Like the drip-drip-drip of a faulty faucet, last month produced yet another reversal of a murder conviction based on problems with eyewitness identifications.
A Cook County associate judge on Sept. 13 ordered a new trial for Jacques Rivera, who served 21 years for a conviction based solely on the eyewitness identification of a boy who was 12 years old at the time. Associate Judge Neera Lall Walsh said she believed the recantation of the eyewitness, Orlando Lopez. Lopez testified that a week after he picked Rivera out of a lineup, he saw a person on the street he suddenly realized was the actual killer. He immediately reported this new fact to the police, but they told Lopez they did not believe him. Lopez went ahead and made an in-court identification of Rivera that led to his conviction. But after hearing Lopez’s recantation, Walsh ordered a new trial.
There has been so much written about serious flaws in eyewitness identification procedures that you may think you have heard it all. But on Aug. 24, the New Jersey Supreme Court issued what is probably the most significant eyewitness identification decision ever produced by an American court. The opinion runs a daunting 134 pages. But no one even remotely connected to the criminal justice system can afford not to carefully read State v. Larry R. Henderson. 2011 N.J. LEXIS 927.
Here is the background. James Womble was with Rodney Harper in a dark hallway in January 2003. Two men approached. Womble knew one man, George Clark; the other was a stranger. While the stranger held a gun on Womble, Clark shot and killed Harper.
Thirteen days later, Womble picked Larry Henderson out of a photo array and identified him as the stranger. Henderson was subsequently charged with murder. The trial judge proceeded to conduct a pretrial hearing to determine the admissibility of Womble’s identification. At the hearing, Womble testified that he felt that the officer who conducted the photo array was “nudging” him to choose the defendant’s photo and that he felt “pressure” to make a choice. Nevertheless, the trial judge ruled that there was nothing improper about the identification and held it admissible.
At trial, Womble made an in-court identification, which was admitted in addition to his out-of-court identification from the photo array. But evidence was also introduced at trial showing that in the hours before the shooting Womble and his girlfriend shared two bags of crack, a bottle of wine and a bottle of champagne. Womble also admitted that it was “pretty dark” in the hallway at the time of the shooting.
The jury acquitted Henderson of murder, but convicted him of reckless homicide. On appeal, the defense made a broad attack on New Jersey’s standards for admitting eyewitness identifications, contending that recent scientific research questioned the validity of New Jersey’s rules. Thus, in 2009, the New Jersey Supreme Court remanded the case for the trial court to conduct a plenary hearing which would be presided over by a special master. The special master at the subsequent 10-day hearing heard from seven expert witnesses and examined about 200 published scientific studies on human memory and eyewitness identification. The special master issued a detailed report, which was then reviewed by the state Supreme Court in this case.
The New Jersey Supreme Court’s opinion completely revamps the state system for dealing with identifications. The opinion begins with some by-now-familiar statistics. Nationally, about 75 percent of the convictions that have been overturned due to DNA evidence have involved eyewitness identifications. In half of those cases, the eyewitness testimony was not corroborated by confessions, forensic science or informants. And 36 percent of these defendants were misidentified by more than one eyewitness. It has been estimated that about 7,500 of every 1.5 million annual convictions in America may be based on misidentifications.
The structure for the admissibility of identifications under the federal due process clause was established by the Supreme Court in Manson v. Braithwaite more than 34 years ago, long before most of the scientific studies were conducted. 432 U.S. 98 (1977). Manson created a two step procedure. First, the defendant has to establish that the procedures used by the state to obtain the identification were suggestive; if the defendant fails in this burden, the identification is admissible. However, if the court indeed finds that the defendant has proven the identification was suggestive, the defendant must then proceed to the second step and establish that the suggestiveness resulted in a very substantial likelihood of irreparable misidentification.
The Henderson opinion begins by dividing the variables that can affect an erroneous identification into two discrete groups. “System variables” are those factors like lineup procedures that are controlled by the state. “Estimator variables,” on the other hand, are factors over which the legal system has no control, such as the lighting conditions at the time of the crime. The problem with the Manson test is that the court is not allowed to even consider the “estimator variables” at the crime scene unless and until the defendant can establish that the state’s behavior has resulted in a suggestive identification.
Henderson rejects Manson by relying on the New Jersey Constitution’s due process clause. And it accomplishes this through two principal ways. First, it holds that a trial court should consider all relevant system and estimator variables as long as there is “some evidence” of suggestiveness. Second, it establishes more effective ways of helping jurors properly evaluate eyewitness identification evidence.
Briefly, here is the system. First, to obtain a pretrial hearing a New Jersey defendant has the initial burden of showing “some evidence” of suggestiveness that could lead to a mistaken identification. Usually this will be based on system variables controlled by the state.
Second, using both system variables and estimator variables, the state must then offer proof that the identification is reliable. The one proviso is that the court is allowed to end the hearing at any time if it finds that the defendant’s threshold allegation of “some evidence” of suggestiveness is groundless.
Third, the ultimate burden remains with the defendant to prove a very substantial likelihood of irreparable misidentification.
Fourth, if the court finds from the totality that the defendant has met its burden, then the court should suppress the identification evidence.
Henderson then offers an extensive, although not exhaustive, list of recommended system variables, including “double-blind” administration (where the officer is not involved in the case and does not know who the suspect is); instructions informing the witness that the suspect may not be present in the lineup; and considering whether the administrator recorded the witness’ statement of confidence immediately after the identification.
Notice that the U.S. Supreme Court’s Manson test does not direct the court to consider estimator variables unless and until the defense first actually establishes suggestiveness. The New Jersey Supreme Court’s new Henderson test, however, directs the trial court to consider estimator variables much earlier in the process — as long as the court believes there is “some actual proof of suggestiveness.” Henderson then sets out a non exhaustive list of 13 estimator variables, including lighting at the scene; time of the confrontation; and degree of the witness’ attention.
Finally, the New Jersey Supreme Court provides for enhanced instructions to be given to the jury both during trial and at the conclusion of evidence. These instructions provide empirical scientific findings to help the jury better evaluate eyewitness identifications. Henderson stated that it anticipated that these enhanced instructions would result in less need for expert testimony on this issue.
Illinois has hardly been a laggard in the area of concern over the reliability of eyewitness identifications. See, e.g., 725 ILCS 5/107A-5 (lineup and photo spread procedure); 725 ILCS 5/107A-10 (pilot study on sequential lineup procedures); Report of the Governor’s Commission on Capital Punishment (2002), Recommendations 10-15. But Henderson is the single most thoughtful appellate opinion I have ever read on the subject of eyewitness identifications. It demands serious attention.