CAAF decided the Army case of United States v. Adams, 74 M.J. 137, 14-0495/AR (CAAFlog case page) (link to slip op.), on Monday, April 27, 2015. The court reverses Appellant’s conviction for larceny after concluding that Appellant’s confession, which was “the government’s key piece of evidence,” was improperly admitted because “virtually none of the facts we have previously articulated as essential were corroborated.” Slip op. at 11 (quoting the CCA’s opinion). CAAF reverses the decision of the Army CCA and the findings and sentence, and authorizes a rehearing.

Judge Erdmann writes for the court, joined by Judges Stucky and Ohlson. Chief Judge Baker dissents, joined by Judge Ryan.

CAAF granted review of a single issue in this case:

Whether the Army Court of Criminal Appeals erred in finding that the military judge did not abuse his discretion in admitting the portion of Appellant’s sworn statement regarding the [theft] of cocaine because the Government failed to corroborate, in accordance with Military Rule of Evidence 304(g), the essential fact that appellant took cocaine.

Appellant was convicted contrary to his pleas of not guilty, by a general court-martial composed of a military judge alone, of one specification of larceny in violation of Article 121 (as a lesser included offense of the charged offense of robbery in violation of Article 122). He was sentenced to confinement for 165 days, reduction to E-1, and a bad-conduct discharge.

Appellant’s conviction was based on an alleged armed robbery of cocaine from a local drug dealer. Acting on a tip from another soldier identified as an accomplice in the theft (but who did not testify at trial because he was in an unauthorized absence status), Army investigators interrogated Appellant, and Appellant  gave a written statement in which he admitted to stealing cocaine from the drug dealer, using a S&W handgun, near a local WalMart.

The statement given by Appellant was admitted at trial over a Defense objection on corroboration grounds. The corroboration rule – M.R.E. 304(g) (2012) (relocated to M.R.E. 304(c) (2013)) – permits use of an admission or confession of the accused against the accused “only if independent evidence, either direct or circumstantial, has been introduced that corroborates the essential facts admitted to justify sufficiently an inference of their truth.” Additionally,

If the independent evidence raises an inference of the truth of some but not all of the essential facts admitted, then the confession or admission may be considered as evidence against the accused only with respect to those essential facts stated in the confession or admission that are corroborated by the independent evidence.

M.R.E. 304(g) (2012) (relocated to M.R.E. (c)(2) (2013)). In this case the Government did not present corroborating testimony from Appellant’s accomplice or from the victim. Rather, the Government presented the testimony of Army investigators. That testimony directly corroborated the existence of a suspected drug dealer matching Appellant’s description of his alleged victim (who was known to the investigators), the existence of the handgun (as seized from Appellant’s home), and the existence of a local WalMart fitting Appellant’s description of the location of the alleged larceny. However, the testimony did not directly corroborate the existence of any cocaine, the object of the alleged larceny.

Nevertheless, the military judge ruled that much of Appellant’s confession was admissible based on the following corroboration:

The description of the handgun the accused admitted to “waiving [sic] around quick” is a “S&W .40 cal.” This matches the description of [the weapon found in the search]. . . . [T]he Court finds that these items found in the accused’s home four days after the alleged crimes coupled with the testimony regarding the location of a Walmart and Microtel in Evans Mills, New York to be sufficient to meet the standard of the slight corroboration required by the rule and case law.

Slip op. at 6. On review, the Army CCA found that the military judge did not abuse his discretion in making that ruling, and the CCA affirmed the findings and sentence. CAAF then granted review in Adams and in another case also presenting a corroboration issue: United States v. Bennett, No. 14-0658/AR (grant discussed here).

Today’s ruling doesn’t just reverse the Army CCA; it also breathes new life into the corroboration rule.

Judge Erdmann explains that corroboration isn’t all-or-nothing, but rather occurs separately for each essential fact:

What constitutes an essential fact of an admission or confession necessarily varies by case. Essential facts we have previously considered include the time, place, persons involved, access, opportunity, method, and motive of the crime. See, e.g., United States v. Baldwin, 54 M.J. 464, 465-66 (C.A.A.F. 2001); Rounds, 30 M.J. at 77-78; United States v. Melvin, 26 M.J. 145, 147 (C.M.A. 1988).

When independent evidence which is sufficient to corroborate an essential fact is provided, that essential fact is admissible. M.R.E. 304(c). If sufficient corroborating evidence of an essential fact is not provided, then the uncorroborated fact is not admissible and the military judge must excise it from the confession. See id. The essential facts which are corroborated may be used against the accused alongside any other properly admitted evidence. See, e.g., Opper, 348 U.S. at 93 (“Those facts plus the other evidence besides the admission must, of course, be sufficient to find guilt beyond a reasonable doubt.”).

There is no “tipping point” of corroboration which would allow admission of the entire confession if a certain percentage of essential facts are found to be corroborated. For instance, if four of five essential facts were corroborated, the entire confession is not admissible. Only the four corroborated facts are admissible and the military judge is required to excise the uncorroborated essential fact. M.R.E. 304(c). This analysis is completed by the military judge examining the potential corroboration for each essential fact the government wishes to admit. Id.

Slip op. at 9-10. Applying this to the facts of Adams, Judge Erdmann and the majority find that:

[I]t is important to determine what was not corroborated. Here, there is no evidence which corroborates Adams’ opportunity or motive to commit the crime, his access, his intent, the accomplices involved, the subject of the larceny (i.e., cocaine), the time of the crime, or the act of the larceny itself (waving the handgun while [DT] grabbed the cocaine). In short, virtually none of the facts we have previously articulated as essential were corroborated. Even if we were to assume that the evidence relied upon below properly corroborated the location of the larceny and the identity of the victim, those facts, combined with the ownership of the handgun, are legally insufficient to support the larceny conviction absent any additional direct evidence of a crime.

Slip op. at 11. The clear takeaway from this analysis is that trial practitioners must parse a confession for its essential components, and then corroborate each such component. Chief Judge Baker’s dissent makes this perfectly clear:

The majority thus requires that every essential fact identified in a confession must be individually corroborated on a one-for-one basis.

Diss. op. at 1-2. Yet the dissenters see the entire rationale for the corroboration rule differently from the majority:

[T]he purpose of the law is to establish the trustworthiness of the statement. In other words, not every element or fact contained in the confession must be independently proved. The goal is trustworthiness. Thus, if substantial independent evidence indicates the statement is trustworthy, then appropriate inferences may be drawn from the statement beyond those for which there is independent evidence including the fact that a crime has been committed.

Diss. op. at 4. And the dissenters see something of a parade of horribles hiding in the wake of the majority’s approach:

If the government were required to have independent evidence of every essential fact in the confession as the majority now concludes, then the confession is no longer independent evidence, it is a redundant supplement to the government’s other evidence. Moreover, the government would be barred from using the confession to fill in essential facts that might not otherwise be known to the government.

Diss. op. at 4. This is rather breathless, as the Government can still use a confession to fill in essential facts – it just must first prove that the confession is accurate. And I think that is a good thing, because the law already recognizes that a confession is the least reliable form of proof, as it is the only form that requires corroboration whatsoever. For this reason I think the dissenters’ assertion that a “confession is no longer independent evidence” is based on a false premise. A confession has never been independent of the other evidence that corroborated its essential facts.

Interestingly, however, CAAF’s majority and minority opinions in Adams split two judge who formerly shared the view that the court’s corroboration jurisprudence was too permissive. In United States v. Seay the court affirmed admission of a confession to stealing a wallet despite no evidence that the victim (who was murdered prior to the theft) actually possessed a wallet, reasoning that:

When a person confesses to participation in the larceny of a wallet, it is reasonable to infer the truth of the confession from the fact that the victim named in the confession knew the Appellant, died as a result of foul play, was found in a concealed place, and did not have a wallet at the time or thereafter.

60 M.J. 73, 80 (C.A.A.F. 2004). Judge Erdmann dissented from the court’s opinion in Seay, joined by then-Judge Baker, writing that:

Relying on these inferences as independent evidence, the majority opinion stretches the corroboration requirement beyond the breaking point. The corroboration rule requires independent evidence upon which inferences can be drawn, not inferences which substitute for evidence. Apart from the confession itself, no evidence suggests that [the victim] ever possessed a wallet at all, much less that he was carrying one at the time of his murder.

70 M.J. at 82 (Erdmann, J., dissenting).

I think the court’s opinion in Adams is the ultimate victory of Judge Erdmann’s analysis from Seay. That the analysis prevails despite losing one of its proponents makes it, I think, all the more compelling.

Case Links:
ACCA opinion
Appellant’s brief
Appellee’s (Government) brief
Blog post: Argument preview
• Oral argument audio
CAAF opinion
Blog post: Opinion analysis

8 Responses to “Opinion Analysis: There is no tipping point of corroboration in United States v. Adams, No. 14-0495/AR”

  1. Jolly Roger says:

    Zach, I’m not singing along with the refrain that confessions are the most unreliable form of evidence known to the law. Certainly it’s the only type of evidence that requires corroboration, but that could well be because of its overwhelming reliability. It could be that confessions are so devastating to the accused that they are nearly dispositive on their subject matter, so they law requires the Government to clear some corroboration hurdles to ensure basic reliability before (essentially) entering judgment against the accused by admitting his confession. There’s probably some history behind the rule I don’t know of but that seems to me the better rationale.
    The idea also doesn’t pass the non-lawyer’s smell test. If you ask a member of the public which is more reliable, the full confession of an accused following an Article 31 waiver or the testimony of an alleged victim with a motive to fabricate and a character for untruthfulness, few would choose the victim. But one requires corroboration before admission and the other requires nothing.

  2. Zachary D Spilman says:

    Well, Jolly Roger, as I wrote over a year ago in my argument preview for United States v. Knapp, 73 M.J. 33 (C.A.A.F. 2014) (CAAFlog case page):

    A confession is the only form of proof that is both presumed inadmissible (when challenged, the Government has the burden to show admissibility) and requires corroboration (separate from the confession itself or another confession). Unlike eyewitness identification, circumstantial evidence, or inconsistent testimony of an alleged victim dripping with bias and prejudice, an accused cannot be convicted on his confession alone.

    I think that speaks not to a confession’s reliability, but rather to its inherent unreliability.

    As for the “non-lawyer’s smell test,” there are a great many seminal principles of law that are would not survive that dubious test.

  3. DCGoneGalt says:

    Mr.Spilman: I understand your point and have seen the literature demonstrating false confessions exist, their known causes, and their prevalence.   However, even after having read this material, I agree with Jolly Roger that the corroboration requirement is necessary in large part because a confession is “case closed” evidence and therefore the government must demonstrate some form of independent evidence to show the confession is not the product of an oppressive interrogation or the product of a troubled mind.  If a confession is more unreliable than eyewitness testimony then I certainly have not seen that research material.
    I certainly believe that the main historical impetus for the rule of corroboration was to prevent governmental abuse in the interrogation process but, even though false confessions and government abuse exist today, I do not believe they remain the primary factor for advocating the retention of the rule.  I believe Jolly Roger correctly points out that most people, including me, give much more weight to a confession than they do any other form of evidence unless there is a great amount of evidence to undercut its reliability.

  4. The Silver Fox says:

    Completely agree, DCGoneGalt and Jolly Roger.

  5. Jolly Roger says:

    I don’t think it’s enough to say that there are procedural or other hurdles to admission, therefore the evidence is considered unreliable.  The burden is on the Government to show the admissibility of evidence that was the product of a search when challenged, too, but certainly not because of reliability concerns.  There are barriers to the admission of many types of evidence, like plea discussions, offers to compromise, liability insurance, subsequent remedial measures, and all the privileged communications and evidence.  There are a variety of policy reasons for those barriers, none of which involves reliability. 
    That doesn’t mean that you’re wrong, it’s just not enough to rest the case on procedural hurdles, especially in the face of the common sense consensus that confessions are super-reliable.  I think the strongest current rationale is that we should pause and make extra sure there’s something there before admitting evidence that is considered so reliable that conviction is nearly automatic, the same way we require a factual basis before accepting a guilty plea (the “strongest” form of proof known to the law).   That being said, I think DCGG is probably right that the original justification for most of the procedural barriers was reliability.
    I’m also not sure

  6. ZeroEl Student says:

    Interesting. I’m curious what the effect this ruling would have had on a recent case I observed.
    Sexual Assault/False Official statements case; six members of a unit gather at one of their coworker’s house for drinking and cards. E-3 female complaining witness, E-6 accused. Case not reported for 18 months. During prolonged interview with special agents, accused admits to providing alcohol to minors and consensual sexual activity with E-3. During trial, E-3 testifies that the last thing she remembers is playing cards in living room, and then waking up with shirt and underwear on in the bed of the E-6 the next morning – nothing in between. No other party guest testifies to seeing or hearing anything which occurred in bedroom, but they did see both enter at different times. Two beds in bedroom; no physical evidence of any sexual activity collected or admitted due to time delay in reporting.   
    During direct exam with the special agent, defense objects to confession statements due to lack of corroboration. The judges recesses for a long time to consider the issue, and ultimately overrules, finding slight corroboration because the confession talks about the general circumstances not in dispute (parties present, physical layout of the house, etc.). If I’m reading this ruling right, it seems like this ruling would leave out the admission of sexual activity on the part of the accused, and ultimately lead to an acquittal on the 120 charge at least.  (And he was ultimately convicted of wrongful sexual contact).

  7. stewie says:

    This is just like 404 evidence.  The concern isn’t lack of reliability, the concern is that the evidence is hyper-relevant and hyper-compelling that we either ban it, in the case of “he’s done this before so he did it this time” evidence, or we require additional protections, as in the case of confessions.
    Added to that IS the concern/possibility of voluntariness/coercion and false confessions, but that is an added concern, it does not mean the evidence is considered “inherently unreliable” it just means that if you have a confession, that’s pretty much it in the minds of the fact-finder (plenty of studies show this) so you better be darn sure the confession is legit before it comes into evidence.

  8. RKincaid3 (RK3PO) says:

    It seems to me that reliability IS in fact part and parcel of the requirement for corroboration.   If one recalls the founder’s were historians and educated–they knew all too well the power of coerced confessions ala torture at the hands of English Royalty, which, once obtained, obviated the need for all other evidence.
    While many may find coerced confessions problematic for reasons other than reliability (such as voluntariness–after all, some ask the obvious: how does one utter an admission/confession without choosing to utter, thus, ipso fato, the utterance is voluntary simply because the utterance occurred after the utterer for whatever reason chose to utter–Q.E.D./Res Ipsa Loquitor), how does one really escape the core issue:  the reliability of the utterred statement? 
    Regardless of how any statement is obtained, be it voluntary or coerced (torture of various sorts), the real issue it seems is reliability.  All else–from asking a simple, pleasant question to pulling fingernails; from depriving one of heat in a cold room to the Iron Maiden; from sleep deprivation to drawing on the Rack; from water boarding to the Breaking Wheel; from slapping or yelling accusatorily to disemboweling; from “good cop/bad cop” to the “Judas cradle”; it is all simply a question of the degree of government power utilized in obtaining an admission/confession.
    As history has evolved, mankind still debates what level of force is acceptable in obtaining a confession, applying different standards for different persons based upon the “status” of the subject (i.e., a suspect for 31B, one who is undergoing “custodial interrogation;” or one who is a “PoW” or an “unlawful enemy combatant.”  But that is the point:  where do we draw the line on obtaining statements when in the end, the fundamental issue is ALWAYS whether the resulting statement was “reliable” and thus should be believed when the government offers it as proof that one should be denied life and/or liberty for the sake fo society–according to government.
    And we will and should continue to debate this for in the process, we are debating our own humanity/inhumanity and tolerance for the least–and wordst–among us.