Two Norfolk women were raped in 1981, and when police showed them suspect lineups, each picked out Arthur
Lee Whitfield and insisted that he was the attacker.
Whitfield was convicted of one of the rapes -- which occurred within an hour of each other -- despite testimony
that he was at a birthday party when they occurred. He took a plea deal in the second rape and was sent to prison. Then, last
August, DNA proved the women were wrong, and Whitfield was freed after 22 years behind bars.
There is a reason many innocent people such as Whitfield are sent to prison, experts say: police lineups.
Mistaken identification by eyewitnesses was a factor in nearly 90 percent of the nation's first 70 convictions
overturned by DNA testing, according to the New York-based Innocence Project, which works to free the wrongly convicted.
Faced with that alarming data, Virginia is working to become one of the first states to keep more innocent people
out of prison by scrapping the traditional lineup. Mistaken identifications in Virginia have helped convict six of eight defendants
later cleared by DNA.
These days, lineups rarely happen as they do on police shows, with suspects standing side by side. Witnesses
more often are asked to identify suspects from a group of mug shots shown all at once.
The Virginia State Crime Commission, a legislative advisory body, is asking law enforcement agencies instead
to adopt a procedure in which witnesses are shown possible suspects or their mug shots one at a time by an investigator who
does not know who the true suspect is. Such simple changes can cut mistaken identifications by half or more, studies have
shown.
"This would put Virginia on the cutting edge of policy to try to eliminate misidentifications," said Sen. Kenneth
W. Stolle (R-Virginia Beach), the commission's vice chairman. New Jersey, North Carolina and Boston have adopted the procedure,
and a smattering of other jurisdictions are testing it.
Starting in October, law enforcement recruits in Virginia will learn only the new procedure, and current officers
will be trained in it, said Kimberly J. Hamilton, executive director of the commission.
Agencies also will be required to have written policies for conducting lineups under two identical bills, sponsored
by Stolle and Del. Brian J. Moran (D-Alexandria), that passed overwhelmingly in the General Assembly.
For years, researchers have said the typical methods encourage witnesses to pick the wrong person. When seeing
suspects all at once, witnesses tend to compare and settle on the one who looks most like the perpetrator, said Gary L. Wells,
a psychology professor at Iowa State University and a specialist in eyewitness identification. Viewing suspects individually
-- called a "sequential lineup" -- curbs that, he said.
Another vital technique is the "double-blind" method, in which the officer conducting the lineup does not know
who the suspect is, researchers say. That prevents the officer from tipping off the witness, even if unintentionally. Witnesses
also must be told the suspect might not be in the lineup, researchers say.
"You legitimize the possibility that the correct answer is, 'None of these guys,' " Wells said.
Moreover, the "filler" suspects also must fit the description given by the witness but not look so much like
one another to be indistinguishable. When used together, the methods can reduce so-called false alarms by more than 50 percent,
Wells said.
The Virginia model will include those methods, though it will simply "encourage" the double-blind technique
because it might be difficult for small departments with few officers on duty, Hamilton said.
Some of the new methods' loudest backers are defense attorneys.
"Eyewitnesses are tremendously persuasive. Juries believe them," said Steve Benjamin, a Richmond defense lawyer
who is immediate past president of the Virginia Association of Criminal Defense Lawyers. "But the problem is they are tremendously
unreliable."
If agencies do not adopt the model policy, Stolle said, the crime commission would back legislation making it
mandatory. Hamilton said she does not expect that to happen -- a handful of state agencies already use the methods, she said,
and most others seem keen to do so.
"The more we can do to ensure the integrity of the investigation and ultimately the case and the benefit to
victims, witnesses and the alleged perpetrator, then the system is that much better off," said Chief Deputy Sheriff Joseph
McLaughlin of New Kent County, where traditional and sequential lineups are used.
Yet some of the commission's recommendations may be met with some resistance. Dana Schrad, executive director
of the Virginia Association of Chiefs of Police, said most of the association's members want to retain flexibility.
And some researchers say police might be right to keep the traditional lineup.
Showing potential suspects one at a time results in fewer identifications overall, which means innocent people
are picked less often -- but so are criminals, said Roy S. Malpass, a psychology and criminal justice professor at the University
of Texas at El Paso.
Still, DNA exonerations have already swayed some law enforcement officials to buck tradition.
The Virginia Beach Police Department, for example, switched to the new methods three years ago after studying
the research and now gives presentations on the methods to other departments. Deputy Chief Greg Mullen said the department
never had a false alarm problem, but investigators feel more confident in lineup results.
John W. Jones, president of the Virginia Sheriffs Association, said the group has agreed to push sheriffs to
use the model lineups because they are more credible in court.
"You have to be willing to change," Jones said.
