Evaluation of Breaking the Cycle

This report was prepared under Grant NIJ#97-IJ-CX-0013 from the National Institute of Justice.

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.

Note: The Portable Document Format (PDF) of this report includes all tables and charts.

EXECUTIVE SUMMARY

This Executive Summary presents the highlights from the evaluation of the Breaking the Cycle (BTC) demonstration projects, conducted in Birmingham, Alabama, Jacksonville, Florida, and Tacoma, Washington, between 1997 and 2001. The BTC demonstrations tested the feasibility and impact of system-wide intervention to reduce drug use among offenders by identifying and intervening with drug-involved felony defendants. BTC reflects the widespread recognition among criminal justice professionals of the link between drug use and crime. Drug users commit crimes to get money for drugs, get involved in violence over drug deals, and commit crimes while under the influence of drugs. Faced daily with drug-abusing offenders, many judges, prosecutors, police, and corrections officers actively support efforts to reduce drug use among offenders and are willing to use the authority of the justice system to this end. The BTC strategy was to screen offenders shortly after arrest and require those found to use drugs to participate in a drug intervention while under criminal justice supervision.

BTC targeted all adult felony defendants and was not limited to those charged with drug offenses. Defendants were ordered to report to BTC for drug screening as a condition of pretrial release; those who reported drug use, tested positive for drugs, or were arrested on drug felony charges were placed in drug testing and, when appropriate, referred to drug treatment or drug education classes. The goal was to expand the scope of earlier programs such as drug courts and Treatment Alternatives to Street Crime (TASC) by incorporating drug reduction activities as part of handling felony cases.

The core system reforms called for by BTC model were early intervention, judicial oversight, graduated sanctions and incentives, and collaboration among justice and treatment agencies.

• Early Intervention. The BTC model called for identifying drug-using offenders immediately after they were arrested. The rationale for intervention at this point is that arrest is a crisis for most individuals and represents an opportunity to assist drug users in confronting and acknowledging a substance abuse problem. To capitalize on this moment of opportunity, BTC plans included drug testing of all offenders as soon as possible after arrest, early clinical assessment, and timely placement in drug treatment or monitoring as indicated by the assessment. Operationally, this meant setting up procedures for (1) screening every defendant entering the criminal justice system to identify drug users and (2) preparing individualized treatment plans and conditions shortly after arrest.
• Judicial Oversight. The BTC model also required close judicial oversight of drug treatment participation. The experience of drug courts suggests that close judicial oversight can help reduce drug use and criminal behavior among participants, a lesson BTC planned to extend to all drug-using defendants under any form of criminal justice supervision. BTC planned to use judicial authority to monitor and sanction violations of conditions of community supervision, focusing on requirements for drug testing and treatment attendance if recommended. Operationally, judicial oversight called for court review of compliance and reports on drug treatment performance and drug testing prior to hearings.
• Use of Graduated Sanctions and Incentives. The agencies involved in BTC were expected to apply steady leverage to require abstinence and retain offenders in treatment. Borrowing from strategies pioneered by drug courts, BTC offender management plans specified consistent and timely use of sanctions. The sanction plans called for immediate and certain response to offender violations, with punishments graduating in severity as needed, and incentives for treatment progress. Operationally, these plans required that BTC specify (1) BTC rules and infractions; (2) sanctions for each type of infraction; and (3) the staff with authority (and responsibility) to respond to compliance and noncompliance. Further, BTC needed to have procedures for ensuring that the rules and policies were understood by the offenders, court staff, treatment staff, and supervising agents or case managers.
• Close Collaboration among Criminal Justice and Drug Treatment Agencies. The BTC sites were asked to engage in cross-agency planning and management. Operationally, key staff members of partner agencies were expected to attend regular meetings to develop policies, solve problems, and engage in setting objectives. Partner agencies were expected to exchange information on offender status and engage in collaborative monitoring of offender compliance in an ongoing and timely manner.

BTC had an important and lasting effect on the demonstration sites. BTC innovations generated considerable local political support and are continuing with substantial local funding in all three sites. A major reason for this support is that BTC is seen as an appropriate and effective alternative to pretrial detention. This was very important to the sites with overcrowded jails. The availability of supervision, drug testing, and penalties for continued drug use increased the willingness of the courts to release arrestees while their cases were pending, and offered the justice system strategies for addressing a factor that contributes to the risk of reoffending. This fact suggests that improvements in the strategies for coordinated intervention to address drug use among offenders were valued and that additional work needs to be done to identify ways to deliver these services efficiently and consistently. This enthusiasm persists despite the challenges, described below, of undertaking such substantial system-wide reforms and appears justified by the significant reductions in drug use and crime found by the evaluation.

IMPLEMENTING BTC

Early intervention proved challenging to the sites, all of which needed to establish new pretrial supervision and case management programs. In Birmingham and Tacoma, defendants were required as a condition of pretrial release to report to the TASC agency offices for a drug test and screening, followed by assessment if needed. In Jacksonville, jail medical staff administered drug tests and the TASC agency conducted screening and assessment in the jail or at the BTC offices. The procedures used are shown in table 1.

The average number of defendants screened, found eligible, and given a treatment plan each month was 365 in Birmingham, 182 in Jacksonville, and 129 in Tacoma. Thus, annually Birmingham assessed about 4,300 defendants, Jacksonville about 2,100, and Tacoma about 1,500.

The demonstration highlighted the potential advantages and disadvantages of conducting the initial screening in jail.

• In-Jail Screening. One advantage of in-jail screening is that results are available at the time of the release decision and can be used in setting release conditions. Another is that results can be used to identify detained defendants who should receive treatment in jail (if available). This procedure resulted in screening the highest proportion of charge-eligible arrestees. However, in-jail screening wastes resources if drug users do not subsequently receive an intervention because charges are dropped, reduced to misdemeanors, or shifted to another jurisdiction. Moreover, in Jacksonville, many defendants with positive screening results were not ordered to report to BTC as a condition of their release.
• Post-release Screening. Advantages of post-release screening are that it does not require jail space and staff. It saved resources in Tacoma and Birmingham, where jail-based treatment was limited to weekly Narcotics Anonymous/Alcoholics Anonymous (NA/AA) groups. Post-release screening and assessment allowed case managers to make immediate referrals to treatment. However, limiting screening to those released may fail to identify the most seriously addicted defendants, who by virtue of their criminal history and lack of cash may be less likely to be released. Another potential disadvantage is that post-release screening gives arrestees advance warning of a test and a chance to avoid detection by abstaining from drug use prior to the initial test. In response to these concerns, Birmingham required defendants arrested on drug felony charges to test negative for drugs on weekly tests for at least one month before concluding that they were not users and releasing them from BTC.

BTC sites used the results of the assessments to assign defendants to a level of drug intervention based on the severity of their drug involvement. There were differences in drug use patterns and severity of abuse across the three sites. In Birmingham, 55 percent of the BTC participants in the impact evaluation sample tested positive on their initial drug test, compared with 65 percent in Jacksonville and 68 percent in Tacoma. Abuse problems, as measured by drug composite scores on the Addiction Severity Index (ASI) (McLellan et al. 1992) above .04 at baseline,¹ were reported by 33 percent of the Birmingham BTC sample, 71 percent of the Jacksonville sample, and 60 percent of the Tacoma sample. However, the treatment placements recorded in the BTC management information systems (MISs) in each site did not parallel the apparent need patterns. The portion of BTC participants placed in residential or outpatient treatment at some time during their participation was about two-thirds in Birmingham, one-fourth in Tacoma, and one in twenty in Jacksonville. It should be noted that Tacoma and Jacksonville had serious problems with their MISs and it is likely that not all treatment placements were recorded.

Drug-involved defendants assessed as needing less intensive intervention were placed in drug testing, sometimes accompanied by drug education classes. Others were placed in drug testing and treatment readiness classes while waiting for treatment spaces. The average number of drug tests per BTC participant was 9.7 in Birmingham, 2.8 in Jacksonville, and 12.2 in Tacoma. Most sanctions were for missing or failing a drug test, were delivered by case managers or pretrial officers, and were relatively mild. The case manager sanctions in Birmingham included an "alert" letter that warned noncompliant clients (received by 60 percent of the violators), case review by the case manager (received by 42 percent of the violators), and termination from BTC (received by 23 percent of the violators). In Jacksonville, the pretrial officer sanctions included warning letters, meeting with the supervising officer, and a violation report to the court. Although the infractions recorded for 20 percent of BTC testing participants were almost always followed by a sanction, only 19 percent of the positive results on tests conducted after BTC entry were noted as infractions. In Tacoma, the impact evaluation sample averaged 5.1 infractions per BTC participant and almost all were sanctioned. About two-thirds of the sanctions were delivered by case managers and involved letters, meetings, or changes in drug testing requirements.

Judicial monitoring was relatively rare in Birmingham and Jacksonville. Few hearings were held to review noncompliance and even fewer judicial sanctions were imposed. A sample of data from the program records showed 28 judicial sanctions in response to 8,601 infractions in Birmingham and 10 judicial sanctions in response to 458 infractions in Jacksonville. In Tacoma, one judge reserved one afternoon a week for review hearings for BTC participants with numerous infractions. About one-third (over 700) of the sanctions given to BTC participants in the impact evaluation sample were administered by the judge.

One problem BTC encountered was that plans did not specify what would happen in cases of sustained noncompliance or conditions under which participants would be terminated from BTC services. There seemed to be an implicit belief that sustained noncompliance would result in judicial sanctions that would graduate in severity and culminate in pretrial detention pending case disposition (prior to sentencing) or in probation revocation (post-sentencing). The influx of cases resulted in much higher than anticipated caseloads for case managers and limited their capacity to provide close monitoring and deliver sanctions in a timely fashion. This problem was exacerbated by the lack of automated data systems linked to other agencies, which would have enabled case managers to distinguish inactive BTC clients (in jail or out on warrant) from those expected to be reporting regularly and to monitor treatment attendance and drug test results easily.

As this description indicates, BTC implementation fell short of the ideal in all three demonstration sites, highlighting the challenge of coordinated efforts to reduce drug use among all drug-involved felony defendants released to the community while their cases are pending.

LESSONS FOR THE FUTURE

The BTC effort to expand lessons from earlier programs produced valuable guidance for future development of programs and policies, particularly in three areas: strategies for identifying and responding to a range of substance abuse problems, the data infrastructure required to track the progress of individuals as they have contact with multiple agencies, and issues in managing interagency collaboration across justice and treatment agencies.

Lesson 1. BTC functioned best when case managers assumed responsibility for both managing treatment referrals and responding to drug testing infractions, backed up by judicial hearings as needed.

Two models emerged for managing treatment and supervision. In Birmingham and Tacoma, BTC expanded TASC case management capacity and asked case managers to adopt more stringent procedures for monitoring compliance with BTC requirements and administering sanctions for noncompliance. In Jacksonville, BTC divided these responsibilities between treatment providers and a newly created BTC supervision unit. In practice, this model split the intervention clearly into coerced abstinence for non-addicted drug users and treatment for those meeting clinical criteria of abuse or dependence. This model was less successful in bridging the gap between justice and treatment agencies. Corrections officers tested, supervised, and sanctioned defendants, maintained records, and requested violation hearings if needed. Treatment agencies provided services for relatively few participants and rarely reported noncompliance to the pretrial officers. The plan for judges to review and sanction persistent noncompliance was consistently implemented only in Tacoma. Future programs need to engage in strategic planning around rules that define infractions, sanctions for responding to infractions, strategies for timely response to noncompliance, and work to engage the courts in sanctioning.

Lesson 2. Automated data systems are essential to ensure offender accountability in interventions aimed at large numbers of individuals. BTC helped sites recognize and address this need.

Criminal justice agencies must know the whereabouts of those under their supervision, and, in the case of BTC, be able to exchange this information with partner agencies. This proved to be very difficult. Lack of access to timely information on defendant release status affected BTC services both at screening and throughout the program. Because arrestees could post bond at any time and could be picked up on other charges at any time, it was difficult to determine when failure to appear for testing or appointments was deliberate noncompliance and when it resulted from incarceration. Efforts to exchange information were only partially successful and relied heavily on faxing, which was difficult, given the large numbers of BTC clients and their constantly shifting criminal justice status. BTC contributed to the solution in Birmingham and Jacksonville by supporting the development of MISs and in Tacoma by supporting the development of a link between the criminal justice data system and the TASC MIS. However, these improvements came very late in the program, and their absence was an impediment for much of the demonstration period.

Lesson 3. A major benefit of BTC, one not measured by the impact or cost-benefit evaluation, was building the structure and relationships to sustain coordinated policy development and ongoing collaboration among criminal justice agencies themselves and with treatment providers.

BTC, like many other demonstration projects, found that effective joint problem solving and strong local leadership were necessary. Two techniques used successfully to facilitate collaboration among the agencies in all BTC sites were (1) the development of detailed strategic plans that specified needed changes, set deadlines, and assigned responsibility for tasks and (2) the preparation of written memorandums of understanding on the purpose of BTC and the role of each agency in implementation. In all three cities, the most challenging management problem was engaging the state agencies that managed probation in making rather substantial changes in supervision practices to accommodate the local BTC vision of combining treatment and monitoring. These large state agencies have statewide policies, data systems, and managerial procedures that make it hard to make changes in a specific jurisdiction. Moreover, the top officials were located in other cities and could not attend policy board meetings. This situation required extra willingness on the part of the regional directors to take on the challenge of training and encouraging their staff to participate in local partnerships.

THE IMPACT OF BTC

The findings from the impact evaluation confirm the benefits of BTC strategies, even when imperfectly implemented, and indicate that efforts to reduce drug use directed at all felony defendants under supervision in the community can produce gains in public safety.

Finding 1. BTC reduced drug use in two sites.

The results show that BTC participants were significantly less likely than similar defendants arrested in the year before BTC to

• Use an illicit drug in the 30 days before follow-up (Birmingham and Jacksonville)
• Use marijuana in the 30 days before follow-up (Jacksonville and black BTC clients in Birmingham)
• Use a stronger drug (mainly cocaine) in the 30 days before follow-up (Birmingham)
• Engage in heavy drug use in the 30 days before follow-up (Jacksonville)

However, no reductions in drug use were found in Tacoma. Several factors may explain this finding. Drug severity scores, as measured by the ASI composite score for drug abuse, were higher for BTC participants in Tacoma than in the other sites. Overall, the demand for treatment exceeded the supply, and long waits were common. Tacoma was the only site to have a significant problem with methamphetamine, a drug that is particularly difficult to treat. Tacoma also had a much higher proportion of white participants, the group that did not show marijuana use reductions in Birmingham.

The observed reductions in drug use were not clearly linked to participation in drug treatment. In Tacoma and Jacksonville, the BTC samples reported significantly more days of drug treatment in the month before the follow-up interview than the pre-BTC sample (2.1 days compared with 0.8 days in Jacksonville and 3 days compared with 1.3 days in Tacoma), but significant reductions in drug use were found in Jacksonville and not Tacoma. The pretrial monitoring was more aggressive and reached a far larger pool of eligible BTC defendants in Jacksonville than in Tacoma, suggesting that the difference in outcomes may have resulted from increased drug testing, monitoring, and pretrial supervision. Both of these sites had higher rates of stronger drug use (cocaine in Jacksonville and cocaine and methamphetamine in Tacoma) than did Birmingham.

In Birmingham, with its lower rate of stronger drug use, the BTC participants were significantly less likely to use drugs. The reductions in marijuana use were larger among black participants than white participants. There was no significant difference in the number of days of drug treatment in the month before follow-up reported by BTC and comparison defendants. If the planned duration of drug treatment was short, BTC participants in Birmingham might have completed treatment prior to the month before follow-up, as the process evaluation indicated that most of those in need of treatment (based on clinical assessment) received it. Delays in treatment entry in Tacoma combined with widespread methamphetamine use may have delayed or derailed the impact of BTC on drug use. A competing (or complementary) explanation is that the heightened pretrial supervision and drug testing in Birmingham and Jacksonville, more widespread than in Tacoma where efforts were focused on a smaller pool of drug abusers, may have reduced drug use among BTC clients placed only in urine monitoring.

Finding 2. BTC reduced criminal activity.

In all three sites, BTC participants were less likely than comparison samples to report committing a crime in the six months before follow-up interview. When only drug offenses are examined, the results show that BTC participants were less likely to report drug sales or possession in the six months before follow-up. These differences were large and highly significant. Our validity analysis found that the comparison samples in Jacksonville and Birmingham reported a smaller portion of the officially recorded arrests than BTC participants in those two cities across the six months prior to follow-up. The difference in willingness to report offending suggests that the impact of BTC based on self-reported offending may be larger than that observed in this study.

In Birmingham and Tacoma, the BTC samples were less likely than the comparison samples to be arrested within a year of entering the program. In contrast, arrest records in Jacksonville showed higher arrest rates (and numbers of arrests) during the year for BTC participants than for the comparison group. This finding in Jacksonville is inconsistent with the analysis of self-reported offending and with findings in the other two sites for reasons that are unknown. One possibility is an overall increase in enforcement during the years of the study, which increased the likelihood of arrest independently of a change in offending; another is that data systems used to check recidivism were improved.

Overall, our conclusion is that BTC reduced criminal activity, with the caveat that Jacksonville did not conform to this general pattern.

Finding 3. BTC reduced family problems in all sites.

BTC participants also reported some improvements in domains of functioning associated with reductions in drug use, particularly improved relationships with family members. In all three sites, the severity of family problems as measured by the ASI composite score was significantly lower at follow-up for the BTC samples than for the pre-BTC samples, controlling for the severity of family problems and other factors at baseline.

Evidence of reduction in other problem areas was mixed. Tacoma BTC participants reported greater reductions than the pre-BTC comparison group in the severity of psychological problems, employment difficulties, and social difficulties. However, these gains were not observed in the other BTC sites. In Jacksonville alone, BTC was associated with a significant reduction in employment problems.

BTC COSTS AND BENEFITS

The analysis of costs and benefits estimated the return to the costs (expenditures and in-kind resources) to participating agencies. The cost estimates do not include the routine costs of operating the jails and courts, but count only the additional costs of BTC interventions. The benefits are confined to outcomes observed during the year after the start of intervention and do not count longer term benefits or those not measured by the impact evaluation. The results show positive returns to the investment in BTC in all three sites. The ratio of the costs averted for each dollar invested was 2.3 in Birmingham, 2.6 in Jacksonville, and 5.3 in Tacoma. Not all of these savings can be readily converted into budget dollars for the agencies participating in BTC since they represent savings to a host of stakeholders, including potential victims, public jurisdictions, insurers, and citizens in general as well as the public law enforcement, court, and corrections systems. However, the benefits are consistent with the public safety, health, and welfare missions of the investing agencies.

The results are conservative. Reductions in arrests, physical and emotional morbidity, and welfare use may lead to longer term savings not measured by this evaluation. In addition, the savings attributable to the reduced arrest rates at two of the sites are very conservatively estimated. If BTC has a very large effect on cutting arrests, it might make it possible to reduce both jail and prison capacity as well as the size of law enforcement and court systems. One large cost saving not captured by this analysis was that in both Birmingham and Tacoma, BTC was used to relieve serious jail overcrowding, thus reducing pressure to construct new jail facilities. Indeed, in Birmingham, the county commissioners voted continuation funds for BTC at the end of federal funding and cancelled plans for a bond referendum on a new jail.

THE BROADER IMPLICATIONS

The BTC findings on the positive impact of interventions with drug offenders are consistent with other recent studies of programs that use case management with treatment and sanctions to encourage abstinence from drug use and crime. For example,

• In the DC Superior Court, graduated sanctions that are certain and swift were found to reduce drug use prior to sentencing and crime in the year after sentencing, independently of the frequency of drug testing and judicial monitoring (Harrell, Cavanagh, and Roman 1999).
• Drug court evaluations with strong experimental or quasi-experimental designs report significantly lower likelihood of rearrest in the year after entry for participants than for control groups. For example, 44 percent of the drug court participants in Baltimore were rearrested within a year of entry compared with 56 percent of the control group (p<.10) (Gottfredson and Exum 2000). In Orange County, California, 22 percent of the drug court participants were rearrested compared with 34 percent of the control group (Deschenes et al. 1999).
• A five-year study of the Clark County, Nevada, drug court showed consistent and significant reductions in rearrest for drug offenses at one, two, and three years after program entry, but inconsistent results for reductions in arrests for other offenses (Goldkamp, White, and Robinson 2001).
• Not all drug courts are equally effective. Differences in the target population demographics and risk factors as well as variations in drug court policies and practices result in differences in the rates of participation, graduation, and treatment retention as well as differences in rearrest rates across courts and over time (Belenko 1999; Goldkamp, White, and Robinson 2001; Truit et al. 2001).

Other studies of using sanctions in criminal justice interventions for drug users stress the importance of two practices used by the BTC sites:

• Making sure defendants understand the sanctions they face. At the heart of increased accountability is the forging of an understanding between the court and the offender on behavioral requirements and consequences. When drug court defendants enter into an agreement with the judge, they accept a "contingency contract" that makes them accountable for participating in treatment; they must comply with a known set of rules that define sanctions and incentives which they can control through their behavior (see Harrell and Smith 1995; Inciardi et al. 1996; Pendergast et al. 1995). The importance of knowing rules in advance is likely to apply to offenders on probation and parole as well as those in drug courts.
• Making sure defendants believe that sanctions earned will be delivered. To be believable and effective, sanctions must be delivered as promised with a high degree of certainty. Somewhat surprising to many people is the finding that severity of punishment does not always influence behavior, and its effects may depend on the certainty of punishment or the salience of the penalty to the individual (Nagin and Pogarsky 2001). Effective responses may include positive incentives or rewards for doing well or punishments for noncompliance, and theoretically rewards are more powerful in eliciting desired behaviors. However, BTC and most programs for offenders relied on sanctions, imposed by case managers or judges. They included assignment to more intensive monitoring or treatment, judicial warnings, or short-term incarceration. Sanctions that are treatment-oriented (e.g., remand to detoxification or more intense treatment) have been shown to hold great promise (Anglin and Hser 1990; Lipton 1994). The critical point is that noncompliance should be consistently followed by a penalty that is severe enough to make the defendant try to avoid it.

In the two sites in which defendant perceptions were measured, BTC participants rated the risk of receiving a sanction for failing a drug test or missing drug treatment significantly higher at follow-up than the comparison group, indicating heightened awareness of the risk of legal consequences. However, defendant views of the severity of sanctions were mixed. BTC participants in Tacoma rated the severity of sanctions for a drug test failure significantly higher than the comparison group did. In Jacksonville, the reverse was true, with the BTC participants rating the severity of probable sanctions for these infractions significantly lower at follow-up than the comparison group did. This finding is consistent with the higher actual risk of judicial sanctions in Tacoma than Jacksonville and the difference in use of sanctions in the two sites.

BTC efforts to combine sanctions and treatment were consistent with other findings on the benefits of using these intervention strategies together.

• Falkin's (1993) evaluation of community-based treatment for offenders found that treatment combined with urinalysis and court monitoring with sanctions had higher rates of success than treatment alone.
• The impact of the Washington, D.C., Graduated Sanctions program was strongest for offenders who participated voluntarily in NA/AA treatment monitoring (Harrell, Cavanagh, and Roman 1999).
• Intensive Supervision Probation programs in California that combined treatment with strict surveillance reduced recidivism by as much as 15 percent more than high levels of surveillance alone, leading to recommendations that treatment be included as part of efforts to reduce criminal activity among drug felony offenders (Petersilia and Turner 1993; Petersilia, Turner, and Deschenes 1992). The authors warned that "Putting drug-dependent offenders in a program that forbids drug use, provides frequent testing, and provides no assured access to drug treatment virtually guarantees high violation rates" 1993, p. 320).

ACKNOWLEDGMENTS

The authors of this report are very grateful to the many people who contributed to this research, providing support, sending data, and answering endless questions.

The staff of the National Institute of Justice provided sustained support and inspiration, supplying funds for the evaluation, guiding the development and implementation of the Breaking the Cycle (BTC) initiatives, and encouraging understanding of the research needs. They include Betsi Griffith, Spurgeon Kennedy, Janice Munsterman, Carolyn Peake, Jennifer Wood, and Edward Zedlewski.

The assistance we received from the BTC staff, partner agencies, and others in the three demonstration sites was outstanding. This research would have been impossible without their willingness to share their time and data.

In Birmingham, special thanks go to Foster Cook, Kenneth Garner, Ralph Hendrix, Suzanne Muir, and Terri Williams of Treatment Alternatives to Street Crime (TASC) for their help and patience; to Glenda Bozle and the late Major Linn Moore of the Sheriff's Department for their willingness to let us interview inmates at the Jefferson County jail; and to Mike Carroll of the Administrative Office of the Courts for providing data. We also want to thank the judges of the District and Circuit courts and David Barber, District Attorney, for answering our many questions on site visits.

In Jacksonville, extraordinary help was provided by Jamie Brooks, the BTC project director, and Assistant Chief Tara Wildes of the Duval County Sheriff's Department. They both provided office space for interviewers, access to records, and data on key measures for the evaluation. We also appreciate the assistance we received from Brian O'Neill of the Pretrial Services Unit, the judges, the Clerk of Court's office, the State's Attorney's office, River Region and Gateway staff, and Judy Truett of the City of Jacksonville, and the willingness of the Florida Department of Corrections to provide computerized data on sample members on probation.

In Tacoma, the Pierce County Alliance staff, especially Jim Boyle, Terree Schmidt-Whelan, and the BTC project directors, Lisa Daheim and Terry Reid, made this research possible by providing data and access to partner agencies. We particularly appreciate the efforts of Roxanne Krieg and her staff at the Pretrial Services Unit in checking criminal history data and of Larry Gezelius of the Pierce County Information Services for providing court records for the evaluation. Thanks also go to Judge Cohoe, the District Attorney's Office, the Sheriff's Department, and the Department of Corrections for their help.

Staff at the Urban Institute (UI) and the Treatment Research Institute (TRI) who made major contributions to this evaluation include Alexa Hirst, UI's research associate, who compiled the data, visited the sites, and participated in all facets of the project before her move to San Francisco; Shannon Cavanagh, who got this project started; Sarah Ring-Kurtz and Patricia Lee, who managed the fieldwork for TRI; and the superb TRI field supervisors, Dianne Nakamura in Tacoma and Rhonda Graybeal in Jacksonville. We also thank John Carver, consultant to the Justice Management Institute, who prepared the early process evaluation report on Birmingham.

Finally, our deepest appreciation goes to the sample members who were willing to share their experiences with us in interviews and focus groups.

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.

Usage, posting and reprint of materials on the UI web site:

Most publications may be downloaded free of charge from the web site in PDF format. This information may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required.

Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact paffairs@urban.org.

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.