Replicating HOPE: Can Others Do It As Well As Hawaii?

When Judge Steven Alm wanted to change the behavior of drug-using persons on probation, he instituted a program that used strict "swift and certain" principles. A rigorous NIJ-funded evaluation in 2009 proved him right. Those on probation in Hawaii's Opportunity Probation with Enforcement (HOPE) program were significantly less likely to fail drug tests or miss probation appointments. They also were sentenced to less time in prison because of probation revocations than were persons on probation who did not participate in the program.

Now, as jurisdictions around the country try to copy Hawaii's HOPE program, one central question arises: Can Hawaii's success be duplicated? To find out, NIJ and the Bureau of Justice Assistance (BJA) are replicating and evaluating the HOPE model in four jurisdictions that vary widely in population density and geographic location: Clackamas County, Ore.; Essex County, Mass.; Saline County, Ark.; and Tarrant County, Texas. To see whether the replications work as well as they did in Hawaii, researchers are conducting process and outcome evaluations and cost assessments.

NIJ asked Angela Hawken, who evaluated Hawaii's HOPE program, to discuss some of the challenges that jurisdictions might face — as well as several keys to success — when implementing a HOPE-style program. Hawken is associate professor of economics and policy analysis at Pepperdine University's School of Public Policy.

NIJ: How did the original HOPE program work, and what were the results?

Angela Hawken (AH): Hawaii's HOPE program begins with a "warning hearing," referred to in some jurisdictions as an "orientation hearing." During warning hearings, the judge clearly lays out the program's rules and structure, and persons on probation are put on notice that they will be punished for violations. The judge emphasizes that their success is entirely within his/her own control.

HOPE involves swift and certain responses to probation violations. The program requires persons on probation with drug conditions — who represent the vast majority of the caseload — to undergo regular random drug tests (six times a month during the first few months). The regular random drug tests remove any "safe window" for undetected drug use. Persons on probation must call a hotline each weekday morning to learn whether they will be drug tested that day. Persons on probation who fail a drug test are arrested immediately. Within a few hours or days, they may be in court, where the judge will modify the terms of their probation to include a stay in jail. Jail terms are brief, typically only a few days, unless the person on probation absconds, in which case a stiffer penalty applies. Each successive violation is met with an escalated response (i.e., longer jail stays).

Unlike drug-treatment-diversion programs and drug courts, HOPE reserves treatment for persons on probation who request it (only a small percentage request a treatment referral) and for those who consistently fail or miss drug tests. In our 2009 NIJ-funded study, we found that when faced with the credible threat of a swift sanction, more than 80 percent of persons on probation stopped using drugs — including even those who had long histories of drug use. Rather than trying to treat everyone and consequently spreading resources thin, the program targets treatment resources to those most in need. As a result, the program can afford to use intensive-treatment services, including long-term residential treatment, rather than relying primarily on outpatient counseling, as most diversion programs do for their clients.

Hawaii's HOPE program has shown that close monitoring — coupled with swift and certain responses to detected violations — improves compliance with probation conditions, including desistance from drug use. About half of the HOPE probationers in our study never tested positive after their initial warning hearing (and thus, did not require sanctions), and about a quarter tested positive only once. Overall, the rate of missed and positive drug tests dropped by more than 80 percent.[1] The other key outcomes included reductions in new crimes and in overall incarceration (primarily due to a reduction in the number of persons on probation who had their probation revoked).

NIJ: Those are impressive results, but Hawaii is a unique place. Will the program work just as well on the mainland?

AH: Whether HOPE works on the mainland is an empirical question. We have seen small-scale trials on the mainland that suggest the approach can reduce drug use and violating behavior — but only if the model's key features are implemented with fidelity: swift, certain and proportionate sanctions. NIJ and BJA currently are supporting implementation and evaluation efforts in four states. These will provide insight into whether HOPE works in other jurisdictions and the local conditions necessary to implement the model successfully.

NIJ: How many localities are now trying out this swift-and-certain-punishment model, and where are they?

AH: We know of at least 40 jurisdictions in 18 states that have implemented similar models. The largest is Washington state. Anyone under community supervision by the Washington State Department of Corrections — which includes more than 100 field offices and more than 15,000 offenders — now falls under the state's Swift and Certain program, which shares many of the key features of HOPE.

NIJ: Most of these mainland programs are very new. Are you seeing any significant differences in how they work or in the results?

AH: Hawaii and Texas have the oldest versions of HOPE-style programs; they both launched pilot versions in 2004. When Texas-based researchers evaluated the Texas model — called the Special Sanctions Court and Supervision With Intensive enForcemenT (SWIFT) — the results were similar to our original findings in Hawaii: Violation rates and new crimes fell.

There are a few differences between the Texas and Hawaii versions. Texas uses hair samples in addition to regular instant-cup [urine] drug tests. Texas also makes next-day arrests for positive drug tests, which are logistically simpler and less costly than immediate arrests. If someone tests dirty, the system in Texas requires that the judge issue a warrant and the warrant be hand-delivered to the sheriff's office. As a result, the logistics of an instant arrest are difficult. Instead, offenders in Texas are instructed to appear in court the following morning. They appear before the judge, and the sheriff takes them into custody. The next-day arrest allows for simplified processing and reduced costs. The offenders know that not showing will be worse and that they will be caught. Those who do not show are given a stiffer sanction when arrested (law enforcement prioritizes these warrants, and most are cleared very quickly). Thus far, the Texas system seems to be working well: 94 percent appear as instructed.

Washington state's Swift and Certain program differs from the original HOPE model in several important ways. First, it targets what most jurisdictions would consider to be parolees, and they are high risk. Washington also uses a clearly elaborated Behavior Accountability Guide, which dictates sanctions. The first violation — for example, a positive drug test or a missed office visit — typically leads to a stipulated agreement. The second through fifth violations result in one to three days of confinement; there is no graduated increase like in HOPE. If the offender commits a serious violation — for example, if they abscond or commit a new offense — stiffer penalties apply.

NIJ: What are the challenges involved in taking a pilot program that works in a relatively small place and trying to apply it in multiple counties or even an entire state?

AH: The HOPE program is easy to describe, but implementing it well takes a great deal of coordination and cooperation. No jurisdiction should plan to launch a HOPE program without first ensuring that all of the key players are on board. This includes local law enforcement partners, jails, prosecutors, public defenders, treatment providers, and, importantly, judges and probation officers. It is essential that the HOPE court and probation office work well together. This is a judge- and probation-centered model.

A committed judge is essential to successful implementation. But without the involvement of other equally essential partners, the program will likely be unsustainable. HOPE is a collaboration. It requires good communication and a leader who solicits and responds to input from all partners.

Washington state provides the best case study of a HOPE-style program rolled out en masse. Within a few months, the state brought more than 10,000 offenders into the Swift and Certain program. Many of us considered this to be an impossible task and worried that law enforcement would be swamped. In HOPE, offending behavior tends to be front-loaded, and so staggered enrollment is better, so as not to overwhelm the system.

Washington experienced a few bumps along the way, but overall, its rollout was impressive. Support from leadership was essential to this success. Leaders at the Washington State Department of Corrections made a smart move early on: They created an implementation team of well-respected and dedicated community corrections professionals to assist with the rollout. The team operated as a mini think tank within the department and was responsible for designing many of the program's details and for training more than 100 field offices. Washington is still modifying the program as leaders learn more, but so far indications point to a successful implementation.

NIJ: Based on all the jurisdictions that are now trying this out, can you isolate some key factors that contribute to success? Are there any pitfalls that criminal justice officials should avoid?

AH: Successful implementation requires the full cooperation and enthusiasm of three essential partners: the designated HOPE judge, the probation department overseeing the HOPE caseload and law enforcement. For jurisdictions in which public defenders and prosecutors will be present for violation hearings, their support also is essential. Failure to secure the support of these key participants could bring HOPE court proceedings to a grinding halt.

Atmosphere matters. It is easier to implement HOPE well in a jurisdiction where the key players want to try something new and are enthusiastic about the behavior-change principles underlying the program.

Relationships matter. The probation office and the judge must have mutual respect. It is essential that other key criminal justice partners — prosecutors, public defenders, law enforcement officers — are brought into the conversation early on and offer their full support.

Communication matters. All jurisdictions will need to modify their programs as they learn from their implementation experience. All partners must be informed of changes and given a chance to weigh in on program reforms.

Discipline matters. HOPE is a program that relies on consistency and should be attempted only in jurisdictions where the judge and the probation office will implement the model with fidelity. Inconsistency in applying discipline undermines the model's behavior-change principles and can lead to resentment among practitioners and persons on probation.

NIJ: What did you learn in your discussions with probationers and former probationers who were involved in the original HOPE program?

AH: HOPE is an adjustment for probationers as well as practitioners. Many probationers and parolees are used to supervision where sanctions are delivered sporadically. HOPE entails closer supervision and swift and certain delivery of sanctions. The sanction, however, is typically modest — a few days in jail compared with the many weeks, months or even years in conventional supervision. Once persons on probation become more familiar with the program, they tend to appreciate the consistency and predictability. In our interviews with probationers, the majority regarded HOPE positively and said that the rules were clear and the program was fair. Many also noted that the consistency and speed of the response to positive drug tests helped them stop their drug use.

NIJ: Do these programs work only with people who have drug problems?

AH: HOPE originally was tested on drug-involved probationers because drug test results provide a convenient measure to test behavior change. Offenders assigned to the HOPE program had dramatic reductions in drug use.

HOPE, however, is not a drug-offender-supervision model. It can be applied to all conditions of probation. In Oahu, all felony sex offenders and felony domestic violence offenders are now supervised under the HOPE model. In other jurisdictions, we are seeing HOPE applied to a broad range of offenders, either as a mixed caseload or, in some cases, through dedicated HOPE courts. For example, there is a dedicated HOPE domestic violence court in Travis County, Texas.

NIJ: What else should criminal justice officials know about these programs?

AH: There are now at least 40 replications of HOPE-style models on the mainland. Evaluations of these programs will help identify the local conditions required to implement HOPE successfully and the characteristics of offenders who respond to the threat of credible sanctions and those who do not. In Hawaii, offenders who are not able to stop their drug use under HOPE are moved into the drug court, which has been retooled to accept high-risk drug-involved offenders who otherwise were prison-bound. A jurisdiction considering a HOPE-style court might want to think about a systemwide approach that provides a continuum of supervision, reducing the intensity of supervision for probationers who demonstrate a willingness to comply and ratcheting up intensity for those who continue to violate.

Jurisdictions planning to implement a HOPE-style program will need to reorganize their current community-supervision practices substantially. Because these practices vary from jurisdiction to jurisdiction, an exact duplication of HOPE across jurisdictions is unlikely. But although the model might need to be tailored to each jurisdiction, it is essential that all programs mirror the model's essential features: a coordinated effort by all parties, increased efforts to detect probation violations, and consistency in responding to violations with swift and certain — but proportionate — responses.

For More Information

• Learn more about the HOPE program.
• Read the final report of the 2009 NIJ-funded evaluation of HOPE, "Managing Drug Involved Probationers With Swift and Certain Sanctions: Evaluating Hawaii's HOPE" (pdf, 67 pages).
• Watch Judge Alm talk about his vision for HOPE.

About This Article

This artice appeared in NIJ Journal Issue 273, March 2014.