Swift and certain probation: Assessing fidelity to the HOPE model 1

Abstract

While outcomes for HOPE and HOPE-like probation programs have received a good deal of attention in the literature, there is arguably less focus on implementation of those programs with fidelity to the HOPE model. This study describes one such program, the Swift and Certain Probation program in Jefferson Parish, Louisiana, and the degree to which it was implemented with fidelity using the list of elements laid out in the Demonstration Field Experiment (DFE) Final Report. We observed the Swift and Certain Probation program for two years and found that it is implemented with fidelity across many, but not all, of the elements of HOPE. We conclude with lingering issues for both the implementation and evaluation of HOPE and, especially, HOPE-like programs.

Keywords

HOPE implementation swift and certain probation

Introduction

Since the Hawaii’s Opportunity Probation with Enforcement (HOPE) model was implemented in 2004 by Judge Steven Alm, of Hawaii’s First Circuit Court, and Probation Section Administrator Cheryl Inouye, promising evaluations of the program (Hawken and Kleiman, 2009) led to the implementation of similar programs across the U.S. (Nagin, 2016). Evaluations of those programs have resulted in conflicting findings, with some reporting positive outcomes similar to the original HOPE model (Hamilton et al., 2016), and others finding less effective, no effect, or even negative effects on participant outcomes (e.g. Lattimore et al., 2016; O’Connell et al., 2016).

The disparate findings of the studies, as well as philosophical contentions have fueled sharp criticisms from some, who contend the HOPE and HOPE-like programs are flawed and should be discontinued (Cullen et al., 2016, 2018; Duriez et al., 2014). Program supporters, including Judge Alm, maintain that the HOPE program, when implemented with fidelity, is a more effective community supervision model in terms of reducing recidivism, drug use, incarceration, and criminal justice costs, compared to standard probation programs (Alm, 2016; Hawken and Kleiman, 2009).

We believe that program fidelity might explain some of the discrepancy in evaluation studies and that a great deal more research is warranted in implementation with fidelity to the end of more thoroughly understanding outcomes for HOPE and HOPE-like programs. In this article, we describe the literature on Hawaii’s Opportunity Probation with Enforcement (HOPE)² and HOPE-like models. We then describe one HOPE-like program, the Swift and Certain (SAC) probation program in Jefferson Parish, Louisiana and assess the degree to which it was implemented with fidelity to the HOPE model. We conclude with a brief summary of changes to the program since the COVID-19 pandemic began, and more broadly, with a comparison of two lists of key elements of HOPE, and possible approaches to measuring the heretofore neglected elements.

Hawaii’s opportunity probation with enforcement (HOPE) and HOPE-like models

The HOPE model is based on deterrence, where punishment is supposed to discourage continued criminal behavior (Paternoster, 2010). Specifically, the HOPE program involves swift, certain, and proportional responses to criminal behavior in a supervised population (Hamilton et al., 2016). Judge Alm refers to HOPE as a “three-legged stool,” involving the collaborative efforts of probation officers, treatment providers, and a patient and caring judge. The third leg represents the swift and certain sanctions for violations, resulting in a highly structured program that he feels is probationers’ best chance at successful completion of supervision without being incarcerated again for violating probation. Judge Alm emphasizes that the rewards system in HOPE, as well as consistent encouragement by the criminal justice personnel, is just as important as the sanctions component (S. Alm, personal communication, 16 July, 2018).

In terms of deterrence, HOPE uses swift, certain, consistent, and proportionate sanctions when violations occur to discourage future misbehavior (Hamilton et al., 2016). Specific deterrence is sought with short periods of confinement for violations, while general deterrence comes from the open hearings whereby other participants attend and see what happens with the other participants (Hamilton et al., 2016). DuPont et al. (2015) note that the group hearings model also promotes a sense of fairness and consistency among the participants, as they see that everyone is treated the same way, though it is important to note that not all other participants, or even any other participants, may witness the violation hearings that occur after noncompliance, potentially undermining general deterrence. Other unique aspects of HOPE compared to standard probation programs include (1) rigorous and frequent random drug testing, (2) a warning hearing at the first court appearance where program rules, resources, and the sanctions for noncompliance are explained, and (3) the credible threat of punishment in the form of certain and immediate brief jail sentences for any violation of the program rules, including and perhaps especially positive drug tests (Alm, 2014; DuPont et al., 2015).

Proportionality in jail sentencing for violations depends not only on the severity of the infraction, but whether or not the participant takes responsibility for their behavior. A short sentence (2 days) is imposed when a participant admits he or she used drugs and then tests positive for drugs. A longer sentence (15 days) is imposed for those who do not admit use and then test positive for drugs, and those who abscond are subject to a 30 day jail sentence.³ Additional orders may include substance abuse treatment, mental health counseling, or other services as circumstances warrant such programming (DuPont et al., 2015). Judge Alm argues that the swift, certain, and proportional nature of the sanctions are more conducive to compliance and successful completion of supervision compared to traditional probation, as traditional programs involve a high level of uncertainty for participants. This is because they are not guaranteed to get caught for violations and are not sure there will be penalties for violations if they are caught (Alm, 2016; DuPont et al., 2015).

Judge Alm contends that the rewards and encouragement elements of the HOPE program are just as important to the model and the success of participants as the sanctions (S. Alm, personal communication, 16 July, 2018). In fact, compliance with the conditions may earn early termination of probation for participants (after two years in the original HOPE model). He also notes the contributions of HOPE to the community by way of less crime, reduced criminal justice system costs, and the boost to the economy that comes with the requirement for participants to work while on probation (Alm, 2014; DuPont et al., 2015), though the requirement to work is common to both HOPE and traditional probation programs.

Evaluations of HOPE and HOPE-like models

Evaluations of HOPE in Hawaii, conducted at three, six, and 12 months, as well as at 10 years, consistently produced encouraging results with regard to the effectiveness of the program compared to standard probation. HOPE participants had significant reductions in positive drug tests and missed appointments over the course of their participation in the program. In addition, HOPE participants were arrested half as much and served about a third of the days incarcerated for violations and new crimes compared to participants in standard probation (Hawken and Kleiman, 2009). The 10 year follow up study found that the HOPE probationers continued to fare better than the control group in maintaining significantly lower rates for new crime committed, particularly for drug-related charges (Hawken et al., 2016). The researchers also concluded that HOPE was perceived positively by participants, and that the program produced significant savings to the government both in terms of the avoidance of incarceration, and in comparison to the cost of standard probation (Hawken and Kleiman, 2009).

Another program emerging at around the same time as HOPE in Hawaii is the 24-7 Sobriety program. As detailed in Bainbridge (2019), the pilot version of this program was started in South Dakota in the mid-1980s when a prosecutor there realized that alcohol use and dependence kept the same defendants coming back into contact with the criminal justice system. In cooperation with a judge, the prosecutor instituted a program in which defendants whose alcohol use constituted a public safety problem, either in terms of drunk driving, domestic violence, or both, were required to alcohol test twice a day every day of the week until their cases were resolved. Those who did not show up for a test or who tested positive for alcohol use were immediately incarcerated for a few days. The pilot program produced a reduction in alcohol use, as well as in drunk driving and domestic violence crimes in the jurisdiction. In 2005, the pilot was expanded to three counties in South Dakota and in 2007, it was introduced statewide.

Following the documented success of the HOPE program, programs modeled after HOPE began quickly emerging across the U.S. (Nagin, 2016). About 40 states now have these programs in criminal justice areas including probation, pretrial, and prison. Evaluation studies of these programs, however, have not produced results as promising as those on HOPE and have been met with considerable criticism (Cullen et al., 2016, 2018; Duriez et al., 2014). In response to the sometimes scathing criticism HOPE-like models, proponents point to lack of fidelity to the HOPE model in the implementation of those programs (Alm, 2016; Hawken and Kleiman, 2009; Kleiman et al., 2014; Latessa, 2018; Vera Institute of Justice, 2010).

The major concerns expressed about the HOPE-like programs concern the lack of cohesiveness with the various actors and agencies within the criminal justice system. Delaware’s “Decide Your Time” (DYT), for example, is a HOPE-like model that implemented swift, certain, and fair sanction in their standard probation model, rather than involving a judge for enforcement and encouragement. In this case, the DYT participants did not have lower rates of drug use or recidivism; indeed, researchers found the opposite in some cases (O’Connell et al., 2016). Similarly, the Demonstration Field Experiment (DFE), (Lattimore et al., 2016), where HOPE-like programs were implemented across four states, produced mixed and generally discouraging results. Two sites had more revocations for participants in the HOPE-like model, one had about the same as standard probationers, and only one had lower revocations. Further, the cost effectiveness of the HOPE-like programs was not supported (Cowell et al., 2019; Lattimore et al., 2016). However, a recent reanalysis of data from the DFE revealed that those assigned to the HOPE conditions were much less likely to test positive for drugs than were those assigned to traditional probation (Humphreys and Kilmer, 2020), which is arguably a very desirable outcome.

Much more promising findings came from an evaluation of the program in Washington. A statewide HOPE-like model was implemented there in 2012 with a central goal of reducing criminal justice costs through lowered confinement. The evaluators found that, after 12 months, two of the Swift and Certain (SAC) groups had about 20 percent reduced odds of being incarcerated as well as shorter lengths of confinement periods compared to standard probationers. The SAC participants were also 20 percent less likely to recidivate, more likely to take advantage of substance abuse and mental health treatment options, and the SAC program was significantly less costly than the standard probation. The evaluators concluded that the SAC model offers “a potential reconceptualization of the deterrence model in crime control policy” (Hamilton et al., 2016: 1041, but see also Cook, 2016). However, it is important to note that the timing of the legalization of marijuana in that state may have impacted these findings. The comparison group in this study was selected before marijuana legalization and the SAC group was selected mostly after legalization. As Lattimore et al. (2016: 1109) state, “the two groups were exposed to different legal environments, which may account for their findings.”

The importance of implementation with fidelity to the HOPE model

Despite positive findings in some HOPE evaluation studies and support from many across the criminal justice field, critics maintain that the program is flawed, even calling for an end to those in place and a moratorium on the implementation of any further such programs (Cullen et al., 2016). The disparate findings across evaluation studies certainly warrants further review on the efficacy of the HOPE-like programs, and we believe what is largely missing from the literature on HOPE-like programs is a thorough description of the degree to which they were implemented with fidelity to the HOPE model. A number of studies have found that programs implemented with fidelity show more success for participants than those implemented without fidelity (e.g. Durlak and DuPre, 2008; Lipsey, 2009). Indeed, one of the studies that showed no desirable outcomes for four HOPE-like programs nevertheless acknowledges that the desirable outcomes observed at one of the sites, that in Texas, may have been due in part to the fact that the program in Texas was implemented with the highest fidelity to the HOPE model. “The site with the only significant positive findings was Texas—positive impacts on revocations and new drug arrests . . . .[A]lthough all sites could be deemed satisfactory in terms of meeting implementation expectations, Texas had by far the best results” (Lattimore et al., 2016: 1129). In order to continue the investigation into fidelity to the HOPE model within HOPE-like programs, we now describe the Swift and Certain (SAC) Probation program in Jefferson Parish, LA, after which we explain our methodological approach to assessing fidelity at this site and our results.

Swift and certain probation in Jefferson Parish, LA

The Smart on Crime Initiative legislation was passed in Louisiana in 2016, and shortly thereafter, a HOPE-like model, Swift and Certain, was implemented in Jefferson Parish. This legislation and programming was a result of two ongoing movements in the state and local criminal justice systems. First, for years, judges in the 24^th Judicial District had been pondering ways to more effectively reduce crime with rehabilitative strategies. Second, a report was released in 2017 by The Louisiana Justice Reinvestment Task Force, revealing deeply troubling patterns of ineffective criminal justice practices in the state. The report affirmed that Louisiana had led the country in incarceration rates. It also determined that the majority of those serving prison time were for non-violent offenses, including more than half due to failed community supervision. Moreover, the system was also deemed financially inefficient.

The portion of the Task Force report focusing on community supervision determined that exceedingly high caseloads for probation and parole officers were to blame for the high rates of failed supervision and consequential reincarceration. More than 70,000 people were under community supervision in Louisiana in 2015, up from under 60,000 a decade prior. There was no corresponding addition of officers in the state during that time, and with people remaining under supervision for longer periods of time, the average caseload ballooned to nearly 150 per officer. In response to these untenable circumstances, the Task Force (2017: 45–46) recommended the following: (1) focus community supervision on the highest-risk period by reducing maximum probation terms and establishing an earned compliance credit incentive, and (2) improve the process for responding to violations of probation and parole conditions with swift, certain, and proportional sanctions. The interests of the judges, coupled with the findings and implications of the Task Force report (2017), paved the way for Swift and Certain (SAC) probation program in Jefferson Parish.

The official start of the SAC program in 2016 was preceded by about a year’s worth of meetings, detailed below, to which the research partners were invited. The meetings provided an opportunity to examine issues around both implementation and the day to day function of the program and to propose solutions to those issues as they manifested at the time. Data collection by the research partners began in early 2016. The SAC program is modeled after the original HOPE program, where deterrence informs a system of swift, certain, proportionate sanctions for noncompliance alongside rewards for compliance, including the opportunity to receive early termination of probation after two years. SAC is notably different from traditional probation in Jefferson Parish in that SAC participants undergo an assessment at the beginning of their participation, where their risks and needs inform an individualized treatment plan. The plans often include substance abuse treatment, mental health counseling, employment and education assistance. Participants are required to drug test frequently, once a week at minimum when they first begin the SAC program, and are required to attend monthly status hearings with the judge to determine progress, issue sanctions if necessary, and to offer encouragement and praise for compliance and successes.

The SAC program started with just two probation officers referring their noncompliant probationers, but now all Jefferson Parish probation officers can refer probationers on their caseload whom they deem noncompliant and believe would be better served by the SAC program. The probationers placed into SAC are at immediate risk of being incarcerated for noncompliance with traditional probation and the probation officers working in the SAC program have caseloads limited to 40 people.

Methodology

Lattimore et al. (2016) and Lattimore et al. (2018) provide a robust methodology to identify and assess key elements of HOPE and the degree to which these key elements have been implemented with fidelity to the HOPE model in HOPE-like probation programs. Lattimore et al. (2018: 25) identified a list of 11 elements that:

[R]epresent essential things [HOPE-like] programs were supposed to do in their day-to-day operation of HOPE and are (1) central to the underlying HOPE model; (2) implicitly set as expectations on the DFE sites through the Bureau of Justice Assistance solicitation that funded the four HOPE DFE sites; and (3) well within the control of the DFE sites. Further, these are measures for which we have enough data to form conclusions about fidelity with respect to the items.

We followed their approaches exactly where possible and include explanations for any deviations on our part from those approaches. What follows in this section is a brief description of each of the items and manner in which Lattimore et al. (2018) assess fidelity with implementation to the HOPE model of the item.

Item 1: Leadership. There is supposed to be clear program leadership at each HOPE-like site and this item was assessed with qualitative interviews with program stakeholders that focused in part on who each respondent identified as the program leader. We conducted qualitative interviews with stakeholders (n = 10) starting shortly after the inception of SAC in 2016 and continuing for about another year; these stakeholders included the judge, probation officers, probation leadership, program coordinators, law enforcement leadership, and other court staff, to determine the number and percentage of stakeholders who identified a clear program leader.

Item 2: Probationers’ High Risk. HOPE and HOPE-like models are designed to target those at risk of being revoked from traditional probation. In order to determine who these probationers are, it is standard practice to utilize a risk assessment tool. Lattimore et al. (2018: 25–26) note that the Ohio Risk Assessment System (OARS) was the tool used at most of the sites they studied. The SAC program in Jefferson Parish utilized the Correctional Offender Profiling for Alternative Sanctions (COMPAS). We determined the number and percentage of participants in SAC (n = 69) who were assessed as high or moderate risk by the COMPAS.

Item 3: Warning Hearing Compliance. There is a formal warning hearing as part of HOPE that is delivered to new participants during their first status hearing. It is supposed to be based on a script that Judge Alm wrote, with appropriate changes for the site at which the hearing takes place. In order to assess delivery of Judge Alm’s warning hearing, we observed the monthly SAC status hearings. Of the 19 SAC status hearings we observed in 2017 and 2018, nine of those included warning hearings for at least one new SAC participant. We determined the number and percentage of these hearings wherein the judge adhered to Judge Alm’s script (with appropriate modifications for the site).

Item 4. Initial Drug Testing Frequency. HOPE-like sites are supposed to drug test probationers eight times at minimum during their first two months of the program. We determined the number and percentage of the 69 SAC participants who received at least eight drug tests during their first two months in the program by counting the drug tests for participants during that timeframe and the percentage who received at least eight during their first two months.

Item 5. Stepped Down Drug Testing Frequency. Assuming that participants are compliant with the requirements of HOPE, they are supposed to be able to take drug tests less frequently as they progress through the program. While Lattimore et al. (2018) used a two month timeframe for the high intensity drug test frequency, we used a four month timeframe, because the SAC Participant Handbook (Smart Supervision, 2017) stipulates that drug test frequency will not decrease until participants move up from Phase I to Phase II, which takes at minimum four months. We determined the number and percentage of SAC participants who had their drug test frequency stepped down after the first four months by counting compliant participants’ drug tests from month five to month nine of enrollment; the ninth month is the earliest time at which a participant could move up from Phase II to Phase III.

Item 6. Exceptions for Missed Drug Testing. In the HOPE model, probationers are not permitted to miss drug tests except under two circumstances, being in the hospital or being in custody. Allowing participants to miss drug tests without sanctions under other circumstances would violate the certainty of consequences on which the HOPE model is based. In order to determine the number of missed drug tests that were met with a sanction, we compared each participant’s positive drug test to receipt of a sanction using the court’s recordkeeping database.

Item 7. Time to Violation Hearing. In the HOPE model, swiftness of sanctions is a crucial element, and the HOPE model specifies no more than three days (72 hours) from the time of violation to appearance before the judge for a violation hearing and the issuing of a sanction. We determined the time to violation hearing for those SAC participants with a violation by examining the court’s recordkeeping database for the date of the violation and the date of the subsequent hearing.

Item 8. Sanction Type. The HOPE model identifies jail as the preferred sanction, as opposed to other possible sanctions, such as community service, essays, or fines. In order to determine the use of jail as a sanction for noncompliance, we examined the court’s recordkeeping database for violations and the resulting sanctions.

Item 9. Sanction Dosage. In the HOPE model, the swiftness and certainty of sanctions are more important than the severity of sanctions and therefore, severe sanctions are to be avoided. As alluded to above, there is some variation in sanctions depending on how the participant chooses to handle a violation. If he or she admits to drug use prior to a drug test, for example, the sanction will be less severe than if the participant denies drug use but then tests positive. In order to assess sanction dosage, we examined the initial violation for each participant using the court’s recordkeeping database as well as the sanction received, and used the sanction grid in the SAC Participant Handbook to determine whether the sanction for a given violation was in accord with it.

Item 10: Sanction Certainty. The HOPE model specifies that sanctions must be applied with certainty for violations of program conditions. In order to determine whether sanctions were applied with certainty, we examined the court’s recordkeeping database for violations of program conditions and the application of sanction as a result. We also surveyed participants on their perceptions of certainty of sanctions. The details of the survey, which is more wide ranging than certainty of sanctions, are provided in (Frailing et al., 2020)

Item 11. Sanction Swiftness. In the HOPE model, sanctions are supposed to begin as quickly as possible after a violation is detected and the warning hearing is held, ideally, the sanction will commence no more than three days after the violation hearing. In order to assess how swiftly sanctions were applied, we examined the dates of violation hearings and the dates that sanctions began using the court’s recordkeeping database.

Results

In this section, we briefly report on our results item by item.

Item 1. Leadership. Each of the stakeholders we interviewed strongly identified the SAC judge as the program leader (10 out of 10, 100 percent). There is additional, anecdotal evidence that the judge is the clear program leader. For example, as we have observed, he calls and sets the agenda for the program’s quarterly meetings and he runs the staffings that precede the status hearings.

Item 2. Probationers High Risk. All 69 SAC participants were assessed with the COMPAS and of those 69, 49 of them were assessed by the COMPAS as being high or moderate risk (71 percent). The COMPAS, which is no longer used by this SAC program, was administered to participants after they joined SAC. The referral process prior to the fourth year of the program involved probation officers referring those on their caseloads whom they deemed at high risk of revocation from probation.

Item 3. Warning Hearing Compliance. Of the nine status hearings we observed that included warning hearings for at least one new participant, the judge delivered Judge Alm’s warning hearing script with 100 percent congruence nine times (100 percent). Upon the ninth observation, we noticed the judge infrequently referred to the script, apparently having delivered it often enough to have partially memorized it.

Item 4. Initial Drug Testing Frequency. Of the 69 participants in SAC, 67 of them (97 percent) received at least eight drug tests in the first two months. Of the two participants who did not, one passed away and the other was transferred to another court.

Item 5. Stepped Down Drug Testing. Of the 61 generally compliant participants at the month four mark (where generally compliant means the participants had earned points toward advancing through the phases of SAC, but not necessarily without, for example, a positive drug test or being late to court), 58 of them (91 percent) were required to drug test once instead of twice a month. However, it is important to note that percentage did not hold from month four to month nine, as six of those 58 participants were ultimately required to resume more frequent drug testing as a result of repeated positive drug tests during Phase II.

Item 6. Exceptions for Missed Drug Testing. Of the 41 participants who missed a drug test during their enrollment without the permissible excuses of being in the hospital or in custody, all 41 of them (100 percent) received a sanction for missing that drug test. Anecdotal evidence indicates that SAC probation officers expect, at least to some degree, that SAC probationers will miss at least one drug test without a permissible excuse, particularly during Phase I. It is interesting that 28 participants did not miss a single drug test during their enrollment (this is not to say they never tested positive for drugs—52 of the 69 SAC probationers did at some point).

Item 7. Time to Violation Hearing. Of the 52 SAC participants who had at least one violation, 47 of them (90 percent) had their violation hearing before the judge within the 72 hour window. The judge has violation hearing time built into his calendar three days a week, which likely facilitates this finding. The five who did not have their violation hearing within the specified timeframe had absconded and had their violation hearings upon their apprehension. It is interesting that 17 SAC participants did not have a single violation while enrolled. This may be tied to the practice of completing the risk assessment after enrollment rather than prior to enrollment. As noted above, 20 of the SAC participants were assessed as low risk, despite being seen by their probation officers as either moderate or high risk.

Item 8. Sanction Type. Of the 52 SAC participants who had at least one violation, 52 of them (100 percent) received jail as their sanction or one of their sanctions for a violation. Other sanctions that were added in some cases included increased frequency of drug testing or community service.

Item 9. Sanction Dosage. Of the 52 SAC participants who had at least one violation, 52 of them (100 percent) received a jail sanction that was in accord with the sanction grid in the SAC Participant Handbook, at least for the initial sanction. For the 33 participants who admitted drug use prior to their drug test, they received the two day jail sanction identified in the sanction grid. For the 19 participants who denied using drugs then tested positive, they received the eight day sanction identified in the sanction grid. This produced a mean number of jail days per sanction of 4.19.

Item 10. Sanction Certainty. Of the 52 SAC participants who had a least one violation, 52 of them (100 percent) received a sanction for that violation at the violation hearing. Moreover, the mean perception of certainty of sanctions among the 31 participants who completed the survey was 4.55 out of 5, where 5 was strongly agree with the statement “you believe that violating the court’s rules will result in punishment.”

Item 11. Sanction Swiftness. Of the 52 SAC participants who had at least one violation, 40 of them (76 percent) began their sanction within three days of the violation hearing. This is very likely a result of a change in practice in the SAC program that occurred around the end of the program’s first year. The judge began allowing participants, particularly those who were employed when they received their sanction, to serve their jail days at a time that would not interfere with their work schedules, for example, on the weekends.

Discussion

Across the 11 items that indicate implementation with fidelity to the HOPE model as identified by Lattimore et al. (2018), it appears that the SAC probation program has been implemented with high fidelity to that model. As Lattimore et al. (2018) note, the threshold of 60 percent fidelity is fairly broadly supported as being able to produce desirable program outcomes. They also note that 80 percent fidelity is indicative of a very well implemented program.⁴ We found that the SAC program met the 80 percent fidelity threshold across nine items and met the 60 percent fidelity threshold across all 11 items. The SAC program rates nearly as highly as the Texas site in Lattimore et al. (2018, 2016) studies, which as noted was implemented with the highest fidelity across the four sites and was the only one to produce desirable outcomes for both recidivism and drug use.

Table 1 replicates Lattimore et al.’s (2018) implementation with fidelity findings and adds our findings for context. In order to better understand outcomes for HOPE participants at the DFE sites, Humphreys and Kilmer (2020) point out that it is useful to examine particular instances where lack of implementation with fidelity may inhibit the program’s ability to produce desirable outcomes for participants. For example, they note that in Oregon, the warning hearing was only delivered per the Alm script 40 percent of the time, which could have a deleterious impact on participants’ understanding of the program. Both Arkansas and Oregon had timely violation hearings less than 40 percent of the time, potentially undermining the swiftness of punishment. Finally, in Arkansas, less than a quarter of participants were deemed medium or high risk of revocation, which is precisely the type of participant HOPE and HOPE-like programs are supposed to serve. While the differences in implementation may seem small across these sites, particularly when observing the Totals row only, doing so may obfuscate important differences that may be at least partially responsible for the findings in Lattimore et al. (2016, 2018).

Table 1.

Percent implementation with fidelity to the HOPE model across 11 items and five sites.

	Saline County, Arkansas	Essex County, Massachusetts	Clackamas County, Oregon	Tarrant County, Texas	Jefferson Parish, Louisiana
1. Leadership	83	83	92	100	100
2. Probationers High Risk	24	88	80	91	71
3. Warning Hearing Compliance	86	95	40	100	100
4. Initial Drug Testing Frequency	88	91	87	90	97
5. Stepped Down Drug Testing	82	83	73	81	91
6. Exceptions for Missed Drug Tests	98	89	99	100	100
7. Time to Violation Hearing	38	56	37	66	90
8. Sanction Type	88	78	92	97	100
9. Sanction Dosage	78(mean = 6 jail days)	85(mean = 4 jail days)	65(mean = 6 jail days)	93(mean = 5 jail days)	100(mean = 4 jail days)
10. Sanction Certainty	97	91	96	99	100
11. Sanction Swiftness	76	60	63	83	76
Totals	9 items at 60%,7 items at 80%	10 items at 60%,8 items at 80%	9 items at 60%,6 items at 80%	11 items at 60%,10 items at 80%	11 items at 60%,9 items at 90%

Source: Adapted from Lattimore et al. (2018: xxi).

COVID-19 adjustments

Research partners on evaluations of correctional programs are no strangers to the fact that the process, the personnel, the structure, or the function (or some or all of these) can change over the course of the evaluation, and if they were strangers to it, they are no longer thanks to the COVID-19 pandemic and the way it has changed the way correctional programs are delivered in the United States. For example, the SAC program under examination here has moved to status hearings conducted over Zoom every week instead of in person once a month; all participants are seen by the judge once a month, but the status hearing wherein all present participants witness each participant’s interaction with the judge is not currently occurring. Second, before the pandemic, there were two drug testing labs open in Jefferson Parish and since the pandemic, only one has remained open. This has likely affected participants’ ability to drug test when required because of the physical layout of Jefferson Parish; Orleans Parish lies in between the two halves of Jefferson Parish and the half with the open lab is on the west side of the Mississippi river. Finally and most relevant to this article, is the fact that point freezes have been implemented largely in lieu of jail sanctions. Participants have the ability to earn points for compliance with SAC requirements; they need these points to advance through the phases to graduation. The point freeze means that even with compliance, they are unable to earn points until the freeze period is over. Moreover, participants can now lose points for serious instances of noncompliance (Smart Supervision, 2020). The loss of points that participants had previously earned was dismissed early on in the stakeholder meetings (see below) as being incongruous with deterrence in general and with the HOPE model specifically. However, the pandemic and the public health danger of relying exclusively on jail as a sanction has created circumstances in which the implementation with fidelity to the HOPE model observed at this program may have had to change, at least in regard to the use of jail as a sanction.

Lingering issues for assessing implementation with fidelity to the HOPE model

While we relied heavily on Lattimore et al.’s (2018) approach to assessing implementation with fidelity to the HOPE model, there is another list of key elements of HOPE that Judge Alm himself helped compose, found in DuPont et al. (2015: 8–24). It is interesting that the Lattimore et al. (2018) study does not make reference to this list, though it is important to note that some of these elements would not have been possible to measure in the DFE study. Take for example integration into the existing probation system. The DFE study used probation as usual (PAU, what we have referred to as traditional probation herein) as the control condition, so examining the degree to which the HOPE model is integrated into the existing system simply was not possible. Take as another example staff training. Lattimore et al. (2018) point out that standard training was provided to staff at the four DFE sites, but that the training was not included as an indicator of successful implementation.

After a number of years observing a HOPE-like program, we believe it is useful to consider the even broader list of key elements of HOPE and how implementation with fidelity to these additional elements may be helpful in better understanding why some HOPE-like programs appear more (or less) successful than others. Table 2 illustrates how elements of HOPE from DuPont et al. (2015) and Lattimore et al. (2018) do (and do not) align.

Table 2.

Key elements of HOPE alignment.

DuPont et al. (2015) (with element number)	Lattimore et al. (2018) (with element number)	Aligned?	If no, possible approaches to assessment
1. Shared Leadership, 4. buy in among key officials	1. Leadership	Yes	N/A (provided in Lattimore et al., 2018)
2. Integrating HOPE into existing probation system	N/A	No	Interviews with stakeholders; observation of stakeholder meetings, particularly at program inception^a
3. Statutory provisions that allow for sufficient Jail terms for non-compliance with conditions of probation	N/A	No	Observation of stakeholder meetings, particularly at program inception; review of program policies and legislative changes^a
5. Staff training in HOPE principles	N/A	No	Observing training sessions for stakeholders; interviews with stakeholders
6. Identifying the Appropriate Target Population and the Means to Make Referrals	2. Probationers High Risk	Yes	N/A (provided in Lattimore et al., 2018)
7. Identifying zero tolerance violations requiring immediate arrest	6. Exceptions for Missed Drug Testing	Yes	N/A (provided in Lattimore et al., 2018)
8. Warning hearing by the judge	3. Warning Hearing Compliance	Yes	N/A (provided in Lattimore et al., 2018)
9. Drug testing strategy with immediate results	4. Initial Drug Testing Frequency, 5. Stepped Down Drug Testing Frequency	Yes	N/A (provided in Lattimore et al., 2018)
10. Means to effect immediate arrest and take into custody	N/A	No	Observation of stakeholder meetings, particularly at program inception; interviews with stakeholders; observation of status hearings^a
11. Expedited warrant service	N/A	No	Observation of stakeholder meetings, particularly at program inception; interviews with stakeholders; observation of status hearings^a
12. Means to hold violation hearings swiftly	7. Time to Violation Hearing	Yes	N/A (provided in Lattimore et al., 2018)
13. Means to timely report to the courts: probation officers’ violation reports and recommendations	N/A	No	Interviews with stakeholders; observation of stakeholder meetings, particularly at program inception^a
14. A sanctions strategy that is consistent and appropriate	8., Sanction Type, 9. Sanction Dosage, 10. Sanction Certainty,11. Sanction Swiftness	Yes	N/A (provided in Lattimore et al., 2018)
15. Continuum of treatment	N/A	No	Review of participants’ files; observation of status hearings^a
16. Quality assurance for fidelity monitoring and feedback	N/A	No	Process evaluation to measure progress toward stated goals; research partner participation in meetings

Indicates we measured these elements using the techniques listed here.

As seen in Table 2, there is a good deal of alignment between the elements of HOPE in DuPont et al. (2015) and Lattimore et al. (2018). Where the elements align well, Lattimore et al. (2018) identify robust assessment strategies for implementation with fidelity to the HOPE model in programs like SAC, and as seen above, these are the ones we followed.

While this is a tremendously useful blueprint for understanding how well HOPE-like programs have been implemented with fidelity, there are a number of elements that may have been neglected thus far in the literature. We provide a number of possible approaches to assessing these elements. For example, in order to assess whether HOPE has been integrated into the existing probation system (element 2 in DuPont et al., 2015), researchers can interview stakeholders, and observe stakeholder meetings, particularly when the HOPE-like program is in the planning or initial stages. We observed all stakeholder meetings as the SAC program in Jefferson Parish was being planned and initially implemented in 2015 and 2016, and we observed the judge and probation leadership in particular taking pains to ensure the SAC program could operate in the existing probation structure of the parish. It was at these same meetings that statutory provisions that allow for sufficient jail terms for non-compliance with conditions of probation (element 3 in DuPont et al., 2015) were discussed at length. The judge took the recommendations from those meetings to the Louisiana State Legislature, which then wrote those recommendations into law for the SAC program. The means to effect immediate arrest and take into custody and expedited warrant service (elements 10 and 11 in DuPont et al., 2015) were also discussed at length at these meetings. The judge took care to solicit input from probation and law enforcement leadership, as well as officers, in order to devise a plan that would allow these elements to be implemented with fidelity. Briefly, both police officers in the cities in Jefferson Parish, as well as officers in its Sheriff’s Department, and probation officers serving in the SAC program were authorized to make arrests and take SAC participants into custody, and a transportation system was developed across these entities. Moreover, stakeholders devised an information sharing protocol within the existing recordkeeping database, as well as some more modern apps, to permit both timely and virtual information sharing about violations (element 13 in DuPont et al., 2015).

We were able to assess continuum of treatment (element 15 in DuPont et al., 2015) in a variety of ways, including periodic reviews of participants’ files, which contained information on drug tests, sanctions, and required conditions of probation such as drug treatment, cognitive behavioral therapy, and employment and educational assistance services. Continuum of treatment was also evident in the liberal use of praise and deliberate communication to participants that he and the other SAC staff cared about them and were committed to providing the support necessary for them to succeed in the SAC program (see especially Frailing et al., 2020, as well as Frailing et al., 2020 for a thorough description of these status hearings).

The implementation with fidelity to the HOPE model in the SAC program may be part of the reason there is evidence that the program is meetings its goals. Frailing et al. (2020) found progress toward the stated goals of SAC in their process evaluation of the program. There was a reduction in drug and alcohol dependence among participants as measured by fewer positive drug tests over the course enrollment, there was a reduction in prison overcrowding resulting from the SAC participants, who as noted were at risk of immediate incarceration for noncompliance with traditional probation, avoiding nearly 9,000 incarceration days. There was an effective use of criminal justice resources—those nearly 9,000 jail days avoided resulted in an approximately $4,000 per participant per year cost avoidance for the parish, even with the costs of SAC probation supervision factored in. Importantly, this cost avoidance did not compromise public safety; SAC participants had fewer revocations, new charges, and less serious new charges when compared to an ad hoc group of felony probationers in the parish. Finally, it appeared as though three sanctions was sufficient to decrease undesirable behavior and improve accountability.

Conclusion

There may be other ways to assess the elements in DuPont et al. (2015) that are not aligned with those in Lattimore et al. (2018), and those ways may permit better quantification that we have provided here. However, we feel the integration of these approaches offers the most robust framework for comprehensive evaluation of HOPE-like programs, such as SAC, which was very deliberately modeled after HOPE, to the point that Judge Alm provided input on the program at its inception. It is important to note that the evaluation framework we have promoted above is most relevant to those programs that have as their espoused theory the HOPE model. Programs that have a very different or perhaps even slightly different espoused theory certainly would not reveal fidelity to the HOPE model following an implementation assessment because the elements of HOPE would not, or even should not, be present in that program. In order to fully understand implementation, it is crucial that the correct elements are being assessed.

Our approaches also indicate the need for research partners to be present at the inception of HOPE-like programs and welcomed by the program stakeholders. Researcher partners are important not only to assess these key program elements, but to provide input on the type of data that will be necessary to (1) measure implementation fidelity, (2) adjust to programmatic changes as they occur, and (3) measure outcomes for these programs. Moreover, having research partners present at the inception of the programs may be useful in crafting goals that are measurable for both process and outcome evaluations (on this point, see Frailing et al., 2020, who note the frustration with measuring legislature-set goals that include words such as increase, decrease, reduce, improve, and so on, which imply that both baseline measurements and consistent measurement approaches are possible). The more thorough assessment possible for these programs, the more clear it will become whether they are effective, and if so, for whom and why.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by a Bureau of Justice Assistance Grant number BJA-2016-SM-BX-0005.

ORCID iD

Kelly Frailing

Notes

Author biographies

Kelly Frailing is an associate professor of Criminology and Justice at Loyola University New Orleans. Her areas of research interest include people with mental illness in the justice system, specialty courts and programs, and crime and disaster. [ Email: klfraili@loyno.edu]

Jahdai Kennedy is an undergraduate student at Loyola University New Orleans who is double majoring in Criminology and Justice and Psychology. [ Email: rjkenned@my.loyno.edu]

Rae Taylor is an associate professor of Criminology and Justice at Loyola University New Orleans. Her research and teaching interests include interpersonal violence, victim services, structural and organizational responses to violence, and mass incarceration. [ Email: rtaylor@loyno.edu]

Victoria Rapp is a master’s student in Criminology and Justice at Loyola University New Orleans. She plans to pursue doctoral education. [ Email: vlrapp@my.loyno.edu]

References

Alm

(2014) HOPE probation and the new drug court: A powerful combination. Minnesota Law Review 99(50): 1665–1696.

Alm

(2016) HOPE probation: Fair sanctions, evidence-based principles, and therapeutic alliances. Criminology & Public Policy 15(4): 1195–1214.

Bainbridge

(2019) Transferring 24/7 Sobriety from South Dakota to South London: The case of MOPAC’s alcohol abstinence monitoring requirement pilot. Addiction 114(9): 1696–1705.

Cook

(2016) Behavioral science critique of HOPE. Criminology & Public Policy 15(4): 1155–1162.

Cowell

Barnosky

Lattimore

, et al (2019) Economic evaluation of the HOPE demonstration field experiment. Criminology & Public Policy 17: 875–900.

Cullen

Pratt

Turanovic

, (2016) It’s hopeless: Beyond zero-tolerance supervision. Criminology & Public Policy 15(4): 1215–1228.

Cullen

Pratt

Turanovic

, et al (2018) When bad news arrives: Project HOPE in a post-factual world. Journal of Contemporary Criminal Justice 34(1): 13–34.

DuPont

Alm

Inouye

, et al (2015) State of the art of HOPE probation. Available at: https://nicic.gov/state-art-hope-probation-2015

Duriez

Cullen

Manchak

(2014) Is Project HOPE creating a false sense of hope: A case study in correctional popularity. Federal Probation 78: 57–70.

10.

Durlak

DuPre

(2008) Implementation matters: A review of research on the influence of implementation on program outcomes and the factors affecting implementation. American Journal of Community Psychology 41: 327–350.

11.

Frailing

Alfonso

Taylor

(2020) Therapeutic jurisprudence in Swift and Certain probation. American Behavioral Scientist 64: 1768–1785.

12.

Frailing

Rapp

Taylor

(2020) Swift and Certain probation as a HOPE-like model: Progress toward goals and lingering challenges. Corrections: Policy, Practice, and Research. Epub ahead of print 18 August. DOI: 10.1080/23774657.2020.1807425.

13.

Hamilton

Campbell

van Wormer

, et al (2016) Impact of swift and certain sanctions: Evaluation of Washington State’s Policy for offenders on community supervision. Criminology & Public Policy 15(4): 1009–1072.

14.

Hawken

Kleiman

(2009) Managing drug-involved probationers with Swift and Certain sanctions: Evaluating Hawaii’s HOPE. No 229023, December. Washington, DC: National Institute of Justice.

15.

Hawken

Kulick

Smith

, et al (2016) HOPE II: A follow-up to Hawai‘i’s HOPE evaluation. May. Available at: https://www.ncjrs.gov/pdffiles1/nij/grants/249912.pdf

16.

Humphreys

Kilmer

(2020) Still HOPEful: Reconsidering a “failed” replication of a swift, certain, and fair approach to reducing substance abuse among individuals under criminal justice supervision. Addiction 115: 1973–1977.

17.

Kleiman

Kilmer

Fisher

(2014) Theory and evidence on the swift-cer tain-fair approach to enforcing conditions of community supervision. Federal Probation 78: 71–74.

18.

Latessa

(2018) Does treatment quality matter? Of course it does, and there is growing evidence to support it. Criminology & Public Policy 17(1): 181–188.

19.

Lattimore

Dawes

MacKenzie

, et al (2018) Evaluation of the honest opportunity probation with enforcement demonstration field experiment (DFE). Final report. Available at: https://www.ncjrs.gov/pdffiles1/nij/grants/251758.pdf

20.

Lattimore

MacKenzie

Zajac

, et al (2016) Outcome findings from the HOPE demonstration field experiment: Is swift, certain, and fair an effective supervision strategy? Criminology & Public Policy 15(4): 1103–1142.

21.

Lipsey

(2009) The primary factors that characterize effective interventions with juvenile offenders: A meta-analytic overview. Victims and Offenders 4: 124–147.

22.

Louisiana Justice Reinvestment Task Force (2017) Louisiana Justice Reinvestment Task Force report and recommendations. 16 March. Available at: https://www.lasc.org/documents/LA_Task_Force_Report_2017_FINAL.pdf

23.

Nagin

(2016) Project HOPE: Does it work? Criminology & Public Policy 15(4): 1005–1008.

24.

O’Connell

Brent

Visher

(2016) Decide your time: A randomized trial of a drug testing and graduated sanctions program for probationers. Criminology & Public Policy 15(4): 1073–1102.

25.

Paternoster

(2010) How much do we really know about criminal deterrence? Journal of Criminal Law and Criminology 100: 765–824.

26.

Smart Supervision (2017) Swift and Certain probation program handbook. Available at: https://www.smartsupervision.us/judge/schlegel/forms-sac

27.

Smart Supervision (2020) Smart Supervision probationers. Available at: https://www.smartsupervision.us/smartprobationer

28.

Task Force (2017) Louisiana justice reinvestment task force report and recommendations. Available at: https://www.lasc.org/documents/LA_Task_Force_Report_2017_FINAL.pdf.

29.

Vera Institute of Justice (2010) More than the sum of its parts: Why Hawaii’s opportunity probation with enforcement (HOPE) program works. Available at: https://www.vera.org/downloads/Publications/more-than-the-sum-of-its-parts-why-hawaiis-opportunity-probation-with-enforcement-hope-program-works/legacy_downloads/HOPE_Policy_Brief.pdf