The Military Justice Act of 2016 (MJA) was this blog’s #1 Military Justice Story of 2016, and will revolutionize the practice of military justice come January 2019. That revolution includes changes to the military justice system’s sentencing regime. One of the most aggressive changes is found in § 5301 of the MJA, which completely rewrites Article 56 to include language designed to temper a court-martial’s exercise of sentencing discretion.

First, § 5301 of the MJA establishes the uncontroversial general rule that:

A court-martial shall impose punishment that is sufficient, but not greater than necessary, to promote justice and to maintain good order and discipline in the armed forces.

When applying that general rule, MJA § 5301 unsurprisingly directs sentencing authorities to consider “the nature and circumstances of the offense and the history and characteristics of the accused” as well as the impact of the offense on victims.

But, § 5301 also codifies, and then expands, Rule for Court-Martial (RCM) 1001(g), which permitted military prosecutors to reference the “generally accepted sentencing philosophies” during their sentencing arguments. Under the MJA, those philosophies are no longer a mere permissible topic of counsel’s argument. They are mandatory considerations for the court-martial. Specifically, once MJA § 5301’s rewrite of Article 53 becomes effective, a court-martial deliberating on a sentence will be required to:

Tak[e] into consideration [] the need for the sentence—

(i) to reflect the seriousness of the offense;

(ii) to promote respect for the law;

(iii) to provide just punishment for the offense;

(iv) to promote adequate deterrence of misconduct;

(v) to protect others from further crimes by the accused;

(vi) to rehabilitate the accused; and

(vii) to provide, in appropriate cases, the opportunity for retraining and return to duty to meet the needs of the service[.]

The codification of those required considerations is a big deal in and of itself, and should heavily shape how government counsel present their sentencing cases under the new UCMJ. But one of those required considerations in particular is worthy of further attention: Considering “the need for the sentence[] to protect others from further crimes by the accused” is functionally a requirement that the court assess the accused’s risk of recidivism. This is not something courts-martial as they currently exist regularly grapple with.

The idea of introducing evidence of risk of recidivism into courts-martial was discussed at length in a recent article published in the Military Law Review by Army Lieutenant Colonel Bradford Bigler. Colonel Bigler is an Associate Professor of Criminal Law at the Army Judge Advocate General’s Legal Center and School, located on the campus of the University of Virginia. His article is entitled Rebalancing Military Sentencing: An Argument to Restore Utilitarian Principles within the Courtroom, 225 Mil. L. Rev. 1 (2017).

Lieutenant Colonel Bigler’s article proposes:

[T]he government should be required to introduce an actuarial risk assessment into evidence where an actuarial tool exists, and seek to develop risk tools where none exist.


Similar to authorities governing the appointment of a sanity board, the convening authority or military judge will order a qualified psychologist to review the case file and provide a sentencing report which scores the offender according to validated actuarial risk models.

Rebalancing Military Sentencing at 26.

Elaborating further, Lieutenant Colonel Bigler notes that:

Although several statistically validated risk models exist, they currently do not cover a sufficiently broad range of offenses to encompass the spectrum of military offenses.


To remedy the problem of having offenses for which no actuarial tool presently exists, Lieutenant Colonel Bigler recommends:

[T]he military should authorize, encourage, and fund—if necessary—criminogenic research to develop new actuarial models for use in the military. New computer modeling techniques have produced significant advancement in actuarial modeling that would help identify which risk factors are most relevant to a specific population.

Id. at 27.

Lieutenant Colonel Bigler’s article calls into question whether the recently proposed amendments to the Manual for Court Martial, discussed on this blog, meet the demands of the new statute. The proposed language for the new RCM 1002 does require the court to consider the new statutory factors, including the need “to protect others from further crimes by the accused.” However, subparagraph (g) of the proposed rule limits the evidence that the court may use to conduct that analysis to the evidence admissible under RCM 1001. Lieutenant Colonel Bigler’s article, at 25-26, argues that this evidentiary barrier may prevent admission of the sort of actuarial evidence necessary to conduct a reliable assessment of the accused’s risk of recidivism.

I am not certain of this. CAAF has allowed the government to present actuarial evidence in sentencing hearings in the past. Specifically, in United States v. Ellis, 68 M.J. 341 (CAAF 2010) (CAAFlog case page), a government expert was permitted to testify in sentencing regarding the Static 99 methodology’s assessment of the accused’s risk of sexual re-offense. Accordingly, at least on first blush, Ellis appears to interpret RCM 1001 in a way that is favorable to Lieutenant Colonel Bigler’s suggestions.

However, it must be conceded that CAAF did take the time in its Ellis decision to express a certain amount of discomfort. The Court make clear that the ruling was confined to the unique facts of that case. Namely, the Court emphasized that Ellis elected to be tried by a military judge sitting alone, and his counsel did not object to the government expert’s testimony on grounds that it was constitutionally unreliable under Daubert.  Ellis, 68 M.J. at 348, n. 8.

In all, the statutory requirement that future courts-martial consider an accused’s risk of recidivism tends to support Lieutenant Colonel Bigler’s recommendation that courts allow the introduction of actuarial evidence. At least on first review, so does CAAF’s 2010 decision in Ellis. But, the caveats CAAF wrote into its Ellis opinion do give cause to some concern. The Court carefully drafted its opinion in that case to leave open, for a future case, the question of whether such evidence is Constitutionally reliable. It seems likely that the MJA § 5301’s rewrite of Article 53 will force that question back into CAAF’s sights in the relatively near future.

17 Responses to “Scholarship Saturday: The government’s coming requirement to show the accused’s recidivism risk during sentencing”

  1. Dustin says:

    First, it seems there is a definite need for change regarding the vast scale of differences in sentencing. 
    However, looking from the panel’s point of view, Possibly an anonymous scorecard with scales of 1 to 10 of appropriate deliberation topics to help them organize their thoughts in a more tangible way. Along with some mandatory reading of scholarly articles and publications on the topic. Make available to them national averages of sentences for similar conduct. 
    Kind of like reading the directions for IKEA furniture.
    I also feel very strongly that sentencing should also weigh the disadvantage of the accused of not having an actual 12 member Jury requiring unanimous agreement on guilt to consider in sentencing as well as any times the member has been in harms way in service to defend the nation and the constitution. It wasn’t that long ago that civilian judges routinely dismissed charges on the agreement the accused would serve in the Armed Forces.
    third thought, a psychological report and sentencing recommendation should/could be provided to the CA to consider while deliberating on affirming a sentence.

  2. Javert says:

    Perhaps we need to move away from doing sentencing right after findings, and get a presentancing report like the feds do. 

  3. Tami a/k/a Princess Leia says:

    Recidivism and the protection of society is something practitioners have dealt with on a regular basis, especially sexual assault and child pornography cases.  Sex offender registration protects society through public shaming, and therefore nail isn’t necessary.  Though I am aware of opinions that say this isn’t allowed, I would say that’s because there is insufficient connection between the accused and where they want to live outside of nail.  Ask client’s where they want to live after trial, print that state’s SOR rules, then ask court to take judicial notice.  That is how you get that information in.  And now I think courts have to let it in.

  4. John Marshall says:

    Sentencing is inherently an individualized assessment, and efforts at uniformity in the federal system have only served to mandate more severe sentences.  To be ethically used, static 99 and other risk assessments require an interview with the accused.  No defense counsel will consent to such an interview unless they know the result is good for their client. It’s been a while since I’ve read Ellis, but I suspect the defense counsel in that case was woefully inadequate for not objecting to the evidence. Also, many of these “actuarial risk assessments” are of suspect validity when applied to a specific accused.  As Mark Twain said, ‘there are lies, damn lies, and statistics.’ 

  5. Vulture says:

    David Rubinstein was on CNN this morning and said that some percentage of the American people think that Judge Judy is on the Supreme Court.
    Ltc Beaker Bigler is a law professor.
    Scholarship Saturday is needed now more than ever.

  6. Alfonso Decimo says:

    When the Trial Guide is rewritten, hopefully it will be easier for the members to understand. Teaching members the rules of sentencing was always nearly impossible back-in-the-day, as evidenced by some of the questions or decisions that would result from their deliberations on sentencing. I would prefer a rule change to make sentencing always judge-alone.

  7. Isaac Kennen says:

    I would prefer a rule change to make sentencing always judge-alone.

    I think the MJA’s invention of allowing convening authorities to refer charges to judge-alone special courts-martial (or magistrate-alone, if the parties consent) will likely facilitate more judge-alone sentencing. It is true that such a court would lack jurisdiction to impose a punitive discharge or more than six months of jail-time, so the most serious offenses would not be handled in that forum.  But, I think there is a lot of misconduct that could be justly handled by six months or less of confinement without a punitive discharge.  
    I’m not convinced the punitive discharge matters all that much, anyway.  During my time as an appellate counsel, as far as I could tell, bad conduct discharges seldom had any enduring effects on my clients.  (Dishonorable discharges are a different story, though.) Folks with BCDs, assuming they weren’t convicted of a sexual or other violent offense, were generally able to get jobs just fine. They were usually also pretty successful at convincing the VA to extend them benefits – including at least part (if not all) of their Post-911 GI Bill benefit.  I asked a VA counselor one time why folks with a BCD who had committed a non-violent offenses were granted benefits despite their punitive discharge, and was told the VA has different priorities. The VA understands that it can pay to help the punitively discharged vet now, or it can pay for it in the future in the court of public opinion – probably around the time that it is trying to justify the need for a budget increase.  The public and their elected representatives have never really distinguished between the homeless Vietnam vet who left with an honorable discharge and the one who was punitively discharged by a court-martial. The VA has traditionally taken the heat for both of those vets in equal measure.  

  8. Vulture says:

    It is to madam justice I dedicate this concerto.
    It used to be that on Scholarship Saturday it was the calls of Siren Song that would crash us on the rocks in search of military justice reform.  So what is it here that we are supposed to seeing as a remedy?  Civlianizing the prosecution of sex crimes is a prospect that has been considered and rejected repeatedly.  Is this appeal to a federal sentencing procedure supposed to be a way to rectify an already shlushy conviction?   What is this article trying to sell me on?  After a contested determination of guilt are the panel members supposed to dispense with the confrontational element to say that the Accused isn’t such a bad guy?  This pawning a kinder, sweeter sentencing is the Swan Song of military justice reform.  What are you saying with all that peeping?
    First the overture, yes the strings, listen carefully can you hear it?  Now the brass: “More judge alone trials.”

  9. Isaac Kennen says:

    I know I sometimes write Scholarship Saturday articles that are designed to persuade. This one is just designed to present an issue that I think folks will need to grapple with in the future, and to highlight Colonel Bigler’s work on the subject.
    I hope I presented the issue in a way that highlights that while the language of the statute may appear straightforward, the devil, as always, is in the details. Applying the statute will be hairy – regardless of whether the sentence is adjudged by a panel or a military jurist.  
    And, I certainly didn’t mean to suggest that more judge alone trials is something I favor. Further, regarding the MJA’s creation of mandatory judge-alone special courts-martial…I have my doubts whether a military member ought to be made to stand trial before an officer, sitting alone, who was appointed by the Judge Advocate General, is accountable to the JAG for future assignments and his or her very appointment as a judge advocate, and who has not been ordained for judicial duty by the People via the consent of their Senate. Such a procedure seems especially undesirable considering the other rights which have been denied to the military accused (i.e., a binding preliminary hearing or grand jury process, a petite jury, a randomly selected panel, a panel large enough to engage in meaningful deliberations, a unanimous verdict, and a prosecuting authority bound by professional licensing standards).  
    Especially given those aspects of the military justice system which are already unfavorable to an accused, I have earnest concerns regarding the MJA’s provisions allowing the Department of Defense to force a military accused to endure a judge-alone special court-martial, where they could suffer a loss of liberty for months on end, when the judge was given their robe by Department authorities, is accountable to Department authorities for their assignments even their designation as a judge advocate, and who lacks the legitimizing mandate of having been consented to by the People via their Senate.
    One might wonder whether a DoD judge, who lacks the tie to the People that would come with Senate confirmation, should to be able to stand in judgement of a citizen without their consent. That seems at odds with the concept that our government is of the People, by the People, and for the People. But, that’s a different discussion altogether. 

  10. stewie says:

    One quibble, when written, it’s Lieutenant Colonel, not Colonel (which is correct when spoken).
    As to the much more important point of the article, recidivism…it’s a soft “science” that really IMO doesn’t have much accuracy. I think it’s fair to say, this person has a history of X or a pattern of Y. I think it also fair to say the offender has done or not done these things to show that he/she is remorseful for their actions. A panel can then use these two pieces of information to decide whether the offender needs to spend more or less time in jail.  But the idea that there is a real scientific way to calculate out a “recidivism level” is just more soft science psychology that slaps an imprimatur of “science” onto something that barely is. And as is noted above, that then tends to lead to harsher sentences or perhaps it’s better to say, more disparate sentences since you generally only have low, medium or high as classifications which isn’t a lot of stratification or individual specificity.

  11. Anon says:

    My first job after being discharged after my court martial (through an Admin Sep board)?  A famous fruit company that makes electronics.  I’m also finishing my undergrad with my Post 9/11 G.I. Bill and I just got my VA Choice Health Care card in the mail today.  Life isn’t as bad as JAG’s and Commanders try to make you think it is.

  12. Isaac Kennen says:

    Thanks for the catch!  Edits made.

  13. Vulture says:

    I don’t disagree with you that the devil is in the details.  I would just add that Hecate is in the subtext.  The article cites to the work of a Judge that I have bought up a few times here that was a Senior Defense Counsel at Fort Hood.  In the article here it determines the requirement that military crimes be ensured reported to the states for repeat offender determination.  What business does an SDC have projecting such a stance?  That’s a rhetorical question because the answer is none.  The fact that he ended up at USALCAS afterwards makes me wonder if chasing dust bunnies gave him so hostile a disposition to defenses.  (a la Ahern).  Now Evan Seamone mentions him also in his article on the rehabilitative ethic.   But if there is a point to be made here, it isn’t about rehabilitation, its not about tempering recidivism, its not about protecting the Accused.

  14. Grey says:

    If these risk factors are going to mechanistically translate into different sentences, this seems similar to aggravating factors that must be proven beyond a reasonable doubt according to the Supreme Court’s Apprendi decision.   In Apprendi the Court held that if a crime carries a more severe penalty when accompanied by aggravating factors (motivated by racial animus in that case), the existence of the aggravating factor functionally becomes an element of the offense that must be proven along with every other element.  It seems that these actuarial factors would function very similar to the aggravating factors addressed in Apprendi.
    Also, actuarial evidence sounds like profile evidence.  Are we comfortable with characteristics of the offender (rather than characteristics of the offense) being used to determine sentence, particularly for characteristics that are innate or nearly so?  For example, in the Static-99 (evaluation used for sex offense recidivism assessment) some of the factors are age, stability of relationships, and history of male victims (which would seem to be aligned with sexual orientation.)  The Static-99 test is only for males, but I wouldn’t be surprised if broader actuarial data would support that males have a higher recidivism rate than females.  It would be a major change in our judicial philosophy if we started to expressly justify harsher sentences based on these factors.

  15. Matt says:

    While I don’t like the changes, I don’t think it would create an Apprendi issue.  They are simply stating that these factors need to be considered for sentencing, not that they require a specific additional sentence.  If they stated that evidence of recidivism must increase the sentence by X years, then I would agree with you.

  16. Philip D Cave says:

    I’m not sure about this change.  I often use a psychosexual evaluation in cases to try and establish the client has rehabilitative potential.  You can do your own rough evaluation by reviewing the STATIC 99 and PCLR online if you are concerned about even asking for one.  It’s my understanding that the STATIC 99 was partly developed in such a way it can be used by non-psych trained probation officers.
    Static99 coding form
    You will see that many clients are going to fare quite well on the STATIC 99.  But then you have to add in the PCLR — you must have both IMHO.  It helps to have a trained psych to also discuss the full meaning and talk about the static and dynamic factors affecting the likelihood of rehabilitation and recidivism which are developed through an interview(s) and reference to collateral information (interview of relatives, spouses, medical and psych records, etc., etc., etc.).
    PCLR outline  I have had some practitioners question the validity of the PCLR, but it’s often an expected evaluation.
    Grey.  I have a query out about the male / female distinction.  As almost all of my sex related clients are male, I’d not considered whether there is a meaningful distinction.  
    I do agree that care should be used to ensure that the evaluation only goes to rehabilitation potential and not to justify additional punishment — kinda like the mendacity issue when a client has testified on the merits.  Perhaps we’d need to fashion some sort of limiting instruction.
    You should also be aware that there is research demonstrating that deniers do not present a more significant risk of recidivism on release compared to admitters if you are thinking of doing this with a client who pleads not guilty and then is found guilty.  One of the factors is that deniers will quite frequently become admitters once their appeal is complete and submit to available counseling.  The question is whether they become acceptors because they accept the need for help or just want to get out of the brig early.

  17. Philip D Cave says:

    I am reliably informed that there is some fairly robust research on this (which I’m not able to share, in deference to my source), but your thought is correct on a difference between males and females and recidivism.  There is at least one military case where this came up and was addressed by the expert.  It appears that the Level of Service Inventory is the preferred tool and not the STATIC 99 for female offenders.

    The devil, of course, is in the details.  Should you have need of this for a particular case I can put you in contact with my source–a very reliable source.