Shortly before accused serial rapist Leon W. Davis went on trial 26 years ago, 60 fervent supporters rallied at an East End church in Richmond in a fight to free him.
As it turns out, the Leon Davis Defense Committee was half right: an innocent man had been swept up and charged in a string of 1984 sexual assaults in the Richmond area. But it was not Leon Davis.
In recent years, DNA has proven that Davis, better known as "the Black Ninja," not only attacked women but also coldly kept silent as Thomas E. Haynesworth, his former neighbor, was wrongly imprisoned for crimes
Davis committed.
Haynsworth, who has been granted parole and who may be released as early as Monday, has been cleared by DNA that implicated Davis in two attacks and has a hearing set March 30 before the Virginia Court of
Appeals in his effort to be exonerated for his remaining convictions.
Capt. Jan Stem of the Henrico County police
department investigated Davis
in 1984 and says Davis, now
serving life in prison, has not changed.
"He could do the right thing by saying, 'Hey, I did that rape that Mr. Haynesworth is serving time for,'" Stem said last week. "But that has not
happened, and the chances of that happening are probably very slim."
Stem, 59, an officer in Henrico for 36 years, said, "I don't think there's a police officer in any position who would ever want to send an innocent man to the
penitentiary."
The Haynesworth/Davis
multiple misidentification case illustrates the dangers faced before DNA testing was widely available in the early 1990s and when witness identification of unknown attackers was often the primary
evidence in rape cases.
However, most crimes then and now do not involve DNA evidence, and an expert warns that identification procedures used by many Virginia police departments are more susceptible to error than they
should be.
"There was a big change in eyewitness procedures after 2005, and it was a big change for the worse," said Brandon Garrett, a law professor at the University of Virginia and an expert on wrongful
convictions.
Garrett, whose book, "Convicting the Innocent: Where Criminal Prosecutions Go Wrong," will be available next month, said, "It seems like we've been going backwards, not forwards."
Misidentifications contributed to 75 percent of the 265 wrongful convictions — 12 in Virginia — documented by the Innocence Project across the country since 1989.
It appears from testimony and media accounts that police used traditional in-person lineups and, in at least some if not all of the cases, photo spreads during the Haynesworth and Davis investigations.
Studies have shown that fewer errors are made if witnesses are shown suspects or photos one at a time, or sequentially, rather than simultaneously in a traditional spread or lineup.
Also, presentations should be conducted by officers who do not know the identity of the suspects and the victim should be told so, Garrett said. Such "double-blind" lineups insure that no
inadvertent signals are given by police or imagined by witnesses, he said.
Virginia enacted a model lineup policy
in 2005 that recommended sequential lineups be used and that they be conducted by independent administrators who do not know the suspect's identity.
However, a Virginia Crime Commission survey of police departments last year suggested that most were using sequential presentations but not independent administrators, much less
double-blind presentations.
"It's the worst of all possible worlds," said Garrett.
That is because it adds more opportunity for error, he said. For example, "you're laying photos down one at a time and there's that witness thinking, 'Did they lay that photo down a little more
slowly, or leave it there longer?'"
The Virginia State Police and Richmond police already use sequential presentations and independent administrators. A bill that would require all police in Virginia to do so failed to win the backing
of the Crime Commission last year.
But Kristen Howard,
director of the crime commission,
said a law passed this year requires the Department of Criminal Justice Services to establish
"best practice" training standards and publish a model policy for law-enforcement personnel.
She said the hope is that the sequential method and blind administration, both among the best practices, would be included in the model policy and training that DCJS develops.
Assuming those best practices are adopted and the training is done well, Garrett said, "it is very good news for Virginia." That is important because memories are malleable and improper techniques
can lead to error, he said.
"You would never handle crime-scene evidence in a careless way because police know if it gets contaminated that it would damage their case," he said. "The same is true of eyewitness evidence, but
there hasn't been that same care taken."
Multiple identifications
The woman raped in a Richmond day-care center on Jan. 3, 1984, said in a recent interview that she was certain Haynesworth was her attacker — only to learn from police in 2009 that DNA
showed that it was Davis. As
she recalls, she picked Haynesworth out of a spread of photographs in 1984.
She said that before Haynesworth was tried, she received anonymous calls, some of them threatening, and one caller said she had identified the wrong man.
That caller told her a time and place where she could see the real assailant. The arranged meeting was staked out by police, but no one appeared.
She assumed the call was from a Haynesworth supporter, and it reinforced her belief Haynesworth was the right man. It is not known if the call was from Davis, who has refused repeated
requests to be interviewed by police and the Richmond Times-Dispatch in recent years.
The 1984 attacks started with that Jan. 3 rape and continued until Dec. 13. They began in the East End of Richmond and also occurred in eastern Henrico.
Haynesworth
became a suspect when a woman who believed he had attacked her saw him walking to a neighborhood store in Richmond on Feb. 5, 1984, and called
police. A total of five women would subsequently identify Haynesworth as their assailant.
Davis was somewhat taller
and heavier than Haynesworth, but the two are close in age and at the time bore a facial resemblance.
DNA has proved that Davis
committed two of the attacks; one charge was never prosecuted. Haynesworth is now seeking exoneration in the Virginia Court of Appeals
for his two remaining convictions, one in Henrico and the other in Richmond.
Prosecutors in Richmond and Henrico, as well as Virginia Attorney General Ken Cuccinelli, are backing his petition for writs of actual innocence.
More attacks after arrest
After Haynesworth's arrest, the 1984 attacks resumed in Henrico and in the Fan and near West End areas of Richmond. The assailant was usually armed with a large knife and sometimes told victims he
was "the Black Ninja."
Davis and Haynesworth had been
East End neighbors, but by August 1984 Davis was living in Henrico near the intersection of Staples Mill and Parham roads.
Some of the 1984 attacks that occurred after Haynesworth's arrest were in the area where Davis had moved and others in Richmond near where his fiancée worked.
Davis became a suspect
after a woman told police in Richmond that she had seen a suspicious man who drove a car with the license plate, LEON-19. The tag was traced to the woman Davis had married in September 1984.
Police obtained a photograph of Davis and showed it to a rape victim, who identified him. He was arrested Dec. 19, 1984, at Eastgate Mall in eastern Henrico.
Stem does not recall what type of mug-shot presentations were made by Henrico during the investigation. It could have been either simultaneous or sequential.
"A lot of things have changed over the years," he said.
Davis' supporters in 1985 had some cause for concern. Like Haynesworth, Davis had no criminal record and maintained he was innocent. Also, he was a black man
identified by white victims — cross-racial identifications being among the more error-prone.
Then two days after Davis' arrest and before some of the victims identified him as their attacker, his mug shot appeared in The Times-Dispatch. As Stem recalls, it was released by accident by the
Henrico sheriff's department.
Davis was identified by at
least three alleged victims in a Dec. 27, 1984, lineup conducted by Richmond police and attended by women and girls believed victims of "the Black
Ninja."
Davis' lawyers and his mother complained that the newspaper photo could influence victim identifications. Police feared it would taint victim identifications in the minds of jurors even if the
victims correctly identified Davis.
The police fears may have been realized in Davis' first trial in which he faced a potential five life sentences but was acquitted. Davis had an alibi in that case, and it is not known for certain why the jury did not
convict.
But the victim, while adamant about her identification, conceded that she picked Davis out of an eight-man lineup six days after seeing his photo in the newspaper. She
had also given an initial description of her attacker that differed in part from Davis.
Davis' lawyer at the time was
Steven D. Benjamin.
Benjamin said the
identification problems in that first trial highlight how important good law enforcement technique is in building a case.
"Bad technique can convict the innocent and can compromise an otherwise good case," he said.
Police investigating the alleged East
Coast rapist, Aaron H.
Thomas, arrested in Connecticut last week and linked by DNA to attacks in Virginia and other states, are not releasing his mug shot.
More victims may come forward and care must be taken, said John Kelly, a Fairfax County detective. "Some of the cases may not have forensic evidence and we would be relying solely on that identification," he said.
fgreen@timesdispatch.com
(804) 649-6340
Soering
Still Claims Innocence
0 Comments
Print
Recommend this page 