WRIT OF HABEAS CORPUS
At the time of its founding in 1999, the California Innocent Project was only the fourth of its kind nationwide. After stints practicing criminal defense in the District of Columbia and Michigan, Justin Brooks relocated to Southern California and established the Project. It is akin to a medical residency program, where law students assist staff in vetting actual applicants and researching new clients’ cases. As a clinical institute, the Project operates out of San Diego-based California Western School of Law. About 60 innocence projects now dot the United States. While they foster a collaborative spirit, they all function as independent entities.
“We have the oddest assortment of funding,” Brooks says. Dodgeball tournaments, auctions, bake sales, sponsored mountain hikes, state and federal grants, and private donations keep the Project afloat. The single biggest donor is the Eagles rock band and its lead guitarist, Joe Walsh. When California’s Penal Code 1405 was enacted in 2001, felony defendants were granted the post-conviction right to forensic DNA testing. The state grant-funded the Project to investigate 1405 claims. “The state recognizes that we screen cases for innocence,” explains Brooks. “So they thought it was a good idea to fund us because we actually cut down on the number of litigations because we talk to the guys, we investigate the cases. And we file on very few cases.” Since its inception, the Project has garnered nine exonerations.
A writ of habeas corpus, or a habeas petition in shorthand legalese, occurs at the conclusion of the appeals stage. The typical sequence is trial, appeal, supreme court petition, habeas petition and clemency. An appeal deals strictly with record issues evident in transcripts, but a habeas petition covers non-record issues. “That’s where we live,” says Brooks. “It’s all the stuff that we find out about that was never brought up at trial.” Once a habeas petition is filed, the court determines whether there’s sufficient merit to proceed. Beyond that hurdle, the district attorney is ordered to show just cause why the petition should not be granted. The district attorney’s return can be countered by the petitioner’s response. Then, the court either grants or denies the petition outright, or it orders an evidentiary hearing to resolve the matter. Essentially, the legal standard for success is the inverse of reasonable doubt. Evidence must point “unerringly to innocence” and undermine the prosecution’s case.
On Aug. 17, 2000, an appellate court upheld Richards’ conviction. Richards completed a Project questionnaire in February 2001. As of November 2001, the Project launched its investigation into Richards’ direct claims of innocence. A 1405 motion was filed in December 2002. Judge Powers approved DNA testing for the forensic material collected at the crime scene in July 2003 through the state Justice Department. Until test results were completed, however, a legitimate habeas petition could not be filed. Despite the Project’s frequent requests for status updates, the DOJ cited the lower priority of Richards’ case for its failure to perform testing in a more timely manner. “Their higher priority is open, active cases as opposed to cold cases or a case that had resulted in conviction,” Stiglitz observes. A partial DOJ report was prepared in January 2006, but a portion of forensic work was farmed out to a private facility as well. Three crime labs and almost two years later in September 2007, all the tests were in. All told, the DNA testing took over four years. On Dec. 5, 2007, the Project filed Richards’ habeas petition in San Bernardino County Superior Court.
According to the petitioner, three principle factors led to the guilty verdict: no evidence of a third party at the crime scene; the blue tuft lodged in Pam’s fingernail judged indistinguishable from Richards’ work shirt; and the alleged bite mark’s purported match to Richards’ unique dentition. The Project debunked these foundations step-by-step. New DNA tests established another individual handled a murder weapon and DNA from hair under Pam’s fingernail implicated a third party. Autopsy photos and videotape footage showed evidence was allegedly manufactured by Gregonis to tie Richards to Pam’s assault. Finally, forensic odontologists on both sides rendered unsound scientific opinions based on limited access to information.
The People’s returns attacked the petitioner’s cause on a host of fronts. After seven years, the habeas petition should have been procedurally barred due to its untimeliness. The petitioner had a habit of beating his wife. Police interviews of Patrick, Pam’s social worker and others were enclosed as return exhibits. Richards failed a lie-detector test. Issues with the low-resolution image used by forensic odontologists were already presented ad nauseam at trial. The petitioner introduced insufficient independent evidence to meet its standard. Allegations leveled against Gregonis were baseless, and Project attorneys should have been reported to the California State Bar for professional misconduct. Several other points constituted a rehash of core prosecutorial positions introduced at trial.
Accounts of domestic abuse and polygraph reports are inadmissible in court. The Project rejected these claims as transparent ploys to prejudice Richards with what the criminal justice system deems unreliable. “Lie detector tests are not evidence in trials for a reason,” Simpson says. “Primarily, the reason is people can often fail a lie detector test when they have not committed the crime.” He says it solely measures the body’s reactive stress to questions, often accusatory in nature. “The only conceivable reason to put that in the opening brief is to poison the well,” Stiglitz added. In the end, Richards’ Project attorneys were equally undeterred by allegations of spousal abuse. Simpson was swayed by Olejnik’s support of her brother-in-law and her refutations of violence within the marriage. Brooks dismissed the hearsay because it was never substantiated. A history of severe abuse, he says, usually precipitates a murder with this level of brutality.
The People filed three briefs before a second amended return was finally accepted by Judge Brian McCarville. Among the Project’s reasons to strike the returns were the blanket generality of denials and nonresponsive treatment of specific facts. “The back-and-forth is not customary because usually the other side presents a filing that you can respond to,” notes Simpson. “It was definitely not gamesmanship or anything like that.” Grover Merritt, the lead deputy district attorney for the Appellate Services Unit, was a signee on all three motions. Stiglitz gives his unvarnished opinion of Merritt. “Some of his work product is terrible to the point of ludicrous. But he’s no fool. He’s no fool. He’s a smart guy. He knows the law, but he’s a quirky kind of lawyer.” Merritt declined to be interviewed for this story.
Ultimately, McCarville found just cause to grant an evidentiary hearing and scheduled a trial date for Jan. 26, 2009. Including oral arguments, there were six days of trial and multiple briefs over a six-month stretch.
“It was never my opinion that the injury on the victim’s hand could only have been created by the defendant and two percent of the population,” Sperber wrote in 2007. In preparation for Richards’ habeas petition, Sperber and Golden each produced recantations of their original testimonies. In particular, Sperber’s declaration included radically new assessments. He testified to scientifically inaccurate percentages at the alleged insistence of the district attorney. Percentages are inappropriate in bite-mark testimony. Risley’s closing argument misstated Sperber’s testimony. “I testified without the benefit of all the evidence that would have been relevant to a bite mark analysis,” he wrote. The single photo should never have been relied upon as conclusive evidence of guilt, he continued. Sperber never received photos of other injuries nor crime-scene images with instruments which could’ve created the crescent-shaped lesion. With the benefit of all the photographic evidence and greater experience, he could not say with any certainty that the injury was a human bite mark at all. Sperber would eventually testify that Richards’ teeth were not consistent with the lesion.
“Norm Sperber’s a big boy, okay?” Risley reacts to Sperber’s account of coerced testimony on percentages. “Norm Sperber, at the time and probably still, is one of the more preeminent forensic odontologists around the country, in the nation. Bill Richards was not Norm Sperber’s first rodeo, okay? If he was uncomfortable testifying to something, then he would’ve let me know, trust me. So, why he said that in his declaration? Hey, he said what he said.”
Gregonis’ case file includes a Scientific Investigations Division general request form date stamped May 23, 1994. The requesting officer is Gregonis. A handwritten entry described the requested examination: “3 photos given to Norm Parent to send to forensic odontologist.” This was 45 days prior to the commencement of Richards’ first trial. Fifteen years later, Stiglitz asked Parent under oath why no record exists of a corresponding odontological report having been generated as a result of this request. Parent remembered the specific form, but not the associated photos.
“I discussed it with my team and the team sergeant,” Parent testified. “It was determined basically when I was at the autopsy, the pathologist, in his opinion, thought – told me in his opinion he thought the bite mark on the arm was a canine.”
“And is this pathologist – do you remember the name of the pathologist?” Stiglitz inquired.
“At the time I believe it would have been Dr. Frank Sheridan.”
On July 19, 1994, Sheridan’s testimony about the lesion on Pam’s right hand, however, seemingly contradicted Parent’s testimony in 2009.
“So human or otherwise,” Risley asked on redirect exam, “what you observed is not even a bite mark?”
“That’s correct,” Sheridan replied. At the convicting trial, Sheridan made no reference to any canine bite mark anywhere on the victim’s body.
Golden testified that bite-mark analysis is considered highly subjective. Less open to interpretation were the American Board of Forensic Odontology’s guidelines in 1994, which dictated that all available evidence must be reviewed prior to rendering an expert opinion. At the time of his declaration in support of Richards’ petition, Golden remained the chief forensic odonologist for San Bernardino County. In 2007, he digitally scanned and enlarged a 35-millimeter slide of the purported bite mark. Golden utilized Richards’ original overlay to compare against the digital image. “The overlay does not line up perfectly with the bitemark,” he wrote, “therefore I would tend to exclude Richards as the suspected biter.” The county expert also stated that the injury was just as likely inflicted by dogs at the crime scene. He also testified he was not provided with requisite police reports or complete autopsy and crime-scene photos at the time of Richards’ prosecution.
“New technology definitively excludes Petitioner as a possible biter,” the Project authors postulated in their habeas petition. Dr. Michael Bowers is a forensic odontologist who employed cutting-edge techniques in digitized enhancement to re-evaluate the bite-mark evidence. This modern method did not exist in the forensic literature of 1997, Bowers wrote in his 2008 declaration. Alongside Dr. Ray Johansen, a forensic odontologist who also testified at Richards’ evidentiary hearing, Bowers literally wrote the book on this digital innovation. They co-authored “Digital Analysis of Bite Mark Evidence Using Adobe Photoshop” in 2000. In addition, the FBI peer-reviewed an article Bowers published on digital correction of photographic evidence. Bowers was able to rectify the angular distortion of the low-quality autopsy photo. He obtained the Styrofoam exemplars from Richards’ dentition used by Sperber in the 1997 trial. Bowers testified that a dental model of Richards’ lower arch, when pressed into Styrofoam, showed the under-erupted canine tooth made an equal indentation as the adjacent teeth. In 1997, Sperber testified the abnormal canine would not make an indentation in any material based on visual comparison. Using an overlay on the rectified photo, Bowers concluded three of Richards’ teeth superimposed onto the hand injury matched, three were misaligned.
Photos of Pam’s other injuries, he found, raised serious doubts that the lesion was a bite at all. In full analysis of the photographic record, Bowers reported two other crescent-shaped injuries on the victim’s arm and right shoulder. Sperber testified at the evidentiary hearing that he never received an underside photo of Pam’s hand to account for the opposite arch. In total, three crescent-shaped injuries were visible on the body. “A single-arch bite mark is rare,” the Project documented in Richards’ petition. “According to Dr. Bowers, because single-arch bite marks are so rare and the odds of one victim sustaining multiple single-arch bite marks are astronomical, it would be unreasonable to consider the injury on Pamela’s right hand dental in origin.”
In addition to his dental credentials, Johansen is also a tool-mark expert. When he enlarged crime scene photos, he determined a number of found objects on the property could’ve created a crescent-shaped injury. The fencing material, he found, was as congruent as Richards’ dentition. In contrast to the three other experts, Johansen compared Richards’ upper arch because he judged it more representative of the injury. He could not include or exclude Richards as the biter.
Whatever the origin of the lesion on the right hand, all four experts agreed that the single poor photo should never have been used at trial.
Pam’s right hand would also produce a severed fingertip with a tuft of blue fibers jammed in the fingernail’s crack. During autopsy, Sheridan testified that he collected Pam’s fingernail scrapings according to protocol “off the tips of the fingers, fingernails, looking for any type of trace evidence, anything that might be adhered to the nails.” No tuft was reported in his autopsy notes. In 2008, Gregonis composed a declaration to address specific issues raised by Richards’ habeas petition. In it, he stated the fibers were first discovered upon stereomicroscopic examination. “Since these fibers were eventually discovered upon a microscopical examination and the fact that there was blood and other debris on the fingertips (Item C-14) it is reasonable to conclude that the fibers would not have been seen with the unaided eye during the autopsy or during the actual severing process.” In the habeas petition, however, the Project indicated that several autopsy photos of Pam’s right hand before and after cleaning showed no tuft.
At Richards’ first trial, Gregonis testified he saw the tuft with his naked eye in the crime lab during his first examination of the severed fingertip.
“Were the blue fibers visible with the naked eye?” Risley asked Gregonis in the third trial.
“Yes, they were.”
When confronted by Stiglitz during the evidentiary hearing about these testimonies versus his later declaration, Gregonis stated that the tuft was visible with the naked eye as well as under his microscope. In retrospect, Gregonis stated he was unsure whether he first inspected the tuft visually or microscopically.
After the autopsy, it became Craig Ogino’s duty to visually and microscopically examine the fingernail scrapings. In addition, when asked about his actions he testified the fingernails themselves were checked for smaller trace evidence. Under stereomicroscopic power, Ogino failed to document any such tuft.
Bowers obtained an autopsy photo showing Pam’s right hand after it was cleaned, before the severing of fingertips. It was a close-up of the fingernail with the purported tuft wedged into a fracture. With Photoshop, Bowers adjusted the image’s color saturation. The enhanced image did not isolate a single blue pixel. Conversely, Bowers performed the same saturation test on a still video image taken from Gregonis’ extraction of the severed fingertip’s tuft. A blue, Z-shaped tuft was visible. Bowers testified he didn’t see the fibers in four autopsy photos he examined. “Considering the size, the amount of material removed or fibers removed from the crime lab at the time of the crime lab examination, I would expect the autopsy photos to have shown blue, but they did not.”
The visual discrepancy between autopsy photography and crime-lab footage led the Project to allege fiber evidence against Richards was planted. “Because of its location in the nail and because the fiber did not appear until Gregonis had access to the fingertips and Petitioner’s denim shirt, there is no innocent explanation for the belated appearance of the fiber.”
“It’s bullshit,” Bradford reacts to the Project’s allegation of planted fibers. “Those types of claims sicken me because I know that shit doesn’t happen here…If there’s actual righteous evidence, things that need to be looked at, we should do that. But for people to make claims of planting evidence….it disgusts me. It’s bullshit.”
Gregonis excised a fiber sample from Richards’ work shirt to compare against the tuft of blue fibers. When his comparison was finished, Gregonis discarded the defendant’s fiber sample. Dr. Patricia Zajac, a former criminalist for the Alameda County Sheriff’s Department’s Crime Lab, also testified on Richards’ behalf. In Zajac’s report prepared for the habeas petition, the criminalist stated that unless items are consumed in testing, it is not standard laboratory practice to destroy or discard evidence.
Testimonies in two death penalty cases were cited in the habeas petition to further impugn Gregonis’ credibility. Before Richards’ conviction, Gregonis testified about urine stains found on a homicide victim’s clothing as a serology expert in People v. Coffman and Marlow. In 1989, Gregonis testified that tests to identify the urine source were inconclusive. Consequently, a defendant could not be excluded as a donor of the urine stains. Zajac studied Gregonis’ bench notes, lab results, testimony and a urine-stain report. In her declaration, Zajac concluded that Gregonis’ tests eliminated Marlow as the source of two stains that yielded definitive results, in contradiction to Gregonis’ sworn remarks.
In 1985, prison escapee Kevin Cooper was convicted on four counts of capital murder and one count of attempted murder in the 1983 stabbings of four members of the Ryen family and their houseguest in Chino Hills, California. Only the family’s eight-year-old son survived. Gregonis performed blood work on the case. From death penalty opponents to prominent legal advocates like Alan Dershowitz, Cooper has become a galvanizing symbol of injustice for his supporters. The high-profile case has been steeped in controversy for years.
When the U.S. Ninth Circuit Court of Appeals upheld Cooper’s conviction, Judge William Fletcher, joined by four of his fellow circuit judges, penned a remarkable 101-page dissent in May 2009. In it, Fletcher openly cast doubt on Gregonis’ testimony about bloodstained paint chips, labeled A-41, taken from a wall in the Ryen home. Gregonis’ lab work matched A-41 to Cooper through blood characteristics.
“There is a strong likelihood that the results of the blood tests performed on A-41, presented at trial, were false evidence,” Fletcher wrote. “There is also a strong likelihood that state actors tampered with A-41 to ensure that it would generate inculpatory results when Cooper’s post-conviction DNA testing was conducted in 2002. Gregonis delayed most of his testing of A-41 until he had information about Cooper’s genetic profile – that is, until he knew what he had to match.” In addition to questionable lab practices, Fletcher accused Gregonis of falsifying evidence before scrambling to conceal the wrongdoing. “When the results of Gregonis’s tests on A-41 were initially inconsistent with Cooper’s expert’s results for a known sample of Cooper’s blood, Gregonis altered his lab notes and claimed that he had misinterpreted his results.”
In his May 2008 response to Richards’ habeas petition, Gregonis defended his professional record. “William Richards’ attack on my credibility by reference to my involvement as a criminalist in the capital cases of Kevin Cooper, Cynthia Coffman, and James Marlow is baseless. For the past 23 years Cooper has attempted to reverse the judgment against him by attacking the credibility of those associated with the prosecution of his case. I am unaware of any court that has been convinced by such claims in the cases of Cooper, Coffman, or Marlow.”
As part of Richards’ 1405 motion, in 2006 DNA tests were conducted on the stepping stone wielded against the victim. Unknown male DNA was detected in a mixture of Pam’s blood. Because the greater quantity of DNA belonged to the victim, the male donation was termed the minor component. There were at least two minor donors. Richards’ profile was not consistent with the unknown male DNA. In two places on the stone, one-tenth and one-sixth of the respective DNA mixtures belonged to unknown third parties. In his 1993 case notes, Gregonis isolated three areas as likely candidates to contain a suspect’s DNA. Minor contributions were found in two of his three designated areas.
Court personnel’s handling of the stepping stone, as well as attorneys talking over it, were explanations proposed by Gregonis for the unknown DNA. The People also speculated the male DNA could’ve been deposited at any point before the crime, and it wasn’t attributable to any one person. Other stepping stones that formed the pathway to the motor home were also tested. No unknown DNA was transmitted to these stones. The Project argued that this indicated the bludgeoning stone was relatively free of historical DNA.
“The staining patterns that are on there are definitely consistent with that being one of the items that came into contact with her skull,” Gregonis testified about the stepping stone during the first mistrial. As he described the prosecution’s carrying scenario, he said that Pam was brought outside and “struck at least twice, probably first by the stepping stone.”
“In your opinion,” Risley asked Gregonis in the last mistrial, “the stepping stone was used to strike the victim’s head, as well?”
“I believe so, yes.”
“There was blood spatter and pooled blood on the stepping stone,” Gregonis wrote in his 2008 declaration. “Based on the presence of the pooled blood, there was the possibility that the stepping stone had been used to hit the victim…The evidence clearly pointed to the cinderblock, a different item found at the crime scene, as the murder weapon.” When confronted during the evidentiary hearing, Gregonis conceded there may have been a difference in his prior testimony. “In this statement as part of my testimony it appears that I was more definite, yes,” he told Stiglitz.
In addition to the DNA on the stepping stone, mitochondrial DNA, or mtDNA, was retrieved from a blonde hair found amongst the victim’s fingernail scrapings. Measuring two centimeters or just under an inch in length, its mtDNA belonged to an unknown third party. Also, the blonde hair had a telogen root. Defined by the mature growth stage of the hair, a telogen root is either ready to fall out or naturally shed. An anogen root is differentiated as a younger, living root.
Ogino, the county criminalist, determined that all the fingernail scrapings taken from Pam were historical, or deposited in the course of daily life. Because the hair had a telogen root, the district attorney argued that it was naturally shed, as opposed to an anogen which is usually forcibly removed. Gialamas did not testify at the evidentiary hearing, but he disagrees with the basic premise that there’s a greater likelihood of finding a hair with an anogen root lodged under a fingernail as the result of a struggle.
“It depends on the struggle and what the issues are,” he says. “It’s very difficult to say…Hairs, even though microscopically small, they’re still a large item to stay under a fingernail. How many times have you just looked under your fingernail and found hair there? Probably not often, rarely if at all. So the fact that it’s even there is an unusual event. So to take an unusual event and make it more unusual than that is a very difficult statistical proposition.”
Zajac had fundamental issues with classifying the hair on the People’s grounds. Most hair collected as evidence has a telogen root, she testified. The type of root had no relevance in her final analysis. Four criteria led Zajac to conclude the hair was non-historical. The length of the hair was such that Pam would likely have noticed and removed it. The crime scene’s location was not a place where one would find a lot of hairs. Hair was found under the fingernail, not simply on its surface, so it would take some kind of action to get it in the place it was found. Lastly, the violent nature of the crime suggested it was deposited during the struggle.
In a post-hearing brief, the Project stated Pam was a woman whose feminine grooming habits were tied to her line of work. “She was a woman who obviously paid attention to her nails – they were painted.” They were also acrylic. “In addition, she was a waitress and was likely not in the habit of serving food with dirty fingernails.” If it was historical, it was argued, the blonde hair would’ve been dislodged as she washed her hands.
No criminal database exists for either the hair’s mtDNA or the stepping stone’s Y-STR DNA profile. Because they have unique profiles, third-party DNA from the stepping stone and hair are incomparable and cannot be narrowed to a single donor. Patrick’s DNA has never been compared against the third-party samples.
In terms of alternate suspects other than Patrick, Hal Smith and Project attorneys were essentially drawing on empty. However, Smith’s first intuition was serial killer Angel Reséndiz.
Known by the infamous moniker the “Railroad Killer,” Reséndiz was a Mexican drifter who targeted residents along train routes across American and Mexican borders. With no fixed abode from the mid-’70s to the late ’90s, in the U.S. alone he hopped freight trains and stole cars from Macon, Georgia, and San Bernardino, California, to St. Louis, Missouri, and Houston, Texas—covering numerous points between. Law enforcement confirmed Reséndiz committed nine murders between August 1997 and June 1999 alone, but he’s been linked to at least 10 more open homicides, as early as the ’80s. However, he was particularly active throughout the ’90s. He was apprehended by a Texas Ranger in 1999, convicted in 2000 for the bludgeoning death of 39-year-old Claudia Benton and executed in 2006 by lethal injection in Huntsville, Texas.
Pamela Richards’ slaying fit his modus operandi. Reséndiz preyed on random victims by train tracks in rural locales, perpetrated extraordinarily violent killings often with blunt instruments, robbed very little in property, raped sometimes as a secondary motive and fled after overkill during predawn hours. According to news reports, former FBI profiler John Douglas said that early in Reséndiz’s homicidal arc, he inflicted blunt-force trauma with weapons of opportunity found at his crime scenes. Before he was taken into custody on the capital murder charge, he had compiled a lengthy criminal résumé. Pam died on Aug. 10, 1993. In March 2000, the U.S. Justice Department’s Office of the Inspector General produced a special report on Reséndiz’s case. It attempted to track the serial killer’s whereabouts through a diverse set of law enforcement records. On June 9, 1993, Reséndiz was arrested after evading police in Carson City, Texas, while operating a vehicle stolen two days earlier in Missouri. He was sentenced to time served on the misdemeanor charge of eluding capture after Missouri authorities declined to extradite for auto theft. It is unknown whether the Immigration and Naturalization Service deported him upon release on July 8, 1993. No other police record is extant until Dec. 7, 1994, when Reséndiz was arrested for operating a stolen vehicle in Albuquerque, New Mexico. The car was stolen a day prior in Arizona.
On Aug. 18, 1995, however, Reséndiz was arrested by railroad police in San Bernardino for trespassing, carrying a loaded firearm and receiving stolen property. Seven days later he was sentenced to 30 days imprisonment. In addition, Reséndiz was the prime suspect in a fatal beating of a transient in a Colton rail yard in July 1997—some 36 miles south of the Richardses’ Hesperia home.
In January 2006, Brooks traveled to Texas to interview Reséndiz on death row. Brooks says that Reséndiz exhibited unmistakable signs of psychosis, and Reséndiz was unsure whether Pam numbered as one of his victims. He said the crime looked like something he would’ve done. Her screen print t-shirt with “Shady Lady” written across the front was disrespectful, he said. And he was well acquainted with Summit Valley, even able to recite train stops throughout the entire region. Reséndiz said that because he was tripping on large quantities of LSD at the time, he couldn’t remember many of his murders.
“After you spent a lot of time [with him],” Brooks recounts, “you realize just how crazy he was. He was just a pure sociopath. He talked about killing one person after the other like he was talking about eating a ham sandwich. So these things weren’t memorable to him because he said he never thought about them 10 minutes later.”
Reséndiz’s dental impressions were collected in October 2005 to reference against Pam’s crescent-shaped lesion on her right hand. The death row inmate could not be excluded as the biter. The Project obtained a DNA sample from him, but he was excluded as a possible donor of tested hairs under Pam’s fingernails. “Some of them [hairs] came back inconclusive and some of them didn’t come back to match,” Simpson explains. “It wasn’t very probative. Since he was in custody, we weren’t really able to get anything more definitive than that.” During the habeas process, the district attorney pointed to these results as emblematic of Richards’ inability to shirk responsibility for his wife’s murder.
At the evidentiary hearing’s onset, Merritt floated the possibility of a plea deal for a lesser charge in exchange for time served. “I talked to Bill about it a number of times and I explained to him that these hearings are a crapshoot,” Stiglitz says. “We thought we had a great case…but you just never know. And if the prosecutor was offering him the opportunity to walk out of prison tomorrow, he ought to be thinking about that.” Richards didn’t blink. “He said, ‘I’ve spent the last several years in prison for a crime I didn’t do, and I’m not going to say I did just to get out of prison.’ That told me all I needed to know.” Yet that wasn’t the first time a plea bargain was presented to Richards. He says he received plea offers in the ’90s for manslaughter and time served. “I would have gone to prison, got a number and been paroled,” he says. Again, it was never a serious consideration. While Risley’s memory is unclear, he says it wouldn’t surprise him if prior to the preliminary hearing or first trial a reduced second-degree charge was proffered. In any event, Richards rejected every plea avenue available.
“When you’re looking at a life term in California,” Smith adds, “punting is always an option you have to look at sincerely. Because life in California means life.”
Near the evidentiary hearing’s midway point, the Project contemplated a prosecutorial misconduct complaint against Risley. According to Stiglitz, Merritt would’ve preferred it had happened sooner. “He wanted to get off the case,” Stiglitz remembers. “He had no interest in litigating this case.” When the prospect of a misconduct claim was formally raised, it was Merritt’s magic bullet to recusal. Despite the fact that the claim was eventually abandoned, on the trial’s fourth day, after the testimony of six forensic experts, new counsel was appointed within the appellate division. Merritt’s defense of Risley would not only have been unpalatable, it presented an inescapable conflict of interest.
According to federal court records, in October 2004 Merritt was served order-of-dismissal papers from the district attorney’s office amid allegations that he leaked two sets of sensitive documents to a local media outlet. One was a death penalty memo, the other contained Merritt’s research into statute-of-limitations laws tied to an open investigation of former San Bernardino County Sheriff Floyd Tidwell. The county’s top lawman, Tidwell was accused of pilfering 523 guns from evidence rooms during his tenure from 1983 to 1991. (Later, he would plead guilty in exchange for misdemeanor charges, probation, the return of over 100 weapons and a $10,000 fine.) In a separation report filed with Merritt’s termination paperwork, the prosecutor’s most recent work performance evaluation was marked “exceeds standards.” In fact, Merritt was instrumental in the passage of California’s controversial three-strikes legislation, which he argued before the U.S. Supreme Court in 2002. Nonetheless, the order enumerated at least six violations. After Merritt’s immediate appeal, the county’s Civil Service Commission sided with the prosecutor, citing deficient grounds for the termination. However, Merritt was not reinstalled as Appellate Services Unit chief. Instead, he was reassigned as a courtroom prosecutor.
Disgruntled with what he considered a demotion, Merritt subsequently sued for wrongful termination in U.S. District Court, naming the county, District Attorney Michael Ramos and Assistant District Attorney Michael Risley in the federal lawsuit. Nine complaints included conspiracy to violate civil rights, wrongful termination, defamation and intentional infliction of emotional distress. In a September 2006 deposition, Merritt presumed that Risley, with Ramos’ tacit consent, leaked the documents in order to frame Merritt, thereby delivering a coup de grâce to Merritt’s chances of opposing Ramos in the 2006 election for district attorney. Risley, in his own declaration, echoed his office’s position that Merritt bore responsibility for the leaked information. A no-fault settlement was ultimately reached in December 2006, which reinstated Merritt to his old post. County Public Information Officer David Wert quantifies the cost to local taxpayers at $444,303, including $146,403 in outside legal expenses and $297,500 toward Merritt’s settlement package.
In retirement, Risley does not have fond remembrances of Merritt. “We can cut to the chase,” he says. “Grover Merritt thinks I’m a liar. And I think he’s a liar. And that’s the way that it is. He recused himself because he thinks I’m a liar. And he’s entitled to his opinion. I think he’s a liar, and I’m entitled to my opinion.”
In Merritt’s absence, the evidentiary hearing moved forward. Post-hearing briefs were filed. Oral arguments were presented. At the conclusion, Judge McCarville addressed both parties. While the petitioner failed to establish Gregonis either committed perjury or planted evidence, the tuft of blue fibers raised factual concerns in the judge’s mind. But it didn’t tip the scales.
“The Court finds that the evidence with respect to the bite mark analysis and the DNA analysis and the hair analysis has established, taken together, that there was a – that there did exist and does exist a fundamental doubt in my mind as to the accuracy and reliability of the evidence presented at the trial proceeding…The Court finds that the entire prosecution case has been undermined, and that the petitioner has established his burden of proof to show that the evidence before me presents or points unerringly to innocence.”
On Aug. 10, 2009, exactly 16 years to the day after Pamela Richards was murdered, William Richards was exonerated.
“It’s hard to decide what I want to do first,” he pondered from behind bars after McCarville’s decision. “I will probably settle for some good food. Ideally, I would move somewhere that nobody knows all of this and try to build a new life. But it’s unlikely that will happen.”