In an attempt to prevent a misidentification early in the proceeding, Monroe County Judge John L. DeMarco in People v. Flowers1 recently directed that certain ground rules be in place for a line-up. Judge DeMarco ordered that when administrating the line-up the police employ a “double blind” technique, where the person administrating the line-up does not know who the suspect is, and that sequential line-up procedures be utilized, where the witness views the persons in the line-up one at a time instead of all at once. Further, he ordered that the process be videotaped or photographed for the court’s review. These procedures help reduce the danger of two of the most common causes of false identifications from a line-up (NYLJ, Feb. 1).

Like the increasing understanding of false confessions which has smashed cherished myths that people never confess to crimes they didn’t commit, eyewitness reliability research has revealed that eyewitnesses, even those “certain” of their identification, are often wrong. Recognizing the staggering injustices brought about by juries’ naïve reliance on eyewitness evidence, New York courts are increasingly inclined to admit expert testimony to educate jurors on the flaws of eyewitness testimony. In addition, some courts are now trying to reduce the number of misidentifications by ordering law enforcement to change decades-old practices-related line-up procedures and interaction with eyewitnesses.

The trend started in 2007 with People v. LeGrand,2 when while reversing a conviction for second degree murder, the Court of Appeals articulated standards for judicial discretion in deciding whether or not to admit—or conduct a hearing on—eyewitness expert testimony. The case turned solely on an uncorroborated eyewitness identification made seven years after the event. The Court decided that when the case “turns on the accuracy of eyewitness identifications and there is little or no corroborating evidence connecting the defendant to the crime,”3 expert testimony is appropriate to explain the testimony and its potential pitfalls.

LeGrand found that studies on the lack of a “correlation between confidence and accuracy of identification, the effect of post-event information on accuracy of identification and confidence malleability” are generally accepted as scientifically valid,4 thus obviating the need for Frye hearings on these factors. Left unanswered by LeGrand was just how much corroborating evidence is necessary for a trial court to appropriately preclude the expert testimony and whether additional eyewitnesses count as corroboration.

The Court next explored this issue in People v. Abney5 which consolidated Abney with People v. Allen. The Court reversed the conviction in Abney and affirmed it in Allen.

Allen is notable because the fact that more than one eyewitness identified the defendant, including at least one witness who knew the assailant before the crime, was implicitly deemed “corroborating” evidence, as it bore “strong indicia of accuracy.”6

The case against Abney was not completely devoid of corroborating evidence. Abney involved a 13-year-old knifepoint robbery victim identifying the defendant. The sole possible corroboration was weak evidence that the defendant had tried to create documentation of an alibi before even being accused. Nonetheless, the Court reversed, highlighting the principle that weak corroboration may not be sufficient to defeat the request to introduce expert testimony.

The concept in Allen and Abney was nuanced in People v. Santiago7 where the Court held that multiple eyewitnesses do not necessarily constitute corroboration of the sort that allows for the rejection of expert testimony. In Santiago, the two “corroborating” eyewitnesses saw only part of the perpetrator’s face and—critically—the Court noted the possibility of taint and eyewitness error of the very sort that would be the subject of typical defense-proffered expert testimony. The Court reversed. In reading both Allen and Santiago, it is apparent that the Court focused on the indicia of reliability of the eyewitnesses.

The Court of Appeals’ rulings are remedial and open the door for the defense to educate a jury on some of the flaws in eyewitness testimony. While important, these measures do not alter the landscape which permits the misidentifications in the first place.

It is not sufficient to attempt to correct the errors after they occur, after an innocent person has been wrongly prosecuted and/or convicted. Thanks to The Innocence Project, which reports that eyewitness misidentification plays a role in more than 75 percent of convictions overturned by DNA testing, we can all now see a hole in the criminal justice system that needs to be mended.

Judge DeMarco’s ruling mandating law enforcement to employ better practices at the outset is an important step to begin to curb the number of errors in eyewitness testimony. It may be time for the Legislature to mandate these safeguards for every line-up across the state. On Feb. 14th, in his State of the Judiciary message, Chief Judge Jonathan Lippman recognized the need for improvements (NYLJ, Feb. 15). He proposed that certain safeguards be employed for photographic displays but he tied the proposal to admitting these arrays at trial. That would be a mistake. As the courts have previously found, jurors may guess the photo in the array is from a prior arrest. We ought not to sacrifice one safeguard in order to obtain another. Further, the safeguards ought to be mandated for line-ups as well as photo arrays.

All parties involved in the criminal justice system including prosecutors, the defense bar and judges alike, have a responsibility to continue to educate themselves, jurors and the public about the dangers of naïve reliance on eyewitness identification and the faulty techniques still commonly employed to secure those identifications.

Bruce Barket and Steven B. Epstein are partners at the Long Island-based firm, Barket, Marion, Epstein & Kearon.


1. 1131B/11.

2. 8 N.Y.3d 449, 867 N.E.2d 374 (2007).

3. LeGrand, at 452.

4. LeGrand at 458.

5. 13 N.Y.3d 251, 918 N.E.2d 486 (2009).

6. People v. Santiago, 17 N.Y.3d 661, 671, 958 N.E.2d 874, 882 (2011).

7. 17 N.Y.3d 661, 958 N.E.2d 874 (2011).