidivism for offenders who received inter-ventions administered in institutionscompared with offenders who receivedinterventions in the community. Nationalsurveys of intensive supervision and af-tercare programs for juveniles completedduring the 1980’s revealed that few pro-grams had been evaluated (Armstrong,1988; Krisberg et al., 1989). Additionally,the evaluations that had been completedwere severely limited in scientific rigor.An exception to this is the Violent Juve-nile Offender Study implemented byOJJDP (Fagan, Forst, and Vivona, 1988).This study found that the group that re-ceived the additional aftercare or super-vision did not have significantly lowerrecidivism rates.

Most recent studies of communityprograms have focused on the increasedsurveillance and restraint aspects andnot on the enhanced services of the pro-grams. It is important to distinguish be-tween increases in control, surveillance,and/or restraints (more contacts withsupervising agents, urine tests, elec-tronic monitoring) and rehabilitation,treatment, and services (meetings forcounseling, drug treatment, family coun-seling, employment training). Whereassome of the programs enhance services,most of the research is designed to com-pare increased surveillance and control,not the services provided. The treatmentand surveillance components of programscannot be untangled, and because the re-search designs focus on surveillance, theoutcomes indicate the effectiveness or,conversely, the ineffectiveness of surveil-lance and control rather than of rehabili-tation. Additionally, when treatment in-tegrity is examined, few differences arefound between the experimental programand the control in either the services de-livered or the impact on risk factors.

The Maryland Report identified sixstudies that compared the recidivism ofjuveniles in intensive supervised proba-tion or parole (ISP) with control groupsthat received other community options:

◆ Land, McCall, and Williams (1990)examined the North Carolina CourtCounselors’ Intensive ProtectiveSupervision Project.

◆ Weibush (1993) compared the perfor-mance of youth on intensive supervi-sion with comparison groups of youthon probation and parole.

◆ Sontheimer and Goodstein (1993)examined an intensive aftercare pro-gram for serious juvenile offenders inPennsylvania.

◆ In two studies, Minor and Elrod (1990,1992) examined the impact of anenhanced treatment program forjuveniles on intensive and moderatelevels of supervision.

◆ Greenwood, Deschenes, and Adams(1993) studied the Skillman after-care program in Michigan andPennsylvania.

The results of these and other analysesare shown in table 6. In general, the re-searchers found no significant differencesbetween the youth receiving ISP and theyouth in the comparison groups. Therewas no consistency in the studies regard-ing which group did better in the commu-nity; sometimes the ISP youth had lowerrecidivism, and sometimes thecomparison group did. Only Land andcolleagues (1990) and Sontheimer andGoodstein (1993) found any significantdifferences between the ISP group andothers. Land and colleagues found thatISP youth, mostly status offenders with noprior delinquent offenses, committedfewer delinquent offenses than the con-trol group. Sontheimer and Goodsteinfound that ISP juveniles had significantlyfewer rearrests than parolees.

Several studies identified problemswith the implementation of the pro-grams. For example, Sontheimer andGoodstein (1993) found that the actualcontacts between youth and supervisingagents were substantially fewer than themandated number and that there was alarge turnover of staff. This turnoverwould be expected to create turmoil foryouth participants and result in unevenstaff training and limited accountability.Combined with an unclear program mis-sion, the turnover led the researchers toquestion whether the unsatisfactory pro-gram results indicated problems in theimplementation of the program treat-ment components, rather than in theprogram’s potential achievement.

19

Table 6: Studies of Juvenile Community Supervision and Recidivism Showing Scientific MethodsScore and Findings

Study Scientific Methods Score Findings1

Land, McCall, and Williams 5 ISP youth (mostly status offenders) with no prior delinquent(1990) offenses had fewer delinquent offenses (12%) than control

group (28%) [S].

ISP youth with prior delinquent offenses had moredelinquent offenses (57%) than control group (33%) [NS].

Weibush (1993) 3 ISP youth had more felony complaints (51%) thanprobationers (38%) but fewer than parolees (57%) [NS].

ISP youth had more adjudications (77%) thanprobationers (62%) but fewer than parolees (78%) [NS].

Sontheimer and Goodstein 5 ISP juveniles had fewer rearrests (50%) than parolees(1993) (74%) [S].

Minor and Elrod (1990) 2 ISP group had more self-reported criminal and statusoffenses [NS].

Minor and Elrod (1992) 2 ISP group had fewer status offenses but more criminaloffenses (68%) than control group (67%) [NS].

Barton and Butts (1990) 5 ISP juveniles had more charges, but control group hadmore serious charges [NS].

Greenwood, Deschenes, and 5 Detroit: Aftercare group (22%) had more arrests thanAdams (1993) control group (18%) [NS].

Pittsburgh: Aftercare group had fewer arrests (49%)compared with control group (48%) [NS].

Gottfredson and Barton (1993) 4 Institutionalized juveniles had fewer arrests thannoninstitutionalized juveniles [S].

1NS, not significant; S, significant.

Similarly, Greenwood and colleagues’(1993) examination of what the Skillmanprograms provided for the youth indi-cated that in comparison with the con-trol group, the aftercare group did notparticipate more in education or workactivities, had little family support, anddid not associate less with delinquentpeers. Thus, despite the fact that theprogram was designed to promotechanges in these risk factors, there waslittle evidence of such change. As wasfound in the previous meta-analyses of

rehabilitation, it appears that the pro-gram did not have the required treat-ment integrity to bring about thechanges in the risk (criminogenic) fac-tors associated with criminal behavior.

The studies listed above comparedISP programs in specific communitieswith other community alternatives. TheMaryland Report examined two studiesdesigned to compare the recidivism ofthose who spent time in communitysupervision with others who had spenttime in training schools: the Barton and

Butts (1990) study comparing treatmentin an inhome ISP program with commit-ment to traditional training schoolsand the Gottfredson and Barton (1993)study comparing commitment to atraining facility with management in thecommunity.

A comparison of those who spendtime in a facility with those who aremanaged in the community is importantbecause the youth who remain in theirown community would be assumed tohave increased contact with agencies

20

and individuals in that community.One of the arguments Altschuler andArmstrong make for aftercare is the im-portance of reintegrating the juvenilesinto their community: “making arrange-ments and linkages with agencies andindividuals in the community that relateto known risk and protective factors,”and “ensuring the delivery of requiredservices and supervision” (p. 2). TheBarton and Butts and Gottfredson andBarton studies provide important in-sight into whether youth who remain inthe community actually receive the de-sired benefits.

Barton and Butts (1990) found thatalthough ISP groups had more charges,the mean seriousness of the controlgroup’s charges was greater; however,the differences were not significant.Gottfredson and Barton (1993) foundthat the recidivism rates of juvenileswho had spent time in the training facil-ity were significantly lower than thoseof the comparison group. The compari-son group was not intensively super-vised, and there is little informationabout what services they may have re-ceived in the community. Gottfredsonand Barton (1993) conclude that youthin the institution most likely receivedmore services and treatment than thosein the community.

SummaryThe Maryland Report’s review of the

juvenile wilderness and challenge pro-grams concluded that these programswere not effective in reducing the recidi-vism of juveniles. Although some ofthese programs did attempt to enhancethe aftercare and reintegration phases,as Altschuler and Armstrong noted, theactual implementations of these phasesmay have been flawed.

Most of the studies comparing youthin ISP with youth in the community re-veal no significant difference betweenthe experimental group and the controlgroup. In part, this finding reflects thesmall number of subjects in each study;there is little basis for detecting any dif-ferences between the groups. Only twoof the studies (Land, McCall, and Will-iams, 1990; Sontheimer and Goodstein,1993) found lower recidivism rates forthe experimental groups. The Land andcolleagues study findings were not en-tirely positive because it was also foundthat youth in the experimental groupwith prior delinquent offenses commit-

ted more delinquent offenses than thecontrol group. The ISP groups in thesetwo studies received more services thanthe comparison groups; thus, the impor-tant aspect may be the amount of reha-bilitation and services, rather than thesurveillance, received by the juveniles.This interpretation supports Altschulerand Armstrong’s assertion that it is im-portant to include appropriate treat-ment during the aftercare phase. How-ever, it is impossible, at this point, tountangle the effects of treatment, sur-veillance, and reintegration services,because the control groups in thesetwo studies received less of all of thesecomponents than the experimentalgroups.

The Gottfredson and Barton (1993)study showing that juveniles who spenttime in an institution had lower recidivismthan those released to the communitysuggests that the quality and amountof treatment the juveniles receive maybe the important factor in reducing re-cidivism. That is, it is not whether thistreatment is delivered in an institutionor in the community, but how much andwhat type of treatment the juvenilesget, no matter where they are located.Again, the research design does not per-mit formation of conclusions about theeffectiveness of aftercare.

Taken as a whole, there is sufficientevidence from the studies cited aboveto conclude that some combination oftreatment methods, with or without sur-veillance in the community, is effectivein reducing the recidivism of juveniles.Whether this treatment must be pro-vided in the community is unclear. Simi-larly, it is unclear whether the increasedsurveillance of the juveniles in the com-munity adds anything to the impact oftreatment and rehabilitation.

The original question posed for thiscommentary was: Is there evidencethat the type of aftercare proposed byAltschuler and Armstrong will be effec-tive in reducing the recidivism of juve-niles? Given the limited amount of qual-ity research, it is difficult to answer thequestion. Certainly, there is sufficientevidence to conclude that effective pro-grams must include rehabilitation andservices to address the needs of indi-vidual juveniles. The research does notpermit conclusions about where thistreatment ought to be delivered. It isalso impossible to draw any conclusionsabout the effectiveness of many of the

programs, because they were not imple-mented as they were designed. This dif-ficulty is clear both in this review exam-ining the scientific merit of the researchand in Altschuler and Armstrong’s re-view of the programs. Problems with theimplementation of programs must beovercome if juvenile justice profession-als are to design effective programs andstudy them. The strong random assign-ment study that is currently in progressto examine the sites where Altschulerand Armstrong’s aftercare model hasbeen implemented is encouraging. Thisstudy should provide information aboutwhether such multifaceted approachesto aftercare and reintegration are effec-tive. The next step will be to untanglethe effects of different program compo-nents to identify the particular compo-nents that are most successful in reduc-ing recidivism. ■

ReferencesAltschuler, D.M. 1984. Community rein-

tegration in juvenile offender program-ming. In Violent Juvenile Offenders: AnAnthology, edited by R. Mathias, P. DeMuro,and R. Allinson. San Francisco, CA: Na-tional Council on Crime and Delinquency,pp. 365–376.

Altschuler, D.M. 1998. Intermediatesanctions and community treatment forserious and violent juvenile offenders. InSerious and Violent Juvenile Offenders: RiskFactors and Successful Interventions, editedby R. Loeber and D. Farrington. ThousandOaks, CA: Sage Publications, Inc.

Altschuler, D.M., and Armstrong, T.L.1991. Intensive aftercare for the high-riskjuvenile parolee: Issues and approachesin reintegration and community supervi-sion. In Intensive Interventions With High-Risk Youths: Promising Approaches in Juve-nile Probation and Parole, edited by T.L.Armstrong. Monsey, NY: Criminal JusticePress, pp. 45–84.

Altschuler, D.M., and Armstrong, T.L.1994a. Intensive Aftercare for High-RiskJuveniles: An Assessment. Report. Wash-ington, DC: U.S. Department of Justice,Office of Justice Programs, Office of Juve-nile Justice and Delinquency Prevention.

Altschuler, D.M., and Armstrong, T.L.1994b. Intensive Aftercare for High-RiskJuveniles: A Community Care Model. Sum-mary. Washington, DC: U.S. Departmentof Justice, Office of Justice Programs,Office of Juvenile Justice and DelinquencyPrevention.

21

Altschuler, D.M., and Armstrong, T.L.1994c. Intensive Aftercare for High-RiskJuveniles: Policies and Procedures. Sum-mary. Washington, DC: U.S. Department ofJustice, Office of Justice Programs, Office ofJuvenile Justice and Delinquency Preven-tion.

Altschuler, D.M., and Armstrong, T.L.1995a. Aftercare in the juvenile justicesystem: New trends and programs. Per-spectives 19(3):24–35.

Altschuler, D.M., and Armstrong, T.L.1995b. Managing aftercare services fordelinquents. In Managing DelinquencyPrograms That Work, edited by B. Glickand A.P. Goldstein. Laurel, MD: AmericanCorrectional Association, pp. 137–170.

Altschuler, D.M., and Armstrong, T.L.1996. Aftercare not afterthought: Testingthe IAP model. Juvenile Justice 3(1):15–22.

Altschuler, D.M., and Armstrong, T.L.1997. Reintegrating high-risk juvenile of-fenders from secure correctional facilitiesinto the community: Report on a four-state demonstration. Corrections Manage-ment Quarterly 1(4):75–83.

Andrews, D.A. 1987. Implications ofclassification for treatment of juveniles.Paper presented at American Probationand Parole Association Meeting in SaltLake City, Utah.

Andrews, D.A., and Bonta, J. 1994. ThePsychology of Criminal Conduct. Cincin-nati, OH: Anderson Publishing Company.

Andrews, D.A., Bonta, J., and Hoge, I.1990. Classification for effective rehabili-tation: Rediscovering psychology. Crimi-nal Justice and Behavior 17(1):19–52.

Andrews, D.A., Zinger, I., Hoge, R.D.,Bonta, J., Gendreau, P., and Cullen, F.T.1990. Does correctional treatment work?A clinically-relevant and psychologically-informed meta-analysis. Criminology28(3):369–404.

Armstrong, T. 1988. National survey ofjuvenile intensive probation supervision:Juvenile intensive probation supervision,Parts 1 and 2. Criminal Justice Abstracts20(2,3):342–348, 497–523.

Baird, S.C. 1983. Report on IntensiveSupervision Programs in Probation andParole. Washington, DC: National Instituteof Corrections.

Baird, S.C., Storrs, G.M., and Connelly,H. 1984. Classification of Juveniles in Cor-rections: A Model Systems Approach. Wash-ington, DC: Arthur D. Little, Inc.

Barton, W.H., and Butts, J.A. 1990.Viable options: Intensive supervisionprograms for juvenile delinquents. Crimeand Delinquency 36(2):238–256.

Byrne, J.M., and Pattavina, A. 1992. Theeffectiveness issue: Assessing what worksin the adult community corrections sys-tem. In Smart Sentencing: The Emergenceof Intermediate Sanctions, edited by J.M.Byrne, A.J. Lurigio, and J. Petersilia. Thou-sand Oaks, CA: Sage Publications, Inc.,pp. 281–303.

Castellano, T.C., and Soderstrom, I.R.1992. Therapeutic wilderness programsand juvenile recidivism: A program evalu-ation. Journal of Offender Rehabilitation17(3/4):19–46.

Catalano, R.F., Wells, E.A., Hawkins, J.D.,and Jenson, J.M. 1988. Transition and after-care services for adjudicated youth. In AGuide to Developing Substance Abuse Treat-ment Programs for Adjudicated Youth, ed-ited by L. Mixdorf, M. Goff, and P. Paugh.Laurel, MD: American Correctional Asso-ciation, pp. 163–189.

Coates, R.B., Miller, A.D., and Ohlin,L.E. 1978. Diversity in a Youth CorrectionalSystem: Handling Delinquents in Massachu-setts. Cambridge, MA: Ballinger PublishingCompany.

Cullen, F.T., and Gendreau, P. 1989. Theeffectiveness of correctional rehabilitation:Reconsidering the “Nothing Works” de-bate. In The American Prison: Issues in Re-search and Policy, edited by L. Goodsteinand D. MacKenzie. New York, NY: PlenumPress.

Deschenes, E.P., Greenwood, P.W., andMarshall, G. 1996. The Nokomis ChallengeProgram Evaluation. Santa Monica, CA:The RAND Corporation.

Elliott, D.S., Huizinga, D., and Ageton,S.S. 1985. Explaining Delinquency and DrugUse. Boulder, CO: Behavioral ResearchInstitute.

Elliott, D.S., and Voss, H. 1974. Delin-quency and Dropout. Lexington, MA: D.C.Heath and Co.

Erwin, B.S., and Bennett, L. 1987. NewDimensions in Probation: Georgia’s Experi-ence With Intensive Probation Supervision.Research in Brief. Washington, DC: U.S.Department of Justice, Office of JusticePrograms, National Institute of Justice.

Fagan, J.A. 1990. Treatment and reinte-gration of violent juvenile offenders:Experimental results. Justice Quarterly7(2):233–263.

Fagan, J., Forst, M., and Vivona, T. 1988.Treatment and Reintegration of ViolentJuvenile Offenders: Experimental Results.San Francisco, CA: URSA Institute.

Fagan, J.A., and Jones, S.J. 1984. Com-munity reintegration in juvenile offenderprogramming. In Violent Juvenile Offend-ers: An Anthology, edited by R. Mathias,P. DeMuro, and R. Allinson. San Francisco,CA: National Council on Crime and Delin-quency, pp. 365–376.

Garrett, C.J. 1985. Effects of residentialtreatment of adjudicated delinquents: Ameta-analysis. Journal of Research inCrime and Delinquency 22(2):287–308.

Gendreau, P. 1996. The principles ofeffective intervention with offenders. InChoosing Correctional Options That Work,edited by A. Harland. Thousand Oaks, CA:Sage Publications, Inc., pp. 117–130.

Gendreau, P., and Ross, R.R. 1979. Ef-fective correctional treatment: Biblio-therapy for cynics. Crime and Delinquency25(4):463–489.

Gendreau, P., and Ross, R.R. 1987.Revivification of rehabilitation: Evidencefrom the 1980’s. Justice Quarterly 4(3):349–407.

Goodstein, L., and Sontheimer, H. 1997.The implementation of an intensive after-care program for serious juvenile offend-ers. Criminal Justice and Behavior24(3):332–359.

Gottfredson, D.C., and Barton, W.H.1993. Deinstitutionalization of juvenileoffenders. Criminology 31(4):591–611.

Gottschalk, R., Davidson, W.S.,Gensheimer, L.K., and Mayer, J. 1987.Community-based interventions. InHandbook of Juvenile Delinquency, ed-ited by H.C. Quay. New York, NY: Wiley.

Greenwood, P.W., Deschenes, E.P., andAdams, J. 1993. Chronic Juvenile Offend-ers: Final Results From the Skillman After-care Experiment. Santa Monica, CA: TheRAND Corporation.

Greenwood, P.W., and Turner, S. 1987.The VisionQuest Program: An Evaluation.Santa Monica, CA: The RAND Corporation.

Greenwood, P.W., and Turner, S. 1993.Evaluation of the Paint Creek Center: Aresidential program for serious delin-quents. Criminology 31(2):263–279.

Krisberg, B., Rodriguez, O., Baake, A.,Neuenfeldt, D., and Steele, P. 1989. Demon-stration of Post-Adjudication Non-ResidentialIntensive Supervision Programs: Assessment

22

Report. San Francisco, CA: National Councilon Crime and Delinquency.

Lab, S.P., and Whitehead, J.T. 1988. Ananalysis of juvenile correctional treat-ment. Crime and Delinquency 34(1):60–83.

Land, K.C., McCall, P.L., and Williams,J.R. 1990. Something that works in juve-nile justice: An evaluation of the NorthCarolina court counselors’ intensive pro-tective supervision randomized experi-mental project, 1987–1989. EvaluationReview 14(6):574–606.

Leiter, V. 1993. Special Analysis of Datafrom the OJJDP Conditions of ConfinementStudy. Boston, MA: Abt Associates.

Lipsey, M. 1992. Juvenile delinquencytreatment: A meta-analytic inquiry intothe variability of effects. In Meta-Analysisfor Explanation: A Casebook, edited byT.D. Cook, H. Cooper, D.S. Cordray, H.Hartmann, L.V. Hedges, R.J. Light, T.A.Louis, and F. Mosteller. New York, NY:Russell Sage Foundation.

Lipsey, M., and Wilson, D. 1998. Effec-tive intervention for serious juvenile of-fenders: A synthesis of research. In Seri-ous and Violent Juvenile Offenders: RiskFactors and Successful Interventions, ed-ited by R. Loeber and D.P. Farrington.Thousand Oaks, CA: Sage Publications, Inc.

Lipton, D., and Pearson, F.S. 1996. TheCDATE Project: Reviewing research onthe effectiveness of treatment programsfor adult and juvenile offenders. Paper pre-sented at the Annual Meeting of the Ameri-can Society of Criminology, Chicago, IL.

MacKenzie, D.L. 1997. Criminal justiceand crime prevention. In PreventingCrime: What Works, What Doesn’t, What’sPromising, edited by L.W. Sherman,D. Gottfredson, D. MacKenzie, J. Eck,P. Reuter, and S. Bushway. Report to theU.S. Congress. Washington, DC: U.S. De-partment of Justice, Office of Justice Pro-grams, National Institute of Justice.

MacKenzie, D.L., and Souryal, C. 1994.Multisite Evaluation of Shock Incarceration.Executive Summary. Washington, DC: U.S.Department of Justice, Office of JusticePrograms, National Institute of Justice.

Markley, G., and Eisenberg, M. 1986.The Texas Board of Pardons and ParoleCase Management System. Austin, TX:Texas Board of Pardons and Paroles.

Minor, K.I., and Elrod, H.P. 1990. The ef-fects of a multi-faceted intervention on theoffense activities of juvenile probationers.

Journal of Offender Counseling, Service andRehabilitation 15(2):87–108.

Minor, K.I., and Elrod, H.P. 1992. Secondwave evaluation of a multi-faceted interven-tion for juvenile court probationers. Inter-national Journal of Offender Therapy andComparative Criminology 36(3):247–262.

Palmer, T. 1975. Martinson revisited.Journal of Research in Crime and Delin-quency 12(2):133–152.

Petersilia, J., and Turner, S. 1991. Evalu-ation of intensive probation in California.Journal of Criminal Law and Criminology82(3):610–658.

Petersilia, J., and Turner, S. 1993. Evalu-ating Intensive Supervision Probation/Parole: Results of a Nationwide Experi-ment. Research in Brief. Washington, DC:U.S. Department of Justice, Office ofJustice Programs, National Institute ofJustice.

Sealock, M.D., Gottfredson, D.C., andGallagher, C.A. 1995. Addressing drug useand recidivism in delinquent youth: Anexamination of residential and aftercaretreatment programs. Paper presented atthe Annual Meeting of the American Soci-ety of Criminology, Boston, MA.

Sealock, M.D., Gottfredson, D.C., andGallagher, C.A. 1997. Drug treatment forjuvenile offenders: Some good and badnews. The Journal of Research in Crimeand Delinquency 34(2):210–236.

Sherman, L.W., Gottfredson, D.,MacKenzie, D.L., Eck, J., Reuter, P., andBushway, S. 1997. Preventing Crime: WhatWorks, What Doesn’t, What’s Promising. Re-port to the U.S. Congress. Washington, DC:U.S. Department of Justice, Office of Jus-tice Programs, National Institute of Justice.

Sickmund, M. 1997. Offenders in Juve-nile Court, 1995. Bulletin. Washington, DC:U.S. Department of Justice, Office of Jus-tice Programs, Office of Juvenile Justiceand Delinquency Prevention.

Sickmund, M. 1997. Percent Change inthe Public Custody Facility Population, 1983–1995. Adapted from Sickmund, M., SnyderH., and Poe-Yamagata, E. Juvenile Offendersand Victims: 1997 Update on Violence.Washington, DC: U.S. Department of Justice,Office of Justice Programs, Office of Juve-nile Justice and Delinquency Prevention.

Sickmund, M., and Snyder, H.N. 1995.Juvenile Offenders and Victims: A NationalReport. Washington, DC: U.S. Departmentof Justice, Office of Justice Programs,

Office of Juvenile Justice and DelinquencyPrevention.

Sickmund, M., Snyder, H.N., and Poe-Yamagata, E. 1997. Juvenile Offendersand Victims: 1997 Update on Violence.Washington, DC: U.S. Department ofJustice, Office of Justice Programs, Officeof Juvenile Justice and DelinquencyPrevention.

Sontheimer, H., and Goodstein, L. 1993.Evaluation of juvenile intensive aftercareprobation: Aftercare versus system re-sponse effects. Justice Quarterly10(2):197–227.

Turner, S., and Petersilia, J. 1992. Fo-cusing on high-risk parolees: An experi-ment to reduce commitments to the TexasDepartment of Corrections. The Journal ofResearch in Crime and Delinquency29(1):34–61.

U.S. Department of Justice, Office ofJustice Programs, Office of Juvenile Jus-tice and Delinquency Prevention. 1996.Children in Custody Census of Public andPrivate Juvenile Detention and Correc-tional, and Shelter Facilities, 1994/95[machine-readable data files]. Washing-ton, DC: U.S. Department of Commerce,Bureau of the Census [producer].

Van Voorhis, P. 1987. Correctionaleffectiveness: The high cost of ignoringsuccess. Federal Probation 51(1):56–60.

Weibush, R.G. 1993. Juvenile intensivesupervision: The impact on felony offend-ers diverted from institutional placement.Crime and Delinquency 39(1):68–89.

Weibush, R.G., McNulty, E., and Le, T.1998. The Intensive Aftercare Program (IAP)Demonstration Project: Interim Implementa-tion Assessment. San Francisco, CA: Na-tional Council on Crime and Delinquency.

Weis, J.G., and Hawkins, J.D. 1981.Reports of the national juvenile justiceassessment centers: Preventing delin-quency. Washington, DC: U.S. Departmentof Justice, Office of Justice Programs,National Institute for Juvenile Justiceand Delinquency Prevention.

Whitehead, J.T., and Lab, S.P. 1989. Ameta-analysis of juvenile correctionaltreatment. Journal of Research in Crimeand Delinquency 26(3):276–295.

Whittaker, J.K. 1979. Caring forTroubled Children: Residential Treatmentin a Community-Based Context. SanFrancisco, CA: Jossey-Bass, Inc.

23

Share With YourColleagues

Unless otherwise noted, OJJDP publica-tions are not copyright protected. We en-courage you to reproduce this document,share it with your colleagues, and re-print it in your newsletter or journal.However, if you reprint, please citeOJJDP and the authors of this Bulletin.We are also interested in your feedback,such as how you received a copy, howyou intend to use the information, andhow OJJDP materials meet your indi-vidual or agency needs. Please directyour comments and questions to:

Juvenile Justice ClearinghousePublication Reprint/FeedbackP.O. Box 6000Rockville, MD 20849–6000800–638–8736301–519–5212 (Fax)E-Mail: askncjrs@ncjrs.org

Acknowledgments

David M. Altschuler, Ph.D., is Principal Research Scientist at the Johns HopkinsUniversity Institute for Policy Studies, Baltimore, MD, and Adjunct Associate Professorin Sociology.

Troy L. Armstrong, Ph.D., is Professor of Anthropology at California State University,Sacramento, and Director of the Center for Delinquency and Crime Policy Studies.

Drs. Altschuler and Armstrong are Coprincipal Investigators on OJJDP’s intensivejuvenile aftercare initiative.

Doris Layton MacKenzie, Ph.D., is Director of the Evaluation Research Group andProfessor of Criminology and Criminal Justice at the University of Maryland, CollegePark. Dr. MacKenzie has provided expertise to Federal, State, and local jurisdictions onshock incarceration, corrections boot camps, corrections policy, intermediate sanctions,research methodology, and statistical analyses.

This Bulletin was prepared under grantnumbers 87–JS–CX–K094 and 95–MU–MU–K016from the Office of Juvenile Justice and Delin-quency Prevention, U.S. Department of Justice.

Points of view or opinions expressed in thisdocument are those of the authors and do notnecessarily represent the official position orpolicies of OJJDP or the U.S. Department ofJustice.

The Office of Juvenile Justice and Delin-quency Prevention is a component of the Of-fice of Justice Programs, which also includesthe Bureau of Justice Assistance, the Bureauof Justice Statistics, the National Institute ofJustice, and the Office for Victims of Crime.

For Further Information

In addition to this Bulletin, the following publications related to reintegration,supervised release, and intensive aftercare are available from the Juvenile JusticeClearinghouse (JJC):

Intensive Aftercare for High-RiskJuveniles: An Assessment.NCJ 144018.

Intensive Aftercare for High-RiskJuveniles: A Community Care Model.NCJ 147575.

Intensive Aftercare for High-RiskJuveniles: Policies and Procedures.NCJ 147712.

Reintegrating Juvenile OffendersInto the Community: OJJDP’sIntensive Community-BasedAftercare Demonstration Program.FS 234.

To order the publications listed above, contact JJC and request the appropriateNCJ or FS numbers.

Juvenile Justice ClearinghouseP.O. Box 6000Rockville, MD 20849–6000800–638–8736301–519–5212 (Fax)E-Mail: askncjrs@ncjrs.orgInternet: www.ncjrs.org

For online access to JJC’s library, search the NCJRS Abstracts Database atwww.ncjrs.org/database.htm. Directions for obtaining documents from the libraryare available at www.ncjrs.org/cgi/help.html.

PRESORTED STANDARDPOSTAGE & FEES PAID

DOJ/OJJDPPERMIT NO. G–91

U.S. Department of Justice

Office of Justice Programs

Office of Juvenile Justice and Delinquency Prevention

Washington, DC 20531

Official BusinessPenalty for Private Use $300

NCJ 175715Bulletin