A staggering new piece of research shows that in almost 20% of criminal trial convictions that DNA evidence has subsequently overturned, the prosecution was able to rely on ultimately false confessions by the accused.* Notably, in virtually every single one of these cases, the false confession was remarkably detailed in its description of the crime, creating the illusion that the confession was both voluntary and reliable. How can this be? How can an innocent person not only confess to a crime they didn’t commit, but actually do so in a manner that is sufficiently detailed to be virtually unassailable?
Doug Mataconis suggests that false confessions in general are so easy to obtain because the police are able to skirt the spirit of Miranda v. Arizona and its predecessor decisions because:
“The police….have been authorized by [other] Supreme Court precedent to engage in lying and trickery in order to obtain information from a suspect. They can tell a suspect that someone has already implicated them in a crime even if that never happened. They can tell a suspect that there are eyewitnesses who don’t exist. They can say that there is evidence implicating the suspect that doesn’t exist.”
This sounds about right to me, particularly given, as Doug notes, that many of these falsely accused suspects are abnormally vulnerable to this type of pressure because they are either young or mentally ill, and are rarely (if ever) acting in the presence of counsel. I’d just add that it is easy to see why an innocent would even be more likely to provide a confession than a guilty party, since a guilty party will often have had ample time to prepare himself for being arrested on the charge in question: few people will know their criminal procedure rights less than someone who never thought they’d be arrested, while few people will know their criminal procedure rights more than someone who fully expects to be arrested.
As for why these confessions are so detailed, the study makes clear that police interrogators quite regularly – and impermissibly, if often unintentionally – “contaminate” the interrogation by introducing details of the crime into the questioning. In some instances, the police were found to even have corrected the accused when he got a particular detail wrong.
All of this suggests that the balance that the Supreme Court tried to strike in Miranda, by requiring that arrestees be aware of their constitutional rights for any confession to be admissible in court but permitting most confessions so obtained to be admissible, has failed. Indeed, it is now well-established that Miranda warnings have “little to no effect on a suspect’s propensity to talk.” Reform is clearly needed, even as it is important to keep in mind that the original balance that the Supreme Court tried to strike was an important one: the deceptive interrogation tools that Doug references can in fact serve a valuable investigative function in permitting the police to obtain valuable evidence and they probably do not, in and of themselves, amount to Constitutional violations.
The obvious solution is, frankly, simple: post-arrest interrogations should not be admissible evidence unless conducted in the confirmed presence of an attorney. This study demonstrates that even detailed confessions are of, at best, limited reliability, while their use as evidence is almost infinitely prejudicial – indeed, in 8 of the convictions included in the study, the DNA evidence that exonerated the accused was actually introduced at trial, but the prosecution was still able to obtain a conviction. In a society that would presumably prefer “that ten guilty men escape than that one innocent suffer,” but which gives its law enforcement enough tools to have the highest per capita prison population on Earth, a solution that would heavily limit the admissibility of confessions but would likely reduce wrongful convictions by at least 15-20%, if not more, seems like a perfectly worthwhile idea.
But, as I said, Miranda tried to strike an important balance that should not be lost in the shuffle. There is a legitimate need for the police to be able to interrogate a suspect and use the tools they have Constitutionally available to obtain information from that suspect that will at the very least lead to the discovery of admissible evidence, whether against the particular suspect or against a co-conspirator. Moreover, I have a hard time arguing that the admissibility of a confession obtained in the absence of counsel (but with a “knowing” waiver of the right to counsel) rises, by itself, to a Constitutional question – instead, it would simply be a rule of evidence resulting from the high degree of unreliability involved in any confession obtained in the absence of counsel. So, keeping this in mind, there would be no “fruit of the poisonous tree” doctrine involved here, except to the extent it’s already inherent in a failure to issue a Miranda warning. That means that if a confession were obtained after the defendant was Mirandized but outside the presence of an attorney, evidence discovered as a result of the confession would be admissible, even though the confession itself would not be.
Unfortunately, the more likely solution for the foreseeable future, mandatory videotaping of confessions, will only marginally improve things in the short-term, and as criminal defense attorney Scott Greenfield points out, in the long run will probably just morph into a new tool for law enforcement and prosecutors. The other plausible solution here – an outright prohibition on the use of lies inflating the evidence against the accused – is equally problematic, as it would be difficult to enforce, even in combination with mandatory videotaping, and as interrogators would simply develop new methods of coercing confessions….that is, after all, their job to no small extent, and there’s no real reason to think that forcing them to alter their methods of interrogation in this manner would make confessions significantly more reliable in the long run.
I recognize that we are never going to have a system that ensures that only guilty people ever go to jail. But in a supposedly “free” country, the fact that we imprison more of our people than any other nation in the world should be an unconscionable embarassment in and of itself. That some unknown percentage of those people are, in fact, innocent should thus be even more unacceptable in the United States than it would be elsewhere. In short, there is simply no justification for the continued use of confessions obtained outside the presence of an attorney as evidence in a criminal proceeding. Frankly, it is difficult for me to conceive of many cases where a defendant’s interrogation-induced confession would be absolutely necessary to a prosecutor’s ability to obtain a conviction where the defendant is, in fact, guilty. I’m sure some such cases exist, but they will be few and far between if the prosecution and law enforcement have obtained all or most of the other evidence out there. The opportunity to eliminate at least 15-20% of wrongful convictions is thus an opportunity to save countless lives from being ruined at litle to no cost to public safety. This will never happen, though: the cost to our collective ego of seriously acknowledging that there is such a thing as too “tough on crime” is too great.
*So far as I can tell, the study focuses only on cases where a falsely-accused defendant’s confession was used at trial. To the extent false confessions were obtained in cases where the exonerated person entered a guilty plea without trial, and were a substantial factor in the decision to enter the guilty plea, this percentage may be significantly higher. Moreover, such cases are likely to be far more common, since the presence of a confession makes it “akin to being a kamikaze” for the defense attorney to allow the case proceed all the way to trial.
NOTE: I apologize for my recent extended absence around these parts. I’ve been dealing with some site access issues that I’m hoping will soon be cleared up.