The Law and Science of Eyewitness Identification

“Several years ago, I sat in a tavern in a city of abandoned textile mills with a client who had been exonerated of several rape convictions by DNA testing. Like the city itself, my client saw better days ahead. The police investigations into the crimes were riddled with errors, and he would later reach a multimillion dollar settlement in his civil rights case. Yet, as we met for the first time to look over the case file, another disturbing aspect became clear: the sheer number of times witnesses had misidentified our client during the investigation.

One area of law in need of profound change is the process by which courts and investigators collect witness identification evidence. Eyewitness identification plays a fundamental role in criminal cases, and also plays a part in civil cases and internal business investigations. But this critical tool is often plagued with reliability problems. While it is important to use identification procedures to solve crimes, it is equally important to exclude innocent people from prosecution.

According to the Innocence Project, since 1989, there have been 349 wrongful convictions overturned via DNA testing. In those cases, eyewitness misidentification was a factor in over 70% of those cases—far more than any other factor.

Last year, I attended the 2016 National Symposium on Eyewitness Identification Reform along with other lawyers and investigators. We studied and discussed the latest research. Historically, many procedures for testing eyewitness identification were not validated in a scientific manner before being implemented by investigators. As a result, many of us have been trained using eyewitness identification techniques that are outmoded and flawed—and those methods have led to innocent people being implicated in an investigation.

However, forward-thinking investigators throughout the country are making changes in their techniques based on new scientific research in this area. Below, I summarize some highlights of the research and provide a link to free training materials.

The Law

The U.S. Supreme Court’s decision, Neil v. Biggers (1972), was one of the first cases where the court made extensive recommendations on evaluating eyewitness testimony. In somewhat of a departure for Supreme Court decisions, the court delved into the facts:

“The assailant directed the victim to ‘tell her [the daughter] to shut up, or I’ll kill you both.’ She did so, and was then walked at knifepoint about two blocks along a railroad track, taken into a woods, and raped there. She testified that ‘the moon was shining brightly, full moon.’”

Biggers was later identified during a live showup at the police station, where the victim viewed Biggers. He was made to repeat the threat uttered by the rapist. Finding the showup here met the criteria for reliability, the court listed five factors for evaluating the accuracy of eyewitness identifications:

“As indicated by our cases, the factors to be considered in evaluating the likelihood of misidentification include the opportunity of the witness to view the criminal at the time of the crime, the witness’ degree of attention, the accuracy of the witness’ prior description of the criminal, the level of certainty demonstrated by the witness at the confrontation, and the length of time between the crime and the confrontation.”

The Neil v. Biggers factors were confirmed in a 1977 case, Manson v. Brathwaite, in which the Supreme Court set a low barrier to admitting shaky evidence, holding that if the circumstances of the identification procedure were suggestive, the Court must then weigh the “totality of the circumstances” to determine whether the identification is still reliable. The law of the land for over three decades now, Manson v. Brathwaite has come under increasing criticism that its holding is inconsistent with recent scientific studies into witness identification and memory.

The Supreme Court has noted the unique power of witness testimony; Justice William J. Brennan, Jr.’s dissent in the 1981 case Watkins v. Sowders summed it up best:

“[E]yewitness testimony is likely to be believed by jurors, especially when it is offered with a high level of confidence, even though the accuracy of an eyewitness and the confidence of that witness may not be related to one another at all. All the evidence points rather strikingly to the conclusion that there is almost nothing more convincing than a live human being who takes the stand, points a finger at the defendant, and says, `That’s the one!’”

The Research

What do we mean when we refer to eyewitness identification?

Eyewitness identification can occur in several formats: a showup, live lineups, or a photo lineup. Showups—where a suspect is apprehended and shown “live” and in-person to a witness, close in time to the event—have been demonstrated to have higher percentages of both correct and incorrect identifications as compared to lineups. Showups tend to be have heightened problems with suggestibility, given the fact that they involve just one person who is surrounded by multiple police officers.

Lineups occur at a time somewhat removed from the event, but because they are planned and staged, a process exists to minimize contamination by suggestive elements (for example, a suspect standing closer to the victim than all lineup fillers; police clapping after a successful identification of a suspect).

How often are witnesses successful in their identification? In one study of 11 peer reviewed studies and over 6,000 actual police lineups, witnesses selected the suspect 41% of the time; witnesses selected a lineup filler 37% of the time (i.e., they selected a person who was known to be innocent of the crime).

Several phenomena have been observed in studies where a crime is staged—say, a person enters a classroom in front of student “witnesses” and steals a purse—and the students/witnesses later are interviewed in a controlled setting about their observations.

Here are a few key findings:

1. Proper eyewitness instructions can reduce false positives.

Research has demonstrated that eyewitnesses tend to identify the person from the lineup who, in the opinion of the eyewitness, looks most like the culprit relative to the other members of the lineup. The problem became readily apparent when experiments were run where the actual culprit was not present in the lineup. Under controlled conditions, a concept called the “relative judgment process” will often yield a positive identification—even when the true culprit had been removed from the lineup. While this may seem to be an obvious, simple observation, research shows the effect is incredibly damaging to accuracy.

But research has demonstrated that this problem can be partly remedied with a simple instruction to the witness that the true culprit may not be in the lineup. Some state courts have adopted a set of instructions to be given to eyewitnesses—such as a caveat that the witness need not feel compelled to make an identification, and that police will continue to investigate this incident, whether or not the witness makes an identification.

“These instructions can reduce pressure on the witness to feel as if they must pick a culprit, and this leads to more accurate results.

2. Double-blind identification procedures can prevent information “leaks.”

A double-blind eyewitness identification is one in which neither the administrator nor the eyewitness knows the identity of the suspect. Research has shown that witnesses were adept at picking up clues from the person administering the lineup when that administrator knew the identity of the suspect. The lineup administrator tended to “leak” information about the suspect in subtle (and not so subtle) ways: leaning forward at certain times; raising his hands; making a comment like “take another look at the third guy;” smiling or nodding in agreement when a positive identification was made; or conversely, frowning when the wrong person was chosen. These kinds of cues lead to inaccurate identifications, as well as false levels of confidence in the witness.

Studies have shown there is far less chance of contamination of the process when double blind testing is implemented.

3. Confidence statements from the witness are most reliable at the time of initial identification.

The role played by eyewitness mistakes in the DNA exoneration cases has helped to create a growing impression that eyewitness memory is inherently unreliable. I have heard several witnesses tell me this on recent cases, in essence doubting their own ability to remember accurately. But this overstates the case. Researchers have discovered—perhaps surprisingly given the general trend— that when eyewitnesses are questioned using proper identification procedures, the confidence they have in their initial identification usually is a highly reliable indicator of accuracy.”

Read More