Innocence Project Files for '94 Killing Evidence

Arkansan Belynda Goff is 20 years into a life sentence for her murdering her husband. She claims wrongful conviction, and she has the backing of the Innocence Project. The legal aid group must fight a hostile district attorney to DNA test evidence that may exonerate Belynda, who was largely convicted on the testimony of a rookie homicide detective and circumstantial evidence.



Innocence Project Files for '94 Killing Evidence

It hopes DNA tests clear victim's wife.

CHEREE FRANCO                                                                                                                                                                                       ARKANSAS DEMOCRAT-GAZETTE

On June 12, 1994, Belynda Goff told police she awoke before dawn to find her husband sprawled in a pool of blood in the entryway to their Green Forest apartment. Stephen Goff, 40, had been bludgeoned to death - with a hammer, police and prosecutors theorized, though no murder weapon was ever identified.

Belynda, 32, passed a lie-detector test, and steadfastly insisted she did not kill her husband of eight years.

Nearly a year after Stephen's death, prosecutors charged Belynda with first-degree murder. She refused a plea deal, saying she could not admit to something she didn't do. In August 1996, a Carroll County jury convicted Belynda and sentenced her to life in prison.

Now 52, Belynda Goff continues to proclaim her innocence and has persuaded a New York-based nonprofit to try to help her prove it.

The Innocence Project describes itself as a not-for-profit legal organization that assists convicted people in gaining access to post-conviction DNA testing. In a motion filed last week, lawyers for the Innocence Project asked the Carroll County Circuit Court to order the release of physical evidence to help them delve into Belynda's case.

The group wants to send Stephen Goff 's clothes, as well as the glasses and keys found near his body, loose hairs collected from his palms and scrapings from his fingernails - all items logged in evidence reports - to a private laboratory in Dallas. It has offered to cover the costs of transportation and DNA testing. It hopes to find DNA evidence that would cast doubt on Belynda's guilt.

But Carroll County prosecutor Tony Rogers is reluctant to open a closed case.

"She's been convicted. Two juries in this county have sentenced her … and she's been through numerous appeals," he said in November.

He had no comment last week about the Innocence Project's court filing, but last month he told the Arkansas Democrat-Gazette that despite his misgivings, he would "go back and look at all the evidence" if the group instituted court proceedings.

Since 2001, Arkansas law has granted people convicted of crimes the right to apply for DNA testing. But that law remains subject to judicial interpretation, and local defense attorneys say judges and prosecutors have been hesitant to allow post-conviction testing.

Even if the Innocence Project is successful in obtaining the evidence stored since Stephen Goff's death, the road to clearing Belynda Goff remains long and difficult, according to experts in the field of post-conviction exoneration.


The Arkansas Crime Laboratory tests DNA in almost every current investigation, Director Kermit Channell said. "We use DNA technology on any case we've got biological fluids on. We even use DNA as a last resort, with touch [skin cell] samples and stuff like that."

He said the Crime Lab rules out about a quarter of police suspects on the basis of DNA testing.

Also, DNA has been used to solve cold cases. Four years ago, questions from reporters at the Arkansas Democrat-Gazette prompted new DNA testing on a young girl's remains that had been unidentified for decades. An earlier test had ruled out a possible identity, but the new tests confirmed that the remains were those of a 12-year-old abducted from her Little Rock home in 1982.

"In a run-of-the-mill case … it's pretty straightforward," said David Kaye, a Pennsylvania State University law professor who has served on a DNA commission headed by the U.S. attorney general.

"But the laboratory could contaminate things. You could have sample mix-ups. It could misinterpret a mixture [one sample with DNA from multiple people]. There could be an incomplete profile that could be explained in several ways, and sometimes, the analyst reaches one conclusion and overlooks the other possibilities … so those would be the things that could be worrisome."

The most problematic cases occur when the DNA sample is tiny, rendering the data more difficult to interpret. But when performed correctly and especially on body fluids, Kaye and Channell estimate that DNA tests approach 100 percent accuracy.

"There are going to be other cases, particularly with touch or degraded DNA, where we can't really say 100 percent. Maybe 90 percent. Still good evidence but not infallible," Kaye said.


At Belynda's trial in Berryville, deputy prosecutor Kenneth Elser opened with an aphorism: "Hell hath no fury like a woman scorned."

He and prosecutor Brad Butler depicted an angry wife who suspected her husband of leaving their apartment to visit a girlfriend just after an anniversary dinner.

Belynda hid in the apartment's dark living room, they said, with a cotton shirt over the silky-pajama set that Stephen had requested that she wear.

When he returned, she caught him off-guard, slamming a hammer against the base of his skull, the prosecutors posited. The initial blow dropped him to his knees. More blows left him slumped in the entryway, legs akimbo.

In Butler and Elser's narrative, Belynda washed the hammer in the tub, ripped her shirt into tiny pieces and flushed it down the toilet. To support this theory, Butler cited a wet plunger, a pile of damp towels and a few diluted drops of Stephen's blood on the underside of the bathtub drain plate.

No blood was found outside the apartment, nor any sign of forced entry. And, according to the prosecutors, there was no way for anyone to leave because Stephen's body was propped against the only door. Belynda and the couple's 3-year-old son were the only ones in the apartment.

The defense team, private attorneys Charles Davis and Stevan Vowell, said Belynda slept through the beating because she was medicated after having a hysterectomy. As for the blocked exit, they countered that the body was near the door, not flush against it, and that plenty of first responders were able to enter and exit the apartment.

The only potential weapons in the apartment, a pair of hammers and some dumbbells, had tested negative for tissue or blood.

The trial lasted five days. Belynda received the maximum penalty possible.


In 1997, Belynda Goff filed a direct appeal with the Arkansas Supreme Court. Her conviction was upheld, but because Circuit Judge Tom Keith had entered the jury room without attorneys present during sentencing deliberations, the high court recommended a resentencing.

A new jury heard attorneys' arguments and the transcript of Belynda's original testimony. Again, she was sentenced to life in prison.

But Arkansas law also provides means of indirect appeal. Known as post-conviction remedies, this approach doesn't seek to prove guilt or innocence. Rather, it seeks to demonstrate that procedural errors occurred during trial, such as ineffective counsel, evidence suppression or judges' mistakes, according to Little Rock defense attorney Jeff Rosenzweig.

In 2001, Belynda fired her original lawyer, Charles Davis, and filed a request for post-conviction relief with a new lawyer, Jack Lassiter.

The request was granted.

Circuit Judge Jay Finch ruled that a life-insurance claim Belynda filed had prejudiced the jury against her and that Davis had created a conflict of interest in paying himself from the subsequent check.

Belynda was the sole beneficiary of a $275,000 policy that Stephen had through his employer. She said she filed the claim per Davis' advice in order to avoid suspicion. The payment was mailed to Davis' office.

During sentencing arguments, the prosecution labeled it "blood money."

Finch ordered a retrial, but the state Supreme Court overturned his decision and once again upheld Belynda's conviction.

For Belynda, the remaining options seem to be executive clemency and exoneration.

In general, exonerations are rare. There have been three in the state since 1989, according to the National Registry of Exonerations, the most comprehensive database available. None of those involved exoneration via DNA evidence.

The state Crime Lab has tested DNA since 1996, but Channell said he has been asked to test post-conviction DNA evidence only a couple of times: "I can't name any, but there might be one or two."

In a practice spanning more than 100 clients, Rosenzweig estimates that he has asked courts for post-conviction DNA testing a dozen times. Mostly the requests were denied. Of the three cases in which testing was allowed, the conviction was affirmed in one. In the second, no DNA was found on the evidence. In the third, the DNA raised questions that helped attorneys secure the defendants' release without exoneration.

The first DNA exoneration in the United States occurred in 1989. According to the Innocence Project, which accepts about 35 cases from several thousand applications each year, there have been 312 DNA exonerations nationally.

Since the organization's inception in 1992, it has worked in 30 states and been directly involved with 172 DNA exonerations. About half of the cases it accepts result in exonerations. The rest usually are affirmed.

Belynda Goff's case marks the first time the group has ventured into Arkansas.


According to Karen Thompson, a lawyer with the Innocence Project, the organization accepts only cases in which there is untested biological evidence. Even then, it doesn't accept every case.

The group researched Belynda Goff's case for nine years before deciding to represent her.

"We weigh each case on its own merits, and in Ms. Goff's case, there is a strong evidential basis that supports potential innocence," Thompson said. "There are behaviors that are clearly pointing to an alternate perpetrator."

During Belynda's trial, her brother, Chris Lindley, a former Marine with explosives experience, was subpoenaed by the defense but was never called to testify.

After her conviction, he wrote a letter to Gov. Mike Huckabee alleging Stephen Goff's involvement in an arson-for-hire. Lindley claimed that a year before his death, Stephen had asked for help burning a warehouse in Flint, Mich.

Lindley wrote that he vacillated over the next few months. When he backed out the final time, a week before Stephen's death, Stephen became hysterical, saying that he'd already spent the $10,000 he'd been paid and would be killed.

According to Lindley, the arson was scheduled for June 10, 1994, two days before Stephen was killed.

On June 14, Lindley said he received a threatening phone call.

"If you open your mouth, you'll be lying right next to Stephen," the unidentified male voice said.

Other testimony that was excluded from the trial included that of a woman who became suspicious of two men asking questions about the residents of two apartments, one of them housed the Goffs, and two people who said they overheard conversations in separate locations about a "blood bath" in Green Forest and a murder.

A year after Stephen Goff's death, a late-night arsonist burned the rental house where Belynda and her children lived. The investigation was closed two months later, without any suspects or arrests.

"There are a lot of things here that are odd to me, but there are a lot of odd things in the world. … We could speculate all day, but the DNA won't lie," Thompson said.

She believes that if unknown DNA, particularly male DNA, is found on the requested evidence, it would undermine Belynda's conviction.

"We know that Mr. Goff was engaged in some sort of defensive fighting posture. He had bruises and scrapes on his knuckles and his hand," she said.

"Obviously, if we got a hit [in DNA databases] on somebody who was already incarcerated for similar crimes, that would be very probative [demonstrative] evidence of innocence. If we don't get definitive results, if we don't get anything that goes against the conviction, nothing changes."

Ralph Gordon, a lieutenant with the Carroll County sheriff's office, recently confirmed that the department still has seven boxes of evidence from the Stephen Goff murder.

The best-case scenario, according to Thompson, is finding the hairs mentioned in the evidence logs and discovering that they still have roots.


In 2012, the Innocence Project used DNA testing on hairs to help Sedrick Courtney of Oklahoma obtain exoneration after serving 15 years in prison for armed robbery.

In April 1995, two armed men kicked in a woman's door and struck her hard enough to cause traumatic brain injury. The victim positively identified Courtney, an acquaintance, as one of the assailants based on the voice and a brief glimpse when the attacker lifted his ski mask.

Hairs found on the ski mask were analyzed before trial. The analyst testified that because of the hairs' unique color Courtney could not be eliminated as a suspect.

Courtney's initial post-conviction request for DNA testing was denied in 2001 because the Tulsa Police Department claimed that the evidence had been destroyed.

After the Innocence Project made repeated requests, the hairs were found and tested for mitochondrial DNA, a science not yet available at the time of Courtney's trial. The tests excluded Courtney as a match for all 15 hairs recovered.

"They didn't have a root, but even so, we were able to exclude him," Thompson said.


The rate of wrongful convictions in capital-murder cases has been estimated at between 2.3 percent and 5 percent, according to Samuel Gross, a University of Michigan law professor who oversees the National Registry of Exonerations.

If the low end of that rate was applied to the entire incarcerated population of 2.3 million, that would be 52,900 people.

The registry lists 1,269 exonerations since 1989, and that is growing by about 200 cases a year. Of those 200, about a third are added to the registry as they occur. Twothirds are old exonerations that come to researchers' attention.

"That's about the rate at which we have the resources to work, and there's no end in sight. We know there's a lot of cases that we haven't found," Gross said.

The registry lists only three exonerations for Arkansas - a murder conviction overturned in Pulaski County in 1989 and two drug-related convictions overturned in Clark County in 2000 and 2010. These convictions were based on false accusations, witnesses' mistaken identification and official misconduct.

Forty-six percent of the cases logged by the registry involve black defendants. Forty percent of the defendants are white, and 12 percent are Hispanic. The majority of the defendants were convicted of homicide or sexual assault and spent an average of 10 years in prison.

Perjury and false accusations contributed to 53 percent of those convictions, official misconduct contributed to 44 percent, and mistaken eyewitness accounts contributed to 39 percent.

A quarter of the exonerations resulted from recanted testimony, most of which involved eyewitnesses.

Another quarter were DNA-based.

Brad Butler, who prosecuted Belynda Goff and was later reprimanded by the Arkansas Supreme Court for inappropriate relations with defendants, stands behind the conviction he helped win.

Two judges involved in Belynda Goff's case disagree on whether her case warrants DNA testing.

Tom Keith, the now-retired judge who presided over Belynda's trial, remains satisfied with the jury's verdict.

"If I was not, I would have done something … reduced the sentence," he said.

But retired judge Jay Finch, who once granted Belynda a new trial, thinks a jury in a more cosmopolitan area might have reached a different conclusion in the case.

"The state couldn't explain the gaps in the case, so they relied on the local prejudice … it shifted the burden of proof," he said.

According to Gross, the justice system tilts toward guilty verdicts. "In the day-to-day processing, we can lose track of the fact that what usually happens isn't what always happens. So that sort of bias is common, just the routine assumption that this is another case of a guilty person. Because it usually is."