Dozens, if not hundreds, of criminal convictions in Ohio could be in jeopardy because a longtime forensic scientist at the state crime lab now stands accused of slanting evidence to help cops and prosecutors build their cases.
The credibility of G. Michele Yezzo, who worked at the Ohio attorney general’s Bureau of Criminal Investigation for more than three decades, has been challenged in two cases in which men were convicted of aggravated murder. One has been freed from prison because of her now-suspect work.
A review of her personnel records by The Dispatch shows that colleagues and supervisors raised questions about Yezzo time and again while she tested evidence and testified in an uncounted number of murder, rape and other criminal cases in the state.
Their concerns included that she presented evidence in the best light for prosecutors instead of objectively, used suspect methods while examining trace evidence from some crime scenes, and made mistakes that, as one former attorney general put it, “could lead to a substantial miscarriage of justice.”
Yezzo, 63, of West Jefferson, told The Dispatch that the accusations about her work being biased are wrong and that she approached her work objectively.
“I have never done anything to overstate analysis of evidence, nor have I done anything, for lack of better a word, to taint the evidence,” Yezzo said. “No, I didn’t appease prosecutors and law enforcement. I bent over backwards to try and find out whatever evidence was there, and that’s the best I can tell you.”
But two former attorneys general, defense attorneys, a judge, a former BCI superintendent and a nationally renowned forensic expert from the FBI all say that Yezzo has credibility issues that may have poisoned cases she touched.
Lee Fisher, who served as attorney general from 1991 to 1995, and Jim Petro, who served as attorney general from 2003 to 2007, both said they didn’t know of Yezzo when they were in office, but they now have concerns about her work.
“I would call for an investigation into every case where her findings and conclusions were instrumental in the final result of a case,” Fisher said. “We have an obligation to the integrity of the criminal-justice system to investigate every case. We have to determine whether her findings or conclusions were suspect.”
Ohio Attorney General Mike DeWine said Friday that his office was alerted to the concerns about Yezzo in 2015 and has since conducted two separate reviews of her work. One involved examining 100 criminal cases where Yezzo’s evidence analysis played a role in a conviction.
DeWine said they found no issues with her work.
Moving forward, DeWine, who did not serve as attorney general during Yezzo’s tenure, said he has no plans for an internal investigation into Yezzo’s history, but he will have open discussions with defense attorneys on a case-by-case basis if they raise questions.
He said the BCI, which handles about 37,000 cases a year, has a “long history of doing good work” and has received the highest level of accreditation.
More than 800 pages of Yezzo’s personnel records paint a disturbing pattern of behavior that started shortly after she walked in the BCI doors in 1976. The concerns escalated over time until she resigned in 2009.
Over the 32 years that Yezzo worked in the crime lab, her bad behavior intensified to the point that colleagues questioned her mental health.
In the records, colleagues and supervisors described these concerns about Yezzo: She threatened to use a gun to shoot her co-workers and herself. She threw a 6-inch metal plate at one co-worker. She exposed her breasts to BCI agents at a bar, flipped off her boss and acted in a hostile manner to almost every lab employee, according to records. She was accused of calling an African-American scientist a racial slur, something Yezzo denies. She frequently broke into crying spells for no apparent reason.
Forensic scientists quit because of her erratic behavior. At one point her union, the Fraternal Order of Police, refused to back her.
Yezzo admits to the majority of the behavior described in her personnel file.
She attributed her erratic and sometimes abusive actions to intense pressure within the BCI to handle an enormous caseload as its lead forensic analyst.
She said the bureau was usually short-staffed and had difficulty keeping up with the workload. She also said she was having problems in her personal life. Those issues related to the loss of her sister and her mother moving in with her after the death.
She doesn’t believe her behavior affected her work.
Yezzo received numerous verbal reprimands and was suspended in 1993. But her analysis of evidence continued to be used in many high-profile felony cases despite the concerns about her work and behavior inside the state’s crime lab in London, where forensic scientists examine and analyze evidence from crime scenes across Ohio.
Yezzo conducted her analysis of evidence without much oversight. Her reports summarizing findings would be reviewed by her supervisors, but her actual work, methods and conclusions rarely were checked by anyone.
Now, defense attorneys in at least two cases have done their own investigations and believe they have proof that Yezzo’s work is suspect. In one of those cases, a judge already has freed a man from prison because of credibility issues described in Yezzo’s personnel file.
The judge in that case and others familiar with Yezzo’s BCI history say that if defense attorneys had known about her work issues during past trials, they potentially could have discredited her as an expert witness.
“I didn’t know of what occurred with Michele Yezzo when I was in that office, but if I had been made aware, I wouldn’t have allowed her to be involved in criminal-justice proceedings,” Petro said. “I am co-counsel in two cases where her work largely convicted men, and her work was shoddy at best. Any case where she provided forensic evidence that resulted in a conviction now comes into question.”
Barbara Parsons was in the bedroom of her home in Norwalk, in northern Ohio, when someone brutally attacked her, striking her 15 times in the head with a large, heavy object.
Her husband, James Parsons, told police that he was working at the local auto repair shop when his wife was murdered in 1981. His alibi held up for about 12 years, until he was arrested and charged with his wife’s death in 1993.
Shortly before his arrest, police resubmitted evidence to the BCI, and Yezzo was assigned to the case. It was mainly circumstantial, but Yezzo’s analysis of the crime scene is what led prosecutors to charge James Parsons with his wife’s murder.
Prosecutors used Yezzo’s analysis of blood patterns to determine that a breaker bar, a long socket wrench made of thick metal, was the murder weapon. Prosecutors said the bar belonged to Parsons.
Yezzo performed an experiment on the crime-scene evidence that compared the head of the breaker bar with blood patterns found on Barbara Parsons’ bedsheets and nightgown. Yezzo concluded that the letters N and S, which were imprinted on the head of the breaker bar, “were consistent with” blood impressions left on the bedsheets. She also concluded a halo-like impression from the breaker bar was consistent with a blood pattern left on both the sheets and nightgown.
Defense attorneys rebutted her findings and said Yezzo did not properly document how she got those results, nor did she properly explain her findings to the jury.
Yezzo was working on the case in 1993 and preparing to testify when she was suspended for threatening a co-worker and placed under investigation.
Parsons was convicted and sentenced to 15 years to life in prison.
He remained in prison for 23 years until Donald Caster, an attorney for the University of Cincinnati-based Ohio Innocence Project, played a hunch and asked for Yezzo’s personnel file.
At a hearing this year in Huron County, Caster convinced Judge Thomas Pokorny that Yezzo’s scientific conclusions were suspect and that her troubled BCI history called into question her credibility.
Pokorny didn’t declare Parsons innocent, but the judge dismissed the murder conviction and released him.
“What has weighed most heavily on the court’s mind is the testimony from Ms. Yezzo’s superior that the integrity of her analysis and conclusions may be suspect as she ‘will stretch the truth to satisfy a department,’” Pokorny said in his ruling, referring to a memo written by an assistant BCI superintendent in 1989. “This opinion together with the other evidence of her troubled behavior in the workplace casts grave doubts about her credibility.”
Yezzo defended herself at James Parsons’ hearing when Caster questioned whether her work was biased toward law enforcement.
“There may have been issues between me and my co-workers, but it was not a circumstance where those issues fell to the analysis of evidence,” Yezzo said. “You’re trying to portray me as a prosecution expert. I testified to the results, not to try and make any points with anybody.”
Yezzo’s direct supervisor, Daniel Cappy, defended her work. Cappy testified that Yezzo had some behavioral issues, but he stood behind the quality of her work as a forensic scientist.
Parsons suffers from congestive heart disease and dementia. He requires constant care and was placed in a nursing home after his release in April.
Prosecutors lost their initial appeal and are now asking the appellate court to reconsider Pokorny’s ruling. It’s unlikely prosecutors would attempt to retry Parsons for his wife’s death.
“It was shocking to me — shocking — that you have a forensic scientist being questioned by her own bosses and no one went back to check her work,” Caster said. “That is grossly negligent at best. It’s clear that she became an advocate for prosecutors, and that’s not what a forensic scientist should be.”
A second chance
On Friday, another man convicted of murder asked a judge to grant him a new trial because of Yezzo’s checkered history.
Kevin Keith was once just 13 days away from being executed.
Prosecutors convinced a jury that he walked into a small apartment in Bucyrus, about 60 miles north of Columbus, on Feb. 13, 1994, and riddled it with gunfire. Three people died and three others were wounded.
The key testimony at the trial that ultimately put him on Death Row came from Yezzo.
Her forensic analysis of tire tracks and a license plate imprint left in a snowbank matched what detectives working the case contended — that both came from the car Keith apparently drove that night.
But in 2010, a retired FBI forensic expert said Yezzo’s conclusions were baseless and her methods were shoddy in Keith’s case. Keith’s legal team, led by Assistant Ohio Public Defender Rachel Troutman, also produced evidence that they say implicates another man, challenges the credibility of eyewitness accounts and shows that police mishandled the case.
All of that prompted then-Gov. Ted Strickland to spare Keith’s life by commuting his sentence to life without parole. Keith remains in the Marion Correctional Institution.
“I can’t thank the governor for giving me life in prison,” Keith told The Dispatch in 2010. “When I found out his decision, it felt like the poison was going through me right then. But the governor gave me some hope by leaving the door open in my case.”
Keith’s attorneys now think he has more reason to hope, and a better chance of proving his innocence. They filed a motion late Friday in Crawford County Common Pleas Court, asking for a new trial based on the problems with Yezzo’s work and her behavioral history.
“Every case she has touched is tainted,” said Zach Swisher, a member of Keith’s legal team who also reviewed Keith’s case in 2010 as legal counsel for Strickland. “The fact that she is considered an expert is a joke. She gave the detectives what they wanted in a capital case, and that is just unconscionable.”
William Bodziak, a forensic expert for 43 years, including 29 of them for the FBI, said Yezzo’s work in the Keith case was below standards for even scientists in training — and Yezzo had decades of experience.
Bodziak said the evidence problems started with the “unprofessional” recovery and documentation of tire-track and license-plate imprints. The lack of photographs and the inability to show dimensions at the crime scene made it nearly impossible to draw conclusions, said the expert, who provided an independent analysis as part of Keith’s case for a pardon.
For the license plate imprint, Bodziak said Yezzo concluded that the numbers “043” could be seen in the snow, and those numbers matched the plate from the car prosecutors said Keith was driving. But Bodziak found multiple problems: He could barely see the 0 or the 4 in the imprint; the license plate was mounted flat, making it unlikely to leave such an impression; and the snow surrounding those numbers remained untouched.
Bodziak said Yezzo’s tire-track conclusions lack even more credibility because she made them by comparing a photograph to a commercial brochure a detective provided. She didn’t inspect the actual tires. Bodziak said the tires on that car were among the most common and popular used on the road.
“There is nothing to support the conclusions she made, nothing at all,” said Bodziak, who remains a consultant on forensic cases for both prosecutors and defense attorneys.
“If I had been working on that case, I would have pointed out all those discrepancies and would not have made any conclusions. But it appears she was giving investigators the conclusions they wanted, and that’s the really bad part of this case,” he said.
Yezzo was unaware that her work was challenged by Bodziak in 2010 and disputes that she made conclusions in the case. She says she never stated that the “043” impression came from the vehicle prosecutors claim Keith was driving, just that she could see those numbers.
She also says she compared the tire tracks to photos the BCI had of the tires on the vehicle Keith used, and she only said the tracks “were consistent with” the treadmarks left at the crime scene.
Yezzo said she couldn’t help it if prosecutors or defense attorneys in her cases overstated or twisted her analysis to help make their arguments in court.
“Attorneys on both sides are going to slant things in the way they want a jury to think about them,” Yezzo said. “I don’t like it when people take what analysts do and stretch the truth. I can only be responsible for my analysis.”