In late 2021, Congress reformed the military justice system in a way that materially alters the traditional division of prosecutorial responsibility between nonlawyer military commanders and uniformed lawyers. That reform is consistent with trends among U.S. allies and the continued evolution of the American military justice system. To aid Congress in deciding whether further changes are warranted, steps should be taken now to gather data on the effect of the reform. Although the services remain surprisingly autonomous in the administration of military justice, it is critical that the data gathered reflect common standards, definitions and reporting periods across service lines.
President Biden signed Public Law 117-81, the National Defense Authorization Act for Fiscal Year 2022 (2022 NDAA), on Dec. 27, 2021. Its hundreds of pages cover a host of topics. Among those that generated the most controversy were the provisions on military justice reform. And among those, attention focused chiefly on the sections that transferred the power to dispose of three categories of “covered offenses” from senior military commanders and other officials who are designated as convening authorities to a special trial counsel (STC) in each service branch.
Now that the guns have fallen silent, it’s possible to make a preliminary assessment of who won and who lost. No one is completely pleased with the end product. Those who opposed any transfer of disposition authority left the field a defeated if not spent force; those who sought transfer of disposition authority over all serious offenses (that is, those punishable by more than a year’s confinement) could claim a substantial, though incomplete, victory; those who chiefly sought transfer of serious sex offenses mostly got what they wanted; and those who sought transfer of serious sex offenses and a broad swath of other offenses involving the victim’s traits and the offender’s intent, such as hate crimes (the Independent Review Commission, or IRC, appointed by Secretary of Defense Lloyd Austin in 2021), got somewhat less than they had hoped for. In short, except for diehard opponents of any transfer of disposition authority, everyone got something and no one got everything they sought.
The reason for this is twofold. First, those seeking change had a variety of agendas, some based overwhelmingly on concern over the unacceptable incidence of sex offenses, some based on concern about racial disparities and some driven by the notion that disposition decisions are quite simply best made by lawyers. Second, at the end of the day, the legislative process evolved into a highly political, highly untransparent exercise in horse-trading, producing a boundary line between charges that would be subject to the new system—and those that would remain with commanders—that can only be described as the criminal-justice equivalent of a gerrymandered legislative district.
The new provisions will take effect in late 2023. Between now and then, the Defense Department and the armed forces will be busy drafting the necessary implementing provisions of the Manual for Courts-Martial and service regulations, as well as preparing to train the judge advocates who will be making disposition decisions and trying cases under the new system. As a practical matter, the fact that it will be two years or more before there can be any reliable sense of the impact of the changes on either level of criminal conduct—or the administration of justice generally under the Uniform Code of Military Justice (UCMJ)—means that Congress and the armed forces may take a breather before returning to the subject of military justice reform. Sheer exhaustion among legislators may also militate against further reform efforts in the next few years, even though a strong case can be made for addressing some changes that did not make the cut—such as bringing the UCMJ’s jury-voting provisions into sync with the unanimity rule the Supreme Court announced in Ramos v. Louisiana and affording military defendants equal access to the Supreme Court.
Perhaps, in two or three years’ time, there will be little interest in revisiting the line-drawing Congress did when it charted the boundary between those cases, the disposition of which will be the responsibility of the STC and those whose disposition would remain within the commander’s bailiwick. This is, however, unlikely given the energy and effort that were expended in negotiating the 2022 NDAA. So, the facts on the ground may also change in the next several years.
For example, will American withdrawal from Afghanistan alter the nature or level of criminal conduct within the armed forces? Or will some new conflict or set of conflicts entail deployments or alter patterns of conduct among serving personnel in other ways? Will there be a reaction against the abandonment of military justice in favor of administrative separations or nonjudicial sanctions, leading to a further reduction in overall court-martial caseloads? Will the mere enactment of the new system have a subliminal effect on crime-victims’ willingness to come forward or on commanders’ decision-making, perhaps even before it goes into effect? Will actual implementation of the new system in late 2023 and thereafter dilute senior officer resistance to further change, or will it only stiffen resistance? Will new data emerge that will strengthen the hand of those who had favored a broader reduction in commanders’ disposition authority on the basis of concerns over racial disparities?
Mindful of these uncertainties, it seems a safe bet that Congress will, in the next few years, assess the effect of the 2022 NDAA changes and consider whether, on the basis of that assessment, the STC-commander boundary should be redrawn or whether commanders’ disposition power over all felony-level offenses should be withdrawn. Steps should be taken now to prepare to gather the rigorous data that will permit Congress to make a wise decision on that question.
Under new Article 24a(c)(2)(A), UCMJ, covered offenses are those whose disposition will be decided by the STCs. They are defined by new Article 1(17), UCMJ. These core, enumerated offenses are wrongful broadcast or distribution of intimate visual images, murder, manslaughter, rape and sexual assault generally or where the victim is a child, other sexual misconduct, kidnapping, domestic violence, stalking, retaliation, and the Article 134, UCMJ offense of child pornography. In addition, the definition sweeps in conspiracy, solicitation and attempt to commit any of those enumerated offenses.
The statute goes further. Under new Article 24a(c)(2)(B), UCMJ, the STCs will have disposition authority over unenumerated offenses that, in the opinion of the STC, are related to an enumerated offense.
Finally, under new Article 24a(c)(2)(B), UCMJ, the STCs will have disposition authority over any other offense, whether enumerated or related, the accused is alleged to have committed. The statute refers to these as “known” offenses, but we believe that category and “alleged” offenses describe the same cohort of offenses.
What Percentage of Serious Offenses Will Be the STCs’ Responsibility?
How much of the serious offense caseload will be disposed of by nonlawyer commanders (as before) and how much will fall to the STCs? The question is not entirely new. The Shadow Advisory Report Group of Experts (SARGE) observed in its response to Section 540F of the National Defense Authorization Act for Fiscal Year 2020:
What fraction of the military justice caseload would come within the ambit of § 540F is impossible to determine from publicly-available data. The Army’s Article 146a, UCMJ, 10 U.S.C. § 946a (2018), report for FY18 states (at 9) that “[o]f the 502 cases in which findings were entered in FY18, 246 of them, or 51 percent, included sexual misconduct related offenses (Articles 120, 120b, and 120c).” The maximum punishment for offenses under each of those punitive articles qualifies them as specified offenses under § 540F(a)(2). Since some non-sex-offense cases in which findings were entered during the same period would also have entailed serious potential maximum punishments under the Manual for Courts-Martial, the alternative system will apply to more than a bare majority of courts-martial[.]
SARGE cautioned that “[w]ithout a pilot program, there is no way of knowing whether the alternative system would affect the number of prosecutions.”
In a later statement on the IRC’s report, SARGE observed:
The three “special victim categories” that would be the responsibility of a lawyer decision maker reach a host of UCMJ violations, depending on the offense, the victim’s traits, and the offender’s intent. See App. B at 10. Taken together, these categories are so expansive that what was plainly intended as an exception to the default principle of commander-centric disposition authority will wind up accounting for a very sizable portion of the military justice system’s entire throughput of serious cases. That exception may even swallow the rule. The residue that the IRC would leave to disposition by commanders may be so small as not to justify a separate system. Moreover, that residue—whatever its size—will include cases (like murder, espionage, and computer or credit card fraud) for which legal training is essential to proper disposition decision making. (Emphasis added.)
While it is possible that the sheer number of covered offenses entering the military justice system’s caseload may increase due to improved confidence in it among complainants, thus driving up the STCs’ share of the caseload (and driving down the commanders’ share), we don’t yet know of a rigorous way to predict the existence, much less the scale, of such a change. The only way to know with any confidence is to wait until the new system has been in effect for some period of time. The pre-2022 NDAA data, however, certainly suggests that cases falling within the authority of the STCs will account for a considerable fraction of the total serious-offense caseload.
We examined a year’s worth of publicly available court-martial data for each service branch. The data are far from uniform across service lines; in fact, it is a statistical Tower of Babel. For example, some of the armed services’ data reflect offenses of which the accused was acquitted; others do not. Some supply sufficient information to form a reasonably reliable estimate of whether the accused’s offenses were covered, while others require judgment calls or guesswork.
The most informative data points are those derived from the Army’s monthly Summarized Reports of Results of Trial. From November 2020 to October 2021, the Army conducted 526 general and special courts-martial. Of these, 56.8 percent—or 299—involved covered offenses, leaving 43.2 percent of the cases subject to disposition by commanders if the new law had been in effect at the time. Only 33 (11 percent) of the 299 covered-offense cases were tried by special courts-martial.
The Air Force’s public website lists 159 general courts-martial and 122 special courts-martial, for a total of 281 trials in fiscal year 2021. Of these, 115—or 40.9 percent—involved covered offenses. For general courts-martial alone, 67.3 percent of the cases involved covered offenses, while for special courts-martial alone it was 6.6 percent.
The Navy has three public sources for court-martial results. The Marine Corps reports 123 general courts-martial, 35 of which included a covered offense (28.5 percent), and 135 special courts-martial, 18 of which included a covered offense (13.3 percent). The Navy reports 107 general courts-martial, of which 82 (76.6 percent) included a covered offense, and 88 special courts-martial, of which 15 (17.0 percent) included a covered offense. An examination of the combined list of records of trial for cases pending before the U.S. Navy-Marine Corps Court of Criminal Appeals for January through October 2021 shows that 295 courts-martial resulted in mandatory appellate review, of which 139 (47.1 percent) included covered offenses.
Four significant caveats apply to the Department of the Navy data: (a) The Marine Corps does not report the charged offenses when there has been an acquittal; (b) the reports do not account for cases in which covered offenses were charged but dismissed or reduced as part of a pretrial agreement (for example, a referred sexual assault charge is negotiated down to an assault and battery guilty plea, thereby avoiding sex offender registration); (c) the data do not account for cases where covered offenses were preferred but not referred for trial; and (d) the data do not account for cases in which a covered offense is referred for trial, but the case is ultimately dealt with through nonjudicial punishment and waiver of administrative separation board proceedings.
The Coast Guard, the smallest armed force, has less informative publicly available court-martial data but, interestingly, provides a wealth of information about summary courts, nonjudicial punishment, and other adverse actions, even those involving civilian personnel. According to its quarterly Good Order and Discipline Reports, the Coast Guard tried only 25 general or special courts-martial during fiscal year 2021. Of those, seven—or 30.4 percent—appear to involve offenses that would have been covered offenses if the new law had been in effect. Of the seven general courts-martial, three—42.9 percent—involved covered offenses; of the 18 special courts-martial, four—22.2 percent—involved covered offenses.
All told, for the most recent 12-month period for which data are available, the armed forces tried 1,285 general and special courts-martial. Of these, 570 (44.4 percent) involved covered offenses, leaving commanders with disposition authority over 55.6 percent of the pie. While the percentage of covered cases tried by general court-martial varied from service to service, overall, it was substantially higher than the percentage tried by special court-martial. Covered offenses plainly represent a substantial chunk of the current overall serious-offenses caseload under the UCMJ—hardly a minor subset.
Moreover, these data almost certainly overstate the percentage of the caseload that would remain with commanders under the new law, as they encompass only cases that wind up in trial. Additional covered offenses may well lurk “upstream” of referral (for example, a case may be pleaded down to a lesser offense, a complainant may decide not to testify, or local authorities may offer to adjudicate the matter) and still other covered offenses might be referred but would be dismissed at trial on motion. Hence, these cases would not show up as either acquittals or convictions.
Because new Articles 24a(c)(2)(A) and (B) both refer to “reported offenses,” it seems clear that Congress’s concern was broader than simply the cases that go to trial, or even those that are “preferred.” It follows that, to be faithful to the congressional intent, analysis should begin with a definition of when an offense has been “reported.” Any such case should be included in the denominator of the fraction of the universe of “cases” that covered offenses constitute. We are skeptical that the armed forces will be able to generate such a number within a reasonable time. Decision-makers may, therefore, have to settle for a second-best calculation that treats preferred cases as the universe, recognizing that the resulting data will understate the fraction of covered cases by excluding those that fall by the wayside upstream of referral.
SARGE suggested in 2020 that data covering a three-year period are desirable as a predicate for future congressional consideration. The authors of the current paper believe three years is a proper benchmark to achieve robust data, but we are open to persuasion that a shorter period is sufficient if Congress were to conclude that more expeditious action is called for in order, for example, to maintain confidence in the military justice system. Such a trade-off may be inevitable.
Conclusions and Recommendations
How much of the serious-offense caseload will remain with commanders under the new law is of no consequence for those who believe the charging decision for any felony-level offense is quintessentially a legal matter that should be made by lawyers, in the interest of fostering public confidence in the administration of justice.
Not everyone will see it this way, of course. For those who do not, the question of zeroing out or further reducing the commanders’ share of serious offenses will remain. There is nothing magical about any particular fraction of the caseload that would tip the scale in favor of removing commanders’ power over serious offenses. Congress could rationally decide that other punitive articles should be added to the list of enumerated covered offenses. Or it could decide that the serious-offense commander-disposition caseload is too small to warrant retention of however much remains of the traditional disposition system caseload. Having two parallel systems performing essentially the same function seems both wasteful and an invitation to confusion and inconsistent results. While it might seem odd given the very considerable political effort that went into the new law, Congress could, of course, also repeal the 2022 NDAA changes or remove some covered offenses from the list. Either of these seems highly improbable, but determined resistance from the services, a shift in control of either or both houses of Congress, and the outcome of the 2024 presidential election might make what is currently unimaginable, entirely possible.
It is difficult to imagine that the impulse to reform the military justice system ends with the 2022 NDAA. For one thing, that legislation did not address two long-running major issues: commander selection of court-martial panel members and equal access for military personnel to the Supreme Court in court-martial appeals. But even as to the transfer of the disposition power over many serious offenses, it seems likely that the contending forces are merely in an armistice, and the issue of where the permanent boundary should be drawn will remain a live one.
To facilitate informed decision-making on that issue down the road, it is critical that steps be taken now to ensure that data regarding cases under the new legislation are captured from day 1. Three years’ data regarding actual operations under the new system, including—importantly—the STCs’ actual exercise of the power to waive exclusive jurisdiction (thereby sending some number of cases back to commanders for disposition), will facilitate thoughtful analysis by Congress in deciding whether to make further adjustments to the UCMJ.
We also recommend that the data be gathered sufficiently far upstream from the proceedings at trial to ensure that the analysis comes as close as possible to taking account of the universe of potential covered offenses. One approach might be to assemble two parallel sets of data—one that uses referred cases as the universe and another that uses preferred cases. Congress could then decide which is the more pertinent, assuming the fraction of cases with a covered offense is materially different between the two. Whichever approach Congress takes with respect to the data, it will be in a position to decide whether to further reduce (or end) whatever remains of commanders’ power to decide the disposition of serious offenses. Continuing with incremental adjustments, however, is inadvisable because piecemeal changes inevitably unsettle what should be a stable institutional framework and further complicate both data-gathering and training.
Finally (and importantly), to be useful, data regarding cases disposed of by the STCs versus commanders should be rigorously compared from armed force to armed force, in contrast to the current welter of divergent standards.
Views expressed in this post are solely those of the authors and should not be imputed to any other person, organization or government agency.