Karen Sue Boes has been insisting she did not kill her 14-year-old daughter Robin since the day her Zeeland, Michigan, house caught on fire almost two decades ago.
As police interrogated her multiple times over six weeks, the 65-year-old maintained she was out shopping with a friend when the tragic fire broke out, killing Robin. But during one grueling questioning on Aug. 7, 2002, police eventually got her to admit she possibly started the fire while in a “dream” or “unconscious” state.
“They had me really disoriented,” Boes told The Daily Beast in a phone interview from the Women’s Huron Valley Correctional Facility on Thursday. “I really started thinking I was going crazy. I knew I didn’t do it. I absolutely knew that, but they kept putting all these scenarios in my head. After hours and hours of that, you start doubting yourself.”
During her murder trial, prosecutors convinced a jury that Boes intentionally set the fire in a hallway outside Robin’s bedroom before leaving the house. They relied heavily on two fire experts, who used “burn pattern analysis,” to argue that the most “intense burns'” were found outside Robin’s bedroom—and thus the fire had to have originated there.
Even after she was given a life sentence without parole, Boes kept insisting she was innocent, filing a series of appeals. By 2017, with all avenues exhausted, she appeared on Netflix’s The Confession Tapes to plead her innocence.
Now, Boes has what may be her best shot at freedom. Her case has been taken up by lawyers with the Michigan Innocence Clinic at the University of Michigan Law School, who are asking a judge to set aside her conviction and order a new trial because prosecutors used junk science to “prove” their case against Boes.
“She never really confessed. She got convinced she must have done it based on what the police were telling her—drilling into her,” David Moran, a law professor and co-founder of the Michigan Innocence Clinic, told The Daily Beast. “This is what is called an internalized false confession. She got convinced she must have done it based on what they were telling her.”
The motion for relief of judgment, to be filed Monday in Ottawa County Circuit Court, argues that the prosecution used two techniques that have since been discredited by experts. Firstly, a mountain of new research proves that some police interrogation techniques often elicit false confessions.
Secondly, the motion adds, “the fire investigation community now rejects the techniques relied upon by the prosecution’s two experts at trial to determine the cause and origin of the Boes fire.”
During the trial, Michael Marquardt and John DeHaan relied on the “burn pattern” technique, which a breakthrough 2008 study concluded is “scientifically invalid and leads to incorrect identifications of fire origin.”
The two experts also used “negative corpus” to conclude that the fire was “intentionally set” because there was no evidence for any other classification. Starting in 2011, however, the “scientific standard of care for the fire investigation profession… fully repudiated the use of negative corpus,” the motion says.
In 2015, DeHaan resigned from the American Academy of Forensic Science (AAFS) after a damning three-year investigation found he had “committed professional misconduct” by giving misleading testimony in another trial. He was “subservien[t] to the wishes of the prosecutors regarding the contents of his reports, and his conclusions in one of his reports that were not based on sound science.”
Boes said it “threw me through a loop” when she heard that old science had been used in her trial. “These are educated people!” she said with a small laugh.
But Moran cautioned that there was no guarantee of a new trial, let alone a hearing. “It’s always an uphill battle,” he told The Daily Beast. “And it is a steeper one since Karen filed a similar motion in 2006 that was rejected. But we have been working on this filing for several years and we have a mountain of evidence that undermines the prosecution’s case. So we are all cautiously optimistic.”
The Day Everything Changed
On the morning of July 30, 2002, Boes woke up around 6:30 a.m. and began a typical day. She made coffee, did some laundry, ironed her shirt.
At around 8:40 a.m., she pulled out of her driveway without seeing her daughter Robin, who was on summer vacation before her freshman year at Zeeland West High School. Boes headed to her husband’s job at a local body shop to grab some money, stopped at Burger King to grab a cinnamon roll and ice tea for breakfast, then picked up her friend Judy Payne for a day of shopping in Grand Rapids.
“All we did was talk about her upcoming vacation. She didn’t seem nervous at all or driving crazy,” Payne told The Daily Beast in a phone interview. “She had no signs on her that she had been around a fire, she wasn’t burned, she didn’t smell like smoke or gas. She seemed totally normal.”
Less than 30 minutes later, while the pair were waiting for a Payless shoe store to open, Boes got a call from her 18-year-old son, telling her their house was on fire. Payne said she remembers the moment: Boes was in “total shock” and drove like “crazy” to get home.
“She was going through red lights and at one point I was like, ‘Karen, stop, pull over. You’re driving crazy,’” she said. “I went with her to the house and I remember lots of people, fire trucks, and Karen wanted to go in the house but they restrained her and didn’t let her go in. She was pretty upset.”
The motion states that Boes was so upset that paramedics had to give her a shot of Valium.
Boes later learned that a neighbor, Gena Zuverink, was the one who called the Fire Department at 9 a.m. When firefighters entered the house, Robin’s bedroom door was closed. Inside, she was found “lying face down on the floor.” She was pronounced dead at 10:05 a.m.
“During a subsequent search of the home, a five-gallon gas can was found in the corner of Robin’s bedroom, a few feet away from her bed, with a small amount of liquid still in the container,” the motion states. The can had originally been elsewhere in the house but had been missing for about three weeks.
Testing of the hallway carpet outside her bedroom came back negative for any trace of liquid accelerants, while the bedroom carpet samples came back positive for gasoline, the motion says. Burned matches, a box of matches, burned candles, and incense were found in Robin’s room.
Desperate to help the investigation, Boes gave police shoes and clothes she was wearing that day, which all came back negative for gasoline. Robin’s clothing came back positive, the motion says.
Investigators also failed to find any trace of accelerant in Boes’ car, but they “immediately focused their attention” on her. On the day of the fire, Boes went to the police station at the request of Zeeland Police Chief William Olney—her neighbor and friend of 13 years—and denied any involvement in the fire.
A week later, she went back under the pretense of a “formality” that would take about two hours. Instead, she was there for 11 hours and was interrogated by “a rotating set of three investigators” from local police, Michigan State Police, and the Bureau of Alcohol Tobacco and Firearms.
“During the interview, ATF Special Agent Eduardo Fernandez lied to Ms. Boes, pretending that investigators had evidence against [her] that did not exist. For example, Fernandez told Ms. Boes that investigators had found her fingerprints on the gas can from Robin’s room—a complete fabrication,” the motion states. “Fernandez admitted that he lied to Ms. Boes on multiple occasions.”
Boes was also interrogated by Olney, who told her she was “probably not [going to] find a better friend here” than him. Boes continually denied starting the fire.
“But after hours of interrogation, Ms. Boes came to believe that she could have set the fire in an unconscious state,” the motion states. “At different points in the interview, Chief Olney told Ms. Boes to concentrate on her ‘dreams’ about what happened with the fire. He also explained that she had a ‘good mind’ and a ‘bad mind’ and that her ‘bad mind’ could have taken over and set the fire.”
Eventually, Boes conceded that she could have gone “insane” and set the fire, but insisted that she had no recollection of doing so. “In my conscious mind I didn’t do this but I could have went temporarily insane,” she told Olney, according to a transcript.
“All I can remember is that it was one after another. Sometimes it was one investigator. Sometimes it was multiple and they were all standing while I was sitting,” Boes told The Daily Beast this week. “It was a very intimidating situation. Very uncomfortable. And for a while, I believed they were going to help me find out what happened to my baby.”
The brutal interrogation left Boes so distraught she was taken to a hospital psychiatric ward, where she told doctors about previous suicidal ideations, her overwhelming grief at losing her daughter, and the distress she felt from the interrogation.
On Aug. 13, Boes again denied any involvement in the fire. She insisted the events police fed her were false. “[M]ost of the stuff that we talked about has been fabricated right here in this room,” Boes told the police.
A month later, Boes was interviewed a fourth time. She insisted that investigators had “suggested things” that made her break down and convince herself she “could have done it.”
“I didn’t do what you said I did,” Boes said before she was placed under arrest for felony murder.
Speaking to The Daily Beast, Boes said it took her a while to realize that her conversations were going from standard questioning of a parent who lost a child to interrogations of a possible suspect. She does, however, remember that during a break after one lengthy interrogation she asked Olney if she needed a lawyer.
“And he said ‘No, I don’t think so yet,’” she recalled. “So I must have had some inkling by then things were not right. My mind was just reeling the whole time.” (Olney could not be reached for comment this week.)
During a 2004 trial, prosecutors argued that Boes “intentionally sprinkled gasoline in the hallway and in Robin’s bedroom before igniting the fire in the hallway and leaving the home,” the motion states. Marquardt and DeHaan backed up this theory with the now-debunked analysis and insisted that a liquid accelerant was used in the hallway—despite the fact that no residue was ever found.
“It was shocking but the prosecution put on a very good show,” Payne said. “They even had a fireman come in in a fire outfit.”
Prosecutor Jon Hulsing, who is now an Ottawa County Circuit Court judge, also played Boes’ interrogations to the jury. Hulsing declined to comment on the case this week.
Payne said the prosecution had “the timeline wrong” but Boes’ defense team had a hard time swaying the jury. She said defense lawyer David Zessin, who had primarily done divorce cases, didn’t even ask Boes about her experience on the day of the fire when she testified. Zessin did not respond to The Daily Beast’s request for comment.
“He never asked me if Karen had any burn marks, or any smudges on her, or smelled of gasoline,” Payne recalled. “And then when the prosecution cross-examined me, they brought up the fact that Karen’s attorney didn’t ask me any of those questions—making it seem like he didn’t ask me questions on purpose.”
Boes testified that she was innocent but “became convinced that she could have gone insane and started the fire because she believed the lies told to her by investigators through hours of interrogation on August 7,” the motion states.
The jury didn’t buy it, finding her guilty of first-degree felony murder, a conviction that was affirmed by the Michigan Court of Appeals. A year later, the Michigan Supreme Court denied her application for leave to appeal. In 2006, Boes filed her first motion for a new trial, but that too was denied two years later.
Revisiting the Science
“This is the first motion in Boes’ case in about 15 years,” Moran said. “And we’ve been working on it for years, making sure it’s filled with new evidence.”
Moran’s team consulted John Lentini, a renowned fire investigator, who says the jury was “misled by now-outdated science” because burn-pattern analysis does not account for ventilation at the location of the fire.
“In this case, Marquardt and DeHaan relied on the intense burn patterns on the bedroom door and adjacent surfaces to conclude that the fire must have started outside the bedroom in the hallway,” the motion states. “Changes in fire science now prove that such a conclusion is unreliable.”
The motion notes that fire experts would today recognize that the Boes fire “most likely originated in the bedroom, making [prosecutors’] theory at trial... impossible.” “[T]he evidence shows that the initial fuel for the fire in the hallway was unburned hydrocarbons produced by the fire starting in the bedroom,” Lentini wrote.
While Marquardt did not respond to a request for comment, a spokesperson for the ATF told The Daily Beast that he “has retired, and we stand by our investigation.”
DeHaan told The Daily Beast that while he does not remember “all the details” of the case, he does remember “doing the scene inspection on the Boes house and spending a lot of time looking at the doorway, clearances, fire damage around it and under it, and discussing it all with the other investigators there before reaching my conclusions.”
“There was also interesting explosion damage to the room that had to be taken into account,” he said.
In response to the AAFSA investigation, DeHaan said the committee that recommended his resignation “relied on some very unreliable information presented by my accuser.” “The testimony that was challenged was based on information provided by another witness that a judge ruled years later was inadmissible,” he said. “When I filed a follow-up report years later, I was forced to change my opinion.”
James Trainum, a 27-year veteran of the D.C. Metropolitan Police Department, who received advanced interview and interrogation training as a detective, wrote in a report for the motion that, since 2006, researchers have found that “certain interrogation practices dramatically reduce the reliability of a confession and can lead to internalized false confessions.”
“[Investigators used tactics] that have the capability of creating a situation where an innocent person may confess to a crime that they did not commit,” Trainum wrote, adding that Boes’ interrogation “meets all of the components that form a process that has been identified in research of internalized false confession cases.”
For Boes and her lawyers, the faulty fire analysis technique, the discredited experts, and the scientific research on false confessions prove the 65-year-old deserves a new trial.
“I always could come back and insist that I didn’t do it because I knew that truth. I knew I could never do that,” Boes said this week.
“For so long there has been a stigma attached to me. This time feels like I finally have more options.”