Professor Mark Kleiman reflects on innovation and trial-and-error, with the recent insight having seen his innovative probation model, now called HOPE, revolutionize the way probationers are treated in Hawaii.
Can you give a brief overview of the HOPE concept?
Traditional drug-diversion programs mandate that offenders go to drug treatment. HOPE mandates that they stop using drugs and backs that mandate with drug testing and consistent sanctions. It turns out that most offenders, even those with severe problems according to conventional risk and needs assessment tools, can and will quit under those conditions. The result is less crime, less drug use, and fewer days behind bars. By reserving formal drug treatment to those who either want it or prove to need it – rather than imposing it on everyone – HOPE avoids wasting scarce treatment resources, and can deliver high-quality and high-intensity treatment to the minority who use formal treatment services.
What need did the model hope to address?
Under the normal system, probation officers would bring revocation motions to the judge for probationers who violated the terms of their probation. Often, though, this was after a dozen or so violations. Due to probation officer caseloads and the amount of time it took to write up a detailed violation report, the first several violations were ignored. But this goes against what we know about changing behavior. You need clear rules and quick consequences. Sporadic sanctioning kept violation rates so high that probation officers simply couldn’t report every violation: a revocation motion could take hours to prepare, and there weren’t enough hours in a work-week. It was a kind of social trap. Judge Alm cut through the problem by starting with a small group of probationers, getting their behavior under control, and then expanding slowly.
What aspects of the program were fine-tuned through trial and error?
The most important innovation was the “warning hearing,” suggested by the public defender. It turned out that for about half of the participants, the warning alone did the trick. Another key change was a simplification of the reporting process for violations from an elaborate report to a two-page check-box form that probation officers could fax in. The program started out just in Judge Alm’s courtroom; then it was expanded to all the felony trial courtrooms in Honolulu. But that created scheduling problems for the lawyers, so the new plan is to make Judge Alm the “HOPE judge,” managing 3,000 probationers all by himself.
Were there any failure traps that the original model and Judge Alm didn’t anticipate?
He got most of it right. For example, he lined up the Marshals Service to serve the bench warrants, and told the probationers at their warning hearings about the consequences of not showing up. As a result, there weren’t many no-shows. The other problem that was overlooked initially was that probationers could simply plan their drug use around their scheduled drug tests, staying clean only the three days before the test. So the program randomized the drug test appointments, so there was never a time they could safely use without consequences.
How were the sanctions themselves determined?
The length and type of sanctions were also figured out by trial and error. At first, probationers were put in jail for a couple of weeks for a first violation. Then it became clear that the sanctions didn’t need to be that severe, as long as they were swift. So Judge Alm cut back to two days for a first sanction. Another judge started with six weeks. But when they looked at the results, more severity didn’t lead to higher compliance. So the more severe judge cut his sentences back. There’s no rigid formula, but two days for the first violation, a week for the second, and two weeks for the third seems to be about average. After a third violation there’s likely to be a mandate to treatment; that’s about one participant in six.
How much variation was there between the judges in how they implemented the program?
The program started with Judge Alm, but eventually expanded to other judges. The probationers ended up doing the same, regardless of which judge was involved. It wasn’t Judge Alm’s charm that made it successful. The common sense of it was all you needed to implement, which anyone could do.
Has the model been tested anywhere else?
Yes, it was tested in the DC Drug Court. One participant group went through the normal process with mandated treatment; the other group got frequent sanctions if they tested dirty. The sanctions track cost one third as much and worked twice as well. But when the short trial period ended, the judges returned to business as usual because business as usual is more comfortable.
What do you think will be the challenges of replicating HOPE, if any?
There are lots of places talking about replicating HOPE, such as Delaware and Las Vegas. I don’t think there are major obstacles to replicating it. There are only two variables: the specific probations and the implementation of the model. There’s nothing special about Hawaii probationers. As for the implementation, you just have to find a judge would wants to do it. The cases in which I’ve seen similar models fail is when the model itself doesn’t deliver the sanctions it promises or if there isn’t adequate buy-in from the judges.
Is there anything unique about Hawaii that might have made HOPE easier to implement?
Hawaii happens to have very competent probation officers. Many are social workers, and some have MSW degrees. They are well-versed in the techniques of behavior modification. Another unique aspect to Hawaii is its relatively simple institutional structure. New York, for example, is not as well integrated. Individuals can be on probation in more than one court, while the probation officer is assigned based on where the probationer lives. There’s not the same tight linkage between the relevant players.
Do you think there are any philosophical or dogmatic obstacles to replicating HOPE?
There are two main barriers: the belief that drug users can’t change their behavior without formal treatment – easily refuted by the data but held by some as an article of faith – and the belief that any probationer who steps out of line ought to be sent to prison instead of being given a chance to behave better in the community.
What are the next steps for the HOPE model?
More trials in more places, expansion to parole and pretrial release, and eventually – if I had my way – national adoption. This is what probation should look like. And it saves enough in reduced prison spending to more than pay its costs.
Do you have any advice for courts that implement the model?
Start small and be patient. Don’t start with thousands of probationers. Judge Alm started with 35 and then scaled up from there. Most importantly, try it. It is remarkable how many programs fail because they were never tried.