This week the Supreme Court decided two cases that highlight the risks of relying on eyewitness testimony to convict people of crimes. On Tuesday the Court overturned the conviction of Juan Smith, who was charged with killing five people during an armed robbery of a New Orleans home in 1995. The only evidence connecting him to the crime was the testimony of Larry Boatner, who was visting the house at the time of the robbery. During the trial Boatner identified Smith as the first gunman to push his way into the house, saying he looked him in the face immediately afterward. As part of the appeals process, Smith's lawyers obtained police records that contradicted Boatner's testimony. According to notes taken by the lead investigator on the night of the murders, Boatner "could not…supply a description of the perpetrators other then [sic] they were black males." Five days after the robbery, the same detective noted, Boatner said he "could not ID anyone because [he] couldn't see faces" and "would not know them if [he] saw them." In a typewritten report, the detective said Boatner "could not identify any of the perpetrators of the murder." Under the 1969 decision Brady v. Maryland, the Supreme Court ruled in an opinion by Chief Justice John Roberts, Smith had a due process right to those records because they were favorable to the defense and there was a "reasonable probability" that they would have changed the outcome. The sole dissenter, Justice Clarence Thomas, conceded that the conflicting statements could have been used to impeach Boatner's testimony but argued that they would not have led to an acquittal. "When weighed against the substantial evidence that Boatner had opportunities to view the first perpetrator, offered consistent descriptions of him on multiple occasions, and even identified him as Smith" out of a photo array, Thomas wrote, "the undisclosed statements do not warrant a new trial."
The other case dealing with eyewitness testimony also produced an 8-to-1 decision, this time against the defendant. Barion Perry was arrested in 2008 for breaking into cars at a Nashua, New Hampshire, apartment building in the middle of the night after a tenant identified him as the perpetrator while looking out a fourth-floor window at the parking lot. At that point Perry was the only black man in the area, and he was standing next to a police officer. Perry argued that the judge should not have allowed the jury to hear this eyewitness testimony because it was tainted by impermissible suggestion. He cited Supreme Court rulings saying eyewitness testimony should be excluded when it is the product of unnecessarily suggestive procedures that create a "substantial likelihood of misidentification." The Court, in a majority opinion by Justice Ruth Bader Ginsburg, rejected Perry's analogy, noting that the circumstances of his arrest were not deliberately arranged by police, so a rule excluding such evidence would not deter improper procedures. Generally speaking, the Court said, the remedy for inaccurate eyewitness testimony, like the remedy for other kinds of potentially unreliable evidence, is convincing the jury that it should be discounted.
That seems reasonable to me, as long as defense attorneys have the means and opportunity to present expert testimony about the perils of eyewitness testimony. I think there remains a substantial gap between public perceptions and scientific research on the question of how much trust to put into someone's confident in-court identification of a defendant. Even when the witness is completely sincere, there are so many ways in which memories can be biased or manufactured, whether intentionally or unintentionally, that such testimony probably should never be enough by itself to convict someone. The research suggests eyewitness identifications are wrong about a third of the time.
In Perry's case, there was other evidence: When a police officer arrived at the parking lot, she heard what "sounded like a metal bat hitting the ground." Then she saw Perry standing between two cars, a metal bat on the ground behind him. He was holding two car-stereo amplifiers, which he claimed to have found on the ground. But in Smith's case, there was no physical evidence and no witness testimony except Boatner's. It seems to me the contradictory statements withheld by the Orleans Parish District Attorney's Office would have created plenty of reasonable doubt. Yet Clarence Thomas disagrees, and I worry that a lot of jurors may likewise be unreasonably impressed by eyewitness testimony.