Further, according to the Dallas News' investigation:
Almost a decade has elapsed since the U.S. Justice Department recommended stricter limits on the use of "showups," as the practice is known in police circles. More than 40 years ago, the U.S. Supreme Court deemed them dangerously suggestive and discouraged their use.
Yet showups have been cited as a critical flaw in at least 20 percent of the 220 DNA exonerations nationwide. Three of the 19 Dallas County wrongful convictions involved a showup, according to court records.
Most agencies do not have written policies on showups or eyewitness ID procedures generally, the News found, but I was pleased to learn the Dallas PD is creating a new, written policy that puts significant limits on the most suggestive uses of this tactic:
Showups continue in Dallas County and elsewhere because police value them, judges seldom suppress them and juries are swayed by the results.
They are done with few rules and scant oversight. Street cops with little formal guidance on identification procedures typically conduct them, and detectives trying to preserve a key part of their case defend them.
No one keeps statistics on their use.
To better understand the prevalence of showups in Dallas County, The Dallas Morning News reviewed more than 20 years of state appellate court opinions. The News found more than 100 felony trial convictions involving showups. Trials represent just a fraction of how charges are disposed; most result in plea agreements.
How often showups result in misidentifications remains a matter of scholarly debate. But it is a given that one-person showups can pose a higher risk of error than the standard six-person photo array or live lineup.
Each of the Dallas County exoneration cases with a showup involved a rape that occurred between 1982 and 1986, when genetic testing was unavailable. Police based the charges almost entirely on the victim's one-on-one identification.
Of more than two dozen law enforcement agencies in Dallas County, only two (Duncanville and Irving) responded that they had written policies governing showups, The News found.
I think the proposed Dallas policy draws the line at pretty much the right place - show-ups should never be used to identify a stranger. What's more, they can be tainted if officers don't properly caution the witness, particularly to tell them the suspect may not be the person who did it.
Ron Waldrop, an assistant Dallas police chief, said he did not know what, if any, training patrol officers receive on showups. The department's policies on identifications do not mention showups, but new guidelines are being written.
In mid-September, after an innocent man was misidentified in a single-photo showup, the Dallas Police Department banned the practice except when a witness knows the suspect by name or face.
"It was inappropriate," Chief Waldrop said, "and we put an end to that."
To be clear: Not everyone identified based on a show-up is innocent, but there's a greater chance they've been falsely accused than if they were selected from a photo array with appropriate fillers, proper witness cautioning and blind administration. And obviously, it would be greatly preferable to corroborate the witness' accusation with some other evidence besides the ID, particularly when the witness didn't previously know the defendant. The tactic dramatically increases the risk of accusing the wrong guy if the witness didn't already know the suspect before the crime.
MORE: From the Dallas News Crime Blog on Dallas PD's recent decision to upgrade its policy on show-ups.