04 October 2019

Alternative Sentencing and the 24 Judges

A group of professors got together and decided to review compliance and attitudes toward alternative sanctions as they have played out in Virginia under the "Non-Violent Risk Assessment" portion of its sentencing guidelines (link here). They, of course, have the dogmatic belief of all academics that anything which keeps a non-violent offender out of prison is a good.1 It colors the paper, but not enough to make it unworthy of a read.

I'm going to offer some observations and critiques. Most are going to be Virginiacentric, but they could still be interesting to those of you with the poor sense to live elsewhere.


Assumptions / Assertions

(1)  To begin with, the article assumes that sentences increased as a result of Virginia's Truth in Sentencing reforms put in place in 1995: "One of the net effects of TIS was to increase the time served by those convicted of felony crimes." This is unattributed and feels like it proceeds from a shallow understanding of TIS. Yes, parole went away which would seem to increase actual sentences. However, at the same time the sentencing guidelines were put in place which lowered sentences back down to about where they actually were under parole. I am a post '95 attorney, but I was close enough to have heard the stories about judges getting talked to for not following the guidelines and sentencing too high (and adjusting their sentences downward to ensure they got reappointed the next time they went before the legislature). I'd really like to see something more than a bald assertion about sentence lengths on lesser felonies.
(2)  "Those who plead guilty to charges were more likely to receive alternative sanctions than those who went to trial."
Absolutely. If you accept responsibility rather than having it thrust upon you then you have taken the first potential step in moving in the right direction toward improving yourself. That voluntary step forward should be rewarded.
(3)  "Males were also found to be less likely to receive a first offender waiver from incarceration than females."
Yes. This is a result of actuarial tables as to the probability of re-offending showing a higher rate for males. It's also a result of the guidelines being entirely discretionary and Virginia statutes forbidding them to be a basis of appeals. If they were mandatory, the fact that the NVRA is openly biased against males would be a serious constitutional issue.



Of the 161 felony trial judges in Virginia, 30 agreed to be interviewed and only 24 actually were. As the authors concede this carries with it the probability of selection bias. They were questioned about the NVRA and their answers were stirred into a pot until certain "themes" were noted. This, of course is inductive reasoning. The academic doublespeak used to describe this by the authors is "an interpretivist approach based on inductive thematic analysis." That gobbledygook could mean anything so the authors are kind enough to explain it further: "Inductive thematic analysis draws from a constructionist worldview in which  [is] meaning and experience are socially-produced and can be uncovered through an examination of the structur[es] al constructs within the collected qualitative data."2  In other words, they're using inductive reasoning.

Inductive reasoning gets a bad rap sometimes. However, in the soft sciences it is the best you are going to get. There are too many constantly shifting variables for deductive reasoning and replicable tests. It's a valid research tool for a study, although you do have to point out weaknesses in a specific study. The specific dings on the study used here would be that the interviewees were self selecting (showing some sort of potentially skewed interest in the subject), the survey pool was too small, and  the data was created instead of mined from a third party data set (increasing the possibility of introduced biases). None of these invalidate the observations made in the paper; they would however increase the margin of error one might attach to any observation. However, in this case I don't think the margin of error would be enough to invalidate all the themes (save perhaps theme 4) the authors pluck from their data.


The Themes

The authors point to seven "themes"3 which they distill from the data set they've created. Next, they claim the "themes were then organized in a conceptual map that offered a general description of the factors influencing the use of risk assessment in sentencing." I think that's more academic buzzspeak surplusage because all they actually do is lay out the themes without any kind of discussion of their interactions and the general weight a judge might give to each theme as one might expect from a conceptual map. In any case, here are the themes:
(1) The NVRA is non binding. 
A strong majority of judges consider it as a factor without it trumping all other factors during sentencing. The professors later in the article suggest making the the NVRA presumptive instead of discretionary. As a practical matter, this would be contrary to the general operation of Virginia's guidelines and give rise to valid constitutionally based appeals as described above.
(2) Deference is given to plea agreements even if they are contrary to NVRA.
Yes, a judge should have a strong preference in favor of a pre-negotiated deal with which both the prosecution and the defense have agreed. A judge knows little or nothing about the facts of a case when he is handed a plea agreement and, at least in Virginia, a judge is forbidden to participate in plea negotiations and should he reject a plea deal the case is assigned to another judge. Should a judge start denying plea agreements because they are contrary to the NVRA, (1) he is arguably participating in plea negotiations, and (2) he might be rejecting a lot of plea agreements and accomplishing almost nothing when those agreements are accepted by a subsequent judge. Well, nothing except having to answer why he rejected two-hundred agreements that were accepted by another judge the next time he is before the legislature trying to get reappointed to his job.

BTW: The authors do a clever thing here where they flip the script and talk about how this could lead to a number of people getting alternative punishments who do not qualify under NVRA. In so doing, they have accidentally tripped over a real issue with NVRA. Males under the age of 30 almost never qualify for alternative punishment. Assuming that alternative punishment is a proper reaction to criminal activity, it cannot be proper to exclude all males from it. Thus, assignment of a male to an alternative program is almost always contrary to NVRA and often the only way they can get there is thru a plea deal.
(3) Lack of alternative programming capacity.
This really is a serious problem. It's a money issue. The problem with this sort of thing is always going to be money. I suspect that if the survey had been completed by all the judges in Virginia this would have been the predominant reason given. There is only so much money a legislative body - be it the General Assembly or the local county board of supervisors - is going to be willing and able to allocate to this as opposed to things such as fixing roads or educating children. Keep in mind, one of the major selling points of these programs is that they will save money, not cause an infinitely growing amount of expenditures.

In Virginia all courts have at least two options available: advisement and CCAP (formerly Detention and Diversion). Rather than reinventing the wheel, here is a link to a post I think is still generally correct from several years back. As the authors point out, many of these programs are limited by available slots and whether the defendant is medically capable of being in the program.
(4) Valuing risk assessment as a way to reduce incarceration rates.
 No. Just No. Nobody thinks about incarceration rates in the courtroom. Their own quotes in this section support this. They quote judges not wanting to send people to prison who are addicted or "low level" offenders. None of them are talking about incarceration rates.

Defense attorneys are trying to get their client the best deal they can with a strong, strong presumption that the client's liberty interest is paramount. Prosecutors are at the very least responsible for protecting the citizenry from destructive behavior first before they start to think about fixing the person who harmed the community; in other words, when they push to incarcerate they are looking toward their community's interest, not how many people are held by the Department of Corrections. The judge is responsible for visiting appropriate consequences on the guilty person standing in front of him. Rate of incarceration is an issue for the legislature, not the people int he courtroom.

A better theme here might have been "Concern that incarcerating low level offenders is not appropriate."
(5) Skeptical that the NVRA is actually predictive.
As the authors point out, judges were particularly skeptical as to the NVRA's ability to be predictive as to the risk of re-offense by addicts; one judge among the interviewed flat out told them that this isn't science. I think this is reflective of a general thought pattern you will find amongst almost anyone that does much work in criminal courtrooms. We see so many recidivists that it's hard to not be colored by it. And almost all of them have had prior opportunities attached to prior convictions.

The problem with things like the NVRA is that they are not predictive as to the individual standing in front of the judge. They are predictive as to trends within a group. Thus, they aren't really scientific when applied to a single person while they might be toward the population in general. Let's put this in perspective. Assume this advertisement is right and 60% of adults have farsightedness (not likely, but don't get sidetracked). You wouldn't want a doctor to immediately stick you in a chair and start lasik because the trend among adults is that the majority have farsightedness. No one would question the fact that the doctor runs specific tests and examinations to determine whether you as an individual are farsighted. Large group data trends may give the doctor an idea of what to look for; they should never be dispositive as to the individual patient.

The same principle applies with an individual in front of a judge. The NVRA may give a judge an idea of which way most convicts will behave in the future. The judge must consider the facts in the presentence report and other evidence presented to make a valid determination concerning the actual person standing in front of him. The NVRA isn't predictive as to the actual person and therefore should not be dispositive.
(6) Judges aren't sure whether  following the NVRA is a downward deviation from the guidelines.
I must admit, I never thought about this before reading the article. Inasmuch as the guidelines say a defendant should get between 7 months and 2 years engaging in alternative sentencing without incarceration seems to be a downward deviation. However, it is an authorized downward deviation and shouldn't count negatively in whatever statistics the legislature maintains on sitting judges. However, I do not know how the legislature actually handles this.
(7) Lack of training on NVRA.
Why would judges require indepth training on this? They understand the legislature has made this a part of the guidelines and therefore a part of the sentencing process if a defendant is being sentenced by the judge. The ones who put thought into it most likely figure out the same thing I reasoned out under theme 5 above. Training as applied here feels more like it is meant as advocacy on the behalf of those who believe the NVRA should be blindly applied to make a better world where actuarial tables determine who should and shouldn't receive alternative sentences without resort to consideration of the particular person being sentenced.


My Thoughts

The NVRA isn't a bad thing or even a good thing. It's just a thing. It's something of a shame that it has been tacked onto the end of the guidelines where it is more likely to be ignored by all unless it benefits one party or the other during arguments over sentencing. If it were at the front of the guidelines it could show the judge the general trend against which he should view the defendant and the evidence provided for her individualized sentence.

I also wish it was more nuanced. As it stands if the points scored in the NVRA are below 32 the defendant qualifies for alternative punishment. I would much prefer, and believe it would be more helpful, that the NVRA were to provide a percentage of potential to re-offend for persons of like circumstance as the defendant without setting a specific line above which persons don't qualify. I realize that the default to simple, easily workable solutions will probably keep this from happening, but I can dream.

And, no, the actuarial tables represented by the NVRA should not become mandatory or even presumptive. The NVRA is a tool among many tools which can be used to decide an individualized sentence. If they were even presumptive the rest of the tools would be ignored the vast majority of the time and that's not the appropriate way to sentence any particular person.


1 Experience in several jury trials, tends to show that the citizens of my county would disagree with this. In my locality there is a long standing joke that the best defense to murder is "He needed to be kilt" and lower level violent crimes tend to get swept up under the term "swarping" which generally means drunk/stupid/fighting and excused as typical Friday night behavior. But God help you if you if you shoplift a $5 item from the local grocer; years have been imposed for such an offense by the jury (jury sentencing, the ultimate barometer of community standards).

2  Yes, yes, I get it. You are terribly modern or post-modern or whatever the current accepted terminology for being an academic hipster is. Your rejection of pre-existent first principles contributes nothing. The basis for the forms in the data (themes in your chosen language) is not relevant only the presence of the forms is.

3  If you've had any sort of training in philosophy or theology you might recognize what they are actually discussing are forms. Out of several similar answers they derive one perfect answer - a form. Referring to it as a "theme" is ambiguous and awkward, but I will use their language because it's their paper.

No comments: