Here’s a 2010 judicial opinion from the Sixth Circuit that deals pretty thoroughly with the Corroboration Requirement for confessions. Here’s the nitty gritty definition:
A dusty doctrine of criminal law—the “corroboration rule”—lies at the heart of this appeal. Rooted in English common law and embraced by the United States Supreme Court in the 1950s, the rule says that no one may be convicted of a crime based solely on his uncorroborated confession. Opper v. United States, 348 U.S. 84, 89, 93 (1954); see also Smith v. United States, 348 U.S. 147, 156 (1954); United States v. Calderon, 348 U.S. 160, 165 (1954). Early versions of the rule developed in 17th-century England when a series of suspects confessed to murders, only to have their alleged victims turn up—alive and well—long after the suspects were imprisoned (or, worse, executed) for the fictitious crimes. See Thomas A. Mullen, Rule Without Reason: Requiring Independent Proof of the Corpus Delicti as a Condition of Admitting an Extrajudicial Confession, 27 U.S.F. L. Rev. 385, 399–401 (1993) (citing Perry’s Case, 14 How. St. Tr. 1311 (1600)). First known as the corpus delicti, or “body of a crime,” rule, the doctrine guarded against convictions for imaginary crimes by requiring prosecutors to demonstrate through independent evidence that the crime occurred before they could use an accused’s own statements to establish guilt. Smith, 348 U.S. at 153–54; see also United States v. Sterling, 555 F.3d 452, 456 (5th Cir. 2009).
It’s a thorough opinion, to be sure. It also spends a lot of ink bloviating about whether the rule is necessary anymore:
The emergence of several new criminal procedure protections over the last several decades . . . has altered the role of the rule. Since 1954, the Court has constitutionalized several areas of criminal law, many to the end of preventing false and uncounseled confessions. Not the least of these are the requirement that law enforcement read custodial suspects their rights before interrogating them, Miranda v. Arizona, 384 U.S. 436 (1966), the requirement that interrogations cease upon a suspect’s request for an attorney, Edwards v. Arizona, 451 U.S. 477, 484–85 (1991), and the multi-factor inquiry courts undertake to ensure the voluntariness of confessions, see, e.g., Withrow v. Williams, 507 U.S. 680, 689 (1993).
Of course, the last of these justifications—ensuring that a confession is voluntary—is irrelevant to the question at hand. Voluntariness goes to a separate issue—admissibility. The court says as much earlier in the opinion:
The corroboration requirement, it bears adding, differs from another requirement—that the confession be voluntary. Voluntariness goes to admissibility,while corroboration goes to sufficiency. See Smith, 348 U.S. at 153.
Notwithstanding the court’s contradictory language, the fact remains that the Court is right—to some extent at least—about the Corroboration rule. The modern version of the Rule is so weak that it’s redundant. Under the modern rule, a prosecutor doesn’t need independent evidence that a crime was committed by the confessor in order to admit a confession to the record. The prosecutor need only prove that a crime was in fact committed. Once you’ve got evidence that a crime took place, you can literally take anyone’s confession in the world, so long as you can demonstrate the confession was voluntary.
The real problem with the Rule as it stands is that, in any situation where this rule would actually grant you relief, you would also have grounds for other relief as well. For example, you could make a motion for a JMOL (Judgment as a Matter of Law) at the end of the prosecution’s case in chief on the basis that the prosecution failed to meet their evidentiary burden as a matter of law. If the jury’s already gone to verdict, you can also make a post-trial motion for a JNOV (judgment non obstante veredicto) on the basis that no reasonable juror could have convicted on the evidence (rarely granted, admittedly, but theoretically possible). Since the standard of proof that governs the Corroboration Rule is so far below reasonable doubt, any situation in which you haven’t satisfied the Corroboration rule is, ipso facto, a situation in which the prosecution has failed to meet its burden to prove each element of a crime beyond a reasonable doubt. If the prosecutor can’t even prove that a crime was committed at all by a standard of evidence that falls below a “preponderance,” then you’ve got issues with your case serious enough to warrant the relief mentioned above.
I honestly can’t think of a situation in which you would have grounds to make a motion for relief on Corpus Delicti grounds without also having basis to move for a JMOL or JNOV on the grounds that the prosecution failed to meet their evidentiary burden as a matter of law. I suppose the advantage is that the traditional Corpus Delicti objection can be made before trial in the form of a motion to suppress. A motion for JMOL or JNOV, on the other hand, is usually filed during or after trial, respectively (since you have to wait for the prosecution to present their case-in-chief before you can move for a directed verdict on insufficient evidence). So I guess that’s probably one reason to keep the rule around. Then again, that’s the sort of thing that can probably be resolved at a probable cause hearing—if you’re in a jurisdiction that allows for that sort of thing.
So there’s really only two options to fix this, as I see it: the Corpus Delicti rule should be strengthened to require independent evidence that shows identity in addition to an actus reus—which would be my preferred option—or, in the alternative, courts should stop leading Defendants on and do away with it completely. At this point, it’s just giving Defendants false hope and taking up space in appellate briefs. Given the Sixth Circuit’s “glowing” review of the Rule, you can take a wild guess as to which one is more likely to happen in the future.