But even as New York City detectives strove to make the lineup more fair by concealing one identifying feature, they left another clue in plain sight: It was lying on the floor, next to the ankle of the man in the No. 5 position.
A shackle. A dead giveaway that No. 5 is the police suspect.
It was not the only potential problem with the lineup, conducted on June 15, 2017, at the 101st Precinct in Queens. The suspect was the only one with twisted, or braided, hair and whose hat had an adjustable strap.
Some lineups do what they are supposed to — engage a witness’ memory and test whether the police have the right person. But others contain clues that could steer the witness toward the suspect whom the police already have in their sights. A judge threw out a witness identification based on the Queens lineup; the robbery charges are still pending.
To be sure, the quality of New York lineups appears to have improved over the years. For a point of reference, consider a 1998 lineup, from a Manhattan manslaughter case. No. 5 seemed to be offering a hint as to the suspect: It was the man in the No. 6 slot, the one to whom No. 5, a police officer, appeared to be pointing.
During the past two decades, the policing profession has become more attuned to the risks of mistaken identifications. Witnesses are wrong with considerable frequency when they claim to recognize a culprit. And detectives have come to understand that suggestive police procedures, such as unfair lineups, raise the risk that an innocent person is prosecuted.
Like many other departments across the country, the New York Police Department has taken steps to improve fairness.
“It is absolutely not in our best interest to do something that is going to be suppressed,” said Dermot Shea, the city’s chief of detectives.
To minimize differences between the suspect and fillers, detectives will cover up tattoos and use hats to obscure differences in hair styles, Shea said. “Fairness is driving everything that we do to make sure that we get the best prosecutable case, and that includes having a fair lineup,” he said.
But suggestive lineups are not yet a thing of the past, as the shackle on the floor in the Queens lineup would seem to indicate.
Detectives, for instance, sometimes still place teenage suspects in lineups alongside grown men, The New York Times found. A year ago, a 17-year-old robbery suspect in Brooklyn was the only teenager in the lineup. Three of the five fillers next to him were in their 30s.
Lineups, say experts who study witness memory, can be thought of as an experiment administered by the police. Detectives have a suspect. Now they want to use the memory of a witness to test whether their suspect committed the crime.
“In the witness’ mind is a recollection, an image,” said William Brooks, the police chief of Norwood, Massachusetts, a town of 30,000, and a longtime advocate for improving lineups. “You can’t see it or touch it, so how do you use it? You give it a stimulus, and you see the reaction, and see if there is a match.”
A confident and quick identification — “That’s him!” or “I’ll never forget that face” — upon first viewing the suspect can be a strong indicator of an accurate identification. But only if the lineup has been administered in a neutral manner, without the police steering the witness toward the suspect.
There are many ways for the police to steer a witness toward a particular suspect. Consider, for instance, the photo lineup — an array of six mug shots — which detectives showed a robbery victim in Queens in 2009. The victim had already told the police her assailant was a goateed man wearing a hooded sweatshirt.
Of the six mug shots, only one — the police’s suspect, on the bottom left — had a goatee and wore a hooded sweatshirt.
In this case, the suspect stood out because he alone fit the description. (The suspect was convicted, but an appellate court ultimately overturned it.) But in other cases, the suspect is more likely to get picked not because there is anything exceptionally obvious about him, but because some of the fillers seem implausible.
By using fillers who bear little resemblance to how witnesses described the perpetrator, detectives increase the odds their suspect is selected, research shows. But not because the suspect is more likely to be guilty. Rather, because he may appear to be the best choice.
That is why, experts say, it is problematic for detectives to place teenage suspects alongside grown men in lineups, as happened in February 2018. The 17-year-old Brooklyn robbery suspect was the only teenager in the lineup. Two of the fillers were in their 20s. Three were in their 30s. Some fillers weighed 200 pounds or more; the suspect weighed 145 pounds.
The Times found three cases from the past decade in which a teenage suspect had a filler in his lineup who was at least twice his age. Because lineup paperwork is not ordinarily made public, there is little data for how frequently such age mismatches occur.
Nancy Ginsburg, a Legal Aid Society lawyer who supervises cases involving adolescent defendants, said it is not unusual for fillers to be noticeably older than the suspect.
In 2012, a 14-year-old suspect who weighed 110 pounds was put alongside fillers who ranged in age from 21 to 31 and were at least 50 pounds heavier. In a 2013 lineup, a 16-year-old suspect appeared alongside three fillers who were twice his age or older.
Shea, the chief of detectives, declined to comment on specific cases.
One 2010 lineup photo that was part of an appeal demonstrated how a mismatch in age can reduce a lineup’s fairness. Robbery victims in Brooklyn described one perpetrator as between 16 and 20 years old. The police arrested an 18-year-old. The fillers alongside him included a couple of men who looked older.
The suspect was No. 4. The filler in the No. 5 position was a 35-year-old police officer. No. 1, another filler, was 29.
A victim who remembers being robbed by a teenager is likely to exclude the older individuals as implausible choices, focusing on the few remaining faces.
“They’re looking for the best match, as opposed to the person who matches their memory,” Karen Newirth, a lawyer with the Innocence Project, said.
In this robbery case, the judge excused the age disparities. “We know it’s difficult for the police to get teenagers into a lineup,” said the judge, Joel Goldberg, who has since retired, according to a transcript. That decision was later overturned by an appellate court.
“The age disparity was sufficiently apparent as to orient the viewer toward the defendant,” the appellate panel wrote.
It is true that detectives struggle to find teenagers willing to act as fillers. In the late 1990s, some parents reported that the police had forced their teenage children into serving as lineup fillers, prompting an apology from the police commissioner.
The challenge of finding suitable fillers is one reason that most police departments around the country have abandoned live lineups. When arranging a photo lineup, detectives can easily find suitable fillers by searching through a mug shot database.
New York law until 2017 prevented photo lineups from being admitted as evidence in a trial, all but requiring detectives to conduct live lineups. The law has since changed.
The New York Police Department conducted 1,353 lineups last year, about half as many as it did even two years ago. Shea attributed the drop to the increasing availability of video evidence, which has lowered the reliance on witness identifications.
Still, lineups remain a key tool.
In much of the city, detectives find fillers in homeless shelters or by enlisting other police officers to act as fillers, although detectives in the Bronx have their own system.
Based on a review of court files and interviews with numerous detectives, it is clear that some detectives put considerable thought into selecting fillers. Others can be less discerning.
In September 2014, it fell to a Bronx detective, Franco Johnson, to administer a lineup in an attempted murder case. He was hoping for five fillers who were male, black and between 30 and 35 years of age, just like the suspect, he testified.
But in the end he settled for fillers who were significantly younger: ages 18, 19, 21, 22 and 24. It was not clear all were black. Still they all went in the lineup.
He was asked in court whether one of the fillers — No. 6 — had a similar complexion to the suspect, who was in the No. 3 position.
“It’s close enough,” Johnson said.
This article originally appeared in The New York Times.