What Works in Offender Supervision

Moderator: Marlene Beckman, Senior Policy Advisor to the Assistant Attorney General, Office of Justice Programs

Marlene Beckman: OK. Good afternoon, and welcome to this panel. My name is Marlene Beckman, and I work at the Office of Justice Programs. Up until recently, I have been a grant monitor in the National Institute of Justice, and the presenters today are the grantees that I've been working with. Just for the sake of truth, I'm actually not with NIJ anymore. I'm working with Laurie Robinson in the Office of Justice Programs, which is an absolute delight for me, and I really love being there with Laurie, now, where I was when she was with the Reno administration.

So this morning were going to hear about two different studies that have in common the fact that they are looking at alternatives to returning people to prison who mess up or might otherwise be revoked and sent back to prison. They're two different approaches to the same problem, and I think that's what makes it interesting — to hear various options because as I'm sure all of you know, today the overcrowding in the prison systems around the country are even more due to the fact that people revoked are being returned to prison than actually new people entering prison. So this is a major issue and we need these options to look at in order to not continue to overcrowd our prisons.

So what were going to do is each of the three speakers are going to speak for 20 minutes. That'll give us plenty of time for questions in the end, and I'm a very strict timekeeper, as they know. First, were going to hear from Professor William Bales who's a professor at Florida State University, and his project has to do with the use of electronic monitoring. Then were going to hear from Judge Steven Alm, who is a judge on the First Circuit, Honolulu Court, and his project is the HOPE Project, which has to do with swift, certain and proportionate sanctions. And Judge Alm will also be appearing tomorrow in a plenary so you all get a sneak, quick preview. And then the third speaker will be Angela Hawken. Angela is a professor at Pepperdine University in California, and Angela has been doing the evaluation of Judge Alms study. Angela has an accent. I've asked her to speak slowly — you get used to it. So well start and have a half an hour for questions at the end.

[Off-mic mumbling audible]

Bill Bales, Associate Professor, College of Criminology and Criminal Justice, Florida State University, Tallahassee, Fla.

Bill Bales: Thank you Marlene, and welcome. I'm going to talk about a study we've been doing for, what, about a year, Marlene ... I think, something like that. Typically we got, like everybody else, we had to get an extension from NIJ. But this study's about evaluating the effectiveness of medium- and high-risk offenders on supervision who are also on electronic monitoring.

[inaudible]

I suspect most of you have heard of electronic monitoring. This is a technique that is only about 25 years old in terms of correctional strategies, and ... I would argue that likely its probably one of the fastest growing methods of community supervision. And I looked at some statistics. In 2000, there were about 30,000 people on EM in the United States. In 2006 — or no, 2005 — it was about 160,000. I suspect today its probably closer to 200,000 — so clearly EM has been accelerating over the last several years. And that's true in Florida; That's where this study is being conducted, is in the state of Florida. One of the reasons, at least in Florida, and this is, I think, true in other states, is the Jessica Lunsford Act. This was an Act that focuses on sex offenders and its mandatory electronic monitoring if they violate while on supervision and the judge decides, I'm not going to put them in prison. I'm going to put them back on supervision.

There's also been a shift away from what was called radio frequency, which is where someone was just tethered to their home, electronically, to global positioning systems. These are methods where someone is tethered to satellites up in the sky, going around the earth, and so officers can literally know where they are every second of the day, and that's recorded. And they know where they are within four feet, something like that. So its pretty amazing technology. In fact, the DOC in Florida put me on it for about three weeks and ... which was an interesting experience. I got some interesting looks in Publix and Lowes and walking around and so forth. So the way it works — just real quickly — at least in Florida is you ... the offender has an ankle bracelet that's immovable unless they decide to cut it off, and then they carry around a box — a black box. And the bracelet is connected to the black box and the black box is connected to the satellites. So if they get away from that black box, just a few feet, then there's ... an alarm is set off. That's how the technology works in very simplistic terms.

The fact is the research really hasn't kept pace with the growth in this technology. There have been several studies, but they're not real comprehensive. There have been some qualitative studies and there have been some empirical studies, but clearly, given the growth in this methodology, more research is needed, and fortunately NIJ was kind enough to fund our study in Florida. And we expect this study to be done in December, and what I'm going to talk about today is very preliminary results, but I ... hopefully you'll find them of interest. And were doing a process evaluation, looking at whether or not, you know, EM is being implemented as intended. Were doing an outcome evaluation and a cost-effective analysis. And the data that were getting is, we've obtained, is from the Department of Corrections in Florida. Anybody here from Florida, by any chance? OK. Florida has what I think is very, very good data — probably one of the best correctional data systems in the country. And they've shared with us dozens of files with millions of records ... and Ill talk a little bit more about the empirical outcome in a minute.

The other thing that's a little bit unique about this study, I believe, is that were doing both an empirical assessment of EM and its effectiveness, and were also doing a qualitative component through interviews with offenders, officers, and administrators. We had done a study at Florida, house arrest cases, and the effectiveness of EM on those particular cases, that was published in [i]Criminology and Public Policy [/i] in 2006, I believe, and we found very good effects. And I'm not gonna talk about that in any detail, but the point is, that while we found a high correlation between being on EM and having success, we don't ... we couldn't answer the question of why. What is it about EM that reduces the likelihood of failure while on supervision? And we surmised that, well, it may be a deterrent effect, it may be a social bonding effect, that offenders on EM are more likely to stay at home, be with their families. They're more likely to go to school and work, and maybe they start to bond to conventional lifestyles, so that's one of the purposes of the qualitative portion.

Again, were looking at the effectiveness of EM on failures on supervision. Also, whether or not EM has a lasting effect. In other words, maybe an offender is successful while they're on EM; once they're taken off EM and not being watched every second of the day, maybe they are more inclined to fail. Whether or not EM works for certain kinds of offenders — does it work better for sex offenders than drug offenders, older offenders, younger offenders, males, females, that sort of thing. And again, how and why EM is effective?

And were also interested — and this is through the qualitative portion — is how does EM effect offenders in terms of their daily lives. How does it affect their families? How does it affect their employment possibilities? The things that we sort of take for granted as being normal parts of our lives.

The quantitative part, were looking at several different types of outcomes: new offenses, technical violations, revocations that result in jail or prison ... so, more consequential outcomes ... and also absconding, when offenders just flee and they're, they cant be found. Obviously, our treatment effect is EM and we do, because of the richness of the DOC data, we have, I think its 128 different variables or factors that are commonly known to influence outcomes on supervision. So we have a tremendous number of control variables, so we can identify the unique effect of EM on these outcomes.

And the research design for the empirical part is ... begins with propensity score matching. And I'm not going to go into the math behind propensity score matching today, but the basic concept is that propensity score matching is a statistical method to get as close to a true experimental random design as possible. Even though its ex post facto, right, these people have already been on supervision. They've either failed or succeeded, but through this method, you can essentially match ... you can take an experimental group, in this case EM people, and find equivalent people on all these 128 variables, right, and get equivalent people.

So, again, its not random assignment, but its as close as we can get, statistically. And were also doing Cox regression, or survival analysis, to look at whether they fail and the timing to failure. I'm going to skip over that part.

So again, that's — I'm not going to repeat these. The basic questions are: does EM work? Who does it work for? Is the program implemented as intended? And, again, what are the ... what's the view of EM from officers, and offenders, and staff?

In this outcome, the empirical outcome study, we have 277,000 cases in total, and these are people placed on supervision of various types from 2000 to 2006, and 6,731 of those people are EM cases. So again, through this propensity score matching, we can find equivalent control cases based on all these factors to those on EM. And that's called a balancing approach, and once you can achieve balance, you know you have equivalent groups. And I wont go into all the details about the factors, but its a host of things about their criminal history, their demographics, their prior employment. Lots and lots of things that we know from prior literature are related to the likelihood of failure on supervision.

So were pretty confident that we have equivalent groups, non EM and EM, and that, of course, if you don't have that, then you're ... what you find is not going to be reliable from a scientific perspective. And we've been building models using different periods of supervision and our preliminary results are statistically significant, OK? So, all things being equal, were finding initially that EM offenders are less likely to fail while on supervision than non EM offenders.

And what we've done so far is kind of the heavy work, and this is sort of process, in terms of putting all this data together and doing propensity score matching. So now were ready to run models on lots of different types of offenders, looking at, you know, do Jessica Lunsford Act cases, ... does EM work well for them as it does for non-JLA cases? Looking at different types of supervision, in Florida, there's about eight different types of felony supervision: probation, sex offender probation, community control, which is house arrest, and actually, through the ... I mean, I've read about these, and I understand them ... their rules and the statutes, but going out into the field and talking to officers about their ... the types of supervision, we've picked up a lot more nuances about the differences in how offenders are treated under different types of supervision.

While this wasn't written in the grant, Marlene, wed also like to look at post-release employment while on supervision and whether EM empirically has an effect on employment. And DOC, fortunately, collects very detailed information about their employment status when they're put on supervision, all the way through until they, they're finished. So were ... Jerry Gace, who is helping me on this project, along with Kelly Barrack from RTI, I talked to him, went to dinner with him last night and was telling him about this idea of employment and he was all excited. He thought that would really be good.

Turning to the qualitative component of this study. Again, we've been interviewing EM offenders, officers, and administrators and we conduct these interviews; Its myself and one other person. The other person records it on a laptop. I take notes, but I conduct the interview. And these things take about 45 minutes for offenders and officers, and about 30 minutes for administrators. And we pilot tested the instrument; we got the instrument approved by the Human Subjects Committee and also the DOC.

Its really been a ... typically, I get databases and I, you know, do analysis, and so to actually go out there and talk to sex offenders and murderers and all kinds of offenders ... and these are, again, medium- and high-risk people; these aren't low-level probationers, typically. Well, they aren't ... is really a pretty fascinating experience. If you ever get the opportunity, I would recommend it.

The, just real quickly, these are the numbers of interviews we've done to date: 62 offenders, 24 officers, and 16 administrators. We have scheduled interviews in four other counties, fairly large counties, and ... that'll end up at the beginning of August. Were also going to go to what's called the Command Center; its in a little town outside of Tampa. The department implemented this Command Center about a year and a half ago because these, this equipment tends to falter pretty often, and there are false alarms all the time. Primarily, offender ... the box loses a signal to the satellites because an offender is in a building, a store, and so forth, and so they have to walk outside for fifteen, twenty minutes and the officer would get an alert. He would say, Walk outside.

So there are a myriad of false alarms, basically. And with this command center, they get the initial call, so it doesn't go to the officer. And they can deal with that call, you know, 95 percent of the time its ... they solve it, a notice is sent to the officer, but the officer doesn't have to deal with it. And we found out from officers that they love that Command Center.

This is just a map of Florida, just to give you a sense of ... that we feel like we have a pretty good geographical diversity in terms of where these interviews are being conducted.
Ok ... I'm just going to go through about four different issues that were trying to address with these qualitative instruments, but ... there's several more. The little boxes at the bottom are quotes from offenders or officers. And so we've been collecting quotes, and some of them are, if you get a chance to breeze through them, I think they're ... I'm not going to read them to you, but I think they're pretty interesting, and I think it adds a little more flavor to some of their perceptions about EM.

But the first issue is sort of officers and offenders perceptions about the effectiveness of EM. And this first part looks at offenders, so the first question being, does being on EM make you more or less likely to, say, abscond? Well, only 20 percent said less likely, 74 percent said it doesn't affect them, and 7 percent said more likely. Violations, a little bit more, 24 percent said it does prevent them from violating the conditions of their supervision. And in terms of new crimes, it goes up even more for offenders. 28 percent said, yeah, being on EM, I'm less likely to commit a new crime. And this was interesting. In terms of getting caught for a crime, 71 percent said they thought they were more likely to get caught for a crime because they were on EM.

Ok, I'm going to read this quote: The GPS gonna tell on you. That's what one person said. And its interesting, too, because what officers do is, they project this GPS equipment as if it actually can see the person 24 hours a day. A lot ... some of them will say, Its an eye in the sky. They know where, what I'm doing all the time. They can see me, so why would I go and do something, commit a crime or violate, because they can see me. Its pretty interesting.

Officers, a little bit more positive about the outcomes: 54 percent said their — offenders are less likely to abscond. 75 percent said they're less likely to violate, so its not surprising. But still, an appreciable amount, say, well EM really doesn't affect their ... the offenders behavior. And in terms of committing a new crime, its actually a little bit lower, 63, thank you, percent versus 37 percent.

We asked a series of questions about their — the effect of, or the impact of EM on the offenders employment, and in terms of obtaining a job, 63 percent of the offenders said it affects their ability to obtain a job. And all of those, 100 percent of those, said it was negative. And some of the common reasons: stigma of being on EM, the device raises questions and concerns, and, a quarter of them said, well, especially in this economy, the employers can find someone who's not on EM. And its basically, this, the visibility of this black box is problematic. You know, you can imagine, someone wants to be — we hear stories all the time. I went to apply at, you know, this restaurant. Well, they don't want me to be a server carrying a black box, cause then people are going to look at me and say, Oh, that's a sex offender. I'm not being served by a sex offender. So, it clearly, employment is a real issue. In almost, well one in five said they've been fired because of being on EM. And the other problem, too, is because the technology is not as good as one would hope because they lose signals all the time ... you know, if you work in a warehouse, and you have to leave every 30 to 45, 60 minutes and go outside for 20 minutes, after a while, your employer is going to say, Heck with this. I cant, I need someone who can do the job on a consistent basis.

And this is interesting. The officers were in very much agreement with the problematic nature of EM relative to employment. 79 percent said its difficult for offenders to find a job. Like the one quote said, Who wants to hire someone wearing a box?

And in terms of the likelihood that they stay employed, 71 percent of the officers disagreed. They think that EM reduces their likelihood of staying employed. This question about, can EM upset employers, causing an individual to lose their job? Almost 80 percent of officers agreed with that statement. The, this is somewhat telling also. We asked the question, ok, once a, an offender gets off EM, is their job affected? Are they, does it change? And — almost 96 percent said yes. And basically, what they're saying is, well, 41.7 percent said the signal loss causes an inconvenience to employer. That goes away; the offender has more chance of getting a job and staying in a job. So again, its, the officers equate it mostly to the equipment.

Do you believe ... the offenders job situation changes once they are off EM? 83 percent said yes. We asked about shaming or embarrassment among offenders in terms of being on this black box, and 54 percent said they were embarrassed. I kinda like this quote, Its like painting a flag on your forehead, saying, I'm a criminal.

If they were ashamed to be on EM, 53 percent said yes. And It serves as a scarlet letter. That was a common quote, actually. That was not just one person saying that.

[inaudible]

Were also interested in terms of whether EM is an alternative to incarceration, and so some questions related to that. We ask offenders, do you think the judge would have given you ... what kind of sentence do you think the judge would have given you if EM didn't exist, the judge had — didn't have that option? And a third of them thought that they would have gone to prison or jail. And although 18.6 percent said they'd rather go to jail. I mean, EM is a — we have another question there about, do you think EM is a punishment? And that's a high percentage of both offenders and officers say EM is a punishment.

We asked officers whether they thought EM was an alternative to incarceration, or an enhancement. A quarter of them said it was an alternative, but three-quarters say, think its an enhancement. One minute. OK.

Here's some other data that shows officers think — 27 percent of the officers think that 50 to 75 percent of their offenders would have gone to prison if not for EM. So ... Ill skip over this. So basically, officers do see it, depending on the offender, as an alternative to prison. If the judge didn't have this option, this offender would have gone to prison. And I wont go through this data, but what we found also from an officers perspective, or officers perspectives, is there are a lot of offenders that they supervise on EM that they say really shouldn't be on EM. And that varies by the type of offender. And on the, ... on the flip side, they have a lot of offenders that aren't on EM that they truly believe should be on EM. So there's some, certainly, incongruity between who's on EM and not on EM and the officers perspective as to who this tool should be placed on to keep them from failing.

[Applause]

And officers, 54 percent thought EM should be expanded, more EM; 25 percent said kept the same, 21 percent say reduced. So officers aren't you, you might expect that they would be unilaterally enthused, supportive of this EM process. That's not quite the case. OK.

Beckman: Thank you.

Bales: Did I make it?

Beckman: Well have more opportunity for questions and elaborate on some of the slides.

[Whispering.]

Unidentified Woman: I'm not starting yet. Everybody wants their minutes.

The Honorable Steven S. Alm, Judge, First Circuit Court, Honolulu, Hawaii

Steven Alm: Good afternoon everybody, I'm Steve Alm; I'm a judge from Honolulu, Hawaii. I think it's an absolute coincidence that the two places Marlene approved for — this was Florida and Hawaii, OK?

[Laughter.]

I was a career prosecutor. I was a United States Attorney from '94 to 2001. I have to say Laurie Robinson's words rang really true about going with the science, going with what works, and that's why I got started in this. I was appointed to the circuit court bench in 2001, and I got assigned to the current assignment, felony jury trial judge, regular case of, you know, calendar of felonies from arraignment, and we do all the pretrial matters, motions to suppress and the like, you know. Its murders, rapes, robberies, theft. Hawaii has one of the lowest violent crime rates in the country. It has one of the highest property crime rates. Were all in the top, like, three or four, and methamphetamine has been the drug of choice for 20 years.

I think our situation is probably very similar to you folks in your own jurisdictions. We have 8,000-plus offenders on felony probation or deferral on Oahu, which is where Honolulu is, the main — 85 percent of the population in Hawaii. Probation officers have these huge caseloads. Most, 80 percent of the cases involve drugs; crystal meth for 20 years has been our biggest drug issue — and I would get motions to revoke probation from the day I started on that calendar, with often two or three pages of violations going back, you know, six months, a year, two years, three years, all, missed appointments, dirty UA's, not going to treatment, [refusals] to go to treatment, often ending with the person absconding, you know? And I realized, the probation officers, like the judges, thought they had a choice between two. If people are doing great on probation, that's fine, but many don't And then the probation officer thought their only choice were, they try to work with the person, you know, talk to them, threaten them, cajole them, and tell them, If you keep this up, if you keep testing dirty or missing appointments or not going to treatment, next year, you're going to go back to court and the judge might send you to prison for five or 10 years.

Well, you know that kind of thinking barely works with us. It certainly doesn't work with the people were working with. You know, next year they could win a lottery, they could get hit by a bus. Their uncle could die and leave them money. You know, its meaningless, that kind of way-down-the-road thing. And who knows, maybe years ago a probation officer actually filed a motion to revoke probation with two violations on it, and the judge said, I'm not going to send somebody to prison for five or 10 years based on one dirty UA and a missed appointment.

So I sat down and I thought, what will work? What can I do? What can we do to change offender behavior? And I looked at the statutes, but mostly just thinking. And then I thought — well ok, what do I do as a father? If my kid screws up, I sanction him right away. I don't send him to jail, but I do sanction him, take away privileges, you know, he can get grounded. Whatever it is, he was a little young to get grounded at that point, but you talk to him and you give him a swift and certain consequence. That way, hell tie in whatever he did wrong with the consequence. And so, I thought, OK, how do we go about doing that in probation? Well, I got together with one of the senior probation officers who's creative and willing to try looking at an old system a new way. I then brought in a supervisor at the prosecutors office, the public defender; we talked for a while. I told the prosecutor, We've got to make this labor non-intensive, or the probation officers will never want to do it. They've been doing LSIR training, motivational interview training, and here's this judge coming with another new idea to try, right? So the, I asked the deputy prosecutor, the supervisor, he came up with a three-page combined motion to modify probation, because I said it should be a modification; Then court staff doesn't have to do a new judgment. They do an order that is very quick, takes two minutes to do, because again, court staff, with judge leadership, if we care about it, the staff will care about it. But its got to be quick and easy to do.

The public defender said, "You know judge, you're going to — now you're talking about enforcing all of the rules on probation," because that's what were doing; were deconstructing the motion to revoke and going back to the first violation. Were going to arrest them as soon as possible and have a hearing a couple of days later. They're held in custody in between. And they get a short stint in jail. That is the basic core of this.
Now, how to make that whole system work was the challenge. As U.S. Attorney, you know, working with Marlene Beckman, working with the folks at Justice Department, we set up a number of weed and seed sites. We set up a weed and seed court in State District Court, so I was thinking along the lines of a swift and certain consequence anyway. We looked at drug courts and thought, you know, is there any parts of that that could be useful? Because we had a drug court at the time, it had between 50 and 100 people, a dedicated judge, to do this, but I was a full time trial court judge, so what could we take from them that would make this work as well? And so it really is this collaborative strategy. Its bringing everybody together, and this is one of the hardest points. That's why what I handed out to you folks is a series of benchmarks we've come up with. That if any place is thinking about doing this, I would be happy to talk to the judges involved, especially on where they're going to have to think differently to do this. There are eight or 10 states that are looking at setting up pilots right now, but if you don't have all these steps — when you read through those steps, you'll realize this is not an easy thing to do. You have to get everybody organized. But the idea is, if you can set up swift and certain consequences, you can change behavior of the offenders.

And the idea — this isn't a magic potion or anything. Its a tool to make probation work the way it ought to work, and probably the way some of the public thinks it works right now, right? Oh, if they test dirty, they must get arrested, right? Or they must get locked up for it. The idea is, if we can keep offenders drug-free, seeing their probation officer and going to treatment, whatever the treatment may be, then they have a real chance to succeed on probation. Because the probation officers can be trained up, and ours are very well trained — motivational interviewing, cog skills — but unless the offender is sitting in the seat and not high, they're not going to get the benefit of that. We targeted the highest risk people, and that is one of the differences with drug court. They wont take the violent guys, sex offenders; the prosecutors have a veto pretrial, which is why we don't have as many people as we should in our drug court. Were after the ones that have the most problems on probation: violent, whatever, the judge, if the people are violent or dangerous at sentencing, the judge should send them to prison. If they're going to be on probation, then we should try to get the hardest folks in here, the most troublesome folks.

We, based on the public defenders request, we actually started this with a warning hearing. So we bring the offenders to court, and I tell them, Because you're not in prison, and you're on probation, you're making a deal with me that you're going to follow all the rules of probation. If you don't want to do that, tell me now, and Ill send you to prison. Because from now on, you're going to have to follow all of the rules of probation. The probation officer is my eyes and ears out there. Whatever he or she says, you have to follow, or you're going to go to jail. If you don't see your probation officer, you know — what I know is, you know you're going to test dirty. That's why you're not going. Or you're doing something you shouldn't do, or you're blowing off the probation officer and me. All three, any three of those, and you're going to go to jail. If you come in and test dirty, you're going to get arrested on the spot, and well have a hearing a couple of days later.

So that's how we set this whole thing up, is to have swift and certain consequences. Now, we have them call a drug test hotline, lets see if I can call it and ...

Recorded Voice: Today is Monday, the 15th, 2009. Thank you for calling the HOPE Hotline, please listen carefully to the entire message for any new instructions. Today's colors will be at the end of this message. All clients scheduled for UA's today will report to the Adult Client Services Office at 777 Punchbowl Street--

Alm: Courthouse.

Recorded Voice: As a reminder, please bring a picture ID. At the reception desk, please sign in on the HOPE sign-in sheet, sit on the bench in the hallway outside of the lobby, and wait to be called. You are instructed to report and provide a sample between 7:45 a.m. and no later than 2 p.m. Please note that you will be subject to an arrest if you are unable to provide a sample by 2 p.m.

Alm: So they don't come at quarter to and say they cant pee.

Recorded Voice: Once again, you will be subject to an arrest if you are unable to provide a sample by 2 p.m. Finally, UA's scheduled for today are for Blue One and Orange Two. Again, UA's scheduled for today are for Blue One and Orange Two. Thank you and have a great day.

[Laughter.]

Alm: You have to be positive!

[Laughter.]

When they start, their colors will come up at least once a week, sometimes two days in a row, six times a month. If they show up all the time and test clean, they're given a new color after two months that's half as often, then half as often again as a reward for being sober and following through with it. Anybody [who] wants to go to treatment can go to treatment. On the other hand, if they can stop using drugs on their own — and some can — its a myth that people, whether its methamphetamine or anything else — are all addicts. And that's one of the beauties of this program is, the ones that don't need to go to treatment don't go. Then the ones who do go and take the precious treatment resources really need it.

Now, we started with 34 people on October 1st of 2004. I did a warning hearing, 18 sex offenders at 8:30, 16 drug offenders convicted of a variety of felonies but failing at probation, headed for revocation. And I told them what all this warning hearing was about. And so from now on at warning hearings, I have a WANTED poster, and I say, If you don't show up for a probation appointments, or you don't show up for a drug test, the fugitive task force-- because due to my law enforcement background, I was able to get the HIDA that we started, and the U.S. Marshall with the Fugitive Task Force serving the warrants from my court. So I say, You've seen these posters at post offices and elsewhere. These guys who make these arrests are going to come and arrest you. They're not going to come alone. People can get hurt. Its embarrassing. So even if you screwed up, you got high or you missed an appointment, come in anyway. Because you will get sanctioned, but not for as long as if we have to use their resources to hunt you down. Because the more responsible you are on probation ... the more freedom you're going to have, and the less I'm going to sanction you.

Now I had a contest among the probation officers to come up with a name for this program. One of the early entries was Yank and Spank. [Laughter.] And I thought, you know, its as ... its pretty accurate, but its not aspirational enough. So when the nomination for Hawaii's Opportunity Probation with Enforcement came in, I thought, yes. We have a good, good name. If you have a good program you've got to have a good name.

And were not, you know, I'm not kidding myself. This is not a be-all and end-all. The short-term goal is to make these folks more ... and I say guys more, its 90 percent men, 10 percent women ... these are observed urine tests. You've got to have testers for both. The idea is to keep them sober, seeing their PO, going to whatever treatment, and then in the long term, they have a chance of changing their thinking, which is what you need to do to change behavior. That's absolutely critical. Now, some of these folks have been in prison before. Some of them have been in ... virtually all have been in jail before. And the thing is, when people at first told us, You know, some of these people have done time. I don't think this is going to work, and I think you know but human nature being what it is, people can do time when they have to, but they don't want to do it today, you know? Its the disruptive nature of it that works. I use today; I might get tested tomorrow and go to jail tomorrow. I don't want to go to jail tomorrow, so I'm not going to use today. You can string together days of that; they can go to their treatment programs if they have that. They can go to AA/NA meetings. They can do whatever it takes. Because we tell them at the warning hearing — we want everyone here to succeed.

And now I'm doing them one at a time, or I'm doing them at sentencing when we put people in HOPE. We want you to succeed. Your public defender does, the prosecutor does, the taxpayers in Hawaii do! It costs $50,000 a year to lock people up in prison in Hawaii. It's $27,000 a year in Arizona, which is why were sending 2,000 inmates to Arizona, disrupting their families, you know, causing them real problems that way.

Ongoing developments — we have six felony judges doing this now. Three misdemeanor felon circuit court judges but doing misdemeanor domestic violence cases. Because a year into this, we started with the DV felons, and I said to the DV folks, Give us the 30 guys that are having the most problems on probation. Out of the list of 30, 20 were on bench warrant status. I guess that's not a big surprise. I gave the list of 20 to the fugitive task force, and I said, you know, Try to go arrest these guys. The other 10, I did a warning hearing with the 10, all young, all male, all violent, a couple had attitudes in court. Well, they behaved just like everybody else. They started seeing their PO; they started going to the domestic violence batterers classes. If they had substance abuse issues, they started testing cleaner. It doesn't seem to matter who that is. And its the same results with all of the judges, because it may come as a shock to you, but not all of my colleagues are quite as enthusiastic about buying into this program as I am. And judges, you know, this expression about herding cats, I used to think that was cute when Id hear it at other conferences. Well, I don't think its cute any more. Trying to, you know, judges are used to having total discretion over things, you know, doing it the way they've always done it, and nobody can tell them what to do. So, this was a challenge. Most of the judges are totally on board with this because they see the results.

We started with 34 people. Now four and a half years later, we have over 1,475 in this, including 1,250 felons. Its a moving target because every week we have folks finishing probation. Its all five years in Hawaii. And we put — have new people starting every week. I personally am supervising 600 offenders on this, in addition to my regular felony load. The only reason this works is number four — virtually no hearings are contested. And, I, there are two reasons for that. One, were not going to throw the book at them. They know that. Second, were talking about what happened two or three days ago, or last week, not what happened a year ago. So we come into court. I say, Ms. So-and-So, the public defender, what does your client want to do? He wants to admit to it. Ok, Mr. So-and-So, you have a right to have a hearing. Prosecutor will bring a witness, the probation officer, they try to prove you violated a condition of your probation. If not, you can admit to it, you did it, that's fine. Then I ask the prosecutor, Your, recommendation? the public defender, Yours? And then I take — we talk to the defendant and then I give him whatever I'm going to give him.

The prosecutor gives me the ... I mean, the probation officer sends an email with their recommendation attached to a template that has the whole history of the case. Then they just have to update it: when the person was on probation, any prior modifications, what the consequences were, what the current violation was, what the probation plan is, and the like.

So, typical sentences, first time violations might be credit for time served. We try to do these two business days later. So, as you know, in the jail world, that's three days, right? Monday, Tuesday, Wednesday. In the rest of the world, that's 48 hours, but that's often what well do. Some of the time, well make them go through the weekend, because it kind of doesn't hurt. They miss whatever they had planned this weekend. Because I told them at the warning hearing, You may have a big event coming up. Your kids first pop warner football game. Well, if you're get ... you get high on ice during the week, you're going to miss that, and that means you chose the ice pipe over your kids football game. So don't kid around. You're an adult. I'm an adult. I cant control what you're going to do, but I can control what I'm going to do, so if you do this, this is going to happen. And I go through it.

We have one of our drug testers, who went to the legislature four years ago and got $1.2 million because we already had stats that the attorney generals office kept. Later on, were getting top-of-the-line research with Dr. Angela Hawken from Pepperdine, and shell be coming up after me. But from the beginning, we had these really great numbers of reduced drug use, increased missed appointments. And part of the money — $770,000 of that $1.2 million — are for drug treatment slots. Were big believers in treatment, but it should be for the folks that really need it. And treatment works, but it works even better if the person in treatment know they're going to get arrested if they walk out the door.

One of our drug testers and three of the positions of the $1.2 million are $25,000-a-year clerks. They do all the drug testing now. So that woman you heard on the phone and two of her male colleagues do all of the observed urine tests. We have same-day hearings if somebody has a paycheck job, at least for the first time, we take him into custody. If he's con ... if he says, you know, I've got a full time job I've got to get to, we take him into custody, bring him to court. Within a couple of hours, we get the attorneys there. Hell admit to it, and Ill give him a weekend or two in jail, but Ill tell him, Don't count on this the next time, because you could be a hazard to other people at work if you're loaded.

Maybe allow passes for people to go to a treatment program to get assessed if they seem to have trouble making it in the front door to get assessed. That means a responsible adult will pick him up at jail, take him to the assessment, take him back to jail. We come back to court a week or two later.

Friends of HOPE is a non-profit looking at the drug court. They've got ... a number of them have nonprofits. They're going to provide child care for anybody in HOPE who needs to go to job training or job interviews and don't ... doesn't have family to watch their kids. They're going to pay for dental work for anybody who needs it.

I would strongly suggest, we don't have time for it now, but if you want to talk to me, the cutoff level for drug testing, there's a federal workplace safety guideline level. That is too high. And it has, who cares if they pass the federal workplace safety guiding level? That's 500 nanograms per milliliter for methamphetamine. The labs with the GCMS can actually test it down to 125. That's what you should be using. So I would ask about it and follow through with that.

I got no extra staff to start. We got no extra money. We did it for a year without any extra dollars. And we had these great results, so it should get started small as well. Anybody contemplating it, on the benchmarks who say under 100, Id suggest 50. Get all of the moving parts together, and if it ... and once you start seeing good results, which I'm confident you will, you can expand it then. We have no due process issues because you have counsel, prosecutor, public defender, private counsel, in court, on the record. They're waiving the hearing and that's why were doing it.

I'm going to turn it over to Angela Hawken. Thank you.

[Applause.]

Angela Hawken, Assistant Professor of Economics and Policy Analysis, School of Public Policy, Pepperdine University, Malibu, Calif.

Angela Hawken: I believe that was the first time Steven Alm has spoken more quickly than I do. Very impressive, Steve.

Good afternoon everybody, and thank you to Marlene Beckman from NIJ for the opportunity to be here today, and also for her for providing us with funding to make this evaluation possible. Its been a thrill to be involved with the evaluation of HOPE ever since, I guess, 2006.

Nonviolent drug offenders are very ... are rarely given jail time. I think many people, when they don't know much about the criminal justice system, assume guys are picked up with a couple of joints in their pockets and they're sent away to prison forever. It doesn't usually work that way. And most people are put on probation or parole for some kind of drug violation, which makes the probation and parole departments extremely important in the fight against drug use. They have this captive audience under their supervision, that they're supervising, that can help us affect these decisions about using drugs. Unfortunately, our parole agents and our probation agents have very high caseloads and very limited resources with which to work.
The accent, by the way, is South African.

Very high caseloads with very limited drug treatment resources, we've seen this across the country, and many of the offenders that were supervising have very serious drug addictions and need good quality care. A big issue, though, is that these are often considered small fish. They're nonviolent drug offenders, so given all the problems we have, who really cares? Well, the small fish grow up to big fish if they're left unchecked. I've studied hundreds of thousands of nonviolent drug offenders. Those ... I kept an eye on those who came in with their first nonviolent drug violation, with no priors whatsoever. Within two years, 15 percent of those individuals had transitioned to high-cost offenders, that is, multiple property violations or crimes against people; so small fish eventually do become big fish.

How HOPE is different from drug courts — one of the problems we have when we talk about HOPE around the country is, everybody says, Well, were doing it already. We have drug courts. And some of you might be thinking the same thing. HOPE is very different from drug courts in the following way: The role of the judge is different. In drug courts, the offenders come before the judge on a regular basis. It could be every couple of weeks, once a month, depending on the program. And at HOPE they only come before the judge if they've violated. You don't see the judge again if you're managing to sort yourself out on your own.

The role of treatment is also very different. Under drug courts, everybody is mandated to treatment while they're being supervised under a drug court, and at HOPE you only go to treatment if you either ask for it or you've demonstrated through your dirty urines that you're unable to desist from drug use on your own.

The result of these two effects — the difference in the use of the judge, the differences in the use of treatments — means that a program like HOPE can scale very differently to what a drug — to how a drug court can scale. There's a limit to how many offenders the drug court system will ever be able to supervise. At the moment, drug courts supervise about 70,000 offenders — drug offenders. Just to give you an example, in California, every year alone we sentence about 70,000 individuals on nondrug ... on nonviolent drug charges, so we have a real problem with scale here.

How HOPE is different from a diversion program, many of you would be familiar with, Arizona moved first with treatment diversion with Proposition 200. California followed with Proposition 36. How these are ... how those are different from a HOPE-style program is the role of sanctions. Under diversion programs, they're very scared of using sanctions. California doesn't allow them. We don't allow jail sanctions in Proposition 36. The role of treatment is also different. They're typically mandated to treatment. Again, HOPE does not mandate treatment. And how the treatment decisions are made are very different. Under a diversion program we do what is called assess and treat. We assess you, which is based on self-reported behavior, and then we send you off to a treatment program. Under HOPE, treatment decisions are based on observed behavior by the offenders.

Why this is a very important distinction is because, this might not come as a surprise, offenders lie. I've just finished a study of offender behavior, and what they say about their drug use versus what they really do. Fifty percent of the lads, they were all lads, 50 percent of the lads that I interviewed said that in order to avoid the referral to jail and to get into treatment instead, 50 percent of them had lied. They had over-reported their drug use to avoid jail and get a referral to treatment. Sixty percent of them had said that they had under-reported their drug use during an assessment to get to a lower level of treatment supervision. So ... you can do the numbers yourself. There's not very credible sources of information from the criminally justice-involved drug offending population. So, given we have such little information on what's really going on with them, using observed behavior is a very strong indicator of what their true needs are.

So we think of HOPE as a behavioral triage model. Wed watch how they behave, and then wed triage them into groups depending on their behavior. If they're testing dirty, we know this person has a serious problem, even when faced with the credible threat of sanctions, if they're still testing dirty, lets get them into high-quality treatment and quickly — a very strong indicator of need.

Well, how does HOPE do as a behavioral triage model? Is this the model of our dreams? Were looking for a model that would do this kind of thing; how did HOPE perform? Well, were very happy to say that, with the exception of a handful of cases, I recorded under five, sanctions were certain. Every time they violated, they were sanctioned under HOPE, which is incredible. The reason swift and certain sanctions programs have failed elsewhere in the country is that the judges were not willing to follow through. They did not credibly and consistently give those sanctions, and at HOPE it did happen.

The sanctions are swift; 70 percent of the hearings are held within 72 hours. Moving the system that quickly is also very unusual; Its an incredible feat on the part of the Hawaiians. These hearings are quick. We recorded all of the court cases; we transcribed all of those court records. The average HOPE hearing, the motion to modify hearing, is about seven and a half minutes. So this is a very quick turnaround time in a court. The warning hearings are also quick because they're typically done en masse. The typical warning hearing is about 14 minutes where were serving many, many offenders at the same time.

The sanctions are parsimonious, and we had some issues here because we have a naughty judge, and judges are often naughty. The average first violation — do we have any judges in the room? [Laughter]

Too late, too late; other than this good, well-behaved judge over here. The first violation, on average, was six days; that had included my naughty judge. The second violation was 14 days, and then on average, the third violation was 48 days. That included my naughty judge who insisted on sending everybody away for a long time. And the point of this program was swift and certain but modest sanctions. Without that naughty judge, the average third violation stay was 26 days in jail.

The study setting. The study was done, men and women over the age of 18, under the supervision at what was called the Integrated Community Sanctions Unit in Honolulu, the average age of our offenders was 37 years old. Eighteen percent of them were female, and just to make sure that you know that were not dealing with the tiny fish here, the average number of prior arrests was 17; the median number was 13. These guys have been around the block before. They all scored very high on LSI scores, and had horrible recent behavior on probation.

Just to give you a sense of drug of choice, the primary drugs for these guys was, and ladies, was methamphetamines, which, I separated it out into two groups: admissions of drug use versus positive UA's What's interesting about the HOPE offenders is they come in and they say, I used. They don't even want to be tested, because they know that the judges will be more lenient if they come in and self-declare.

The probation officers, in the study that, Marlenes study at NIJ, was in integrated community sanctions units. They were a very special unit that dealt with the highest of the high-risk offenders. They had an average caseload of about 87 clients, and the POs had just over four years of experience in, as probation officers. At the same time, thanks to Marlene, we were able to set up a randomized control trial of this program that was funded by the Smith Richardson Foundation. Those POs have a more typical caseload of 176 probationers in their caseloads.

The probation officers, if there's any in the room, please raise your hand. If you're from Hawaii, those POs are incredible. They are well-trained in cognitive behavioral therapy, motivation, interviewing skills. They were also given training on how HOPE works: how to fill in the blanks, how to make ... how to interact with the courts, et cetera, to move the program forward.

We had a control group for the randomized control trial, and a comparison group for the other unit, and they were subject to probation as usual. And the, just, keep this in mind as I explain the differences in outcomes. The probation as usual involved no random testing whatsoever. They had scheduled appointments with their POs, typically once a month, not every month but most months, that then be given a UA. They knew a month in advance when they would be tested. They knew a month in advance when they'd have to pee in a cup.

The HOPE probationers, by contrast, were randomly tested, you heard. During the first two months, they'd be tested six times a month, often multiple times, multiple days in a row. So they had no advance notice of when they'd be taking their urines.

We had five aims with this study. The first was to look at differences between HOPE and probation as usual on drug use. The second was on attending appointments, as in miss ... rates of missing. We looked at jails days served, prison sentences, and on revocation rates. The big ones, the last two turned out to be the big two. Differences in revocation and differences in prison, if you're interested in cost, and I am; I'm an economist — dominate everything else. If you save prison days, you've saved the system, because of the high cost of incarceration.

Here are our findings. Urine analysis, just a reminder, again, HOPE was tested randomly, comparison group was not. We see ... now this is, this is funny. When I was first invited to do this evaluation, Id seen some of the numbers by the AGs office, and I didn't believe them. Then separately and independently we did the same analysis and came within a percent ... a few percentage points of each other.

For the HOPE group, we found, over the first three months in their follow-up, it was an 83 percent reduction in their drug use. Over six months there was a 93 percent reduction. Notice the baseline values for HOPE and for the comparison group were slightly different. The HOPE offenders were a tougher crowd. Now, based on our models, we can show that the best predictor of your follow-up drug use is your prior drug use. So this should have been ... its relatively strong biased against HOPE. Anything we would have seen them doing better ... worse than the comparison group, and if anything, they've shown impressive improvements over time. For lack of time, and I have four-year follow-ups on some of these guys, and that just flattens. So that downward trend you see that red bar? Those are the HOPE guys. It collapses, and it continues to collapse all the way to four years out.

This is the ... now look at this. Look at how beautiful and elegant this is as a behavioral triage model. [Laughter.] When brought into court and told with credibility that you will be sanctioned every time you mess up, look at this. Now this is a high-risk group. They were violating all the time. Sixty-one percent of them never tested positive again. That is an impressive number! Look at the triage! You see? Now we know who can do it on their own, which means we only have to now even care about the other 40 percent who couldn't when facing a consistent sanction.

So 20 percent of them tried Judge Alm and his colleagues. They said, OK, lets see if they really mean it. So they peed dirty once. And look at that. It just keeps collapsing. Two, three, four, five — we have very few of them going on into the right end tail. Multiple, multiple, multiple, multiple drug users now with dirty tests. Notice what they can do now. They can take the three, four, five, and six repeat dirties, and send them — whisk them all away to good-quality treatment. So they're not spending it on that big bar in the front end. They're spending all their treatment resources on those who really need it. And that's what we've seen. The HOPE offenders who fail the dirty drug tests are being sent off to very impressive quality care in Hawaii. It often involves multiple years of residential care because the treatment resources can do it. It would have been impossible without this type of triage. Missed appointments — we see the same kinds of reductions there come into line. We see that by six month follow-up, only 1 percent are missing their appointments. Taking it out over time, it just collapses to, by a year I think were at 0.003 missing. Same thing in terms of the number of missed appointments. Seventy percent of them never missed an appointment. This was a group that never showed up — often didn't show up — zero. 70 percent of them never missed once they knew we meant business, or they meant business.

If there's an economist in the room, I can essentially go home now. These are revocation rates. We have a three-to-one revocation rate against ... you get the idea. Control group, comparison groups are being revoked at three times the rate than the HOPE group. This has huge implications for prison costs.

Here's the differences in incarcerations, here's the number of days incarcerated. The bottom bar ... this is fascinating. The bottom bar in that ugly, anemic color. Those tiny little bars at the bottom, those are jail stays. I would absolutely, a priori, going into this, having to decide how to test this hypothesis, I would have expected to test a one-sided hypothesis in favor of HOPE, that HOPE would have spent more time in jail, because they get sanctioned all the time. They go if they mess up. Look at that. The HOPE offenders actually have a slightly lower number of days in jail over the course of the study. The reason is, when they go to jail, they're going for a couple of days at a time. When the control group or comparison group are going away to jail, they're going for six months, 365 days. It does ... you could have a lot of HOPE guys spending short stints in jail for every one of the comparison group that's going for the longer haul.

The other nice thing is that we haven't separated them from the communities for very long. We haven't disrupted their work. So ... essentially, there's no statistical significance in the number of jail days between the two groups. So, essentially, HOPE is jail-neutral. Look at the prison bar. We find that its saving about 192 days per, averaged over all the offenders, saving about 192 days of prison days per offender. And by the way, I'm not going to do this again. Dr. Bale talked about all the statistical bells and whistles, trust me. They're there, they're in the background. When you see the reports, all the nasty statistical stuff is there controlling for everything under the sun, and we have a randomized control trial. Were seeing huge differences in prison stays.

Bench warrants. Well, there's a lot of bench warrant activity because now every time they don't come for an appointment, if they don't come for a UA, were going to have to see a bench warrant being issued. We were very concerned about bench warrants. Eighty-three percent of them never needed to be served a bench warrant. Again, this is a high-risk, high-violating population. When they were told, the word got out pretty quickly that this program meant business, and they quickly stepped into compliance. Eighty-three percent of them never had a bench warrant, and you see it goes on. Twelve percent had one warrant. Four of them ... 4 percent had two, and so on.

All of this bench warrant activity ... now this is very important for other jurisdictions who are considering a similar program; all of the boo-boos are front-loaded. The first three months that they're in the program is when they'll test you, when they might have a couple of missings, when they might not show up and they might need a bench warrant. Once they settle into the program, everybody starts behaving, so most of those bench warrants were issued during the first three months. Only 4 percent of the probationers had a bench warrant issued after the three month marker. After six months, only 2 percent. After one year, it fell to 0.03. So we had very little bench warrant activity for people who were maturing in the program, and its only the newbies that were still testing the system. Which is a very ... again, reiterates the case that Judge Alm has made. If you're going to try something like this, start small and scale it up. Let them all catch up. Let the system settle in, and then bring them on so that you're never overwhelming your jails, you're never overwhelming your cops who have to run around with the bench warrants. So starting small is the way to go.

HOPE requires everybody to work differently and to work much more quickly. Its not that there's so much more work. Its just how you do it that really makes a — that really matters. This graphic here shows, for the groups that we interviewed — prosecutors, public defenders, judges, POs from the two units as well as court staff. By the way, we have a census. Were missing one probation officer. We have a census. Everybody involved with the system is included in this data. That's the nice thing about doing research in Hawaii; they're so incredibly cooperative with returning surveys. Good luck in California if I can get a 25 percent response rate.

What we found across the board is that HOPE does ... its more work. They ... self-report that its more work. What's interesting is looking at the second to last, the group that's second to last in this graphic. Those are the probation officers in the intensive ... basically, this is the probation officers that first started using HOPE. They've had the longest experience with the program. And notice, they didn't find it much more work. And their whole caseload is HOPE. The reason they don't find it much more work is that they've settled then. They know how to do it. And so its — again, even on the workload issues, its front-loaded, figuring out how the system works. Once we know how the system works, everything looks OK.

General perceptions across the board, again, and we have a near census here — across the board, there were positive ... the green bars are people who liked the program. Yellow is neutral. And red was ... not very happy, negative responses. Notice how happy the POs are. The probation officers across the board love this program. They keep telling us its fundamentally changed how they feel about their job. They're very happy with their jobs under HOPE. They have much more power; they're not fighting with their clients any more. The judges are also very happy.

The most dissent came from court administrators. They had to do more work and unlike the POs and the judges, they didn't really see any of the benefit of this. They weren't seeing the clients doing — this is court staff. They weren't seeing the clients doing better. Some of the prosecutors had some negative comments which are important to focus on. About 25 percent of them said they had some concerns with the program, and their concerns were valid. They were very concerned about net-widening — that people would get into HOPE. There were no — there was no mechanism on the front end to determine eligibility for the program. And they were worried that some people would make it into the HOPE who should otherwise be long in prison, and they'd like to see some eligibility criteria set up for who — maybe based on recent via priors or something, that certain individuals would be excluded from entering into HOPE.

We were interested in how the probation officers thought their job had changed. Notice 95 percent of them thought that their job had improved, that they were more effective using HOPE. This is not ... probation-as-usual. What is also very interesting — we thought that probation officers actually would be slightly resentful of HOPE because they lost a lot of discretion. Under probation-as-usual, probationers can decide if they violate the guy. If there's a dirty urine, they decide what to do with it, either I do something, I write you up, or I don't If they miss an appointment, they can decide what happens.

Under HOPE, there's none of that option. It all just happens automatically. Everybody gets violated all the time. And we thought that they would resist ... would resist giving up their discretion. The majority of them actually, when we interviewed them and surveyed them, said No, we don't want that discretion because that discretion causes fights with our clients. If you take away the discretion — I'm just doing what I have to do — its much easier for me to manage that probationer when you remove that discretion. For us, that was a fascinating finding, to see them willingly give up that discretion.

This is a fascinating result. We surveyed about 250 probationers, asking them how they thought about the program, and the left group there were those probationers who had been sent to long-term residential treatment under HOPE. Look how happy they are. Just under 70 percent of them were very positive about the program. Just over 10 percent of them were negative. But the most interesting group, I think, is those who are in jail. These were interviewed while they were serving time on a HOPE violation. If anybody's going to be unhappy, this group should be miserable. Well, they were overwhelmingly positive about the program also. And in our open-ended responses, they just said things that, like, you know, The judge said he was going to do it, and he did. It was my fault. I tried him. And he did exactly what he said. The most common comment was, Its tough, but its fair. And we keep hearing that over and over and over in our interviews with these probationers, the fairness of the program.

I'm nearly done, Marlene!

The general observations — we've now tested HOPE in two different probation offices. We've tested it with different POs. We now have it with different judges. The results are consistent across groups. We have a very robust finding. We found that the application by judges was uneven, but the outcomes were not, which meant that some judges were sending — remember I mentioned Judge Star, the naughty judge, would send people away for a long time? Well, his outcome data was exactly the same as another judge who was sending them away for just a couple of days. So all he was doing was adding to taxpayer costs. He was just making us pay for jail time that we really didn't need. Now, were going to have over ... what is a large tax payer savings? To a project like HOPE, the large reductions in prison days served has a meaningful implication for the taxpayer, for budgets.

Study limitations, just to wrap up, we had an interesting problem, and were writing proposals now. We don't know what happens to them when they graduate. What happens when they fall off of probation, and they're no longer being kept under the strict supervision? Does the fact that we've kept them on three or four years, sometimes, of this kind of behavior change their long-term outcomes? There was a spillover effect of HOPE, which was very interesting, which biases our study results, and that is that offenders in the control group and the comparison group would go to the same probation office to receive ... for their meetings. They saw HOPE guys testing dirty and being taken away in handcuffs, and they didn't know that that was not what was going to happen to them. So the POs, in interviews with them, said they saw some improvement in outcomes of the control group because they didn't realize that they were not in the same program.

Two big questions. And we have a proposal in now with NIH. How does this generalize to other jurisdictions? Marlene was really good at spearheading an initiative at NIJ to try this on the mainland so we are now going to have multiple studies rolling out to test this on the mainland, because whether or not this is unique to Hawaii is an important question.

We also have a proposal going in to find out what's going on. Why are these people behaving this way? Have we changed how they respond to stimuli? Have we made their cravings go away? What's going on? So we have a proposal now to start studying their brains. Wed like to do a brain imaging study to really understand what's going on with these individuals, why were getting such really good results.

So, congratulations to Judge Alm for this innovative program. I think its ... he's getting so much attention across the country now, because it has been, had such impressive results, and again, thank you to Marlene Beckman. Thank you.

[Applause.]