Wiley Fountain was the obvious choice among the six Polaroids police assembled for the rape victim to view.
He was the only man wearing a dark baseball cap and light-colored warm-up suit, similar to what the attacker had on. He fit the rapist's description "to a T," a Dallas police officer later testified. The victim was sure. Prosecutors believed her. So did the jury. But all of them were wrong.
In September 2002, after Mr. Fountain had spent 15 years in prison, DNA testing proved his innocence. Today, he is free but homeless, scrounging for aluminum cans on the rugged streets of South Dallas.
The story of his wrongful conviction and that of 18 others is lifting the curtain on criminal justice in Dallas County, which has led the nation in DNA exonerations since 2001. In every instance but one, a Dallas Morning News investigation found, police and prosecutors built their case on eyewitness accounts, even though they knew such testimony can be fatally flawed.
Research has long shown that memory is both fallible and malleable. Initial recollections can be wrong, especially when a victim must identify a stranger, and even more so if the stranger is of a different race. If a victim later believes that a suspect, such as someone seen in a lineup, was the attacker, the brain can rewrite history and create a false memory that is as vivid and convincing as if it were real.
In such cases, witnesses can give false testimony with complete confidence. And nothing convicts like a confident eyewitness.
Officially, law enforcement has said the false convictions were tragic aberrations. No one has been charged with lying or disciplined for incompetence or negligence in connection with the DNA exonerations. An eight-month review by The News of previously closed prosecution files found, however, that the faulty identifications were the predictable consequences of a criminal justice system that ignored safeguards meant to protect the innocent. The files reveal a law-and-order machine that focused on securing and bolstering eyewitness testimony, regardless of the victim's doubt or the lack of corroborating evidence.
That may be starting to change. In January 2007, Craig Watkins became the state's first elected black district attorney and quickly focused on wrongful convictions. The issue resonated with the former defense lawyer. He was also the first district attorney in memory with no ties to the prosecutor's office and has shown he is not afraid to reexamine its past.
So earlier this year, Mr. Watkins granted a request from The News to review prosecution files to analyze the root causes of the wrongful convictions. Reporters also consulted more than 70 current and former prosecutors and police officers, defense lawyers, judges, jurors and exonerees, as well as legal scholars and those who pursue wrongful conviction cases.
In addition to an almost slavish reliance on eyewitness testimony, a review of the Dallas County DNA cases showed that:
- 13 of the 19 wrongly convicted men were black. Eight of the 13 were misidentified by victims of another race. Police investigators and prosecutors in the cases were all white, as were many of the juries of the 1980s.
- Police officers used suggestive lineup procedures, sometimes pressured victims to pick their suspect and then cleared the case once an identification was made.
- Prosecutors frequently went to trial with single-witness identifications and flimsy corroboration. Some tried to preserve shaky identifications by withholding evidence that pointed to other potential suspects.
- Judges, governed by case law that has not kept pace with developments in DNA testing or research on eyewitness testimony, routinely approved even tainted pretrial identifications as long as an eyewitness expressed certainty in court.
As a result, victims who sought only justice sent innocent men to prison while the real criminals went free and committed other violent crimes. Taxpayers spent more than $3 million in compensation and incarceration for the Dallas County cases alone. (17 exonerations have occurred elsewhere in Texas.) And some of the discredited police practices continue to this day.
"It's almost like it's the whole system," Terri Moore, Mr. Watkins' top assistant and a former federal prosecutor, said when presented with the newspaper's findings. "Everybody drops the ball somewhere, starting with the police investigation. And we just take the case and adopt what the police say."
Golden testimony
Eyewitness testimony is the crack cocaine of the criminal justice system.
Law officers know the potential risks but are addicted to its power to convict.
"Eyewitness testimony was gold," said Kevin Brooks, who heads the district attorney's felony trial bureau. "If the witness said they saw it, they saw it."
Misidentifications have been cited as a key factor in an estimated 75 5 of the 220 wrongful convictions exposed by DNA testing nationwide since 1989. No local jurisdiction other than Dallas County has had as many surface since 2001, when state law allowed testing for prisoners.
2 things are clear: Dallas County did a better job than most of preserving biological evidence, and all but 5 of the wrongful convictions occurred under the late District Attorney Henry Wade.
Mr. Wade felt crime victims deserved their day in court. If a victim was positive of an identification, that was usually good enough for him and his prosecutors.
"No one ever thought a one-eyewitness case was good," said Joe Kendall, a Wade prosecutor from 1980 to 1982. "But if you had a one-eyewitness case, and it was a rape case, and the victim said that's the one, you couldn't dismiss."
The Wade era, from 1951 to 1986, was marked by take-no-prisoner trial tactics, conviction rates that topped 90 percent and record-length punishments.
Attorneys and investigators who work to free the innocent contend that such an adversary mentality contributes to wrongful convictions.
"We are dealing with a deeply entrenched institutional attitude towards criminal justice that works on an us-vs.-them philosophy," said Amarillo attorney Jeff Blackburn, chief counsel for the Innocence Project of Texas. "It cares about convictions because that gets you a bigger budget and re-elected."
Bill Hill, the only surviving Wade protg who was district attorney before Mr. Watkins, said he was confident his assistants verified the accuracy of all eyewitness identifications.
"I had no reason to believe any of my prosecutors ever did anything that would subject an innocent man to jeopardy," Mr. Hill wrote in an e-mail to The News. "Matter of fact, I would be devastated to find out they had done that."
When told his office prosecuted one of the 19 DNA exonerees, Andrew Gossett, Mr. Hill said the 2 prosecutors on the case were incompetent holdovers from the previous administration.
Dallas County is not unique in its approach to eyewitness testimony. Most prosecutors across the U.S. have maintained an equally deep devotion despite the swelling number of wrongful convictions exposed by DNA testing.
"These are systemic problems, absolutely," said Rob Warden, executive director of the Center on Wrongful Convictions at Northwestern University law school. "And they are closing their eyes to it. They are not doing so in bad faith. They are doing so because they really believe it."
Eyewitness testimony is not always flawed, but no one really knows how often it's wrong. But because the same identification procedures were routinely used in cases that lacked biological evidence to test, Mr. Watkins said, the number is "a lot more than 19."
It is likely, he said, that an innocent person has been convicted by faulty eyewitness testimony during his first two years in office. He is trying to instill in his prosecutors the need to be more skeptical and to seek out corroboration, Mr. Watkins said.
"We know that eyewitness identification is faulty," he added.
Troubled lineups
Mr. Fountain's case epitomized many of the shortcomings The News found.
His was one of 14 DNA-based exonerations that relied on a photographic lineup, by far the most widely used identification method by police.
Photo lineups gained popularity after the U.S. Supreme Court ruled in 1967 that suspects in live lineups had the right to have an attorney present. Photo arrays are usually shown outside the presence of defense attorneys and with no audio or video recordings.
Typically, 6 mug-shot photos are arrayed in 2 rows of 3 for a victim to view. Dallas and many other police departments use computer programs to generate "filler" photos for lineups based on general descriptions entered by a detective.
As in Mr. Fountain's case, questions persist about the fairness police officers demonstrated in obtaining identifications from photo lineups.
Mr. Fountain's photo was shown to the victim 4 hours after the rape. The case was marked "cleared" by the lead detective the following day. The only other evidence presented at trial was a blood-typing test that concluded no living person could be eliminated as a suspect.
A review of the photo lineups in the other DNA exonerations showed that 1 wrongly convicted man from Dallas County was a former neighbor of the victim. Another worked at the same business. A 3rd had the same name as a man the victim knew was a suspect. 4 men were suspects in other unsolved crimes.
All but one of the exonerated men had prior arrest records and booking photos that could be used in lineups.
The lead detectives were nearly all seasoned veterans. 2 had 2 cases each with faulty photo identifications. Another had shown photo spreads in the Lenell Geter case, one of Dallas County's most infamous wrongful convictions.
Police insist that faulty identifications from photo lineups are almost always the result of human error, not intentional bad faith.
"Everyone gets sloppy," said P.E. Jones, a retired Dallas robbery and homicide detective. "You think you have a slam-dunk case, and so you don't go in there and dot your I's and cross your T's. And all of a sudden, it comes back to bite you."
Critics contend photo arrays are inherently vulnerable to manipulation: the detective's body language, the instruction the witness receives, feedback provided by officers if an identification is made.
"I call them misconduct identifications," the Innocence Project's Mr. Blackburn said. "I have yet to see one of these eyewitness IDs that goes wrong that didn't have some element of deliberately suggestive behavior going on by the cops."
After one witness hesitated to identify exoneree Patrick Waller from a photo lineup, prosecution notes show, the lead detective pointed to Mr. Waller's picture and said two other eyewitnesses had identified him. The witness remained uncertain and was not called to testify at trial. Mr. Waller spent 16 years in prison before DNA exonerated him in July. The real criminal later was imprisoned for burglary.
Jim McCloskey, a nationally known prisoner rights' advocate, said his investigations of wrongful convictions nationwide often involve photo lineups that appear to be the result of police officers coaxing witnesses.
"Who looks most similar to the suspect?" said Mr. McCloskey, executive director of Centurion Ministries Inc. of Princeton, N.J. "These conversations go on, and they evolve until before you know it they have got themselves an ID. Now they've got the guy [witness] locked in, and he's afraid to go back."
Little training
The use of arrest photos with height lines in the background or departmental identification placards in front can suggest involvement in crime.
The angle at which a photograph is taken can distort a subject's true size. A photo's age can obscure physical changes. The exposure can make skin tones darker.
Exoneree Donald Wayne Good, in a federal civil rights lawsuit, accuses an Irving police detective of deliberately underexposing his lineup photo to obscure a facial scar and tattoo. Mr. Good spent 21 years in prison before being freed in 2004.
Attorneys for the detective and the city have denied that Mr. Good's placement in the lineup was retaliation for his lack of cooperation in another case.
Five other exonerees have sued the cities of Dallas, Irving and Garland, claiming that a lack of formal identification policies and inadequate police training and supervision demonstrated "deliberate indifference" to their civil rights.
Most police agencies don't have written policies on identification techniques, and police officers receive little formal training. Only five police departments in Dallas County, including the Dallas Police Department, provided The News with written policies regarding eyewitness investigations.
Dallas Police Chief David Kunkle said investigators sometimes were too focused on their suspect and ignored signs that someone else could be guilty. That, in combination with bad witness identifications and prosecutorial misconduct, contributed to the wrongful convictions, he said.
Law enforcement agencies have generally resisted calls to overhaul their identification procedures and to use a sequential blind method. In it, photos are shown one at a time, instead of simultaneously, by an officer not involved in the case.
Social scientists have touted the sequential procedure for 20 years, and a U.S. Justice Department task force recommended it in 1999.
A bill that died in the Texas Legislature last year would have created a task force to draft voluntary guidelines for sequential blind procedures.
Dallas police have said for nearly 2 years they would participate in a pilot project to use the sequential blind approach in some felony cases, but the project has yet to begin.
Brad Lollar, a Wade prosecutor from 1978 to 1982, said he did not grill police investigators about how they obtained a victim identification.
"You'd ask them just to make sure there wasn't any deviation from the script," he said. "The script was always, 'No, I didn't tell her. No, I didn't indicate. No, I didn't point out any particular photograph.' "
By the time a case got to trial, it was not unusual for the eyewitness to have seen the defendant multiple times either through photos or in person. Social scientists describe this as "confirmation bias," saying it shores up a shaky identification.
"There is nothing as dramatic as being in front of a jury and that witness says 'That's the man. I will never forget his face,' " said Mr. Brooks, a veteran Dallas prosecutor. "The impact that has on a jury is unreal."
Assembly-line justice
If an eyewitness exhibits certainty, records and interviews show, judges do not suppress a prior identification no matter how the photo lineup was conducted.
"The fact that some particular process might not have been as pristine as we'd like it to be does not in and of itself make them wrong," said state District Judge John Creuzot.
U.S. Supreme Court decisions in the 1960s and 1970s set a high bar for withholding an eyewitness identification from a jury. And in 1975, the Texas Legislature amended state law to make it easier to prosecute sexual assaults without evidence to corroborate eyewitness testimony.
The law was part of a national effort by women's groups to remove provisions deemed unfair barriers to the prosecution of sexual assaults. Each of the 19 Dallas County DNA exonerations occurred after that law change.
In Mr. Fountain's 1986 trial, state District Judge Jack Hampton sanctioned the victim's identification, even though defense attorney Mike Rodgers called it "the most suggestive photo spread I've ever seen in my life."
It didn't matter that Mr. Fountain wasn't wearing blue jeans under the warm-up suit as the victim had described. Or that there was no incriminating medical evidence. Or that Mr. Fountain had an alibi affirmed by his teenage cousin.
Once he was identified, police declared him guilty and prosecutors built a case around the victim's testimony. That, says Ms. Moore, first assistant to Mr. Watkins, is "the assembly line" of criminal justice.
The lead prosecutor in the case said she didn't doubt the victim and was not bothered by the way the photo lineup was constructed.
"I would say, knowing what I know now, that's not good procedure," said Lana Myers, now a felony court judge. "But then, she was believable to me. Obviously, she believed he was the one."
One juror, Michael Reeb, said any doubts about the prosecution's case were overcome by the testimony of the victim, a young clerk pregnant with her 3rd child.
"It came to the point where she's identified him. I don't know what he's done with the pants, but let's go with the identity," Mr. Reeb said.
The jury convicted Mr. Fountain and sentenced him to 40 years in prison.
Once behind bars, Mr. Fountain said he devoted himself to regaining his freedom. He wrote letters to the trial judge, the district attorney, anyone he thought might listen.
No one did until the results of a March 2002 DNA test finally proved Mr. Fountain's innocence.
A life unraveled
Life outside prison has been bumpy.
Mr. Fountain said he collected $190,000 in compensation from the state, gave most of it to his family, and a girlfriend drained the rest from his bank account.
"I was a fool for letting her know the number," he said.
Mr. Fountain took to the streets after his mother died in 2005. He sleeps in an abandoned house and spends his days collecting cans to sell to scrap yards along South Lamar Street. On a good day, he said, he can make $40, which he spends on fast food and beer.
Sometimes he stays at his sister's home in Seagoville, but never long. He said he refuses to live ever again under someone else's rules.
Looking back, the gaunt 52-year-old in ripped pants and a grimy T-shirt said the justice system never gave him a chance.
"Back then, all they needed was testimony of the victim," he said. "If the jury believed that, they didn't need nothing else."
AT A GLANCE: OVERTURNED CONVICTIONS
Of the 19 Dallas County convictions overturned by DNA testing, 18 were based on faulty witness identifications. One murder case had no eyewitness.
A review of the case files by The Dallas Morning News found that:
8 of the wrongly identified men lived, worked or socialized near the witnesses.
4 were suspects in unrelated crimes.
All the cases involved sexual assault though not all men were charged with rape.
14 cases were prosecuted during former District Attorney Henry Wade's tenure, which ended in 1986 and was marked by hardball trial tactics and high conviction rates.
5 faulty convictions were recorded during the terms of 2 Wade lieutenants: John Vance, who had 4, and Bill Hill, who had 1.
Mr. Hill was district attorney in April 2001 when Texas prisoners won the right to file for post-conviction DNA testing. His prosecutors opposed genetic testing in 10 cases where that evidence ultimately invalidated the conviction.
Testing delays in 1 armed robbery-rape case made it legally impossible to prosecute the true culprit.
Mr. Hill said his prosecutors opposed DNA testing only when they were convinced of the defendant's guilt.
"I would never, ever knowingly refuse an innocent person's freedom," he said. "I was a tough prosecutor, but I only wanted the guilty to be charged and tried."
*****************
Eyewitnesses still play key roles in cases where DNA, other evidence is lacking
The fallibility of eyewitness testimony revealed by DNA exonerations in Dallas County and nationwide is not a relic of the past. Police and prosecutors still depend on the same discredited identification procedures to ensure convictions today.
Police use these techniques in a variety of crimes from murders to robberies. The difference between today's cases and the 19 exonerations involving sexual assaults is that often there is no DNA to ensure guilt or innocence.
"We've shown how unreliable eyewitness testimony is in sexual assault cases," said Rob Warden, executive director of the Center on Wrongful Convictions at Northwestern University law school. "But now the system itself is pretending that all of these armed robbery cases are just hunky dory when we know, if anything, it's no doubt less reliable in an armed robbery case than in a sexual assault case."
It's impossible to estimate how many wrongful convictions might be occurring in cases without genetic evidence, experts say.
"There is no question that there are many more mistakes that we will never know about because there is no DNA in those cases," said Edwin Colfax, an Austin researcher with the nonprofit reform group The Justice Project.
To examine whether the flawed practices highlighted by the DNA exoneration cases were still in use, The Dallas Morning News examined robbery trials in Dallas County from 2006 and 2007 the last year of former District Attorney Bill Hill's tenure and the 1st year Craig Watkins was in office. Eyewitness testimony is the most crucial element in robbery cases.
The newspaper found that law enforcement still relies heavily on eyewitness testimony, even if corroborating evidence is weak and despite decades of research showing its shortcomings.
' Eyewitness testimony can be unreliable for a variety of reasons. Victims often get only a brief look at their attackers, especially in robberies, and they tend to focus on the gun rather than the face, research shows. Accurately identifying strangers, especially of a different race, is difficult, and the stress of being a crime victim can distort memories.
Nonetheless, police and prosecutors say eyewitnesses are 2nd only to genetic evidence in convincing jurors of a defendant's guilt.
"Nothing beats a victim or a witness standing there when the crime was committed," Dallas homicide Sgt. Larry Lewis said. "The only thing that would beat that is DNA."
Troubles persist
The News reviewed 31 robbery trials over two years in which defendants were found guilty and appealed their sentences. Only cases on appeal have trial transcripts, which allow for an examination of what happened at trial and, in part, during the investigation.
The cases reviewed represented about two-thirds of all appealed robberies. However, more than 1,700 defendants during the same two years pleaded guilty and were given prison time, probation or deferred adjudication probation. As seen in some of the 19 exoneration cases, defendants sometimes accept a plea bargain to ensure a shorter sentence.
The newspaper did not reinvestigate the cases to determine whether an innocent defendant had been found guilty of the crime. But it found several of the same practices discredited in the DNA exonerations, including cases in which:
- Witnesses often selected people from photo lineups when they could not provide detailed descriptions of the robbers' faces. In a 2007 Dallas robbery case, a pizza deliveryman did not provide details of the robbers' faces or mention noticing a permanent grill on the defendant's teeth, but he identified a suspect with a grill during a photo lineup.
- Photo lineups were not always conducted according to best practices. In two Dallas cases, officers asked two witnesses to identify a suspect at the same time. Best practice dictates that witnesses view suspects separately.
- Victims sometimes picked suspects from lineups by eliminating other photos and then selecting the remaining one who most resembled their attacker. In a 2007 Coppell robbery case, the victim told jurors that she selected the defendant because he had the darkest complexion and the largest forehead of anyone in the photo lineup.
- Police focused on suspects in spite of conflicting evidence. A man was convicted in 2006 after a police dog tracked a scent from a store several doors down from a Duncanville crime scene. The dog led officers to an apartment complex where a man known to police lived. Investigators found a gun in the man's bedroom that didn't match the robbery weapon. But he was arrested after the victim identified him in a photo lineup.
Although defense attorneys questioned the unreliability of eyewitness identification, they did not call experts to testify about studies proving that point.
Victims in 6 cases could not identify their attackers, sometimes because the robbers wore masks. In those cases, DNA or robbers using victims' credit cards led to arrests.
DNA was found in only 3 cases: One involved a sexual assault; in the other 2, defendants left behind gloves with DNA.
The identification practice commonly known as "showups" occurred in about 20 % of the robbery cases examined. That tracks with the percentage of showups involved in DNA exonerations nationwide.
Showups allow police to show witnesses a suspect 1-on-1, either in person or by photo. The U.S. Supreme Court has discouraged but not banned the practice.
In one 2006 trial involving a showup, a woman was robbed at gunpoint while parking cars at the State Fair of Texas. She said the man who robbed her was wearing a white tank top and jeans and took from her $15 or $16 in cash, a cellphone and some wadded up toilet paper she had in her pocket.
Police put the woman in a squad car and drove her around the area until police spotted a man on the street wearing the same kind of clothes as the robber. After she identified him, police found $51 in his pockets but no gun or the woman's cellphone. He did have some crumpled toilet paper in his back pocket and was arrested. The man, who had a prior record, was later convicted and sentenced to 70 years in prison. No one tracks how often police conduct showups, but officers believe strongly in their value. One officer in a Dallas County robbery case tried last year testified that he preferred using showups to photo lineups.
Addressing problems
Mr. Watkins has made righting wrongful convictions a hallmark of his nearly 2-year administration. Presented with The News' findings, he acknowledged that his office needs to do much more to improve how it handles eyewitness testimony.
"This is just another indication of how far we need to go," Mr. Watkins said in an interview. "I was thinking, 'Yeah we'll get there,' but we have a long way to go."
However, he said, he cannot stop trying cases even if they are based on little else but eyewitness testimony.
"You can't just throw the whole process out of the window," he said. "The thing is to try to improve the process."
Mr. Watkins and his top assistant, Terri Moore, said they want to better train prosecutors to ask questions of witnesses and police officers about how identifications were obtained.
"Teaching them how to rethink some things, challenging ourselves," said Ms. Moore, a former federal prosecutor.
"I don't think you are going to stop wrongful convictions today because of a DNA test," Ms. Moore said, referring to how few cases have genetic evidence that can be used to verify the verdict. "I think it goes well beyond that."
Mr. Watkins said he would not support a law requiring all police departments to follow a specific policy on the use of eyewitness testimony. But he said he would lobby the Legislature to mandate that all police departments adopt standards for how they handle witness identifications. Only five Dallas County police departments could provide The News written policies on how to conduct witness identifications.
The Dallas Police Department in recent years began requiring witnesses to sign a form, acknowledging that the officer conducting the lineup warned them that the photo spread may not include the actual suspect. The department also has written guidelines for how to conduct photo lineups, but police officials say there is little oversight.
Mr. Watkins said he also would not support a special jury instruction warning that eyewitness identification can be unreliable. That point, he said, is more appropriately handled by presenting expert witnesses on the subject.
Texas law allows for convictions based on the testimony of a single eyewitness. Potential jurors who cannot convict based on the testimony of that lone witness if they believe the testimony beyond a reasonable doubt cannot legally serve on a jury panel.
Longtime Dallas defense attorney John Read said that since Mr. Watkins took office, county prosecutors have been more likely to question eyewitnesses' reliability. But police departments in Dallas County and prosecutors in other counties still base many of their cases on an eyewitness picking out a suspect, he said.
Mr. Read said he also believes that police, intentionally or otherwise, are still suggesting in lineups which suspect the victim should identify.
"I believe they do what they do whatever it is they do with their suggestions because they think they've got the right person," Mr. Read said. "They believe he's guilty."
'The revelation'
Earlier this year, Richardson police changed how they handled photo lineups after wrongly charging a man in a 1985 burglary and rape. The man, Thomas McGowan, was cleared in April through DNA testing after serving 22 years in prison.
Richardson police Chief Larry Zacharias called the city's only DNA exoneration "the revelation."
The photo lineup that led to Mr. McGowan's arrest and conviction included color photos, black-and-white photos and photocopies of photos a violation of the best practice that mandates all photos should be similar in composition. Mr. McGowan was pictured in a color photo, wearing a placard that read Richardson police.
Now, instead of the case investigator showing six photos at one time to a witness, a detective not involved in the case shows witnesses one photo at a time. The practice is known as a sequential blind lineup.
Investigators also video record all witness identifications conducted at the station, which previously had not been done. Field identifications are at least audio recorded.
Those seeking eyewitness identification reforms say videotaping reduces the likelihood of suggestion by the detective and allows prosecutors to show juries that the photo lineup was conducted properly.
Recording of some kind, Chief Zacharias said, is an "absolute" in identifications.
By contrast, Dallas police, who had 13 of the 19 exonerations, have been considering for nearly 2 years whether to participate in a limited study using sequential blind identification procedures. The study which would focus primarily on robbery cases could begin as early as November, departmental officials said.
When the new procedures begin, Dallas police Assistant Chief Ron Waldrop said, officers will probably make an audio recording of the identifications in photo lineups. Unlike Richardson, however, there are no plans for Dallas police to videotape its lineups.
"It's not going to eliminate mistakes," Chief Waldrop said of the proposed changes. "We just want to make sure the practices we put in place don't aid the mistakes."
Source for both: Dallas Morning News
He was the only man wearing a dark baseball cap and light-colored warm-up suit, similar to what the attacker had on. He fit the rapist's description "to a T," a Dallas police officer later testified. The victim was sure. Prosecutors believed her. So did the jury. But all of them were wrong.
In September 2002, after Mr. Fountain had spent 15 years in prison, DNA testing proved his innocence. Today, he is free but homeless, scrounging for aluminum cans on the rugged streets of South Dallas.
The story of his wrongful conviction and that of 18 others is lifting the curtain on criminal justice in Dallas County, which has led the nation in DNA exonerations since 2001. In every instance but one, a Dallas Morning News investigation found, police and prosecutors built their case on eyewitness accounts, even though they knew such testimony can be fatally flawed.
Research has long shown that memory is both fallible and malleable. Initial recollections can be wrong, especially when a victim must identify a stranger, and even more so if the stranger is of a different race. If a victim later believes that a suspect, such as someone seen in a lineup, was the attacker, the brain can rewrite history and create a false memory that is as vivid and convincing as if it were real.
In such cases, witnesses can give false testimony with complete confidence. And nothing convicts like a confident eyewitness.
Officially, law enforcement has said the false convictions were tragic aberrations. No one has been charged with lying or disciplined for incompetence or negligence in connection with the DNA exonerations. An eight-month review by The News of previously closed prosecution files found, however, that the faulty identifications were the predictable consequences of a criminal justice system that ignored safeguards meant to protect the innocent. The files reveal a law-and-order machine that focused on securing and bolstering eyewitness testimony, regardless of the victim's doubt or the lack of corroborating evidence.
That may be starting to change. In January 2007, Craig Watkins became the state's first elected black district attorney and quickly focused on wrongful convictions. The issue resonated with the former defense lawyer. He was also the first district attorney in memory with no ties to the prosecutor's office and has shown he is not afraid to reexamine its past.
So earlier this year, Mr. Watkins granted a request from The News to review prosecution files to analyze the root causes of the wrongful convictions. Reporters also consulted more than 70 current and former prosecutors and police officers, defense lawyers, judges, jurors and exonerees, as well as legal scholars and those who pursue wrongful conviction cases.
In addition to an almost slavish reliance on eyewitness testimony, a review of the Dallas County DNA cases showed that:
- 13 of the 19 wrongly convicted men were black. Eight of the 13 were misidentified by victims of another race. Police investigators and prosecutors in the cases were all white, as were many of the juries of the 1980s.
- Police officers used suggestive lineup procedures, sometimes pressured victims to pick their suspect and then cleared the case once an identification was made.
- Prosecutors frequently went to trial with single-witness identifications and flimsy corroboration. Some tried to preserve shaky identifications by withholding evidence that pointed to other potential suspects.
- Judges, governed by case law that has not kept pace with developments in DNA testing or research on eyewitness testimony, routinely approved even tainted pretrial identifications as long as an eyewitness expressed certainty in court.
As a result, victims who sought only justice sent innocent men to prison while the real criminals went free and committed other violent crimes. Taxpayers spent more than $3 million in compensation and incarceration for the Dallas County cases alone. (17 exonerations have occurred elsewhere in Texas.) And some of the discredited police practices continue to this day.
"It's almost like it's the whole system," Terri Moore, Mr. Watkins' top assistant and a former federal prosecutor, said when presented with the newspaper's findings. "Everybody drops the ball somewhere, starting with the police investigation. And we just take the case and adopt what the police say."
Golden testimony
Eyewitness testimony is the crack cocaine of the criminal justice system.
Law officers know the potential risks but are addicted to its power to convict.
"Eyewitness testimony was gold," said Kevin Brooks, who heads the district attorney's felony trial bureau. "If the witness said they saw it, they saw it."
Misidentifications have been cited as a key factor in an estimated 75 5 of the 220 wrongful convictions exposed by DNA testing nationwide since 1989. No local jurisdiction other than Dallas County has had as many surface since 2001, when state law allowed testing for prisoners.
2 things are clear: Dallas County did a better job than most of preserving biological evidence, and all but 5 of the wrongful convictions occurred under the late District Attorney Henry Wade.
Mr. Wade felt crime victims deserved their day in court. If a victim was positive of an identification, that was usually good enough for him and his prosecutors.
"No one ever thought a one-eyewitness case was good," said Joe Kendall, a Wade prosecutor from 1980 to 1982. "But if you had a one-eyewitness case, and it was a rape case, and the victim said that's the one, you couldn't dismiss."
The Wade era, from 1951 to 1986, was marked by take-no-prisoner trial tactics, conviction rates that topped 90 percent and record-length punishments.
Attorneys and investigators who work to free the innocent contend that such an adversary mentality contributes to wrongful convictions.
"We are dealing with a deeply entrenched institutional attitude towards criminal justice that works on an us-vs.-them philosophy," said Amarillo attorney Jeff Blackburn, chief counsel for the Innocence Project of Texas. "It cares about convictions because that gets you a bigger budget and re-elected."
Bill Hill, the only surviving Wade protg who was district attorney before Mr. Watkins, said he was confident his assistants verified the accuracy of all eyewitness identifications.
"I had no reason to believe any of my prosecutors ever did anything that would subject an innocent man to jeopardy," Mr. Hill wrote in an e-mail to The News. "Matter of fact, I would be devastated to find out they had done that."
When told his office prosecuted one of the 19 DNA exonerees, Andrew Gossett, Mr. Hill said the 2 prosecutors on the case were incompetent holdovers from the previous administration.
Dallas County is not unique in its approach to eyewitness testimony. Most prosecutors across the U.S. have maintained an equally deep devotion despite the swelling number of wrongful convictions exposed by DNA testing.
"These are systemic problems, absolutely," said Rob Warden, executive director of the Center on Wrongful Convictions at Northwestern University law school. "And they are closing their eyes to it. They are not doing so in bad faith. They are doing so because they really believe it."
Eyewitness testimony is not always flawed, but no one really knows how often it's wrong. But because the same identification procedures were routinely used in cases that lacked biological evidence to test, Mr. Watkins said, the number is "a lot more than 19."
It is likely, he said, that an innocent person has been convicted by faulty eyewitness testimony during his first two years in office. He is trying to instill in his prosecutors the need to be more skeptical and to seek out corroboration, Mr. Watkins said.
"We know that eyewitness identification is faulty," he added.
Troubled lineups
Mr. Fountain's case epitomized many of the shortcomings The News found.
His was one of 14 DNA-based exonerations that relied on a photographic lineup, by far the most widely used identification method by police.
Photo lineups gained popularity after the U.S. Supreme Court ruled in 1967 that suspects in live lineups had the right to have an attorney present. Photo arrays are usually shown outside the presence of defense attorneys and with no audio or video recordings.
Typically, 6 mug-shot photos are arrayed in 2 rows of 3 for a victim to view. Dallas and many other police departments use computer programs to generate "filler" photos for lineups based on general descriptions entered by a detective.
As in Mr. Fountain's case, questions persist about the fairness police officers demonstrated in obtaining identifications from photo lineups.
Mr. Fountain's photo was shown to the victim 4 hours after the rape. The case was marked "cleared" by the lead detective the following day. The only other evidence presented at trial was a blood-typing test that concluded no living person could be eliminated as a suspect.
A review of the photo lineups in the other DNA exonerations showed that 1 wrongly convicted man from Dallas County was a former neighbor of the victim. Another worked at the same business. A 3rd had the same name as a man the victim knew was a suspect. 4 men were suspects in other unsolved crimes.
All but one of the exonerated men had prior arrest records and booking photos that could be used in lineups.
The lead detectives were nearly all seasoned veterans. 2 had 2 cases each with faulty photo identifications. Another had shown photo spreads in the Lenell Geter case, one of Dallas County's most infamous wrongful convictions.
Police insist that faulty identifications from photo lineups are almost always the result of human error, not intentional bad faith.
"Everyone gets sloppy," said P.E. Jones, a retired Dallas robbery and homicide detective. "You think you have a slam-dunk case, and so you don't go in there and dot your I's and cross your T's. And all of a sudden, it comes back to bite you."
Critics contend photo arrays are inherently vulnerable to manipulation: the detective's body language, the instruction the witness receives, feedback provided by officers if an identification is made.
"I call them misconduct identifications," the Innocence Project's Mr. Blackburn said. "I have yet to see one of these eyewitness IDs that goes wrong that didn't have some element of deliberately suggestive behavior going on by the cops."
After one witness hesitated to identify exoneree Patrick Waller from a photo lineup, prosecution notes show, the lead detective pointed to Mr. Waller's picture and said two other eyewitnesses had identified him. The witness remained uncertain and was not called to testify at trial. Mr. Waller spent 16 years in prison before DNA exonerated him in July. The real criminal later was imprisoned for burglary.
Jim McCloskey, a nationally known prisoner rights' advocate, said his investigations of wrongful convictions nationwide often involve photo lineups that appear to be the result of police officers coaxing witnesses.
"Who looks most similar to the suspect?" said Mr. McCloskey, executive director of Centurion Ministries Inc. of Princeton, N.J. "These conversations go on, and they evolve until before you know it they have got themselves an ID. Now they've got the guy [witness] locked in, and he's afraid to go back."
Little training
The use of arrest photos with height lines in the background or departmental identification placards in front can suggest involvement in crime.
The angle at which a photograph is taken can distort a subject's true size. A photo's age can obscure physical changes. The exposure can make skin tones darker.
Exoneree Donald Wayne Good, in a federal civil rights lawsuit, accuses an Irving police detective of deliberately underexposing his lineup photo to obscure a facial scar and tattoo. Mr. Good spent 21 years in prison before being freed in 2004.
Attorneys for the detective and the city have denied that Mr. Good's placement in the lineup was retaliation for his lack of cooperation in another case.
Five other exonerees have sued the cities of Dallas, Irving and Garland, claiming that a lack of formal identification policies and inadequate police training and supervision demonstrated "deliberate indifference" to their civil rights.
Most police agencies don't have written policies on identification techniques, and police officers receive little formal training. Only five police departments in Dallas County, including the Dallas Police Department, provided The News with written policies regarding eyewitness investigations.
Dallas Police Chief David Kunkle said investigators sometimes were too focused on their suspect and ignored signs that someone else could be guilty. That, in combination with bad witness identifications and prosecutorial misconduct, contributed to the wrongful convictions, he said.
Law enforcement agencies have generally resisted calls to overhaul their identification procedures and to use a sequential blind method. In it, photos are shown one at a time, instead of simultaneously, by an officer not involved in the case.
Social scientists have touted the sequential procedure for 20 years, and a U.S. Justice Department task force recommended it in 1999.
A bill that died in the Texas Legislature last year would have created a task force to draft voluntary guidelines for sequential blind procedures.
Dallas police have said for nearly 2 years they would participate in a pilot project to use the sequential blind approach in some felony cases, but the project has yet to begin.
Brad Lollar, a Wade prosecutor from 1978 to 1982, said he did not grill police investigators about how they obtained a victim identification.
"You'd ask them just to make sure there wasn't any deviation from the script," he said. "The script was always, 'No, I didn't tell her. No, I didn't indicate. No, I didn't point out any particular photograph.' "
By the time a case got to trial, it was not unusual for the eyewitness to have seen the defendant multiple times either through photos or in person. Social scientists describe this as "confirmation bias," saying it shores up a shaky identification.
"There is nothing as dramatic as being in front of a jury and that witness says 'That's the man. I will never forget his face,' " said Mr. Brooks, a veteran Dallas prosecutor. "The impact that has on a jury is unreal."
Assembly-line justice
If an eyewitness exhibits certainty, records and interviews show, judges do not suppress a prior identification no matter how the photo lineup was conducted.
"The fact that some particular process might not have been as pristine as we'd like it to be does not in and of itself make them wrong," said state District Judge John Creuzot.
U.S. Supreme Court decisions in the 1960s and 1970s set a high bar for withholding an eyewitness identification from a jury. And in 1975, the Texas Legislature amended state law to make it easier to prosecute sexual assaults without evidence to corroborate eyewitness testimony.
The law was part of a national effort by women's groups to remove provisions deemed unfair barriers to the prosecution of sexual assaults. Each of the 19 Dallas County DNA exonerations occurred after that law change.
In Mr. Fountain's 1986 trial, state District Judge Jack Hampton sanctioned the victim's identification, even though defense attorney Mike Rodgers called it "the most suggestive photo spread I've ever seen in my life."
It didn't matter that Mr. Fountain wasn't wearing blue jeans under the warm-up suit as the victim had described. Or that there was no incriminating medical evidence. Or that Mr. Fountain had an alibi affirmed by his teenage cousin.
Once he was identified, police declared him guilty and prosecutors built a case around the victim's testimony. That, says Ms. Moore, first assistant to Mr. Watkins, is "the assembly line" of criminal justice.
The lead prosecutor in the case said she didn't doubt the victim and was not bothered by the way the photo lineup was constructed.
"I would say, knowing what I know now, that's not good procedure," said Lana Myers, now a felony court judge. "But then, she was believable to me. Obviously, she believed he was the one."
One juror, Michael Reeb, said any doubts about the prosecution's case were overcome by the testimony of the victim, a young clerk pregnant with her 3rd child.
"It came to the point where she's identified him. I don't know what he's done with the pants, but let's go with the identity," Mr. Reeb said.
The jury convicted Mr. Fountain and sentenced him to 40 years in prison.
Once behind bars, Mr. Fountain said he devoted himself to regaining his freedom. He wrote letters to the trial judge, the district attorney, anyone he thought might listen.
No one did until the results of a March 2002 DNA test finally proved Mr. Fountain's innocence.
A life unraveled
Life outside prison has been bumpy.
Mr. Fountain said he collected $190,000 in compensation from the state, gave most of it to his family, and a girlfriend drained the rest from his bank account.
"I was a fool for letting her know the number," he said.
Mr. Fountain took to the streets after his mother died in 2005. He sleeps in an abandoned house and spends his days collecting cans to sell to scrap yards along South Lamar Street. On a good day, he said, he can make $40, which he spends on fast food and beer.
Sometimes he stays at his sister's home in Seagoville, but never long. He said he refuses to live ever again under someone else's rules.
Looking back, the gaunt 52-year-old in ripped pants and a grimy T-shirt said the justice system never gave him a chance.
"Back then, all they needed was testimony of the victim," he said. "If the jury believed that, they didn't need nothing else."
AT A GLANCE: OVERTURNED CONVICTIONS
Of the 19 Dallas County convictions overturned by DNA testing, 18 were based on faulty witness identifications. One murder case had no eyewitness.
A review of the case files by The Dallas Morning News found that:
8 of the wrongly identified men lived, worked or socialized near the witnesses.
4 were suspects in unrelated crimes.
All the cases involved sexual assault though not all men were charged with rape.
14 cases were prosecuted during former District Attorney Henry Wade's tenure, which ended in 1986 and was marked by hardball trial tactics and high conviction rates.
5 faulty convictions were recorded during the terms of 2 Wade lieutenants: John Vance, who had 4, and Bill Hill, who had 1.
Mr. Hill was district attorney in April 2001 when Texas prisoners won the right to file for post-conviction DNA testing. His prosecutors opposed genetic testing in 10 cases where that evidence ultimately invalidated the conviction.
Testing delays in 1 armed robbery-rape case made it legally impossible to prosecute the true culprit.
Mr. Hill said his prosecutors opposed DNA testing only when they were convinced of the defendant's guilt.
"I would never, ever knowingly refuse an innocent person's freedom," he said. "I was a tough prosecutor, but I only wanted the guilty to be charged and tried."
*****************
Eyewitnesses still play key roles in cases where DNA, other evidence is lacking
The fallibility of eyewitness testimony revealed by DNA exonerations in Dallas County and nationwide is not a relic of the past. Police and prosecutors still depend on the same discredited identification procedures to ensure convictions today.
Police use these techniques in a variety of crimes from murders to robberies. The difference between today's cases and the 19 exonerations involving sexual assaults is that often there is no DNA to ensure guilt or innocence.
"We've shown how unreliable eyewitness testimony is in sexual assault cases," said Rob Warden, executive director of the Center on Wrongful Convictions at Northwestern University law school. "But now the system itself is pretending that all of these armed robbery cases are just hunky dory when we know, if anything, it's no doubt less reliable in an armed robbery case than in a sexual assault case."
It's impossible to estimate how many wrongful convictions might be occurring in cases without genetic evidence, experts say.
"There is no question that there are many more mistakes that we will never know about because there is no DNA in those cases," said Edwin Colfax, an Austin researcher with the nonprofit reform group The Justice Project.
To examine whether the flawed practices highlighted by the DNA exoneration cases were still in use, The Dallas Morning News examined robbery trials in Dallas County from 2006 and 2007 the last year of former District Attorney Bill Hill's tenure and the 1st year Craig Watkins was in office. Eyewitness testimony is the most crucial element in robbery cases.
The newspaper found that law enforcement still relies heavily on eyewitness testimony, even if corroborating evidence is weak and despite decades of research showing its shortcomings.
' Eyewitness testimony can be unreliable for a variety of reasons. Victims often get only a brief look at their attackers, especially in robberies, and they tend to focus on the gun rather than the face, research shows. Accurately identifying strangers, especially of a different race, is difficult, and the stress of being a crime victim can distort memories.
Nonetheless, police and prosecutors say eyewitnesses are 2nd only to genetic evidence in convincing jurors of a defendant's guilt.
"Nothing beats a victim or a witness standing there when the crime was committed," Dallas homicide Sgt. Larry Lewis said. "The only thing that would beat that is DNA."
Troubles persist
The News reviewed 31 robbery trials over two years in which defendants were found guilty and appealed their sentences. Only cases on appeal have trial transcripts, which allow for an examination of what happened at trial and, in part, during the investigation.
The cases reviewed represented about two-thirds of all appealed robberies. However, more than 1,700 defendants during the same two years pleaded guilty and were given prison time, probation or deferred adjudication probation. As seen in some of the 19 exoneration cases, defendants sometimes accept a plea bargain to ensure a shorter sentence.
The newspaper did not reinvestigate the cases to determine whether an innocent defendant had been found guilty of the crime. But it found several of the same practices discredited in the DNA exonerations, including cases in which:
- Witnesses often selected people from photo lineups when they could not provide detailed descriptions of the robbers' faces. In a 2007 Dallas robbery case, a pizza deliveryman did not provide details of the robbers' faces or mention noticing a permanent grill on the defendant's teeth, but he identified a suspect with a grill during a photo lineup.
- Photo lineups were not always conducted according to best practices. In two Dallas cases, officers asked two witnesses to identify a suspect at the same time. Best practice dictates that witnesses view suspects separately.
- Victims sometimes picked suspects from lineups by eliminating other photos and then selecting the remaining one who most resembled their attacker. In a 2007 Coppell robbery case, the victim told jurors that she selected the defendant because he had the darkest complexion and the largest forehead of anyone in the photo lineup.
- Police focused on suspects in spite of conflicting evidence. A man was convicted in 2006 after a police dog tracked a scent from a store several doors down from a Duncanville crime scene. The dog led officers to an apartment complex where a man known to police lived. Investigators found a gun in the man's bedroom that didn't match the robbery weapon. But he was arrested after the victim identified him in a photo lineup.
Although defense attorneys questioned the unreliability of eyewitness identification, they did not call experts to testify about studies proving that point.
Victims in 6 cases could not identify their attackers, sometimes because the robbers wore masks. In those cases, DNA or robbers using victims' credit cards led to arrests.
DNA was found in only 3 cases: One involved a sexual assault; in the other 2, defendants left behind gloves with DNA.
The identification practice commonly known as "showups" occurred in about 20 % of the robbery cases examined. That tracks with the percentage of showups involved in DNA exonerations nationwide.
Showups allow police to show witnesses a suspect 1-on-1, either in person or by photo. The U.S. Supreme Court has discouraged but not banned the practice.
In one 2006 trial involving a showup, a woman was robbed at gunpoint while parking cars at the State Fair of Texas. She said the man who robbed her was wearing a white tank top and jeans and took from her $15 or $16 in cash, a cellphone and some wadded up toilet paper she had in her pocket.
Police put the woman in a squad car and drove her around the area until police spotted a man on the street wearing the same kind of clothes as the robber. After she identified him, police found $51 in his pockets but no gun or the woman's cellphone. He did have some crumpled toilet paper in his back pocket and was arrested. The man, who had a prior record, was later convicted and sentenced to 70 years in prison. No one tracks how often police conduct showups, but officers believe strongly in their value. One officer in a Dallas County robbery case tried last year testified that he preferred using showups to photo lineups.
Addressing problems
Mr. Watkins has made righting wrongful convictions a hallmark of his nearly 2-year administration. Presented with The News' findings, he acknowledged that his office needs to do much more to improve how it handles eyewitness testimony.
"This is just another indication of how far we need to go," Mr. Watkins said in an interview. "I was thinking, 'Yeah we'll get there,' but we have a long way to go."
However, he said, he cannot stop trying cases even if they are based on little else but eyewitness testimony.
"You can't just throw the whole process out of the window," he said. "The thing is to try to improve the process."
Mr. Watkins and his top assistant, Terri Moore, said they want to better train prosecutors to ask questions of witnesses and police officers about how identifications were obtained.
"Teaching them how to rethink some things, challenging ourselves," said Ms. Moore, a former federal prosecutor.
"I don't think you are going to stop wrongful convictions today because of a DNA test," Ms. Moore said, referring to how few cases have genetic evidence that can be used to verify the verdict. "I think it goes well beyond that."
Mr. Watkins said he would not support a law requiring all police departments to follow a specific policy on the use of eyewitness testimony. But he said he would lobby the Legislature to mandate that all police departments adopt standards for how they handle witness identifications. Only five Dallas County police departments could provide The News written policies on how to conduct witness identifications.
The Dallas Police Department in recent years began requiring witnesses to sign a form, acknowledging that the officer conducting the lineup warned them that the photo spread may not include the actual suspect. The department also has written guidelines for how to conduct photo lineups, but police officials say there is little oversight.
Mr. Watkins said he also would not support a special jury instruction warning that eyewitness identification can be unreliable. That point, he said, is more appropriately handled by presenting expert witnesses on the subject.
Texas law allows for convictions based on the testimony of a single eyewitness. Potential jurors who cannot convict based on the testimony of that lone witness if they believe the testimony beyond a reasonable doubt cannot legally serve on a jury panel.
Longtime Dallas defense attorney John Read said that since Mr. Watkins took office, county prosecutors have been more likely to question eyewitnesses' reliability. But police departments in Dallas County and prosecutors in other counties still base many of their cases on an eyewitness picking out a suspect, he said.
Mr. Read said he also believes that police, intentionally or otherwise, are still suggesting in lineups which suspect the victim should identify.
"I believe they do what they do whatever it is they do with their suggestions because they think they've got the right person," Mr. Read said. "They believe he's guilty."
'The revelation'
Earlier this year, Richardson police changed how they handled photo lineups after wrongly charging a man in a 1985 burglary and rape. The man, Thomas McGowan, was cleared in April through DNA testing after serving 22 years in prison.
Richardson police Chief Larry Zacharias called the city's only DNA exoneration "the revelation."
The photo lineup that led to Mr. McGowan's arrest and conviction included color photos, black-and-white photos and photocopies of photos a violation of the best practice that mandates all photos should be similar in composition. Mr. McGowan was pictured in a color photo, wearing a placard that read Richardson police.
Now, instead of the case investigator showing six photos at one time to a witness, a detective not involved in the case shows witnesses one photo at a time. The practice is known as a sequential blind lineup.
Investigators also video record all witness identifications conducted at the station, which previously had not been done. Field identifications are at least audio recorded.
Those seeking eyewitness identification reforms say videotaping reduces the likelihood of suggestion by the detective and allows prosecutors to show juries that the photo lineup was conducted properly.
Recording of some kind, Chief Zacharias said, is an "absolute" in identifications.
By contrast, Dallas police, who had 13 of the 19 exonerations, have been considering for nearly 2 years whether to participate in a limited study using sequential blind identification procedures. The study which would focus primarily on robbery cases could begin as early as November, departmental officials said.
When the new procedures begin, Dallas police Assistant Chief Ron Waldrop said, officers will probably make an audio recording of the identifications in photo lineups. Unlike Richardson, however, there are no plans for Dallas police to videotape its lineups.
"It's not going to eliminate mistakes," Chief Waldrop said of the proposed changes. "We just want to make sure the practices we put in place don't aid the mistakes."
Source for both: Dallas Morning News
No comments:
Post a Comment