Since reinstatement of the death penalty, over eighty people have been exonerated after spending years on death row. Something has gone wrong in these cases, but what failings of the criminal justice system explain the wrongful conviction of innocent defendants and their sentence to death row? Others have examined capital cases. Barry Scheck and the law students of the Innocence Project have investigated wrongful convictions, newspapers and journalism students have brought these issues to light, and non-profit groups have offered proposals to reform the criminal justice process. This report, however, offers a new twist. Not only were the researchers undergraduate students, but our motivation was not exoneration but analysis. Representing multiple views on the death penalty, we have come together on this project to investigate shortcomings in capital prosecutions and to recommend changes to the investigation and trial of death penalty cases. What we advocate are common sense solutions, politically saleable and in accord with our Constitution, which hopefully most Americans of all perspectives may support, for we ourselves represent a broad array of perspectives, ideologies, and backgrounds.

Accuracy Project Honors Clinic

In the fall of 2001 ten honors undergraduate students at George Mason University took aim at the many capital cases in which a defendant had been wrongly convicted and later exonerated. Working under the aegis of the Accuracy Project at George Mason University, and supervised by Professor Jon Gould of the Administration of Justice Program, the students reflected a variety of views on the death penalty. Some supported capital punishment, others would eliminate it, but all agreed on a single point: no civilized nation should countenance the execution of an innocent person. The students worked together to investigate why defendants had been wrongly convicted and to propose recommendations to prevent innocent persons from bearing the consequences of heinous acts they did not commit.

The first step in the analytical process was to select cases on which to base the study. We started by compiling a list of cases in which the defendant had been convicted in a court of law and later exonerated, be it scientifically or by some other form. Over the last several years DNA testing has become available, and an increasing number of defendants have used the technology to exonerate themselves. We considered over 30 cases, eventually narrowing the list to three for extensive examination.

As part of the investigation, each of the research teams gathered contact information for people involved in the cases, pored over newspaper articles, court transcripts, and made plenty of phone calls. The teams tried to conduct as many personal interviews as possible, whether on-site, by phone, or over e-mail.

The Cases

What we saw among the cases was remarkable and should concern anyone who seeks a fair, accurate, and legitimate justice process. Police officers lied or were inept, defense counsel ignored their duties or were overloaded, and questionable witnesses were offered and allowed to testify unchecked. By no means do we believe our sample is representative of the majority of well-handled cases, but the fact that these grievous errors exist – and in the most serious of all prosecutions – raises significant questions about the accuracy and legality of the criminal justice system.

Louisiana v. Graham

The first case reviewed was that of Michael Ray Graham, Jr. Graham was a drifter charged and convicted of the 1986 murder of an elderly couple, William and Callie Frost, in their home in Union Parish, Louisiana.

Six weeks after the crime a woman, Janet Burrell, came forward claiming to have information about the murders. She said her soon-to-be ex-husband, Ronnie Burrell, and an accomplice, Graham, had committed the murders. She told police that Burrell had shown her a wallet with William Frost’s driver’s license, along with a rifle and twenty-seven $100 bills. When she confronted Burrell about the money he allegedly confessed the murders to her. Janet Burrell, however, would turn out to be an unreliable witness. After trial she changed her story twice, but even beforehand the sheriff’s office should have doubted her account; they had already found William Frost’s wallet under a pile of clothes at the crime scene.

Kenneth St. Clair and Amy Opal, a 14-year-old girl living at the St. Clair home, also claimed to have seen Graham and Burrell the night of the murders. Opal said they were counting stacks of money and that she saw Graham with blood on his clothing. But St. Clair had reason to lie, being a possible suspect himself, and Opal later admitted that she was pressured into implicating Graham. The other chief witness against Graham, a jailhouse snitch named Olan Wayne Brantley, had a long history of mental illness and pathological lies that should have been known to prosecutors. Brantley had already been committed to the Central State Hospital in Louisiana on the basis of insanity, and his record includes many instances of con games.

The forensic investigation was also full of holes. Deputies at the crime scene did not know how to use the available camera, and only one picture of the scene survived. A footprint mold was taken outside of the victim’s house but later was inexplicably destroyed. No other physical evidence was collected. The sheriff’s office did not ask for any outside help, perhaps because the Sheriff at the time was embezzling $44,000 of the parish’s money for personal use.

Many of these problems could have been addressed if Graham had been received adequate representation. Instead, he was appointed a novice real estate lawyer and a separate attorney who dropped out of the case before sentencing. The whole case reads like the plot from a bad movie, but the consequences were all too real. Michael Ray Graham, Jr. was sentenced to die and spent 14 years on death row before a new trial was ordered in March of 2000. The prosecution chose not to retry the case for lack of evidence, and Graham and Burrell were subsequently released from prison in January 2001.

Florida v. Smith

The second case investigated was that of Frank Lee Smith, a convicted felon who was arrested for the brutal rape and murder of eight-year old Shandra Whitehead in 1985. Whitehead was attacked in her home and died nine days later of head injuries brought on by a severe beating.

There were three witnesses in the case: the victim’s mother, Chiquita Lowe, and Gerald Davis. After Broward County police showed the three a photo lineup containing a picture of Smith, each identified the defendant as a man who entered the Whitehead’s home on the night of the murder. None, however, claimed to have seen the murder take place, nor was there forensic evidence linking Smith to the crime. Eddie Lee Mosley was a relative of the victim who lived only a few blocks away. He was being investigated for similar crimes at the time of the murder, but police did not even question him about the Whitehead case. Based on three witnesses that put Smith at the crime scene – and without a shred of physical evidence – Smith was convicted and sentenced to death. His execution was set for January 16, 1990.

Later investigation suggested a different killer. One month before Smith’s scheduled execution, Chiquita Lowe was finally shown Mosley’s picture. She immediately changed her story, saying that Mosley was the man she had seen outside the Whitehead home. She claimed detective Richard Scheff pressured her into identifying Smith as the suspect. Because of Lowe’s recantation, Smith’s execution was temporarily stayed.

Detective Richard Scheff is primarily responsible for keeping Smith behind bars. He told numerous lies about his investigative techniques, one lie contradicting the other, all to ensure that Smith’s conviction was not overturned. Along the way, prosecutors engaged in improper ex parte conversations with the judge who would review Smith’s case.

Frank Lee Smith remained on death row for fourteen years. In January 2000, Smith died of cancer while still incarcerated. Almost a year after his death DNA tests exonerated Smith and implicated Eddie Lee Mosley in the rape and murder of eight-year-old Shandra Whitehead.

Idaho v. Fain

Finally, we examined the case of Charles Fain from Idaho. In 1982, nine-year-old Daralyn Johnson was walking to school when a man in a mysterious car abducted her. Her body was found in a nearby creek three days later. A young girl, Christina Raass, who went to the same elementary school claimed to have seen Daralyn in a gray car the morning of her abduction, but detectives failed to take proper notes of the interview and later were unable to confirm details given by Raass. At trial her testimony was shown to be unreliable because she kept changing her story, but that was only after the jury heard her testimony.

The rape kit and autopsy would also be called into question due to missing vaginal swabs, which could have possibly contained the DNA of the killer. The chain of custody surrounding the rape kit was unclear and during the process items were lost. Pubic hairs were collected from Fain to determine whether they matched the ones found at the crime scene. Fain’s hairs were found to be consistent with the ones from the crime scene, but without DNA technology in the early 1980s that conclusion was hardly dispositive.

Fain claimed that he was living with his parents in Washington State at the time of the murders. He frequently visited the library there, and the librarian said she was almost positive she had seen Fain in the library on the day of the murders. His parents also vouched for him, saying that he was living with them at the time of the murder.

Ricky Lee Chilton, a jailhouse snitch, claimed to have had a conversation with Fain in which Fain admitted his guilt. Yet Chilton was also being represented by Fain’s attorney, which provided a conflict of interest. That attorney, Vanroy Bishop, was over-committed and could not give Fain an adequate defense. Another jailhouse informant, Bobby Allen Roberson, said Fain had bragged about how easy it was to get Daralyn in the car. Roberson, however, was related to a supervisor at the jail and received special treatment for his story.

Charles Fain was sentenced to death for the killing of nine year-old Daralyn Johnson. Years later, DNA testing was performed, and the results proved the pubic hairs could not have come from Fain. After spending 18 years on death row, Fain was finally released on September 6, 2001.

Common Problems

Despite their specific facts, each of the three cases shared common problems that led to the wrongful convictions. Investigative errors plagued all three cases. In the Smith case investigators failed to check out all possible suspects. In the Graham case, the investigators did not know how to use a camera and broke a footprint mold. Autopsy swabs from the rape kit were lost in the Fain case.

Faulty eyewitness identification also played a large factor in the three cases, particularly the use of jailhouse informants. Studies show how unreliable human memory and sight can be at times, and these were the primary evidence used to convict in the cases.

Prosecutorial misconduct was also an issue in the cases. In the Smith case, for example, the prosecutor was found to have had improper ex parte communications with the judge. But improper legal work was hardly limited to the prosecution, as two of the three defendants had ineffective representation. Inexperience and high caseloads impaired the attorneys from providing a decent defense for their clients.

Recommendations

Given the seriousness of these problems, and the stakes of capital punishment, the report concludes with recommendations and options for improving the investigation and prosecution of death penalty cases. Recommendations are presented with the complete unanimity of the ten students, while options reflect the views of at least two thirds of the researchers. These proposals cover the problems of investigative error, eyewitness identification, prosecutorial misconduct, and ineffective assistance of counsel.

With these problems in mind, the report offers the following recommendations and options to improve the investigation and prosecution of capital cases:

Law enforcement officers must be properly trained in collecting evidence, the use of equipment at hand, and must follow the chain of custody correctly.
Counties that do not conduct in-depth background checks need to establish guidelines to be followed when hiring new officers.
In certain circumstances it may be advisable to appoint an independent prosecutor to investigate allegations of police misconduct.
Civil suits are a worthy option to address investigative errors.
Prosecutors and defense counsel should validate eyewitness testimony before introducing it as evidence.
Prosecutors may wish to reconsider the use of polygraphs to check the motives and backgrounds of witnesses. This is particularly true for “jailhouse informants,” those witnesses who have already been accused of a crime and may seek to offer questionable testimony in exchange for better treatment.
As an additional check against unreliable testimony, courts may wish to perform their own evaluation of testimony before it is presented to jurors, much like a search warrant requires judicial review before evidence is seized.
In response to allegations of prosecutorial misconduct, each state should require all practicing attorneys in the state to become members of the state’s bar and follow rules of professional conduct.
The states should publish guidelines for filing a complaint against a lawyer’s unethical behavior.
Each state should require continuing legal education for active members of the State Bar.
Apart from court action, each state may wish to establish an oversight body to review complaints of ineffective assistance of counsel. At the very least, lawyers found liable for ineffective assistance should be prohibited from receiving court appointments.
We recommend the establishment of statewide standards for the qualification and performance expectations of capital defense attorneys. Among these requirements, every lead defense attorney should have previously sat “second chair” on a capital case.
Every capital defendant should be allocated adequate time and financial resources to mount a reasonable defense.
Each capital defense attorney should successfully complete rigorous training on the peculiarities and intricacies of capital defense (similar to the current Capital Defense Workshop presently held annually by the Virginia Bar Association). The State Bar should supervise such training.

·Finally, we recommend federal legislation requiring new jury instructions in all capital cases. In addition to the sentence of death, juries should be told the alternative punishments available.

Conclusion

In publishing this report, the students of the Accuracy Project hope that our hard work and dedication to justice will be looked upon with serious consideration. The solutions proposed require little additional funding and include recommendations that should find agreement from both supporters and opponents of the death penalty, for the students of the Accuracy Project reflect the widest range of views. Acknowledging the obvious problems that occur in the justice system is the first step to making a change. We hope to prod policymakers to the second step by offering simple solutions that can only better the American way of justice. It is now time to take the third step, which implements these changes so the innocent are not convicted and the guilty are forced to take responsibility for their actions. This responsibility is in the hands of people who have the power to make the necessary changes. As this report indicates, the stakes are too high not to act.

FINAL REPORT

Introduction

[top]

The death penalty is one of the most controversial facets of the American criminal justice process. Since the very beginnings of our nation, founding philosophers and policy makers have debated whether the state has the right to deprive an individual of life as punishment for a heinous crime. Whether one supports or opposes capital punishment on moral, legal, religious, or other grounds, there is no question that it is controversial. In recent years, the number of exonerated death row inmates has increased dramatically, largely as DNA technology has shown that their convictions were erroneous. As a result, the criminal justice system has come under greater scrutiny and questioning.

In response to growing concerns about the current state of capital punishment, George Mason University (GMU) has instituted the Accuracy Project, a multi-disciplinary clinic that seeks to understand and mitigate an inherent fallibility within the American criminal justice system, namely the problem of erroneous convictions. The Accuracy Project is a novel undertaking at GMU comprised of a select group of undergraduate students, all of whom are committed to ensuring equal access to justice and equality before the law.

In the fall of 2001, ten honors students from the Accuracy Project took on a significant task: to investigate cases in which a defendant on death row was later exonerated. The students were well prepared for this task, having been selected by the faculty from a pool of capable applicants. They also represented a spectrum of views and perspectives on capital punishment as well as bringing varied life experiences to enrich the seminar. The students disagreed over the merits of capital punishment, but none of them wanted to see an innocent person convicted or sentenced to die.

Working under the direction of Professor Jon Gould, Assistant Director of GMU’s Administration of Justice Program, the students were divided into three investigative teams, each responsible for researching a capital case in which an innocent defendant was convicted and later exonerated. After identifying the general problems that precipitated and perpetuated these injustices, we were able to propose recommendations and/or options for change based on our findings. It is imperative that something be done. For while the country is beginning to awake to the potential frailties of capital prosecutions, our findings suggest that the present approach to capital punishment in America is fraught with many concrete dangers.

Though this class is believed to be one of the first times that undergraduate students have undertaken clinical research on capital cases, others have paved the way. The Chicago Tribune and Northwestern University were involved in projects that led to Governor Ryan's moratorium on executions in the State of Illinois. Barry Scheck and the Innocence Project have both worked to correct what they perceive as errors committed by the system by helping to exonerate those who were wrongly convicted. The Constitution Project has also published a report containing similar calls and options for reform. Many other groups, including the Justice Project, the National Association for the Advancement of Colored People, and the American Civil Liberties Union, continue to work tirelessly to improve the system and remedy any errors they see.

Unlike some of these groups, though, our motivation was not exoneration but analysis. We had no disqualifying motive, no defendant to save from impending execution. We chose to analyze cases that were already concluded, so that we might provide a new and objective perspective on what went wrong. We based our recommendations on the principal of unanimity; if all participants agreed, then a given solution was recommended. Possible solutions that achieved majority support, but lacked unanimity, were listed as potential options to be considered by policymakers.

Now is exactly the time for such analysis. Here in Virginia the Joint Legislative Audit and Review Commission of the General Assembly has recently completed a study of the death penalty. Though the report has been attacked for failing to represent what critics believe is a racial influence in sentencing, the report’s very existence illustrates the growing national attention to capital punishment and underscores the need to analyze the administration and implementation of the death penalty in our legal system.

In this report we examine in depth three cases in which a defendant was convicted and later exonerated of capital murder: Louisiana v. Graham, Florida v. Smith, and Idaho v. Fain. The cases come from across the country, making our charge of serious shortcomings all the more troubling. These are not cases from an anomalous jurisdiction with flawed procedures; they reflect frailties from across the criminal justice system. While we certainly do not intend to cast aspersions on the vast majority of professional and capable justice officials, the findings from these cases should give all of us pause. But for the tireless efforts of those outside of the criminal justice system, these three defendants would have been executed. Indeed, one of them died in prison and only later was exonerated.

In 1986, Frank Lee Smith was tried, convicted, and sentenced to death for the 1985 rape and murder of Shandra Whitehead, an eight-year-old Broward County girl. Smith’s conviction was based largely on false eyewitness identifications and coerced testimony, the most damning of which was later recanted. On appeal, Smith encountered more formidable resistance. Richard Scheff, the detective who interrogated and charged Smith, and testified against him prior to his conviction, offered false statements aimed at keeping Smith behind bars. In December 2000, eleven months after Smith died of cancer in prison, DNA testing posthumously cleared him of the 1985 rape and murder and implicated Eddie Lee Mosley, a relative of the victim.

In 1987, Michael Ray Graham, Jr. was tried, convicted, and sentenced to death for the 1986 murders of an elderly couple in Union Parish, Louisiana. During the trial, prosecutors withheld key information from the defense, failed to produce any physical evidence, and relied on witness testimony, which later was discredited. Graham was released in December 2000 after the state acknowledged there was a lack of credible evidence linking him to the crime. Recent DNA tests proved that blood found at the victims’ home did not belong to Graham.

In 1983 Charles Irvin Fain was tried, convicted, and sentenced to death for the 1982 kidnapping, rape, and murder of Daralyn Johnson, a nine year old girl who was snatched off the street in Nampa, Idaho. A team of forensic experts from the FBI said microscopic examination - the standard test at the time - showed three hairs found on the victim’s body were probably his. In August 2001 Fain was exonerated on the basis of DNA tests that indicated the hairs were not in fact his.

In the following report we advance a particular argument about the proper workings of the American criminal justice system. We offer the view not as a gospel but rather to promote critical thought. The aim throughout is to help in the development of discourse, so as to provide a way to facilitate change. The report is a comprehensive empirical analysis of three individually distinct cases that collectively reflect the most problematic institutional impediments innocent capital defendants encounter in their quest for vindication. Although we recognize that in so doing some topics must be dealt with cryptically, this seems better than to leave them untouched on the one hand, or to try to exhaust them on the other. If this inquiry can provide a conceptual framework to consider the failings that occasionally convict innocent persons, its objectives will have been largely achieved.

The remainder of the report is divided into three sections. In part one we describe the cases we investigated and explain how the defendants were wrongly convicted. These reports were based on exhaustive research by team members, the process for which is also described. In two of the three cases we provide appendixes detailing the chain of events.

In the second section we explain the common problems from the cases that led to capital convictions and offer impassioned explanation for why these failings must be addressed. In turn, the third section presents a series of recommendations and options from team members to address these very problems. Finally we close by calling on policymakers to make these reforms a priority.

I. The Cases

[top]

A. State of Louisiana V. Michael Ray Graham, Jr.

Case Overview

On the night of August 31, 1986, an elderly couple, William and Callie Frost, were murdered in their home in Union Parish, Louisiana. According to the police, the motive for the murders was robbery, but this conclusion seems unlikely. The Frosts were far from wealthy; they sold vegetables from their garden at a roadside stand to supplement their government assistance checks. However, there was testimony at trial that the Frosts did not maintain bank accounts and preferred to keep their money in the house. Regardless of the motive, their murders shocked the community.

In this small community people were distressed by these brutal murders. With no obvious suspects and a lack of physical evidence, the recently elected sheriff was under tremendous public and political pressure to solve the crime. The people of this close-knit community sought justice and wanted to exact retribution for this transgression and attempt to re-establish their own peace of mind.

Monty Forbess, the chief deputy in Union Parish, headed the investigation. He theorized that the Frosts were both shot to death by a small caliber rifle from a point outside the house, probably through a window. The weapon was later identified as a .22 caliber, although no determination could be made whether it came from a rifle or a pistol. The deputy charged with taking pictures at the crime scene was not properly trained in how to operate the camera and only one picture from the roll could be recovered. Another piece of physical evidence, a foot print impression from outside the house, was destroyed during evidence collection.

The police also failed to thoroughly investigate all of the suspects. Michael Rogers, the Frosts’ nephew, had visited them on the night of the murders. Rogers owned a .22 caliber pistol, which was seized by the police but not tested. Another witness, Willie Wilson, stated that he saw a white car at the Frosts’ house about 11:00 pm, the same time Rogers told police he was at home. Rogers drove a cream or off-white colored car. Rogers’ alibi for his whereabouts on the night of the murders conflicted with his parents’ recollection of events. Additionally, there was a dispute between the Frosts and Rogers over property, which provides a possible motive. The local police dismissed all of this information as unimportant.

The investigation turned its focus to Michael Ray Graham, an outsider to this small community, and Albert Ronnie Burrell, a mildly retarded mechanic. Their inclusion in the investigation was based solely on the questionable testimony of three witnesses, each of whom had possible ulterior motives for identifying Graham and Burrell. The first witness, Janet Burrell, who was Ronnie Burrell’s wife at the time, was also dating Ronnie’s brother, James. She was fighting with Ronnie over the custody of their son and invented a story about Graham and Ronnie’s involvement to get her soon-to-be ex-husband in trouble. She later recanted her testimony on two occasions.

The second witness, Amy Opal, a 14-year-old girl who sometimes stayed at the same home as Graham, not only gave questionable testimony but also recanted her testimony later. Opal stated she was forced to give false information to the police and later at Graham’s trial.

The third witness, Olan Wayne Brantley, nicknamed “Lyin’ Wayne” by local police agencies, was a con man who was in the Union Parish Jail at the time. Brantley had a documented history of mental illness and had been prescribed mood-altering medications. He volunteered to testify in exchange for extra privileges and a reduction in his forgery sentence.

The prosecutor, Dan Grady, stated after the trial that he had recommended against prosecuting the cases because the evidence was too weak. However, Grady was directed by his boss, District Attorney Tommy Adkins, to proceed and present the cases to the grand jury.

Graham’s testimony before a U.S. Senate Judiciary Committee in June 2000 suggests that his lawyers were not prepared to defend a capital case. For example, Graham stated that they “didn’t investigate the snitch’s deal with the prosecution. They didn’t know the rules of evidence. They didn’t object to a jury instruction that I later learned was totally illegal under Louisiana law. And they did nothing to prepare for my sentencing phase.” He told the Judiciary Committee that his defense team consisted of two lawyers: One had some criminal law experience but had never tried a death penalty case, and the other lawyer had just graduated from law school. After Graham’s conviction, his experienced lawyer did not return to court to handle the sentencing phase of trial. Instead, Graham had to rely on the recent law school graduate to cover the sentencing hearing by himself. Both lawyers were ill prepared to handle a case of this magnitude.

Graham and Burrell were tried separately; both were found guilty and sentenced to death. After 14 years of appeals, they were both granted new trials. Judge Cynthia Woodard finally heard an appeal from Burrell in 1999 and concluded that all of the trial witness testimony used to convict Burrell was tainted. Since prosecutors used the same witnesses to convict Graham, he was also granted a new trial. In reviewing both cases, the prosecutor’s office declined to retry either case due to a lack of credible evidence - this lead to their eventual release in 2001.

Case History

Michael Ray Graham worked as a roofer in Virginia Beach, Virginia, during the summer of 1986. He met and worked with Kenneth St. Clair, a resident of Union Parish, Louisiana. After their current roofing job had ended, St. Clair convinced Graham to go to Louisiana with him. By mid-September of 1986, both St. Clair and Graham had been arrested in Downsville, Louisiana (in Union Parish) for writing fraudulent checks. The checks in question were stolen from a woman who had hired Graham and St. Clair to do some yard work in Lincoln Parish, a few miles away. Both St. Clair and Graham were incarcerated in the Union Parish Jail in Farmerville, Louisiana. Graham has described his behavior prior to his arrest by saying that he was “no angel.”

Ronnie Albert Burrell was a functionally illiterate and mildly retarded mechanic in Union Parish, Louisiana, in 1986. He and his wife, Janet, and six-year old son, Charles, lived in a trailer park a few doors away from the St. Clair family. However, Janet had been dating Ronnie’s brother, James, for the past year. At the time of the murders, she was locked in a heated battle with Ronnie for custody of their son.

The victims, William and Callie Maude Frost, lived in a three-room house in rural Downsville near the Union and Lincoln Parish line. They made a living by selling vegetables from their garden at a roadside stand in their front yard. Callie Frost was disabled and could not walk without assistance. William Frost distrusted banks and reportedly kept what money he had in a box inside the house. A few days before the murders the Frosts had received their combined monthly Social Security checks, worth a total of about $504, and William had just received an insurance payment of about $650.

The area where the crimes took place, Union Parish, Louisiana, covers an area of about 877 square miles in north central Louisiana. It is a largely white, rural and/or semi-rural area. The average educational level is 11^th grade and the average household income was only $18,000 during 1995. The unemployment rate was 12.7% at that time, and 93% of the parish residents were considered to be below the poverty line.

Sometime between 8:00 P.M. on August 31, 1986 and 10:00 A.M. on September 2, 1986, the Frosts were shot to death with a small caliber weapon. Michael Rogers, Frost’s nephew, and his girlfriend, Ruth Toney, visited the Frosts on the night of the murder. Rogers later told investigators that he arrived at the Frosts' home around dark and left at about 8:30 or 9:00 that evening. He said he arrived home 35 miles from the Frosts' home at 10:00 or 11:00 P.M. Rogers stated that he was not wearing a watch that evening and could not be sure of the time.

The Frosts were discovered by their neighbors just after 10:00 A.M. on September 2, 1986. Investigators found William Frost’s body lying on the floor near the front door of the house. They found Mrs. Frost’s body seated in a chair. Among other things, investigators found that the box containing money, as had been speculated about earlier, was missing from the house. They also found William Frost’s wallet, with money still in it, under some clothes on the bed.

There were problems with the investigation from the start. Only one crime scene photograph is available. The deputy charged with taking pictures did not know how to properly use the camera. Another piece of physical evidence, a foot print impression outside the house, was destroyed during the evidence collection process. Although blood evidence was collected at the scene, DNA analysis of blood evidence was not available to local police departments in most parts of the country in late 1986. DNA testing was being conducted on an experimental basis for a few years, but wide-scale use of the technique had not yet been established. Nevertheless, the blood evidence could have been identified by type using methods available at the time, but those tests were never performed. There is no indication that any other evidence was collected at the crime scene. In 1999, the blood evidence from the scene was finally tested; it was found to be the victims’ own blood.

As soon as the murders occurred, the investigation stalled. There were no witnesses, no leads, and no real evidence. The community was nervous. People wanted the police to take action to solve these murders. For more than a month, nothing happened.

Six weeks after the Frosts were murdered, Janet Burrell told investigator Monty Forbess from the Union Parish Sheriff’s Office that she met with Ronnie Burrell twice on the night of August 31 in order to discuss whether she could have custody of their son for a weekend. According to Janet Burrell's description, she and Ronnie arranged to meet at a designated spot at the side of a road. Janet was also dating Ronnie’s brother, James, at that time. She and Ronnie were at odds about who should have permanent custody of their son.

According to Janet Burrell’s testimony, during the first meeting around 8:00 p.m., Ronnie Burrell said that he could not stay and asked Janet to come back at 11:00 p.m. When Ronnie Burrell arrived for the later meeting at a few minutes past eleven o'clock, Janet said she sat down in the front passenger seat of Ronnie’s car. She moved some clothes which were in her way and discovered a wallet. She told investigators that the wallet contained a driver’s license and social security card, each bearing the name of William D. Frost. Janet Burrell’s credibility should have been questioned right there. This interview took place several weeks after the crimes, and investigators already knew that they had collected Frost’s wallet at the scene. Even a superficial check of known facts about the case should have sent up a red flag to investigators about the credibility of this potential informant.

Janet went on to say that she asked Ronnie about the wallet, and Ronnie eventually told her he had taken the wallet from the Frosts' house. According to Janet, Ronnie told her that he had gone to the Frosts' house, shot through their window, and had seen William Frost lying in a puddle of blood near the front door. Ronnie then reportedly showed Janet twenty-seven $100 bills. Janet also told investigators that she saw a rifle in the back seat of Ronnie’s car.

A few days later, Kenneth St. Clair, who was arrested along with Graham after the Frost murders on charges of writing fraudulent checks, told police he saw things that made him believe that Graham was involved with the murders. Although it is unclear from available information exactly what he told investigators, St. Clair's younger sister and his mother backed up his story as well.

Kenneth St. Clair may have been worried that he would become the focus of the murder investigation when Ronnie Burrell became a suspect. St. Clair lived a few doors away in the same trailer park, and investigators began to speculate that St. Clair and Michael Rogers, the Frosts’ nephew, were tied to the murders. After Graham was identified as a suspect, investigators turned their attention away from St. Clair and Rogers.

Another witness, Amy Opal, a 14-year-old girl who stayed part-time at the St. Clair trailer, said that she saw Graham and Burrell together on the night of August 31, both before and after the time when the murders apparently took place. She said that Burrell and Graham were seen on the night of the murders at the St. Clair trailer, where Graham was staying. Opal further stated that on the night of the murder she saw Graham and Burrell sitting on a couch and in front of them were two or three stacks of money and a brown suitcase. She said that at some point on that night she also saw blood on Graham.

Two weeks after Opal’s statement, Olan Wayne Brantley, a middle-aged con man serving time in the Union Parish jail for writing fraudulent checks, told Deputy Forbess that Graham had confessed to him that he had committed the murders. Brantley made a request for a lesser sentence based upon his cooperation and testimony in the murder investigation.

Despite testimony of some questionable informants, Dan Grady, the Assistant Prosecuting Attorney for Union Parish, took the investigation before a grand jury in Farmerville, Louisiana where the jurors returned a true bill. Grady would later be quoted as saying that he never thought the case had enough merit to bring it before a grand jury but that his supervisor told him to do it anyway.

Graham’s trial lasted only three days. He was appointed lawyers who had little experience with death penalty cases. The court appointed John Sheehan, a real estate lawyer, and Robert Earle, a lawyer whose general legal practice had been part of the community for the past few decades. In a recent interview, Earle explained that the Graham case was just one of several cases he was assigned to handle and that he did not have much time to devote to the case. Earle even failed to appear at several of Graham’s hearings. As further evidence of their inability to properly defend their client, there is no indication that they ever filed a motion for change of venue, which would have been important since most of the facts about the murders and Graham’s arrest would have been known to virtually everyone in town. It is unclear if the defense ever conducted an independent investigation of the facts of the case.

At trial, the prosecution did not present any physical evidence. The jury heard testimony from Janet Burrell, Amy Opal, and Olan Wayne Brantley. Because of misrepresentation by the prosecution, the jury never heard that Brantley had been found incompetent to stand trial and not guilty by reason of insanity in two previous cases. They did not hear that Brantley was currently taking lithium carbonate medication in order to control mood swings. The jury also never knew about Janet Burrell’s conflicting testimony.

In retrospect, Janet Burrell’s earlier statement that she saw William Frost’s wallet in Albert Burrell’s car on the night of the murders was not consistent with the evidence. Investigators found Frost’s wallet under some clothing in the bed. Deputy Forbess was allowed to testify that he thought that Albert Burrell had returned to the crime scene to plant the wallet after Janet saw it in his car. Janet Burrell also testified that the wallet she saw was green; William Frost’s wallet was black.

Graham, in a later statement to the U.S. Senate Judiciary Committee, stated that the jury heard illegal instructions. Under Louisiana law, the jury could find Graham guilty of first-degree murder, and therefore make him eligible for the death penalty, only if the jury found that he had the specific intent to kill. But Graham’s jury did not receive this instruction, believing instead that Graham could be found guilty of first-degree murder without having had a specific intent to kill. John Holdridge, the defense lawyer who helped Graham appeal his case, stated that since Graham never filed a direct appeal brief (the case remained in the trial court on remand from the Louisiana Supreme Court before his direct appeal was completed), the issue of the illegal jury instruction was not reviewed. Had his case ever reached the Louisiana Supreme Court in a timely fashion, the justices may well have had to consider this issue since the Court reviewed all errors in capital cases regardless of whether there was an objection at the trial level. However, by the time that Graham was released the Court had changed this rule. That means the Court would have been unable to rectify the errors of Graham’s trial lawyers had the case been appealed at that time.

Instead, Graham was released after an appeal in Ronnie Burrell’s conviction caused Louisiana District Judge Cynthia Woodard to review both cases. She ruled that the juries should have been told that a key witness against the defendants, Brantley, had been found not guilty by reason of insanity in a prior case and that his own criminal charges were reduced.

Amy Opal eventually came forward to recant her story. She told investigators that she had been coerced by the St.Clair family to lie. In fact, she said that she had seen blood on Kenneth St. Clair’s boots on the night of the murders. She also saw Kenneth St. Clair and another man counting money from a suitcase in the St. Clair trailer on the night of the murders. Ms. Opal admitted she lied at trial and has maintained that her second version of events was in fact true.

After Graham was sentenced, Janet Burrell began writing letters to Ronnie Burrell stating that his conviction was entirely her fault. She wrote to Graham saying that she would get him out of prison and that they should start a life together after he got out. Janet Burrell stated in her letters that she was coerced by her grandmother and a friend to make statements about Ronnie Burrell in an effort to get custody of her son, Charles Burrell. She believed that if Ronnie were convicted, she would stand a much better chance of getting custody. At a hearing held on July 6, 1988, as part of Ronnie Burrell’s appeal, Janet Burrell admitted that she had lied at the trial in order to get custody of her son. Janet's recantation of her trial testimony was recorded on tape. She was recorded in the office of Ronnie Burrell's attorney, where Janet answered questions and signed a typed transcript of the interview. Janet Burrell stated that she had tried to change her story before the trial and that she had been warned by Forbess that she would lose custody of her child if she changed her story.

A few weeks later, Janet Burrell changed her story once again. This time she stated that at the time she made her previous statement recanting her courtroom testimony she had actually been coerced at gunpoint. She said that Dorothy Ambrose, a friend of Ronnie Burrell's mother, threatened her. Mrs. Ambrose told her that "the Welfare" would take Janet's children away from her and that the FBI was going to charge Janet with perjury if she did not change her original testimony. Mrs. Ambrose allegedly offered Janet "immunity papers" and money in exchange for Janet changing her testimony in order to exculpate Ronnie. Janet claimed that Mrs. Ambrose was carrying a gun in her purse when she accompanied Janet to the office of Burrell's attorney. Janet's version of these events was supported by recordings made of two conversations between Janet and Mrs. Ambrose, which took place after Janet's recantation. These recordings were made by the Union Parish Sheriff's Office without the knowledge of Mrs. Ambrose.

Problems in the Case

In reviewing the Michael Graham case, our research team has identified six significant problems with how the case was handled: the prosecution improperly introduced witnesses and testimony, the prosecution withheld exculpatory information from Graham’s lawyers, Graham’s lawyers did not provide effective representation, investigators and prosecutors felt political pressure to convict someone for the crimes, the sheriff was involved in a continuing criminal enterprise during the investigation and trial, and investigators failed to obtain outside help. Any one of these problems should have created reasonable doubt about Graham’s guilt. That they combined in one case represents a serious miscarriage of justice.

a. The prosecution improperly handled witnesses and testimony

This broad category holds most of the keys to what went wrong in the Graham case. As we have already seen, the case was stagnant for weeks after the murders. Newspaper reports at the time indicated that townspeople were scared and wanted the police to do something to make the community safe from murderers. At the same time, a woman with obvious emotional problems, Janet Burrell, decided that she could benefit from making her soon-to-be ex-husband into a murder suspect. There is no evidence to support her motive beyond the custody battle for the child. Perhaps she only expected to bring Ronnie Burrell under suspicion of a violent crime, or perhaps she was actually trying to get him convicted of murder. Either way, it started a chain of events that led to the conviction of Graham and Burrell.

Kenneth St. Clair became a suspect soon after Burrell was arrested. Although there is no reason to believe that St. Clair was involved with the murders, he chose to implicate Graham. From this implication, several of St. Clair’s relatives “suddenly remembered” incriminating facts about Graham and came forward to testify.

Perhaps the most harmful thing that happened to Graham was being jailed in the same facility as Olan Wayne Brantley. Only a few days after Janet Burrell accused her husband of murder and Kenneth St. Clair began to tell his stories, Brantley decided that he might be able to “cash in” with some accusations of his own. It appears that Brantley decided that he could get his sentence reduced, or possibly get better status within the jail, if he turned in one of his cellmates as another possible suspect. Brantley has done this same thing at least four times. This unlikely turn of events became the basis for the case against Graham. The most egregious error in this part of the investigation was that the informants were never properly questioned.

Apparently, none of the investigators in this case took the time to check the informants’ statements against known facts. William Frost’s wallet is perhaps the most obvious of these errors. Investigators recovered William Frost’s wallet from the murder scene; according to the report, the wallet was found on the bed under some clothing. In Janet Burrell’s initial statement, however, she placed the wallet in her soon-to-be ex-husband’s car just after the murders and stated that he showed her money that he had apparently taken from the Frosts’ home. When investigators found William Frost’s wallet, it still contained money. All of this was explained away by Deputy Forbess at trial. Forbess testified that Burrell had placed the wallet at the scene after he showed it to his wife, but before the murders were discovered.

If this theory is true, Ronnie Burrell, who is mildly retarded, committed a double-homicide and then had presence of mind to show up not once, but twice, on the same night as the murders to talk with this soon-to-be ex-wife about taking their son for the weekend. He then would have shown his wife - the one who had been looking for some reason to take his son away from him - a wad of cash and a wallet from the crime scene and even taken the time to explain to her its source. Then, Ronnie Burrell would have taken an incredible risk; he would have returned to the crime scene - not knowing if the crime has been discovered yet, as well as being unaware when or if he would be spotted at the scene. He would have put money back into the wallet, though robbery was the supposed motive for the murders, and replaced it carefully into the house - not just inside the door, but also under some clothing on the bed - all without leaving any evidence that he was there either time.

b. The prosecution withheld exculpatory information from Graham’s lawyers

Much of the information about the prosecution’s most powerful witnesses was withheld from both the defense and from the jury. If there was no intent on the part of the prosecution to rush the case through the courts and convict Graham and Burrell, then this can only have happened due to a total breakdown in communication between the prosecution and the defense. Evidence suggests that prosecutors were aware of Brantley’s past. Investigators and prosecutors are required to disclose any part of the investigation to the defense that would be considered exculpatory for the defendant. In this case, there is considerable evidence to believe that everyone who claimed to have knowledge of Graham’s involvement in the murders also had ample reason to use Graham’s, or Burrell’s, conviction for personal gain.

If the prosecution intended to railroad Graham, then withholding information such as Brantley’s past would have been vital to their case. If the information had been released to the defense, it would have tainted the prosecution’s star witness and called all of the other witnesses’ credibility into question. If this theory is true, then the prosecutor may be added to the list of people who would have seen personal gain when Graham was convicted. Either way, the jury should have been notified of the following:

Brantley’s previous mental history

Any reasonable prosecutor should have known that he was basing his case on a highly effective con man with an extraordinary criminal and medical history. In 1978 Brantley first started writing bad checks. At about the same time, he became increasingly subject to serious mood swings, sometimes hyperactive and energetic, sometimes seriously depressed. During the hyperactive phases, he was talkative and able to tell fantastic stories; during his depressed phases he was restless, isolated, and unable to sleep. He was placed on medication, lithium carbonate and Trofranil, to keep his condition under control.

The first time Brantley was prosecuted for a worthless check, he claimed insanity. The doctors concluded that the illness prevented Brantley from distinguishing right from wrong at the time of the offense. The trial court found him not guilty by reason of insanity. Brantley was committed to the Central State Hospital in Alexandria, Louisiana for treatment in February 1981. The reports from Central State Hospital indicate that Brantley seemed well enough to be discharged within 60 days of admission, but he was held until August 1981, a period of approximately five and a half months.

In 1982 Brantley was again prosecuted for writing worthless checks. He was jailed on two counts, and, because he again claimed insanity, another sanity commission was ordered. Doctors found that Brantley was not competent to proceed. The trial court ordered him committed once again to Central State Hospital. The record does not show the exact length of this stay.

As soon as he was released from the state hospital, Brantley resumed writing bad checks. This time, Brantley was jailed once in Morehouse Parish for several months in 1985 and again in Ouachita Parish from January 1986 to September 1986.

Brantley was at trial again in Farmerville in September 1986, just days after the Frosts were murdered. He was sentenced to serve time in the Union Parish Jail. It seems unlikely that the prosecutor’s office, in such a small jurisdiction, would have a suspect like Mr. Brantley in court one week and then forget about his past only a few days later when he came forward to provide information about the Frost murders. Travis Ellison, the Assistant Attorney General for Louisiana, stated in 1986 that the Union Parish and Lincoln Parish prosecutors’ offices were supervised jointly, at that time by Tommy Adkins, and that there were “four or five” prosecuting attorneys in the Union Parish office at the time of the Brantley and Graham trials.

Both Morehouse Parish and Ouachita Parish are only a few miles away. It is equally unlikely that detectives from Union Parish would have been unfamiliar with Brantley’s past, especially since many of his fraud crimes appear to have been committed in Union Parish and the several adjoining parishes. It is not as if Brantley had suddenly appeared just in time to provide this evidence against Graham and Burrell. Brantley had been in and out of the criminal justice system in many of the parishes in north central Louisiana for the past decade.

Brantley’s use of mood-altering drugs was never revealed

If Brantley had been taking prescribed medication, then the jail that provided his housing would have been required to maintain a record of that medication. Both lithium bicarbonate and Trofranil® (propaninol) are prescription medications. Inmates are not allowed to bring their own medications into the jail; instead the jail contracts to provide these medications through a service provider, usually a local pharmacy or hospital. Thus, if Brantley’s mental illness was being treated at the time he supposedly spoke with Graham in jail, the sheriff’s office would have known that their star witness had serious limitations. Even so, Brantley’s history should have been common knowledge to the sheriff’s office, especially in a small community like Union Parish.

Brantley’s deal for leniency was never revealed

According to Louisiana Assistant Attorney General Travis Ellison, Union Parish officials have never admitted that Brantley was given any sort of deal for his information in the Graham and Burrell investigations. During an interview with the research team, however, Ellison expressed his doubts that a deal had never taken place. He cited the fact that Brantley was given special consideration in his sentencing hearings and that he acquired special trustee status within the jail. He commented that his investigation showed no basis for Brantley’s change in status if it were based solely on Brantley’s record.

This argument again raises questions about the intent of the prosecutor’s office in convicting Graham. If a deal was made legitimately, then that information should have been given to the defense counsel under Jencks. By withholding the information, it appears there was an under-the-table deal that prosecutors could not admit. Later, Brantley was arrested in Florida in 1990 for conducting a fraud scheme with another con man. The two were representing themselves as talent scouts for a country music producer and stealing money and services from people all across the South. While in jail in Florida, Brantley again told his jailers that another inmate confessed to him about a murder and offered to testify in return for leniency in his sentence. He stated to investigators that he was connected to the Gambino crime family and that he regularly met with John Gotti.

c. Graham’s lawyers did not provide effective representation

Just as the prosecution failed to perform its job adequately, Graham’s own defense lawyers failed him. We know that his lawyers, Sheehan and Earle, were assigned by the court to Graham. According to John Holdridge, one of the two lawyers who defended Graham post-conviction, Sheehan and Earle did not do much in representing Graham during the trial or sentencing. The obvious question arises: Did Sheehan and Earle have Graham’s best interests in mind when they were working on the case?

After Graham was found guilty at trial, one of his lawyers, Robert Earle, left the case. More accurately, he just stopped showing up for the hearings. In Graham’s testimony before the U.S. Senate Committee on the Judiciary, Graham stated that Earle was shaken by the outcome of the trial and failed to show up at Graham’s sentencing hearings. When interviewed for this report, Earle failed to comment about his actions at that time.

Sheehan, the other lawyer assigned to Graham’s case, had never seen a capital case to trial before. He was left to defend Graham at the sentencing hearing by himself. Even if Sheehan had more experience, it is unacceptable for his co-counsel to simply leave the case. Graham, like any defendant, deserved a more adequate and active defense.

If Earle and Sheehan had been thorough, they would have known to object to the jury instructions that stated that the jury could find Graham guilty of first-degree murder without showing specific intent to kill. If challenged at the trial level, it could have saved Graham from ever facing the death penalty. The controlling case, State v. Holmes, was decided by the Louisiana Supreme Court in 1980. If Earle and Sheehan had spent even a small amount of time conducting legal research, they could have served their client’s best interest by limiting Graham’s exposure to second-degree murder and protecting him from a conviction for first-degree murder and the death penalty. Since this error was not challenged in a direct appeal, the Louisiana courts never had the chance to rectify the error. Graham’s lawyers failed to follow through with the appeals process.

Other issues with Graham’s defense surfaced as well. Like the prosecutor’s office, Olan Wayne Brantley was probably known to the defense bar. In this small community, Brantley had been in trouble and periodically in the criminal justice system for nearly a decade before the Frosts were murdered. In September of 1986, just days after the murders, Brantley was at trial in Farmerville. Being indigent, Brantley was most likely represented by an appointed attorney - in this case by Earle or one of his colleagues. In support of this argument, Travis Ellison, the Assistant Attorney General for Louisiana, stated that in 1986 there were no more than “two or three” attorneys available in Union Parish to handle Brantley’s case.

Amazingly, the defense team never filed a motion to change venue from this small town to another jurisdiction for the trial. Graham was an outsider who had been arrested for another crime when he became a suspect in the Frost murders. There was little press coverage outside Union Parish, but the general feeling inside the parish was that everyone knew about the murders and of Graham’s arrest. Graham was identified by name in two separate newspaper articles just before his trial. At the trial, a photocopy of one of the newspapers was given to Brantley just before his testimony.

d. Investigators and prosecutors felt political pressure to convict someone for the crime

Dan Grady, the prosecutor who handled the Graham and Burrell cases from the grand jury to trial and conviction, said that after the arrests of Graham and Burrell he had recommended against prosecuting the cases because the evidence was too weak. Grady, however, received pressure from his supervisor, Tommy Adkins, who directed him "to present the cases to the grand jury and try them to avoid embarrassment." Adkins said he was attempting to show support for Larry Averitt, who was sheriff at the time. Grady stated that Adkins told him, “If it's going to be dropped, let the grand jury drop it. We're not going to drop it and embarrass the sheriff.”

e. The sheriff was involved in a continuing criminal enterprise

During the time that the Graham case was under investigation, Sheriff Averitt was involved in a fraud scheme that continued for several months after Graham’s conviction. Investigators found evidence that Averitt used $44,000 of Parish funds to purchase items for his personal use. On one occasion, Averitt submitted an invoice indicating the purchase of ten pairs of boots for his deputies. The actual items purchased included a felt hat, a bit, reins, blankets, and other horseback riding accessories. He also charged his office for clothing, tires, hunting supplies, and motel rooms. Averitt was arrested and pleaded guilty to one count of mail fraud. Thirteen other counts of his federal indictment were dropped when he entered into a plea bargain. He was sentenced and served three years in a federal prison for his crimes. Even though Averitt had no direct connection to the Graham investigation, he supervised lead investigator Monty Forbess. Indeed, Averitt’s actions give us insight into the minds of people associated with the investigation and highlight the “good-ole-boy” attitude that seemed to permeate the investigation.

f. Investigators failed to obtain outside help

Information provided by the Attorney’s General Office indicates that the FBI field office in Monroe offered its help with the Frost murder investigation when the case had stalled in October 1986. An Assistant Attorney General told the research team that the Union Parish Sheriff’s office declined the offer of assistance. At first, this refusal of assistance seemed counterproductive; the sheriff’s investigators had a murder investigation that was growing more difficult to solve by the minute. However, their refusal makes more sense when one considers that the sheriff, and possibly others inside the department, was participating in a scheme to bilk the parish government of nearly $44,000. It is likely that the sheriff declined the offer of help from the FBI in order to prevent them from snooping around in the sheriff’s back yard.

It is unclear whether state officials made similar offers, but any other offers probably would have been declined as well. The refusal to accept outside help did not focus the investigation on Burrell or Graham, but it is likely that other suspects could have been thoroughly investigated and other leads followed if more than one agency had been involved in this important and complicated investigation.

Research Methodology

This case provided some interesting challenges to standard research methods. Much of the necessary information about the case was unavailable through normal channels. The team divided the research task into five work areas: court transcripts and official publications, media coverage, the prosecution, defense, and personal interviews. The team identified potential leads by determining what aspects of the case could be explored and whether personal interviews could be conducted effec