14 FEDERAL PROBATION Volume 66 Number 2 Supervision— Exploring the Dimensions of Effectiveness

Faye S. Taxman, Ph.D.* Bureau of Governmental Research, University of Maryland

WITH OVER 4.2 million adults under of a punishment than other interventions. As limited impact on recidivism unless they in- criminal supervision and over one-third of we will discuss in this paper, a discussion clude a therapeutic component. The question the new intakes to prison a year being fail- about the effectiveness of supervision must that looms is how to incorporate the thera- ures from criminal supervision, the effective- ultimately require a revised model of how peutic component within the fabric of cor- ness of supervision is frequently questioned. supervision can impact offender outcomes. rectional programming to ensure that Meta-analysts had concluded that much in In this paper, we will review the existing lit- behavior change is a goal. Recent efforts have the correctional arena did not work (e.g., boot erature, outline the components of a model aimed at improving the capacity of the su- camps, intensive supervision, and case man- of supervision based on the evidenced-based pervision agencies to handle the offender be- agement) and some interventions work for literature in corrections and psychological havior in the community (Petersilia, 1999; select offenders (e.g., cognitive behavioral interventions, and identify some of the issues Taxman, 1998; Harrell et al., 2002), whether therapy, intensive supervision with treatment, that supervision agencies must address as they through drug court, systemic efforts, or treat- therapeutic community with aftercare). But move towards an evidenced-based model of ment as part of supervision. overall, the vast majority of correctional in- supervision. Researchers have generally concluded that terventions fall into the “don’t know” cat- intensive supervision is ineffective (Mac- egory, where we are unsure about the I. What Works in Supervision? kenzie, 2000; Sherman, et al., 1997). This effectiveness due to a lack of quality evalua- leaves open the question about the effective- For the past nearly two decades, incarcera- tions (e.g., drug courts, supervision, drug test- ness of general supervision, since it is gener- tion (overcrowding) and intermediate sanc- ing, outpatient programming, etc.) ally presumed that general supervision is tions have dominated the discussions in (MacKenzie, 2000; Taxman, 1999; Andrews different from intensive supervision. Byrne corrections. Intermediate sanctions devel- & Bonta, 1996; Martinson, 1974). The field and Pattavina (1992) point out that most of- oped as an approach to address overcrowd- of supervision is one area where very little is fenders complete supervision without a tech- ing, although it was widely recognized that known, primarily due to the scanty number nical violation or new arrest—nearly 60 intermediate sanction programs add intensity of studies that have been devoted to measur- percent according to the latest Bureau of Jus- to the seemingly stark community supervi- ing the effectiveness of overall supervision. tice report (Bureau of Justice Statistics, sion. Intermediate sanction programs were Since supervision is often considered to be 2001a,b)—and therefore supervision is viable. conceived and implemented as short inten- in the background of other programming But few studies have assessed the varying sive programs—such as day reporting centers, (e.g., outpatient therapy, cognitive behavioral frameworks for supervision that reflect dif- boot camps, intensive supervision programs, skill building, drug courts, day reporting pro- ferent missions/goals, different theoretical and drug courts—that use control and pun- grams, etc.), few studies have been devoted frameworks, and different operational com- ishment techniques to handle the correction to understanding what “works” in supervi- ponents. The available studies have not tried populations. Petersilia (1999), in a recent re- sion. The nature and activities of supervision to measure the differential effects of various view of the lessons learned during the decade are often considered inconsequential to effec- types of supervision. From a research perspec- of intermediate sanction programming, con- tiveness. The general impression is that su- tive, a series of randomized experiments or cluded that control-oriented programs have pervision is “in lieu of incarceration,” or less clinical trials is needed to understand the im-

*We are indebted to the Maryland Department of Public Safety and Correctional Services under grant 98-07-298149 for the project that allowed for the development of this model. All opinions are those of the authors and do not reflect the opinion of the sponsoring agency. Special acknowledgement to Brad Bogue & Teresa Herbert (JSAT); Judith Sachwald, Rick Sullivan & Ernest Eley (DPP); and Eric Shepardson & Lina Bello (BGR) for assisting in this project. All questions should be directed to Dr. Faye S. Taxman at [email protected] or (301) 403-4403. Do not cite without permission of Dr. Taxman. September 2002 DIMENSIONS OF EFFECTIVENESS 15 pact of different sanctions and a mixture of general assumption that more contacts are caseload with better outcomes if they had a supervision services, such as a comparison of needed for high-risk offenders to provide ex- smaller number of offenders to supervise. weekly to monthly contact, where it is hypoth- ternal controls on their behavior. With the average agent having over 100 of- esized that such incremental differences (four A number of studies have been conducted fenders to supervise, it was widely recognized contacts vs. one contact a month) could make to test the effectiveness of contacts on offender that such a caseload did not allow for appro- a difference in outcomes. outcomes, as shown in Table 1. Contacts have priate monitoring, oversight, or rapport Overall, there have been few rigorous as- been operationalized in two different ways: building. The preferred caseload size (rang- sessments of the effectiveness of different in- 1) increasing the number of times that there ing from 25 to 40 to one agent) in theory al- terventions in the field of supervision. The is personal exposure between the offender and lows the staff to devote more time to each case. majority of studies have related to caseload agent, generally referred to as intensive su- A series of studies on caseload size, beginning size and intensive supervision. Little has been pervision; and 2) reducing the span of con- in the 1970s, illustrated that the caseload size done on case management, risk assessment, trol of agents to a manageable caseload size did not make a difference in offender out- or models testing different philosophies of to allow the agent and the offender to inter- comes. Smaller caseloads did not reduce re- supervision. Overall, supervision is consid- act both more frequently and more directly arrest rates. In a recent study conducted by ered atheoretical in that it is the process of on criminogenic issues. These two concepts Latessa and his colleagues, offenders super- monitoring. It is typically based on no theory of supervision programs—number of con- vised in smaller caseloads had similar rearrest other than social control. Monitoring is rec- tacts and size of the caseload—have been rates to offenders on normal caseloads (75 or ognized as a form of external control by the studied in a number of experiments dating more). The researchers also found that offend- provision of an authority figure to monitor back to the 1960s. Results from the initiatives ers in smaller caseloads and traditional caseloads the adherence to certain restrictions (e.g., tended to indicate that increased contacts or received similar services (Latessa, Travis, Fulton curfews, drug use, gun possession, etc.). Es- smaller caseloads did not result in reduced & Stichman, 1998). Smaller caseloads did not sentially the external control model presumes recidivism (Petersilia & Turner, 1993a,b; result in fewer arrests or greater participation that the offender has the capacity and skills Petersilia, 1998; Gottfredson & Gottfredson, in treatment services. The studies tended to find to internalize the required change as part of 1985; Taxman, 1982; MacKenzie, 2000). that agents with reduced caseloads tended to be the compliance process. It also assumes that The most widely recognized evaluation of involved in more administrative duties than in the external controls will be perceived as lim- intensive supervision was conducted by Joan supervision of offenders. iting in the eyes of the offender, which will Petersilia and Susan Turner in the late 1980s/ The search for the magic number of con- ultimately improve offender compliance. early 1990s. The fourteen (14) site evaluations tacts and the appropriate caseload size has Supervision services are built on the of intensive supervision randomly assigned resulted in some disappointments, because framework that “contacts,” or the relation- offenders to intensive supervision (ranging the research continues to find that the quan- ship between the offender and the supervi- from 4 to 20 contacts a month) or general su- titative nature of contacts does not impact sion agent, are the cornerstone to managing pervision (refer to Table 2). Table 2 also illus- offender outcomes such as rearrest rates. and/or changing offender behavior. (Even in trates the focus of the supervision programs MacKenzie in her review of correctional pro- the control model, the anticipated change is that were generally surveillance oriented, with gramming comments that: compliance with the rules of supervision, in- a few sites devoted to a brokerage model of Although research has not revealed a sig- cluding being crime-free). Contacts provide referring to services. In this multi-site ISP nificant relationship between levels of sur- the means to monitor the performance of of- evaluation, the frequency of contacts varied veillance and recidivism, there was some fenders and to provide direction to the of- from weekly to a monthly, yet there was no evidence that increased treatment of of- fender. As defined by most agencies, a appreciable difference in the rearrest rates com- fenders in ISP programs may be related to supervision contact refers to the number of pared to routine probationers that were being significant reductions in rearrests. Follow- times that an offender meets (e.g., the expo- supervised at much lower contact rates (an up analyses by the RAND researchers sure rate between a supervision agent and an average of thirty-eight (38) percent rearrest rate (Petersilia & Turner 1993a,b) and also re- offender). Contacts can also take the form of for ISP compared to thirty-six (36) percent for searchers evaluating ISP programs in Mas- face-to-face interactions, telephone calls, col- routine probationers). The increase in con- sachusetts (Byrne & Kelly 1989), Oregon (Jolin & Stipack 1991) and Ohio (Latessa, lateral contacts (e.g., employer, family mem- tacts, however, was helpful in closer surveil- 1993a,b) had found evidence that rearrests ber, sponsor, etc.), and notification from lance of the offender and therefore uncovered are reduced when offenders receive treat- service agencies (e.g., drug treatment, men- more technical violations. Petersilia and ment services in addition to the increased tal health, etc.). Generally contacts are cat- Turner (1993a,b) report that seventy (70) per- surveillance and control of the ISP pro- egorized as direct (face-to-face) or collateral cent of the ISP offenders and forty (40) per- grams. For example, Petersilia and Turner (with someone other than agent). Contacts cent of the control offenders had technical (1993a,b) reported a 10 to 20 percent re- became accepted in the supervision field be- violations, with more ISP offenders returning duction in recidivism for those who were cause they are easily quantifiable and can be to prison or jail after one year. These study most active in programs while they were in measured in a workload formula. In the risk findings replicate prior studies on intensive the community. However, the research de- management literature, the assumption is that supervision that found increasing the number signs used in these evaluations did not reach the number of contacts will increase as the of contacts did not improve outcomes the experimental rigor of the random as- signment study by RAND that examined offender is deemed to be more of a risk to (MacKenzie, 2000; Sherman, et al., 1997). the effect of increasing the surveillance and recidivate (O’Leary & Clear, 1984). Contacts The second concept, reduced caseload size, control of ISP participants. (MacKenzie, are generally considered an important com- hypothesizes that agents could manage the 1997:447) ponent of the supervision process, with the 16 FEDERAL PROBATION Volume 66 Number 2

TABLE 1 Summary of Major Studies on Supervision

Scholars Year Emphasis Methods* Findings

Adams, Welch, & Bonds 1958 Caseload Size Randomized No Difference in Recidivism Eze 1970 Caseload Size Randomized No Difference in Recidivism Havel & Sulka 1964 Caseload Size Randomized No Difference in Recidivism Havel 1965 Caseload Size Randomized No Difference in Recidivism Department of Corrections 1960, 1961 Caseload Size/ISP Randomized No Difference in Recidivism Himelson & Margulies 1965 Caseload Size Randomized No Difference in Recidivism Sing 1967 Caseload Size Randomized No Difference in Recidivism Burkhart 1969 Caseload Size Randomized No Difference in Recidivism Robison, Wilkins, Carter & Wahl 1969 Caseload Size Randomized No Difference in Recidivism Lohman, Wahl, Carter & Lewis 1967 Caseload Size Randomized No Difference in Recidivism Oklahoma Department of Corrections 1972 Caseload Size Randomized No Difference in Recidivism Fallen, Apperson, Hall-Milligan & Aos 1981 ISP Quasi-Experimental (3) No Difference but Lower for ISP Erwin 1986 ISP Quasi-Experimental (3) No Difference but Lower for ISP Mitchell & Butter 1986 ISP Quasi-Experimental (3) No Difference. Higher for ISP Pearson 1987 ISP Quasi-Experimental (3) No Difference. Lower for ISP Byrne & Kelly 1989 ISP Quasi-Experimental (3) No Difference. Lower for ISP Jolin & Stipack 1994 ISP Quasi-Experimental (2) No Difference. Higher for ISP Petersilia 1993a,b ISP Randomized (5)— No Difference. Higher for ISP & Turner 14 Sites in 10 sites/lower in 4 Austin & Hardyman 1991 ISP-Elect Monitor Quasi-Experimental (3) No Difference. Higher for ISP NCCD 1991 ISP Quasi-Experimental (3) No Difference. Higher for ISP Probationers but lower for ISP Parolees Latessa 1992 ISP Quasi-Experimental (3) No Difference. Higher for ISP Latessa 1993 ISP Quasi-Experimental (2) No Difference. Drug/Mental Health Offenders had higher recidivism rates than others Moon & Latessa 1993 ISP Quasi-Experimental (3) No Difference. Lower for ISP Latessa 1993b ISP Quasi-Experimental (3) No Difference. Higher for ISP Latessa, Travis, Fulton 1998 ISP Randomized— No Difference for ISP & Stichman 2 Agencies

*The rating reflects the rank assigned by the University of Maryland in their review of the literature on effective interventions. See Sherman, et al., 1997 for a description of the scale and MacKenzie (1997) for a discussion of the studies. September 2002 DIMENSIONS OF EFFECTIVENESS 17

TABLE 2 Summary of 14 Sites of the ISP Experiment

Supervision Components Caseload Face-to-Face Number of Rearrest Technical Size Contacts Drug Tests Emphasis Rate (%) Violations

California Contra Costa 40:1 4 4 Drug Testing 29 64 33:1 16 * Active Elec Monitor 32 61 Ventura 19:1 16 4 Police Coordination/Job 32 70

Washington Seattle 20:1 12 8 Surveillance/TX Referrals 46 73

Georgia 25:2 12 8 Passive Elec Monitors 12 65 Macon 25:2 12 8 Active Elec Monitors 42 100 Waycross 25:2 12 8 TX referrals 12 18

New Santa Fe 35:2 12 4 Counseling 48 69

Iowa Des Moines 35:3 16 8 Active Elec Monitors 24 59

Virginia Winchester 24:1 12 X Substance Abuse Evaluation 25 64

Texas Dallas 25:1 6 * Employment/Graduate 39 20 25:1 6 * Sanctions 44 81 Employment/Graduate Sanctions

Oregon Marion 30:2 20 * Surveillance 33 92

Wisconsin Milwaukee 40:2 12 X Passive Elec Monitors 58 92

Very few of the studies discussed the is- the attention to supervision objectives gree to which they are under the control of sues surrounding the qualitative nature of (Clear & O’Leary, 1983; Dembo, 1972; an authority figure, the degree to which the the contacts that occur in the supervision Duffee, 1975; Katz 1982). That is, monitor- offender feels an obligation to conform, and setting. In fact, the lack of studies on the ing or contacts are believed to be a form of the degree to which the offender is vested contextual nature of contacts suggested that “control.” Correctional scholars and prac- in the goals of supervision or a myriad of the concept of a contact, generally the core titioners have not defined a theoretical other hypotheses about the impact of sur- of supervision, is atheoretical. The relation- model for monitoring other than that it is veillance on offender behavior. Studies on ship that occurrs between the offender and a form of “control.” Insufficient studies intensive supervision and reduced caseload the agent is presumed to be the basis for have been conducted to determine whether size indicate that unless the contacts are the offender to change due to the controls the impact of the control or surveillance more than “check-ins,” it is unlikely that that the agent places on the offender and affects the offenders’ perception of the de- they will impact offender outcomes. 18 FEDERAL PROBATION Volume 66 Number 2

A. Unanswered Questions control on the offender (e.g., drug tests, cur- diction use a risk/needs assessment tool to About Supervision fews, house arrest, more reporting and face- make decisions about contact level and service to-face contacts, etc.). The enforcer role needs based on this instrument. This process In trying to understand the effectiveness of placed less emphasis on providing services to also called for reassessing offenders every six supervision, two other issues need to be con- address underlying criminogenic risk and months to realign supervision contacts based sidered: purpose of supervision and case man- need factors. A de-emphasis on brokering on performance in supervision. The classifi- agement protocols. As previously mentioned, resulted as many agencies assumed that the cation and reclassification process was deemed Petersilia (1999) has indicated that one of the traditional social work role was outside of the a critical component of effectively managing critical lessons learned in the intermediate scope of supervision. Similarly, the rise in case the offender in the community (O’Leary & sanction era (late 1980s-1990s) was the im- management as a function separate from su- Clear, 1984). Similar processes were recom- portance of clinical approaches. These clini- pervision (e.g., Treatment Alternatives to mended for offenders being supervised by pre- cal approaches are more likely to lead to Street Crime, etc.) promoted the enforcers’ trial and parole agencies. Risk and need factors reductions in recidivism. In fact, in the 14- role of supervision. Case management func- were considered important to drive caseload site study of ISPs, the researchers found that tions were characterized as antithetical to su- management and the use of scarce correctional the offenders with some counseling services pervision services since agents were ultimately resources on offenders that had the greatest (e.g., employment, substance abuse, etc.) “freed up” from the responsibility of address- likelihood of recidivating. Today, few systems tended to have better ISP outcomes than those ing the services articulated in the court or- have systemic screening and assessment pro- with merely surveillance functions (Petersilia der. The fragmentation of supervision services cesses to identify target populations for more & Turner 1993a). This suggested the impor- into risk management/control and case man- intensive services, even though a key to good tance of the use of therapeutic techniques to agement basically allowed supervision agen- correctional programming is to target the high- reduce involvement in crime-free activities. cies to emphasize the monitoring functions risk offender to more intensive controls and characteristics of the “enforcer” role. This services. A few tests of the responsivity, assign- Purpose of Supervision added to bifurcation of the brokerage func- ing high-risk offenders to more intensive ser- Supervision agencies have generally been per- tion ascribed to supervision. vices, have generally found high-risk offenders plexed about their actual role in the criminal Interestingly, few evaluations have been who have lower recidivism rates when they are justice system. Historically agencies have tried conducted of the basic premise of supervision, assigned to more intensive services (Thanner to achieve two purposes—enforcer and so- whether it reflects a social work or enforcer & Taxman, 2001; Andrews & Bonta, 1996). cial worker—and have found the polar na- philosophy. As shown in Table 2, more em- ture of the two tasks often conflicting. During phasis has been placed on surveillance tech- New Frontiers in Supervision different periods in the history of supervision, niques as the focus of the supervision. One Since little in the supervision literature has one or other of the two roles has tended to can conclude, as did Petersilia (1999), that been empirically tested, it is important to ex- dominate. Social worker orientation tended monitoring did little to yield better offender amine the potential theoretical frameworks to dominate the field until the mid-1970s, outcomes. But overall, the research literature for supervision that could translate into of- when the enforcer role took precedent. The has not tested the different components of fender behavior change. Taxman, Soule, and social worker role emphasized the brokerage supervision—monitoring, brokerage, direct Gelb (1999) argue that a system of procedural model, with agents responsible for referring service, etc.—on offender outcomes. The em- justice would improve compliance with the offenders to needed services in the commu- phasis has been more theoretical—about the rules of supervision and therefore translate nity. In some select agencies, agents used their general mission of supervision and how this into changes in offender behavior. Applying counseling skills and ran group sessions. The translates into functions for staff. The impact the constructs of procedural justice, the schol- social work approach focused on brokering of these different staff operations on offender ars contend that providing a setting where the available services in the community from outcomes is unclear. rules/expectations are clear and uniformly other agencies (e.g., alcohol and drug treat- applied would increase the overall compliance ment, employment, mental health services, Risk Assessment rate and improve outcomes. As part of this community services, etc.), instead of provid- Assessment provides critical information model, the supervision contact is the means ing supervision agencies with the capacity to about the characteristics of the offender that to ensure uniformity to the general rules of directly offer these services to offenders. impact effective supervision of the offender. In supervision. Taxman and Bouffard (2000) The emergence of the enforcer role oc- the 1980s the National Institute of Corrections have extended this theory to define supervi- curred as “nothing works” promulgated in the (NIC) recommended that correctional agen- sion as an intervention where the purpose and field, stressing the historical foundation of cies should assess the offender on risk factors intent of the contact is to motivate the of- supervision as part of the judicial arm man- (propensity to commit crimes) and need fac- fender to change his/her behavior. The con- dated to monitor compliance with court-or- tors (sociological needs that impact criminal tact is equivalent to “brief interventions,” dered conditions. Dissatisfaction with behavior). NIC developed a risk tool and pro- which are short in duration but empower the rehabilitation efforts, along with a growing in- cess that provided the framework for many offender to address criminogenic factors that terest in retributive justice, focused on com- correctional agencies (Van Hoorhis & Brown, contribute to offender behavior (e.g., sub- munity supervision as an enforcer of the 1997). The suggested design required the risk stance abuse, educational deficits, etc.). conditions of release. The enforcer role fo- tool to be validated on the population of a ju- Similarly Sachwald (2001) has identified cused on the offender complying with condi- risdiction. NIC recommended that each juris- how supervision can borrow from the “what tions of release, and placed more external September 2002 DIMENSIONS OF EFFECTIVENESS 19 works” literature to include more cognitive TABLE 3 behavior strategies to achieve desired changes Mapping “What Works” with Supervision Components in offender behavior. In her model, she illus- trates how supervision can meet the tenets of What Works Maryland’s Supervision Model sound intervention programming. As shown below, Sachwald maps the convergence of the Identify • Identify high-risk offenders by place or drug use different concepts to illustrate how supervi- Criminogenic • Develop new risk tool to guide risk decision sion can be construed as an intervention. Risk/Needs • Utilize tools to identify needs Factors • Use drug testing to identify drug-using offenders B. Evidenced-based Practices Target • Target places with heavy criminal activity in Changing Offender Behavior Interventions • Use Drug Court/Correctional Options for high-risk drug The larger body of literature in the field of to High-Risk offenders corrections, addictions, and psychological Offenders • Use Break the Cycle methods to monitor use and target for interventions provides evidence about prac- services based on use tices that could be applicable to the field of • Monitor sex offenders supervision. These practices could be incor- porated into the field to develop a theoretical Minimize • Use monitoring tools for low risk model of supervision that contributes to Services for • Differential caseloads: ratio of 50-55 for high risk and ratio changes in offender behavior to maximize Low-Risk of 200 for low-risk, low-need recidivism reduction. Offenders • Use Kiosk for low risk Use Cognitive • Integrate cognitive behavioral therapy into Use of Informal Social Controls Behavioral supervision contacts Family, peers and community had been shown Interventions • Utilize Motivational Interviewing as part of the to have a more direct effect on offender be- supervision contacts havior than formal social controls such as law • Identify interventions that are appropriate for enforcement or supervision (see, e.g., different offenders (e.g., ASAM, review of treatment Gottfredson & Hirschi, 1990; Byrne, 1990; interventions, etc.) Sampson & Laub, 1993). In a series of studies that is part of the understanding of the life • Use graduated sanctions for behavioral monitoring course, Sampson and Laub (1993) and Warr Engage Social • Utilize community team strategies involving police, (1998) document that offenders tended to re- Support in treatment, and other community agencies spond more positively to the needs and desires Intervention • Use home contacts of natural support systems such as family, peers and the community. In most correctional pro- gramming, informal social controls are per- ceived as being insignificant to the controls widely recognized that sustained behavior Dosage Units are Important exerted by authority figures such as judges, change cannot be achieved in a short period of Besides duration, the amount of treatment probation/parole officials, and police officers. time. Sustained change is expected in the val- provided is also important. While research Correctional programming is generally pre- ues, attitudes, and behaviors of the offenders. has demonstrated that changing the atti- mised on formal social control, but the involve- The duration of the intervention then becomes tude, values, and behavior of offenders is a ment of the family and community in offender important as a means to reinforce the change process that is unlikely to occur over a short behavior is critical to ensure long-term process. The recommended treatment process period of time, the amount of services pro- changes. The value of using the natural sys- is approximately 18 months in duration vided and the versatility in the nature of the tems to address law-abiding behavior is that (Simpson & Knight, 1999; Taxman, 1998) in- services is also important. Most offenders correctional agencies can provide the spring- volving multiple stages. For example, in one have multiple needs (e.g., housing, sub- board to stabilizing the offender in the com- of the renowned Key/Crest therapeutic com- stance abuse, mental health, etc.) and the munity. The process also focuses the attention munity programming in Delaware, Maryland, most successful interventions attend to the on the natural system, and provides needed the program consisted of several levels: in- issues that impact outcomes (Etheridge, et support for such changes. prison treatment, work release center, and al., 1997). The nature and type of treatment community-based services. The programming provided is critical to ensure that they ad- Duration of the Intervention transcends all levels of services and provides a dress the psychosocial needs of the of- Length of treatment has been a consistent multi-stage programming. While research on fender. Intensive services should be finding in the research literature on effective program duration is inconclusive, a growing followed by support services provided dur- interventions. While few studies have exam- consensus finds that programs 90 days or ing stabilization and maintenance periods ined the optimal length of interventions, it is longer have better programmatic outcomes. to reinforce treatment messages (NIDA, 2000; Surgeon General, 2000). 20 FEDERAL PROBATION Volume 66 Number 2

As Prochaski and DiClemente (1986) dis- models the principles of procedural justice, has to bring about changes in the offender’s be- cuss in their stages-of-change model, chang- generally found that addicts tend to improve havior. There are three key areas of the su- ing requires a nonlinear process involving their outcomes when a behavioral contract pervision process: 1) engagement of the different steps—precontemplation, contem- exists and the contract is uniformly applied offender in the process of change through the plation, action planning, maintenance, and (Silverman, et al., 1996). Overall, the strength assessment of criminogenic factors and de- relapse. To assist the offender in this process, of the literature in contingency management velopment of a plan to address these factors; interventions that are suitable to the needs of supports the basic principles of procedural jus- 2) involvement in early behavioral changes the offender and are likely to require different tice that deterrence is possible if the offender through the use of targeted services (e.g., dosages of services are needed. That is, the population has a clear understanding of the treatment, etc.) and controls; and, 3) sus- once-a-month or once-a-week “contact” may rules and the system (e.g., supervision, treat- tained change through compliance manage- be insufficient to achieve the goal of supervi- ment, judiciary, etc.), and if officers respond ment techniques. The glue of the process is sion. Intervention units should be matched to swiftly and with certain clearly articulated re- deportment or the manner of being between offenders’ risks and needs, and their mental sponses and graduated responses. the offender and the agent. The contact is the state regarding the readiness to change. Often, key because it is the means to focus the pur- intensive interventions are more effective when II. Towards An Evidence-based pose of supervision and it allows the offender they are preceded by treatment focused on Model of Supervision and agent to develop a rapport. Like in the building offender motivation and advancing therapeutic setting, the degree of rapport be- their readiness for change (Taxman, 1999; Supervision has been dominated by surveil- tween the offender and agent is an important Simpson & Knight, 1999). lance and control strategies, with some efforts component for the supervision process to towards brokering treatment and employment achieve better outcomes. To make supervi- Continuum of Care services. The approach has generally been to sion the most successful, contacts must have rely upon the treatment interventions that Continuity in behavior-change interventions a function that exceed the mere exchange of serve offender populations to incorporate the is critical to achieving gains in offender behav- information. The contact is more of an en- research principles instead of developing ior (Taxman, 1998; Simpson, Wexler & gagement process that is designed to achieve within supervision such evidenced-based prac- Inciardi, 1999). Interventions offered in prison, desired outcomes. tices. Yet, supervision, by its nature, is designed the community, or community-based facilities to work on “…the offender’s attitudes, by should be built on each other. Since offenders Engagement in Pro-Social strengthening the offender as a person, by are likely to be involved in various interven- Values & Behaviors reducing various external pressures and by tions during the different phases of the legal increasing supports and opportunities, and by Initial impressions are usually very important, process, it is important for the approaches to helping the offender become more satisfied and in fact can define the relationship. As part be compatible. Continuity can be achieved on and self-fulfilled within the context of of the process of wedding the offender to be- two levels: 1) similar philosophy of care, and society’s values” (Palmer, 1995). Using the havior change, the first stage of the supervi- 2) continuation of the treatment programming procedural justice and behavioral interven- sion process should be devoted to through either stepping-up or stepping-down tions, a model of supervision can be achieved understanding the criminogenic risk and the services. Continuity provides for a con- to garner greater compliance with the condi- needs of the offender. Usually referred to as tinuum of care that assists with cost-effective tions of release, and therefore increase the intake, the introduction to supervision is strategies of services that increase the duration specific deterrence impact. In essence, super- more than a mere formality. It provides the in treatment but also serve to incorporate com- vision is a means to engage the offender in a setting to diagnoses factors contributing to patible services for the offender. process of improving compliance with gen- criminal behavior, to outline the ground rules eral societal norms, including the conditions and expectations for supervision, and to en- Offender Accountability Through of release. Supervision has the following ob- gage the offender in assuming responsibility Contingency Management and jectives that focus on offender compliance: for the success on supervision. The engage- Graduated Responses ment process requires the use of diagnostic ● To use the supervision period to engage processes to put together a case plan and/or Both in the behavioral management and jus- the offender in a process of change; behavioral contract that respond to the crimi- tice literature, scholars have promoted a se- nogenic factors. The six general areas that ries of consistent rules that guide participation ● To assist the offender in understanding should be addressed are: anti-social person- in social interventions. Clearly laying the his/her behavior and becoming commit- ality, low self-control, deviant peers, sub- ground rules reduces the mystique of super- ted to behavioral change; stance abuse, antisocial values, and family vision, and clearly applying the rules rein- ● To assist the offender in learning to man- issues. These are dynamic factors that change forces the expected behavior of the offender. age his behavior and comply with societal over time and are less likely to be static (e.g., A favored tool is a behavioral contract that norms. less susceptible to change). identifies the expected behavior, the conse- The engagement process should be de- quences of non-compliance, and the benefits A model of supervision can be found in voted to getting the offender ready to address of compliance. Specifying the rules and their Exhibit 1, which identifies how the supervi- these criminogenic factors by illustrating how consistent application increases the offender’s sion process works. That is, supervision must the factors contribute to legal troubles. The awareness of his/her responsibility. Studies in be perceived as a process that involves a se- “honeymoon” period is basically designed to the area of contingency management, which ries of steps and progress measures in order September 2002 DIMENSIONS OF EFFECTIVENESS 21

EXHIBIT 1 importance of support systems in minimiz- Model Supervision Process ing criminal behavior (Sampson & Laub, 1993) provide convincing support for agents Early Early Sustained to develop these natural systems for the pur- Engagement Change Change pose of ensuring that the offender works to- wards the goal of being a contributing member of the community during the super- Formal Behavioral Assessment • Better Case Controls Change • Reduced vision period. & Case Information Crime A new mechanism used by some correc- Sufficient • Less Services • Reduced tional agencies, particularly when the offender Retention Defensive Drug Use does not have a natural support system, is Offender community advocates or guardians. These ad- Informal Psycho- • Improved Expectations vocates are citizens who volunteer (or are & Ground Rules Controls Social Family/ Change Community paid) to be vehicles to communicate with the offender, and provide daily guidance in liv- ing in the community. The guardian provides Deportment for a community companion who is available • Express Empathy to assist the offender in acquiring and main- • Avoid Argumentation • Roll with Resistance taining employment and services (e.g., health, • Deploy Discrepancy mental health, social, drug or alcohol treat- • Support Self-Efficacy ment, etc.). The advocate is a companion to the offender, similar to a sponsor in a self- 4 Months 12 Months 18 Months help group. Formal Controls/Services. The supervision plan should also include a mixture of clinical engage the offender in the change process by low need. The placement should determine and control services. Informational controls preparing them for dealing with issues that af- the degree of services and controls that are are needed as part of the service matching to fect criminal behavior. This preparation is criti- needed to maximize public safety. The risk ensure that offenders are maintaining the in- cally important because it addresses the two factors should also respond to the special tegrity of the case plan. Most of these formal main factors that prevent people from mak- needs of offenders based on their typology: controls can actually be considered as infor- ing the commitment to change, namely defen- drug involved offender, alcohol involved of- mational controls—that is, they provide feed- siveness and ambivalence. Defensiveness refers fender, mental health needs, sexual deviant back to the agent on the progress of the to the walls that are put up around dealing with behavior, disassociated offender (e.g., not offender. Drug testing, curfews, electronic issues that affect criminality. It is hypothesized connected to the community), and persistent monitoring, progress reports, etc. are needed that by using different rapport and commu- offending. The supervision plan addresses the to provide objective information about the nications strategies the contact can be used criminogenic factors through the use of services degree to which the offender is internalizing to break down the defensiveness and increase and control of the offender’s behavior. The plan the behavior change. The formal controls the offender’s commitment to change. The needs to have three major components to should complement the informal social con- second is ambivalence, or the process of be- address evidence-based practices: trols and services. ing non-committal. To break down the de- Informal Social Controls. Involving the Table 4 illustrates the integration of the fensiveness and ambivalence of the offender community, a support group, and/or family different services and (formal and informal) requires skills focusing on moving the of- is part of the process of building the offender’s social controls to control and change the be- fender into recognizing that certain issues sense of responsibility and sense of belong- havior of the offender (Taxman, Young & (e.g., family, employment, substance abuse, ing to the community. The informal social Byrne 2002). The degree of social control etc.) are problem behaviors and that there are controls will transcend the justice system to should depend on the severity of the crimi- means to address these behaviors. Similar to provide the natural protectors when the jus- nogenic risk factors. More restrictions are the Prochaski and DiClemente’s (1986) tice system is no longer involved. Informal warranted for more serious behavior and stages-of-change model, this moves from pre- social controls can also be trained to under- criminogenic risk factors. For example, tech- contemplation to contemplation. stand the offender’s deviant behavior (e.g., in nology can provide enhancements to moni- The goal of the engagement period is a case the case of sex offenders) that can be instruc- tor offender’s behavior and provide objective plan that moves the offender into an action tive for the natural system that must work measures of behavior. The electronic moni- plan to address criminogenic factors. The ac- with the offender to minimize harm to the toring device is one tool to limit the behavior tion plan should use the controls and services community. The agent is then working with of the offender when area restrictions or cur- to prepare the offender to begin to make the the offender and the natural system to develop fews are insufficient. That is, offenders who psycho-social behavioral changes. The assess- controls that can transcend the justice system. have more difficulty controlling their behav- ment should place the offender in one of the From the restorative justice perspective, this ior may need the electronic monitor to pro- following boxes: high risk/high need; high capacity building is advantageous for the of- vide the external controls. Drug testing is risk/low need; low risk/high need; low risk/ fender and the victim. Research studies on the another tool to determine whether the of- 22 FEDERAL PROBATION Volume 66 Number 2

TABLE 4 Examples of Different Controls for Different Types of Offenders

Type Clinical Formal Informal of Offender Services Social Controls Social Controls

All Offenders Educational/ Area Restrictions or Curfews Guardian Vocational Electronic Monitors Transitional Housing Drug Testing Police-Supervision Contacts Face-to-Face Contacts Graduated Sanctions Drug Dependent or Substance Abuse Drug Testing Self-Help Groups Involved Offenders Treatment Treatment Guardian/Advocate Curfew Restrictions Transitional Housing Graduated Sanctions Guardian Mentally Ill Counseling Treatment/Counseling Self-Help Groups Offenders Psychotropic Psychotropic Counseling Medications Medication Advocate Transitional Housing Sex Offenders Counseling/Therapy Curfew and Area Restrictions Family/Support Plethysmography System Monitor Polygraph Behavior Medications Area Restrictions Counseling Victim Awareness Graduated Sanctions Repeat Offender Therapy Area Restrictions or Curfews Advocates Electronic Monitors Guardians Drug Testing Transitional Housing Alcohol Monitoring Victim Awareness Community Service Graduated Sanctions

fender is using illicit substances. Plethysmog- is illustrated by two variables: 1) compliance During the period of early commitment, raphy is a technological tool to measure the with the case plan; and, 2) retention in the rec- the tools of graduated responses should be arousal behavior of sex offenders, which has ommended services. An agent can determine used to address problems of noncompliance been successfully used by a number of com- the offender’s level of commitment to behav- and ambivalence. The use of ground rules is munity correctional officials in monitoring ior change by assessing how well the offender needed to clarify the expectations during the serious sex offenders. This technology can be is adhering to the case plan. Critical issues sur- supervision period and to identify the conse- used to reassess the performance of the of- round retention in recommended therapeutic quences of compliance and noncompliance. fender in the community for the purpose of services and employment. In the drug treat- The presentation of ground-rules is part of a adjusting the supervision plan. ment literature, treatment retention has repeat- procedural justice process whereby the rules edly been found to be a determining factor of are clearly articulated and implemented. The B. Making the Commitment to better outcomes. The same is true for supervi- offender must be aware that the ground rules Change (Early Change) sion—offenders who retain in treatment ser- will be applied swiftly, with certainty, and with vices are more likely to do better in terms of graduation in responses based on a pattern The second part of the model is the commit- outcomes from supervision. They are less likely of consistent behavior. Graduated sanctions ment to change. The case plan will detail the to be noncompliant with the conditions of re- had been attributed to be critical in ensuring formal controls, services, and informal controls lease, since technical violations drive negative compliance in that they resemble contingency that are used to guide the offender in the offender outcomes. management and token economies where the change process. The commitment to change September 2002 DIMENSIONS OF EFFECTIVENESS 23 offender is rewarded for positive behavior and enhance communication. Communication ciations—all in the big six areas of crimino- sanctioned for negative behavior. The key to can be achieved in the following way: 1) ex- genic needs. Improvements in these domains success is consistency in the application of the press empathy for the offender’s situation and will improve the offender’s family and com- model that is one of the tenets of procedural the difficulty of achieving small gains (e.g., munity commitment while reducing crimi- justice. The ground rules should be used to being crime-free, being drug-free, obtaining nal behavior and drug use. It is during this focus on retention and continued commit- and retaining a job, etc.); 2) avoid arguing stage that a revised case plan is needed that ment to the case plan. with the offender on any conditions or re- focuses more on relapse prevention or main- During this period of time, the focus quirements of supervision. Argumentation is tenance goals—sustaining the change. The should be on gauging where the offender is in generally a threat to the power of the agent focus of the contacts is on rehearsing with the the process of change, the degree of compli- and begins to erode the validity of the case offender the skills gained to prevent problem ance, and modifications of the case plan to fur- plan. It is critical, as part of the process, that behaviors. ther engage the offender in the change process. the agent review the case information and This is a period where compliance manage- risk/need factors that support the agreed- III. Moving Towards A ment should drive the next steps, with the agent upon case plan; 3) roll with resistance by rec- Proactive Model of Supervision using the informal social controls and services ognizing that some negative attitudes and to maximize commitment to the change pro- rigidity are part of the defense mechanisms. Prior research on supervision has not clearly cess. The role of the agent in this period is to The process of behavioral change is difficult defined the purpose and intent of the core facilitate the change. The key during this pe- for the offender and therefore some resistance component of the monitoring or contacts. riod of time is to continue the offender in the is considered part of the ambivalence. By fo- The supervision process identifies three func- process of change. It is often during this pe- cusing on the case plan and commitment to tions that a contact has: 1) to allow the agent riod of time that offenders begin to test the the components, and ignoring the offender’s to develop a relationship with the offender to system by beginning the process of noncom- negativity, an officer can focus attention on focus on offender change and compliance; 2) pliance. Emphasis on the relationship between measurable outcomes; 4) deploy discrepan- to use assessment and case planning to imple- the offender and agent will maintain the com- cies that may occur but focus on the compli- ment Responsivity or matching of services mitment to the goals of supervision—behav- ance issues. In many ways, the offender may and controls to maximize outcomes; and 3) ioral change. Use of the ground rules is the use several discrepancies to divert the atten- to implement the ground rules of supervision. most visible component, but equally impor- tion of the agent on less important issues. In- The proposed model positions the contact as tant is the focus on deportment. stead of examining the discrepancies, focus the core element to bring about change in the Deportment becomes a key component of on the case plan and progress towards the behavior of the offender. Contacts were con- the process. The stronger the rapport between goals and objectives; and 5) support self-effi- sidered to be brief intervention therapy where the offender and agent, the greater the degree cacy by providing the offender with some of the agent uses motivational interviewing skills of compliance. During this period, the keys of the skills to review his or her behavior. A criti- to achieve one of the three desired functions. effective communication are critical to im- cal part of this process is building the The question that agencies have had to prove rapport and address the issues that offender’s skills at self-efficacy. All of these address is the steps that are needed to threaten retention. The goal of deportment is are identified in effective practices using the reengineer the supervision to be a process of for the agent to build a trusting relationship motivational interviewing techniques where offender change. That is, how does the super- with the offender. Deportment has four main the goal is to utilize effective communication vision agency transform itself so that super- components: 1) eye contact, which is a stan- with the offender to retain commitment to vision is recognized as a process with clear dard protocol to give respect to the offender the case plans, and crime-free goals of super- steps? A beginning point is to realize that the as well as to learn to assess the offender’s body vision. The communication tools are part of transformation of the contact is really a trans- language during the different phases of super- the overall strategy of strengthening the con- formation of the staff in terms of one of the vision; 2) social graces, such as shaking hands, tact—by making the contact a means to main- most difficult aspects—style or interaction. It being prompt for appointments, and other tain commitment to the case plan. Then, the is relatively easy to put a process in place, but typical signs of mutual respect are used to sig- contact becomes more meaningful. it is harder to get the agents to use the pro- nify to the offender that he/she is a member of cess and to use it effectively. The behaviorist the community; 3) candid review of offender C. Sustained Change approach moves from the “social worker vs. information, without ascribing blame, where for the Long Term law enforcement” conflict to imagining su- the agent informs the offender of results from pervision agents as change agents. The con- assessments, informational controls, and per- Underlying this approach is the theory that cept of a change agent is appropriate, because formance; and 4) empathy or the use of active the engagement and early commitment to it both symbolizes the new role (from enforcer listening skills to acknowledge the offender’s change will result in sustained change. The or compliance manager to change agent) and perspective yet identify the ground rules. The change process will require the use of differ- new responsibilities (from directing the deportment process depends on the commu- ent psychosocial processes, develop social offender to providing guidance and allowing nication skills of the agent to build the rela- networks, and develop competencies in key offenders to make their own choices with tionship that will work to move the offender areas (e.g., employment, family, etc.) and ac- known consequences). To effect the change, from ambivalence to action. countability. Sustained change will be evident the following need some attention: 1) tech- To facilitate the change process, the con- through improvements in key areas such as nical skill set of agents to diagnoses, to craft tacts between the offender and agent must employment, family, housing, and peer asso- plan (responsivity), and to maintain expec- 24 FEDERAL PROBATION Volume 66 Number 2

tations; 2) philosophies and attitudes of have been shown to influence the type of strat- open question asks the offender to provide a agents on their role in the supervision pro- egies therapists use in working with offend- broader description of the living arrangement. cess; 3) communication skills agents; and 4) ers; it is probable to assume that agents would Second, the agents did not use communication management oversight of the implementation respond differently depending on their view techniques to engage the offender in a change of the new strategy. We will use some data as to the culpability of the offender. process. A number of different techniques are and experiences from one state to discuss Communication Skill Set of Agents. Face- useful to demonstrate empathy and to elicit in- these points. to-face contacts required the agent and the terest in behavioral outcomes. Instructions, Technical Skill Set of Agents. Few states offender to communicate about the condi- motivating statements, summarizing offender have a correctional academy that requires the tions of release, as well as the status of the of- information, and positive recognitions are tech- agent both pre-service and in-service to de- fender. Agents by nature are required to niques to enhance communication. Yet, as velop expertise in four critical areas: 1) crimi- inform the offender of the conditions of re- shown in Table 5, agents infrequently used suc- nogenic need and risk factors; 2) assessment, lease, and then monitor these conditions. All cessful communication techniques to build rap- diagnosis and case planning; 3) compliance- of this requires communication, and the style port with the offender. Finally, overall skill set gaining strategies such as contingency man- of communication is directly related to the of the agent to motivate offenders through com- agement, ground rules, etc.; and 4) offender’s willingness to disclose information munication is a 55.2 out of 100 using the stan- interviewing and observational skills. Most to the agent. The manner in which questions dard clinical skill rating scale employed by the training is the “on-the-job” training that fo- are posed has been shown to relate to the type scholars motivational interviewing techniques. cuses on the required protocols, forms, and of responses that offenders provide. Table 5 This is a low rating which indicates that the ba- procedures. While these are critical to under- illustrates the communication skills of seven sic functioning of an agent, and the ability to standing every job, it is the staff development hundred-twenty (720) agents, with an aver- communicate ground rules and expectations, areas in offender management that are ne- age age of forty-four (44) and an average of as well as use the case plan as a behavioral con- glected in the current training and staff de- fourteen (14) years employed as an agent. tract, is fairly minimal. velopment efforts. For example, in the recent The skill data illustrates that few agents had Management Oversight of the Process to training in the state of Maryland for over 720 basic interviewing skills needed to elicit in- Implement the New Strategy. Implementing agents, it was found that the average score on formation from offenders. The data reveals new initiatives is always challenging. Don a pre-training examination of technical su- three trends about the communication skills Cochran (1992), in his review of the attempts pervision skills was fifty-five (55) percent. of agents. First, fewer than a quarter of them to implement intermediate sanctions in the This was well below the expected knowledge, use open questions, posed in such a manner 1990s, found that the agents would generally even though most of the agents had been that the offender must provide an explana- work to undermine new initiatives to protect employed as agents for more than fourteen tion. Instead of asking whether an offender the existing programs and services. This is al- (14) years. lives with his or her significant other, which ways the challenge to ensuring successful Philosophies and Attitudes of Agents on allows the offender to answer yes or no, an implementation. A critical component is their Role in the Supervision Process. Few sur- veys for supervision agents have been com- pleted on their attitudes and values towards TABLE 5 their job, towards offenders, or towards their role in the criminal justice system. One in- Communication Skills of 720 Agents strument has been developed to measure the Percent of Agents orientation of the agents. This instrument, applied in the same state, found that agents Percent Questions Open Ended Questions 28.35 tend to view themselves as resource brokers (49 percent), law enforcement (26 percent) Percent Questions Closed (yes/no) 23.10 and social workers (24 percent) (Shearer, Percent Positive Recognition of Offender Response 10.09 2001). Similarly, when the agents were given the Understanding of Alcoholism Scale Percent Provide Reflections on Offenders 12.76 (UAS), a tool used in a number of research studies to measure attitudes to causality for Percent Summarize Offender Information 7.18 drug/alcohol use, it was found that nearly half of the agents (42 percent) subscribe to a dis- Percent Use Motivating Statements 5.72 ease model and another 48 percent subscribe to a psychosocial model where the offender. Percent Use Instructions 9.93 The impact of these varying philosophies and perspectives on the performance of the job Percent Confrontational Interactions 3.30 (e.g., rearrest rates, technical violations, etc.) is unknown, although it is believed that law Percent Questions Interact with Offender and Agent 18.86 enforcement orientation will lead to higher program failures. The varying philosophies Percent Skills of Agents 55.20 September 2002 DIMENSIONS OF EFFECTIVENESS 25

identifying management oversight and tech- Conclusion Byrne, J.M. & L.M. Kelly. (1989). Restructuring Pro- niques to inculcate the new initiative into the Prior research supports the use of cognitive bation as an Intermediate Sanction: An Evalua- fabric of the agency. The proposed model would behavioral and treatment interventions as the tion of the Massachusetts Intensive Probation alter the basic foundation of the supervision most effective strategy to reduce recidivism for Supervision Program: Final Report to the National agency by focusing on supervision as an inter- hard-core offenders (Andrews & Bonta, 1996; Institute of Justice. Lowell, MA: University of vention, instead of a tool to monitor compli- Petersilia, 1999; Taxman, 1999; MacKenzie, Lowell, Department of Criminal Justice. ance. With a subtle change, the ability to test an 1997; 2000; Latessa, et al., 1998). For the most Burkhart, W. R. 1969. “The Parole Work Unit interventionist model, focused primarily on part, supervision has been defined from a sur- Programme: An Evaluation Report.” British enhancing the core of the business between the veillance function and the research findings Journal of Criminology, 9: 125–147. agent and the offender (contacts), would affect continue to find that these approaches are in- the basic values and philosophies of the agent. effective. The question that has not been an- California Department of Corrections, 1961. “Nar- Without support from management, at all lev- swered is whether supervision can be redefined cotics Treatment—Control Program. Progress els, but particularly the front-line supervisors, a as an intervention. A model has been presented Report—First Year’s Program Experience, Oct change in the work product is doomed. Tools that defines supervision in such a manner, 1, 1959–September 30, 1960.” Sacramento, CA: to integrate the change into the management mapping supervision to a process that involves California Department of Corrections, Adult process are critical to ensure that the agent fo- three main components: engagement, early Parole Division. cuses on using the intervention model. change, and sustained changed. The glue California Department of Corrections, 1960. “Nar- The core of the intervention model is to among the components is the use of deport- cotics Treatment—Control Project Directive.” change the nature of the contacts that occur ment strategies that focus on improving the Sacramento, CA: California Department of between the offender and agent. This can only contacts between the offender and agent. This Corrections, Adult Parole Division. be done by training staff and then by man- model incorporates the principles of evidence- dating that staff use the new strategy. Several based practice into the fold of supervision. The Clear, T.R. & V. O’Leary. 1983. “Controlling the management techniques are required to en- looming question is whether the supervision Offender in the Community: Reforming the sure the later—changing the offender’s be- agencies can incorporate an interventionist Community-Supervision Function.” Lexing- havior. The favored ones: 1) conduct an model as part of a strategy to protect the pub- ton, MA: D.C. Health. evaluation of the new initiative that measures lic. While the use of deportment strategies will Cochran, D. 1992. The Long Road From Policy Devel- agent change; 2) change the tools to measure humanize the supervision experience to fulfill opment to Real Change in Sanctioning Practice, pp. the supervision outcomes by focusing on behavioral objectives, it is unknown whether 281–303, in J. Byrne, A. Lurigio & J. Petersilia more offender positive outcomes (e.g., em- traditional, surveillance-oriented agencies can (eds.), Smart Sentencing: The Emergence of In- ployment, achievement of case plan); 3) hold move in the direction of employing evidenced- termediate Sanctions. Newbury, CA: Sage. middle management staff responsible for based practices of interventions. agent compliance to the model; and 4) change Dembo, R. 1972. “Orientation and Activities of the staff performance measurement tools to References the Parole Officer.” Criminology 10(2) 193–215. move away from counting contacts and in- Duffee, D. 1975. Correctional Policy and Prison stead focus on measuring offender progress. Adams, S., A. Welch & J. Bonds, 1958. “The spe- Organization. New York, NY: Sage. Part of the strategy for implementation must cialized Intensive Parole Unit: Supplementary address the management of the initiative to Report.” No. 1. Sacramento, California: Bureau Erwin, B. 1986. “Turning Up the Heat on Proba- ensure that the process does not fall by the of Criminal Statistics, Research Section. tioners in .” Federal Probation. 50:17–29. wayside of other initiatives where the staff Andrews, D.A. & J. Bonta. 1996. The Psychology Etheridge, R. M., R.L. Hubbard, J. Anderson, S.G. undermined the implementation efforts. of Criminal Conduct. Cincinnati, OH: Ander- Craddock & P.M. Flynn. 1997. “Treatment In Maryland, where the initiative is being son Publishing Co. Structure and Program Services in the Drug implemented, a performance measurement Abuse Treatment Outcome Study (DATOS).” tool is measuring use of the new strategy. The Bureau of Justice Statistics—Special Report. 2001a. Psychology of Addictive Behaviors, 11:244–260. four components—deportment, assessment, “Trends in State Parole, 1990–2000.” NCJ treatment, and ground rules—were translated 184735. Washington, D.C.: U.S. Department Eze, E. 1962. “Specialized Intensive Parole Unit, in measurable components. The Quality Con- of Justice. Phase III.” Sacramento, California: Bureau of tact Standards (QCS) was created to be used Bureau of Justice Statistics. 2001b. “U.S. Correc- Criminal Statistics, Research Section. by frontline supervisors to measure the agents’ tional Population, 1999.” NCJ 183508. Wash- Fallen, D.L., C.G. Apperson, J. Hall-Milligan & use of the four components of the new model. ington, D.C.: U.S. Department of Justice. S. Aos. 1977. “Report: Intensive Parole Super- The QCS form (available by request) is used vision.” Washington, D.C.: Department of So- by the middle managers to monitor agents’ use Byrne, J. 1990. “The Future of Intensive Supervision cial and Health Services. of the techniques. The tool is one strategy to Probation.” Crime and Delinquency 36: 6-39. provide the agents with the incentive to em- Byrne, J. M. & A. Pattavina 1992. The Effectiveness Gottfredson, S.D., & D.M. Gottfredson. 1985. ploy the motivational interviewing strategies Issue: Assessing What Works in the Adult Com- “Screening for Risk Assessment Parolees: during the contacts with the offender. The QCS munity Corrections System, pp. 281–303 in Policy, Practice, and Method,” in D.P. is completed as part of the employee review J. Byrne, A. Lurigio & J. Petersilia (eds.), Smart Farrington & R. Tarling (eds.) Prediction in with a focus on measuring contacts to deter- Sentencing: The Emergence of Intermediate Criminology. Albany, NY: State University of mine how agents utilize the new strategy. Sanctions. Newbury, CA: Sage. New York Press. 26 FEDERAL PROBATION Volume 66 Number 2

Gottfredson, M. & T. Hirschi. 1990. A General Works, What Doesn’t, What’s Promising. Wash- Prochaska, J.O. & C.C. DiClemente. 1986. “To- Theory of Crime. Stanford, CA: Stanford Uni- ington, D.C.: Office of Justice Programs. ward a Comprehensive Model of Change,” in versity Press. W.R. Miller & N. Heather (eds.), Treating Ad- Martinson, R. 1974. “What Works? Questions and dictive Behaviors: Processes of Change. New Harrell, A., O. Mitchell, A. Hirst, D. Marlowe & J. Answers About Prison Reform.” The Public In- York: Plenum Press. Merrill. 2002. “Breaking the Cycle of Drugs and terest, 35: 22–54. Crime: Findings From the Birmingham BTC Robison, J., L.T. Wilkins, R.M. Carter & G.A. Wahl, Mitchell, C.J. & C. Butter. 1986. “Intensive Super- Demonstration.” Criminology & Public 1969. “A Review and New Findings: Offender vision/Early Release Parole.” Utah Department Policy1(2): 189–215. Classification, Sentencing, Disposition, Super- of Corrections. vision Intensity, Violation Rate.” The San Fran- Havel, J., 1965. “Special Intensive Parole Unit Phase Moon, M.M., & E.J. Latessa. 1993. “The Effective- cisco Project Research Report Number 14. IV.” Research Report No. 13. Sacramento, Cali- ness of an Outpatient Drug Treatment Program Berkeley, California; University of California, fornia: Bureau of Criminal Statistics, Research on Felony Probationers.” Paper presented at the School of Criminology. Section. annual meeting of the Academy of Criminal Sachwald, J. 2001. Proactive Community Supervi- Havel, J. & E. Sulk, 1962. “Special Intensive Parole Justice Sciences, Kansas City. sion: A Plan for Making Maryland Communi- Unit Phase III.” Research Report No. 3. Sacra- National Council on Crime and Delinquency. ties Safer. Department of Public Safety and mento, California: Bureau of Criminal Statis- 1991. Evaluation of the Florida Community Correctional Services, Baltimore, MD. tics, Research Section. Control Program. San Francisco: Author. Sampson, R.J. & J. H. Laub. 1993. “Crime in the Himelson, A.N. & B.C. Margulies. 1965. “Narcot- National Institute on Drug Abuse (NIDA). 2000. Making: Pathways and Turning Points ics Treatment Control Program Phase III.” Sac- Principles of Drug Addiction Treatment. National Through Life.” Cambridge, MA: Howard Uni- ramento, California: Bureau of Criminal Institutes of Health: Rockville, MD. versity Press. Statistics, Research Section. Oklahoma Department of Corrections, 1972. “A Schearer, R. A. 2001. “Strategic Alignment in Com- Jolin, A. & B. Stipack. (1991). “Clackamas County Study of Probation and Parole Supervision, munity Supervision of Offenders.” Perspectives: Community Corrections Intensive Drug Pro- 1969-1972.” Oklahoma City, Oklahoma: Okla- 18–21. gram: Program Evaluation Report.” Portland, homa Department of Corrections. OR: Portland State University, Department of Sherman, L.W., D. Gottfredson, D. MacKenzie, J. Administration of Justice. O’Leary, V. & T. Clear. 1984. “Managing the Eck, P. Reuter & S. Bushway. 1997. “Prevent- Offender in the Community.” Washington, ing Crime: What Works, What Doesn’t, What’s Katz, J. 1982. “The Activities and Decisions of Pro- D.C.: National Institute of Corrections. Promising.” Washington, D.C. bation Officers.” Criminal Justice and Behav- ior, 9: 455–475. Palmer, T. 1995. “Programmatic and Nonpro- Silverman, K., S. Higgins, R. Brooner, C. Montoya, grammatic Aspects of Successful Intervention: E. Cone, C. Schuster & K. Preston. 1996. “Sus- Latessa, E. J. 1993a. “An Evaluation of the Lucas New Directions for Research.” Crime and De- tained Cocaine Abstinence in Methadone County Adult Probation Department’s IDU and linquency, 41(1): 100–131. Maintenance Patients Through Voucher-Based High Risk Groups.” Unpublished manuscript. Reinforcement Therapy.” Archives of General Cincinnati, OH: University of Cincinnati. Pearson, F. S. 1987. “Research on New Jersey’s In- Psychiatry 53: 409–415. tensive Supervision Program.” Washington, Latessa, E. J. 1993b. “Profile of the Special Units of D.C: U.S. Department of Justice, National In- Simpson, D. D. & K. Knight. 1999. “TCU Model the Lucas County Adult Probation Depart- stitute of Justice. of Treatment Process and Outcomes in Cor- ment.” Unpublished manuscript. Cincinnati, rectional Settings.” U.S. Department of Justice, OH: University of Cincinnati. Petersilia, J. 1999. “A Decade with Experimenting National Institute of Justice. with Intermediate Sanctions: What Have We Latessa, E.J., L. Travis, B. Fulton & A. Stichman, Learned?” Perspectives, 23(1): 39–44. Simpson, D.D., H.K. Wexler & J.A. Inciardi. 1999. 1998. Evaluating the Prototypical ISP: Final “Drug Treatment Outcomes for Correctional Set- Report. Cincinnati, Ohio: University of Cincin- Petersilia, J. 1998. “Parole and Prisoner Reentry in tings, Part 1.” The Prison Journal, 79(3): 291–293. nati and American Probation and Parole Assoc. the ,” pp. 470–530, in M. Tonry & J. Petersilia (eds.), Crime and Justice, , Sing, G.E. 1967. “The Relationship Between the Lohman, J.D., A. Wahl, R. M. Carter & S.P. Lewis, IL: University of Chicago. Average Number of Supervising Agent-Nar- 1967. “The Intensive Supervision Caseload: A cotic Addict Outpatient Contacts and the Num- Preliminary Evaluation: Research Report Num- Petersilia, J. & S. Turner. 1993a. “Evaluating In- ber of Successful Weeks in the California ber 11.” Berkeley, California: University of Cali- tensive Supervision Probation/Parole: Results Narcotic Addict Outpatient Program.” Corona, fornia, School of Criminology. of a Nationwide Experiment.” Research in Brief. California: California Rehabilitation Center, Washington, D.C.: National Institute of Jus- MacKenzie, D.L. 2000. “Evidence-based Correc- Research Division. tice. tions: Identifying What Works.” Crime and Surgeon General. 2000. “Mental Health: A Report Delinquency, 46: 457-471. Petersilia, J. & S. Turner. 1993b. “Intensive Proba- of the Surgeon General.” Washington, D.C. tion and Parole.” Crime and Justice, 17: 281–335. MacKenzie, D.L. 1997. “Criminal Justice and Crime Prevention,” pp. 447, in L.W. Sherman, D. Gottfredson, D. MacKenzie, J. Eck, P. Reuter & S. Bushway (eds.), Preventing Crime: What September 2002 DIMENSIONS OF EFFECTIVENESS 27

Taxman, F.S. 1999. “Unraveling “What Works” Taxman, F.S., D. Young & J.M. Byrne. 2002. “Tar- Thanner, M.T. & F. S. Taxman 2001. “The Mul- for Offenders in Substance Abuse Treatment geting for Reentry: Matching Needs and Ser- tiple Career Paths of Drug Offenders.” Unpub- Services.” National Drug Court Institute Review, vices to Maximize Public Safety.” Unpublished lished paper. College Park, Maryland: Vol. II (2): 93–134. paper. College Park, MD: University of Mary- University of Maryland, College Park (under land, College Park (under review). review). Taxman, F.S. 1998. “Reducing Recidivism Through A Seamless System of Care: Compo- Taxman, F.S. & J. Bouffard. 2000. “The Importance Van Voorhis, P., & Brown, R. 1997. Risk Classifi- nents of Effective Treatment, Supervision, and of Systems in Improving Offender Outcomes: cation in the 1990’s. National Institute of Cor- Transition Services in the Community.” Col- Critical Elements of Treatment Integrity.” rections. Washington, D.C. lege Park, Maryland: University of Maryland, Justice Research and Policy 2(2): 9–30. Warr, M. 1998. “Life Course Transitions and De- College Park. Taxman, F.S., D. Soule & A. Gelb. 1999. “Gradu- sistance from Crime.” Criminology 36: 183–218. Taxman, F.S., 1982. “Need and Risk Classifica- ated Sanctions: Stepping into Accountable Sys- tion.” Newark, N.J.: Rutgers University-School tems and Offenders,” Prison Journal: 182-205. of Criminal Justice. Unpublished Dissertation.