Your lyin' eyes
A long time ago, in a galaxy far away, I had a client named Larry Kaiser. He and his brother decided to rob a steakhouse one night. There was an off-duty cop who sought to intervene; he was killed, but managed to shoot Larry's brother. Larry, ever the solicitous sort, took his brother to the emergency room, opened the door and pushed him out onto the sidewalk, then sped off. The story of his capture a few days later coincided with the story of the officer's burial, and the front page of the newspaper was adorned with a large picture of the funeral procession, and a smaller mug shot of Larry. That day, the police got a call from a bar owner identifying the newspaper picture of Larry as the same guy who'd robbed his tavern a month earlier and killed a patron. Larry subsequently gained the distinction of being the only non-serial killer in Ohio to get consecutive death penalties for two wholly unrelated killings.
A year later, the US Supreme Court struck down Ohio's death penalty, and Larry's sentence was commuted to life imprisonment. I didn't have anything to do with the case after that, but I always thought that the identification in the bar shooting was bogus. If the police had shown the bar owner Larry's picture and told him, "We arrested this guy for killing a cop. Is this the same guy that shot up your place?" any court in the country would've suppressed any subsequent in-court identification as being the product of a blatantly suggestive procedure. That was in fact the information communicated to the owner by the newspaper article, but that doesn't mean anything if there's no police involvement.
Or does it?
That's the question the Supreme Court is going to have to answer next term in Perry v. New Hampshire. Perry had been stopped by the police on suspicion of burglary; at the moment he was standing in the parking lot in the company of a police officer, a woman who claimed to have witnessed the robbery looked out the window of her apartment and saw him. Her identification of him at that point was suspect; she'd given no description of the burglar to the police, she failed to identify him in a subsequent lineup and at trial. Had the police shown her Perry under those circumstances, the out-of-court identification would likely have been suppressed; a "show-up" has long been recognized as highly suggestive. But the witness' observation of Perry resulted from happenstance, not police instigation.
That was sufficient to end the matter for the New Hampshire Supreme Court, but other courts have reached opposite conclusions. Two years ago, the 1st Circuit in US v. De Leon-Quinones rejected the argument that a confrontation couldn't be deemed suggestive because the police hadn't been involved:
Because the due process focus in the identification context is on the fairness of the trial and not exclusively on police deterrence, it follows that federal courts should scrutinize all suggestive identification procedures, not just those orchestrated by the police, to determine if they would sufficiently taint the trial so as to deprive the defendant of due process.
That's a critical distinction. Suppression of evidence for violation of the 4th Amendment or Miranda is intended to deter police misconduct; in neither is the reliability of the evidence a factor. (And, at least in 4th Amendment cases, the evidence is suppressed despite the fact that it reliability is indisputable.) Suppression of identifications, however, is based solely on concerns as whether it is reliable. In fact, the Supreme Court's seminal decision on identification, Manson v. Brathwaite, said as much: "reliability is the linchpin in determining the admissibility of identification testimony."
The problem I have with the New Hampshire decision is that it defines the goal of the criminal justice system very narrowly: the concern is limited to ensuring that the government does not infringe upon a defendant's right to a fair trial. But substance is more important than process; we should derive little solace from the fact that the proceeding was fair if its conclusion was wrong. And no one denies that incorrect identification evidence plays a large role in wrongful convictions. In fact, one can argue that this indeed goes to the fairness of the trial: how can a trial be deemed fair if unreliable evidence is allowed to be introduced? That's the heart of Perry's claim of a due process violation.
So what do to about it? Sociological and psychological research in the 34 years since Manson was decided has shown that the five factors it lists for determining the reliability of an identification are of questionable significance. (One of them, how certain the witness is in the identification, has been shown by research to have virtually no correlation to how accurate the identification is; what's worse, jurors will place inordinate importance in the testimony of a witness who professes to be "110% certain" of his identification of the culprit.) The Telfaire instruction, developed five years before Manson and intended to inform the jury of the problems with identification testimony, has been shown by research to be "completely ineffective at sensitizing jurors to eyewitness evidence." Cross-examination is of little use; it's quite effective in ferreting out lies, but it's harder to shake a witness who honestly believes that he saw what he says he saw. Expert witnesses have been increasingly employed, but that can be an expensive proposition, and many courts have rejected such testimony on the grounds that problems with identification testimony are within "the common knowledge of the ordinary juror," and thus expert testimony isn't needed. (And, of course, appellate review of those decisions is under the obsequiously deferential "abuse of discretion" standard.) Many states, including Ohio, have passed laws or regulations requiring that police follow certain procedures in obtaining identifications which are intended to eliminate suggestiveness. The problem is that this presumes that suggestiveness is the major problem with identification testimony. It's not. The major problem is that people's eyes and memory don't work nearly as well as we think they do. In fact, laws mandating "double-blind" procedures for lineups or photo arrays may actually make the situation worse: a jury may conclude that simply because the procedures were followed, the identification is even more reliable than they otherwise would have believed.
The issue before the Supreme Court is a relatively narrow one: can an identification be suppressed if it is the result of a suggestiveness, even if the government had nothing to do with that. A broader ruling would entail looking at identification testimony itself, an issue which has bedeviled criminal justice for years. Given this Court, I don't anticipate a broad ruling, but we'll know more after the oral argument this fall.