Authors: Edward Humes
“What we find most troubling about this case is not [the prosecutor’s] initial transgression,” Justice Kozinski wrote, “but that he seemed to be totally unaware he’d done anything at all wrong and that there was no one in the United States Attorney’s Office to set him straight. . . . The overwhelming majority of prosecutors are decent, ethical, honorable lawyers who understand the awesome power they wield, and the responsibility that goes with it. But the temptation is always there: It’s the easiest thing in the world for people trained in the adversarial ethic to think a prosecutor’s job is simply to win.”
58
Prosecutorial misconduct is handled more gingerly than any other transgression in the justice system (except for judicial misconduct, the most sacred subject of all). Prosecutors are rarely cited by name for misconduct in published court opinions—Justice Kozinski’s opinion in
the heroin case being an unusual exception. Indeed, the response of the offending prosecutor’s supervisor in the Los Angeles U.S. Attorney’s Office was not to issue a reprimand, but to petition the court to remove the prosecutor’s name from the written opinion to spare him embarrassment.
59
The appeals court eventually agreed to do so.
Appeals courts routinely omit the names of prosecutors accused of misconduct, going so far as to “depublish” lower court opinions to avoid stigmatizing the government lawyers, a consideration never offered individuals charged with a crime—even ones later proved innocent. In one stunning case that occurred in early 1992, one Titus Lee Brown of South Central Los Angeles was released from prison for a murder and robbery conviction in which key evidence against him included jewelry and other items that he had supposedly stolen from the victim. Only the Los Angeles County deputy DA on the case knew—even as she argued otherwise to the jury—that those supposedly stolen items had actually been found by emergency-room workers on the murder victim’s body and had been returned to the victim’s family. It was a situation remarkably similar to the case against Pat Dunn, in which Deputy DA John Somers insisted that Sandy Dunn always wore expensive jewelry on her predawn walks, despite his own witnesses telling him the opposite. Such misstatements are routinely tolerated. In Titus Brown’s case, not only did the state appeals courts deny him a new trial, but the California Supreme Court ordered a lower court’s opinion discussing the misconduct “depublished.” And then, once again, the Ninth Circuit Court of Appeals had to step in to correct the errors of the California state courts.
“The prosecutor’s actions in this case are intolerable,”
the federal justices wrote. “Possessed of knowledge that destroyed her theory of the case, the prosecutor had a duty not to mislead the jury. Instead, she kept the facts secret . . . and then presented testimony in such a way as to suggest the opposite of what she alone knew to be true. . . . Such conduct perverts the adversarial system and endangers its ability to produce just results.”
60
Yet, even as Titus Brown got his new trial and the Los Angeles District Attorney’s Office got a black eye because one of its own lied to a jury to win a conviction, the individual prosecutor on the case was spared any public humiliation. The Ninth Circuit opinion, though brutal in its criticism, never mentioned her by name.
• • •
There is no agency that tracks prosecutorial misconduct, no outside scrutiny of prosecutors, no mechanism for policing them other than the prosecutors themselves and the chance discoveries of defense lawyers, judges and journalists—which is to say prosecutors are almost never reprimanded for their courtroom misconduct. However, a few measures that do exist suggest that prosecutorial transgressions are on the rise. In 1993, for instance, twenty Justice Department lawyers left their jobs while under investigation for charges of misconduct, a record unmatched since the birth of the nation. Overall, allegations of prosecutorial misconduct that year had climbed 78 percent from the previous year within the Justice Department, the only prosecuting agency in the country that disseminates such statistics for its lawyers. By 1995, that number had soared another 71 percent.
Meanwhile, a 1992 national study of four hundred people wrongfully convicted of death-penalty offenses since the 1940s found 15 to 20 percent of them had been
set free because of unethical and illegal conduct by prosecutors. That translates to a total of sixty to eighty innocents sentenced to death because of false or hidden evidence or other government misbehavior over a half century.
61
A congressional report published in 1993 listed forty-eight men released from death row between 1973 and 1993. All were later found to be innocent; many had been convicted as a result of prosecutorial misconduct.
62
Nationally, the justice system’s reaction to these findings exactly mirrored the reaction in Kern County when its high-profile prosecutions crumbled one by one. The tendency was not to take a hard look at prosecutors and their accountability, or to question the many laws passed in the 1980s and early 1990s that, in an understandable desire to crack down on crime, had shifted power in the justice system from judges to prosecutors. Instead, the solution has been to make it harder for convicts to present new evidence of innocence and to appeal their cases. New federal laws and Supreme Court rulings have sharply limited the number of federal appeals allowed in criminal cases, and the Supreme Court also created a standard that requires near-absolute proof of innocence before a conviction could be overturned for “factual” reasons—not even the strongest of doubts are enough in the absence of “legal” and “procedural” errors in the case.
63
At the same time, many states have imposed strict time limits on the presentation of new evidence. Texas, for example, which leads the nation in executions, set that limit at thirty days. After that period, a man convicted of murder could theoretically discover on day thirty-one a videotape of someone else committing the crime, and still be legally executed—he literally would have no
avenue of appeal under Texas law. (This scenario is not fantasy—it was presented by Supreme Court Justice Anthony Kennedy during oral arguments in
Herrera vs. Collins,
a death-penalty case. Kennedy could not have been too concerned, though—he still voted to uphold a law that would allow just such an execution, deciding that it was a matter of states’ rights.)
64
Opponents of such strict deadlines argued in vain that the time limits—intended to reduce frivolous appeals in criminal cases and to speed up executions—also give prosecutors an incentive to hide evidence. An unscrupulous prosecutor would know that he or she had to hang on for only a month after conviction before winning by default, no matter what new evidence—or evidence of misconduct—surfaced after that point.
Just such an argument was made in the case of a Virginia coal miner named Roger Keith Coleman, convicted of killing his sister-in-law and scheduled to be executed in May 1992. After his conviction and death sentence, Coleman and his lawyer learned that the police and prosecutors had covered up a host of facts useful to the defense: evidence of forced entry at the victim’s house (at trial, they had argued she had known the killer—Coleman—and let him into the house); knowledge that the knife allegedly used by Coleman didn’t fit the victim’s wounds; and information that a pair of scissors and a blood-soaked sheet had been found in a trash can outside the victim’s home by a neighbor. The blood evidence not only had been kept from the defense—it had not even undergone forensics tests. Finally Coleman’s team learned that the police hid a time card that they had obtained which would have corroborated Coleman’s alibi that he had been at work at the time of the murder.
Meanwhile, the jailhouse informant who was the chief witness in the case, and who had gotten four separate prison sentences reduced to probation for his trouble, had recanted and maintained that Coleman had never confessed to him. On top of all this, new evidence had surfaced linking another man to the crime. That man had lived in the victim’s neighborhood, had been accused of raping other women in a similar manner, and had allegedly confessed to one of his rape victims that he had killed Coleman’s sister-in-law.
In short, while Coleman did not have a videotape of another man committing the murder, he had assembled a mountain of new evidence that seemed almost as strong, with clear and seemingly irrefutable evidence of prosecutorial misconduct to go with it. Before the Supreme Court, his lawyers argued that prosecutors should not be rewarded, and an innocent man punished, for such misdeeds simply because the prosecutors successfully covered up their conduct long enough to elude a deadline. But the Court not only refused to stop the execution, it wouldn’t even consider the new evidence. Coleman’s lawyers had blundered and missed a filing deadline of their own—by three days. If the State of Virginia wanted to stick by its deadlines, the federal government would not intercede, Justice Sandra Day O’Connor wrote for the majority.
“This is a case about federalism,” O’Connor reasoned. “It concerns the respect that federal courts owe the states.”
The State of Virginia appreciated the Supreme Court’s deference; prosecutors there used exactly the same sort of “technicality” that they normally accused defense lawyers of exploiting—a deadline missed by seventy-two
hours—to deny a hearing to explore Coleman’s new evidence. And thus, at 11:00
P.M
. on May 20, 1992, Roger Keith Coleman was put to death.
65
In his dissent in
Herrera vs. Collins,
a subsequent, similar Texas death case, a disgusted Justice Harry Blackmun wrote, “The execution of a person who can show that he is innocent comes perilously close to simple murder.”
66
That opinion, however, was in the minority.
10
C
OMPARED TO THE EVIDENCE DISCOVERED—AND
ignored—in Roger Coleman’s case, the new evidence unearthed by Laura Lawhon to justify a new trial for Pat Dunn was rather threadbare. Even Laura had to admit that. Try as she might, she just couldn’t find the ammunition she desperately sought and hoped for. She never did find out about the missing report from Detective Banducci or the new check forgery charges against Jerry Coble. She could not know of Kate Rosenlieb’s full role in the case. Only the district attorney’s office and its witnesses could provide that information, and they did not do so.
What Laura did find was this: Jennifer Dunn, Pat’s daughter, belatedly recalled that when she had helped gather up Sandy’s clothes to donate to charity many months after the murder, there had been several dozen nightgowns, some worn, some unused. To Laura, this suggested that Sandy did not always sleep in the nude, as Deputy DA Somers argued. It was a small point, but a point nevertheless.
Another piece of Laura’s case came from Pat’s mechanic, who had serviced the Dunns’ Chevy Blazer and changed its two front tires the day after Sandy disappeared. Somers had called this evidence of a cover-up, and, Laura knew
from her post-trial interviews, the jurors had taken it very seriously. But the mechanic told Laura that it had been his idea, not Pat’s, to replace the front tires, which were badly worn and in danger of blowing out. Pat had been out searching for Sandy and gotten worried that the car was shaking so badly, he might not be able to keep looking. The mechanic also recounted to Laura that the oil seals on the wheels were leaking and needed replacement. When he checked them, he saw that the fluid behind them was clean. This would not be the case, the mechanic said, if Pat had recently driven on dirt roads. Thus, the service performed on the Blazer indicated that Pat could not have driven up into the mountains—over rugged dirt roads—to dispose of Sandy’s body the day before.
This was somewhat more significant than the nightgowns, but still, the problem with these “new” pieces of evidence was that they weren’t really new, for the defense had access to the mechanic and Jennifer before and during the trial. Pat’s team simply failed to ask the right questions. The judge would almost certainly reject appeals focusing on the nightgowns and auto repairs out of hand for that reason alone.
There was, however, a new witness. Donald Dean Unsell, a self-described “Dumpster diver,” recalled picking “treasures” from trash bins at the College Center mall in July 1992, looking for things to sell at garage sales, his primary source of income. One night while Unsell was at work, a man came by the mall between one and three in the morning, identified himself as Pat Dunn, and asked Unsell if he had seen his wife walking around. While awaiting trial, Pat had recalled such an encounter with an unidentified trash-picker, and the defense team had
looked for the man, hoping to use him as a witness at trial to show that Pat really had searched for his missing wife. But they had been unable to find him: Unsell, it seems, had been incarcerated himself, and hadn’t been around the mall for many months. He had not heard or read anything about the Dunn case, and so hadn’t given another thought to his encounter with a man looking for his wife.