New Jersey is a leader in addressing problems with eyewitness testimony

McKinley Cromedy spent five years behind bars in New Jersey after a rape victim testified she was certain he was the one who attacked her.

Cromedy's lawyer questioned the ability of the woman, who is white, to differentiate among black men like the defendant, but the jury convicted Cromedy on the strength of the victim's memory.

DNA eventually showed he didn't do it, and the New Jersey Supreme Court responded with a bold move: It ordered trial judges across the state to instruct juries about the difficulties of cross-racial identification.

More than a decade later, New Jersey remains a leader nationally in efforts to deal with the problem of misidentification. A report filed last week with the state high court said that even more steps were needed to take advantage of the wealth of scientific studies casting light on the issue.

"New Jersey is a vanguard state," said Duquesne University law professor John T. Rago, among those working on the issue in Pennsylvania as part of an examination of the underlying causes of wrongful convictions.

Rago and others said Pennsylvania has a way to go to catch up with its neighboring state on the possible pitfalls of eyewitness testimony, which is among the most powerful evidence in criminal cases.

"There hasn't been much in the way of reform in Pennsylvania at all," said Gary Wells, a psychology professor at Iowa State University who was one of seven experts to testify before the special master who filed the New Jersey report on the issue.

A final report to the Pennsylvania Senate, due in late summer, is expected to explore eyewitness identification.

"It's not an easy issue," said Rago, who said the 51 committee members were examining scientific advancements focusing on a number of avenues of criminal investigation.

The problem with eyewitness identification has come to light because of DNA testing, the great equalizer in the criminal justice system because it helps to convict the guilty and clear the innocent.

About three-quarters of the 254 defendants exonerated nationally by DNA testing, including Cromedy, had been convicted largely on the basis of eyewitness testimony.

"It was him," the young rape victim testified in 1994, telling a jury that Cromedy had attacked her in her basement apartment near the Rutgers University campus in New Brunswick.

Cromedy's lawyer asked the judge to tell the jury about studies showing that people of one race often have difficulty identifying a perpetrator of another race. The judge declined, and the jury convicted Cromedy, who was sentenced to 60 years in prison.

In the 1999 decision on the appeal, the state Supreme Court cited the scientific studies on eyewitness identification when it ruled that the cross-racial instruction should be given when the eyewitness is uncorroborated.

Two years later, New Jersey again addressed the issue when the state Attorney General's Office issued guidelines aimed at preventing police from making suggestive comments during a photo array or lineup.

Now, the state's high court will start to weigh the 88-page report submitted by the special master, retired appellate judge Geoffrey Gaulkin, who heard testimony from seven experts and examined more than 200 studies, articles, and books on the evolving science about eyewitness identification.

Gaulkin recommended that judges adopt the scientific studies on the issue and that prosecutors be required to prove the reliability of an eyewitness during a pretrial hearing, much as they already must show the integrity of trace evidence collected at crime scenes.

It's not that eyewitnesses are lying. Studies attribute mistakes to a variety of causes. Memory changes with time, police may be too suggestive during lineups about who they believe is the perpetrator, and if weapons are used in crimes, they draw victims' attention away from the criminals. Also, even the most-certain eyewitness can be wrong.

"A new framework is urgently needed to address what the science has told us," said Ezekiel R. Edwards, a lawyer with the Innocence Project in New York who participated in the New Jersey inquiry on the issue.

Prosecutors have expressed concern that even valid eyewitness identifications might be cast in doubt and that eyewitnesses will feel they are on trial.

Rago said he, too, worried about the victims. The goal, he said, is to come up with sound practices "in a way that respects the very, very difficult journey that a victim goes through."

The sheer magnitude of the New Jersey inquiry was hailed by some as an example of how seriously the state high court views the matter. Oral arguments on the issue are expected in the fall.

The review was ordered in a Camden case in which the defense contended that the police had violated the 2001 guidelines on lineups.

Those guidelines say that an administrator not involved in the case should conduct the lineup and that individual photos of suspects should be shown sequentially, instead of as part of an array of photos shown all at once.

The guidelines also say that witnesses should be told the perpetrator might not even be among the suspects in a photo array or lineup.

Pennsylvania, meanwhile, has no statewide guidelines on lineups.

Pennsylvania also is one of only two states that have barred expert testimony on the evolving science of memory and identification, said Rago, who said Tennessee and a federal appeals court in the West also prohibit such testimony.

The Pennsylvania Supreme Court said in a 1954 case that jurors can view an eyewitness with caution if the witness did not have a good view or had equivocated on the identification. But the court had not broached issues such as cross-racial identification.

Veteran criminal-defense lawyer Jules Epstein, a law professor at Widener University, examined the issue of mistaken identification in a research paper comparing how Pennsylvania, New Jersey, and Delaware have dealt with the problem.

He concluded that, except in New Jersey, too little has been done to prevent miscarriages of justice.

But, he noted, the problem is nothing new.

Epstein, among the experts called to testify in New Jersey, dug up a Pennsylvania Supreme Court case from 1896.

In that case, the court called identification "one of the least reliable" kinds of evidence and one of the "more difficult subjects" for the justice system.