Private: Eyewitness Identifications and Due Process

By Brandon L. Garrett, a professor at University of Virginia School of Law and author of "Convicting the Innocent: Where Criminal Prosecutions Go Wrong.”

While there had been great interest in the oral arguments in Perry v. New Hampshire yesterday, because the case is the first case dealing with due process standards for reviewing eyewitness identifications that the Court has heard since deciding Manson v. Brathwaite thirty-four years ago, and because there has been a drumbeat of revelations about wrongful convictions based on eyewitness misidentifications, the proceedings suggested that this was an odd case for the Court to have accepted in the first place. 

The case involved a report of a black man breaking into a car in a parking lot to steal a stereo. Police arrived at the scene and asked Barion Perry to wait with a uniformed officer. Another police officer went into a building to talk to an eyewitness, who promptly looked out the window and identified Perry, the only black man standing there, next to the officer in uniform, as the man who had stolen the stereo. The eyewitness was later unable to identify Perry either in a subsequent lineup or at his trial. The trial court rejected Perry’s claim that this was a show-up identification, where the eyewitness was shown a single suspect, that should be reviewed under the Manson v. Brathwaite due process standard.  The trial court, and then the New Hampshire Supreme Court, concluded that the identification happened by happenstance, and since the police did not intend to arrange a show-up, there was no state action, and due process did not apply.  Since the case was about whether to even apply the Manson v. Brathwaite standard, the larger and more pressing issues about improving that standard, regulating presentation of eyewitness evidence in the courtroom, and improving lineup procedures in the first instance, are not squarely raised.

The American Psychology Association amicus brief is a must-read since it addresses some of the larger problems that the Court likely will not address. It describes how crucial it is that law enforcement and courts get lineups right – as many as a third of identifications are known to be likely incorrect (because “fillers” are selected). Those errors due to faulty lineups hinder police investigations far more often than they cause wrongful convictions. When a single-person show-up rather than a lineup is used, the dangers of error are far greater. Show-ups are not only inherently suggestive, but they are even more problematic when not conducted in a neutral location (e.g. not the crime scene) – and even an “apparent connection” to the police can be highly suggestive to the witness. When Justice Kennedy asked whether there would be anything wrong with asking the witness, “is this the man,” the answer was emphatically yes – that is terribly suggestive – a line-up should be done instead, and even when a show-up is permitted, it should be done in a neutral location and making clear to the witness that the person may or may not be the perpetrator.

The New Hampshire Supreme Court has in effect added a new requirement, that police intentionally orchestrate or manipulate the identification in order to trigger the due process analysis. The Justices expressed some much warranted skepticism that the due process test should be modified, particularly since it would require courts to inquire into the “intent” or “mens rea”  of the police officer. The police in Perry may have been only negligent – but they nevertheless produced a suggestive identification. And we know that it was unreliable, because when later shown a photo array, the eyewitness was unable to identify the defendant – despite having identified him before.

The Justices also made comments about eyewitness memory generally and the need to regulate it. Some seemed to suggest that police suggestion should not be problematic – that somehow lawyers can just “tell it to the jury” if police contaminate an eyewitness identification.  Justice Scalia asked, “Why is unreliable eyewitness identification any different from unreliable anything else?” Justices raised other forms of potentially unreliable evidence and asked whether they should receive due process protection as well. Certainly other forms of evidence can be unreliable. I have studied how faulty forensics, contaminated confessions, and jailhouse informant testimony contributed to wrongful convictions, and I have argued that we can and should more carefully regulate certain other types of evidence. 

But what makes eyewitness memory so different from those other forms of evidence is that the memory of an eyewitness itself is highly malleable and fragile in a way that the jury cannot observe. Justice Scalia asked an important question – “what is magic about suggestiveness”?  The answer is that power of suggestion, even unintentional suggestion, has been documented by a massive body of empirical research. Suggestive police procedures can actually change an eyewitness’s memory of the person they saw. Even comments like “good job, you identified the suspect,” that are intended to reassure the eyewitness, can cause tragic errors, as in the well-known case of Ronald Cotton. That contamination, even if the product of negligence or good intentions, may be permanent.  The jury will see an eyewitness confidently identify the defendant – confidence that may be the product of suggestion. Jury studies show that jurors place great weight on that confidence. 

In my work studying DNA exonerations, I found that 76% of the first 250 DNA exonerations involved eyewitness misidentifications. I found that in court, eyewitnesses were almost without exception certain that they had picked the right person. Yet in each case, they were wrong. The unsound and suggestive identification procedures my research identified in most of these trials likely played a troubling role in generating their false confidence.  

The Court adopted “reliability” factors in Manson v. Brathwaite in order to water down prior due process holdings that unnecessarily suggestive police conduct should lead to the per se suppression of eyewitness identifications. Those so-called reliability factors do not correlate at all with reliability – they should be revamped. A now-massive body of social science evidence has discredited the current due process standard. The truly groundbreaking Henderson decision in New Jersey adopted a comprehensive social science framework for regulating eyewitness identifications in the courtroom. That case provides the national model that has long been needed so that the law can catch up to what social science has shown for years. Moreover, there is a deep need to ensure that sound lineups are conducted in the first instance. But we should not expect the Justices to be ready to carefully engage with the national issue of eyewitness misidentifications – as Dahlia Lithwick put it well, the Justices decide cases, and typically do not reach out to decide issues. The Perry case may simply be the wrong the case to address the larger problem.

Civil rights, Constitutional Interpretation, Criminal Justice, Equality and Liberty