Investigating Wrongful Convictions: Getting Started

A Resource for Journalists

By Edward Humes,
author of Mean Justice: A Town’s Terror, A Prosecutor’s Power, a Betrayal of Innocence

You’ve just received an emotional letter from a woman who swears her son has been convicted of a murder he didn’t commit. The allegations of official mistakes or misconduct she describes sound rational, even plausible. But old clips on the case suggest evidence of guilt is strong. How do you decide if there is something worth pursuing? Was the prosecution case open and shut, or full of holes?

One approach is to look for clusters of the warning signs of injustice — factors that have often appeared in documented cases of wrongful conviction. The presence of several of these factors in a single case can help you decide if there is a story — and a possible wrongful conviction — to pursue. And if it seems there may be a story there, this list can provide starting points for in-depth digging into police reports, court records, autopsies, forensic reports, and your own interviewing of witnesses and others.

1. No smoking gun: The absence of physical evidence linking the defendant to the crime/crime scene. (Examples: No blood where it should have been; no gunshot residue where it should have been; no fingerprints or the presence of unknown fingerprints at the scene; fruitless searches of the suspect’s person, house or car.) For the reporter, this is the threshold point to consider — the presence of strong, unambiguous physical evidence that cannot easily be explained innocently is a good reality check on claims of wrongful conviction. However, the absence of such evidence, or the existence of a reasonable interpretation of that evidence consistent with innocence, opens the door to consider other factors on this list.
2. Jailhouse informants who are freed or who receive leniency from prosecutors, then testify that the defendant incriminated himself while incarcerated. Probing the history of such witnesses — and whether that history was disclosed by the prosecution (see item #5 below) is critical, particularly when such witnesses are key to the state's case or when they materialize late in the game.
3. Leniency deals: Witnesses or codefendants granted leniency in their own criminal cases in exchange for testimony, particularly when there is no corroboration of their stories. Pay attention to witnesses whose stories improve (from the prosecution point of view) with time — i.e., become more detailed and more incriminating when compared to initial statements.
4. Theory before evidence: In complex and difficult cases where the identity of the perpetrator is not immediately apparent, model criminal investigations start with a thorough gathering of evidence before focusing on a single suspect. Others start with a likely suspect, then gather evidence that fits the official theory — the approach most common in wrongful convictions. Such investigations readily — sometimes uncritically — accept witnesses and evidence that fit the theory. But other seemingly credible witnesses or evidence may be ignored or minimized if they fail to incriminate — or tend to exonerate — the chosen suspect. Such witnesses — even ones initially viewed as credible by authorities — may be subjected to criminal investigation, asked to change their stories, or asked to concede that they might be wrong. Sometimes such information about witnesses favorable to the defense is buried or kept hidden (see #5 below). These officially “wrong” witnesses and pieces of evidence can provide most fertile ground for investigative reporters. Note: This selective view of the evidence by the authorities almost always arises from genuine belief in the guilt of the suspect (as opposed to a corrupt desire to frustrate the search for truth). The tendency is to view even the most powerful and convincing contrary information as an obstacle to overcome, a mistake, a lie, or a defense attorney trick.
5. Questionable prosecutorial tactics or misconduct: The withholding of evidence helpful to the defense is the most common problem in this category — and the most difficult to detect, as prosecutors control the flow of information in criminal cases, with virtually no oversight. Other problems somewhat more visible include the use of false evidence or testimony; the use of surprise evidence or witnesses disclosed at the last minute; courtroom misbehavior; and substantial changes in the prosecution’s theory of the case over time or contradictions between multiple trials.
6. Use of questionable scientific analyses, particularly involving physical (non-DNA) comparisons of hair samples; determinations in arson cases as to whether accelerants were used to start fires; early DNA tests (RFLP testing) that may overstate the likelihood of a match; use of scent dogs or “scent machines;” unproved scientific methods (example: the “wink test” to detect sexual abuse in children, responsible for dozens of false convictions).
7. Questionable eyewitness identifications in which the witness initially expressed uncertainty or described a suspect different in appearance.
8. Confessions in which coercion by the authorities is asserted (of particular concern with juveniles — see false/tainted confessions in Chicago’s recent Ryan Harris murder case and San Diego’s Stephanie Crowe murder case).
9. Witness statements in which coercion by the authorities is asserted (of particular concern with child victims of sex crimes where medical evidence fails to corroborate the allegations — see the Bakersfield, Wenatchee, Amirault, Little Rascals, McMartin and other flawed molestation prosecutions, where coercion and “brainwashing” of victims by the authorities has been cited by appellate courts as grounds for overturning convictions).
10. Existence of viable alternative suspects not thoroughly investigated by police, prosecution or defense. Potentially fertile, but often very difficult, ground for reporters.
11. Little apparent work by the defense: Look for few motions filed, few witnesses called, little or no investigative work prior to trial, little cross-examination of prosecution witnesses, absence of expert witnesses to counter experts called by the prosecution.
12. Emphasis on “state of mind” or “suspicious conduct.” Such evidence is so subjective, it can always be viewed as evidence of guilt. (Examples: Defendant was insufficiently grief-stricken or the converse, defendant was uncharacteristically emotional, so must have been faking it; defendant was angry and uncooperative with investigators or the converse, defendant was too willing to talk, must have been trying to lie his way out of trouble.)

A Final Note: These factors are just a starting point. The mere presence of one or more of these indicia proves nothing — many cases that result in appropriate guilty verdicts may include some of these elements. Indeed, one would be hard pressed to find a criminal case that did not include at least one of these elements. Further, any of these factors can be rendered insignificant by other evidence — such as use of a jailhouse informant whose testimony is corroborated by reliable DNA analysis (as opposed to an informant's testimony when physical evidence is absent, a combination often present in cases of wrongful conviction). But constellations of these elements — in the context of each unique case — can indicate possible wrongful convictions worthy of further reporting. One last caution: Proving a wrongful conviction should never be confused with the much more rare and difficult task of proving someone innocent.

---

Investigating Wrongful Convictions: Case Study

The People of Kern County vs. Patrick O’Dale Dunn
A Resource for Journalists

From MEAN JUSTICE, by Edward Humes

Background: Patrick Dunn was sentenced to life in prison without possibility of parole in June 1993 for murdering his wife, Sandy. Financial gain was the alleged motive. He reported her missing after she failed to return from a predawn walk. Her body was found ten days later in a remote desert grave sixty miles from home. No physical evidence linked Dunn to the crime.

What the prosecution relied on:

Suspicious conduct showing a guilty state of mind. Examples: Dunn bantered with the sheriff’s dispatcher who took the missing persons report; he failed to tell acquaintances that his wife was missing; he asserted that his wife suffered from the early onset of Alzheimer's Disease, something Sandy’s friends denied; the first person he called about the disappearance of his wife was his civil attorney, in an alleged attempt to gain illegal control of Sandy's money.

Financial motive to kill. Sandy Dunn had millions from her first marriage; Pat Dunn, a retired high school principal, had no assets. One witness said Sandy planned to divorce Pat, leaving him penniless; a neighbor testified the Dunns had an angry exchange of words on the day she disappeared; a close friend swore the Dunns often had vicious arguments.

Opportunity to kill. It was alleged that Sandy always wore expensive jewelry on her walks and always wore her glasses. But after her death, none of her jewelry was missing and her glasses were in the house. This meant she must have died at home, the prosecution argued, otherwise there would be missing jewelry and glasses. Thus, only Pat Dunn had the opportunity to kill her.

Informant. A career criminal and drug addict named Jerry Coble, the only prosecution eyewitness, swore that, while driving at 1 a.m. looking for a lost bundle of heroin, he inadvertently spotted Pat Dunn dragging his wife’s shrouded body from the house and throwing it into the back of a pickup truck. Coble swore this was the only time he was in the Dunn neighborhood. Coble got a plea bargain guaranteeing his freedom; he had been facing six years in prison. He claimed to be a reluctant informant who came forward only when it appeared Dunn was getting away with murder.

What the defense relied on:

Absence of physical evidence. No blood in the home, despite microscopic search by the authorities and stab wounds that produced copious bleeding; no weapon or sign of a struggle; fiber evidence found on the body could not have come from Pat Dunn's vehicles or house.

Impeachment of the informant’s story. Jerry Coble initially described vehicles in the Dunn driveway that were not there on the night of the murder, but were there one month after — following the police seizure of Dunn's very distinctive and huge vintage Cadillac (Coble never mentioned seeing this car; instead, he described a car Dunn rented after the seizure. Later, in front of the jury, he changed his account to describe the correct cars.) A month after the murder, a visiting friend saw a car repeatedly drive by Dunn’s home, its driver peering into the house; the friend followed the car and wrote down the license number. In court, the friend identified the driver as Coble; the license number matched the informant’s mother’s car. The defense asserted Coble cased the “crime scene” after the news broke that Pat Dunn was a suspect, so that he could cook up phony testimony to leverage a plea bargain. Coble's brother, Gary, another career criminal, testified Jerry admitted framing Dunn.

No motive. A financial planner met with the Dunns a few hours before Sandy disappeared. Sandy asked him to prepare a trust giving Pat joint control over her money and property, and that would allow him to inherit everything without probate. She disappeared before these documents could be prepared and signed by her. This showed no divorce was contemplated, and that Pat had a strong motive for Sandy to live — if for no other reason than to sign those papers. Friends described the Dunns’ relationship as loving, with no talk of divorce. The neighbor who witnessed the loud argument on the night of Sandy’s disappearance admitted being partially deaf; he never actually heard what was said.

No opportunity to kill. The secretary for the Dunn's accountant swore she had a telephone conversation with a disoriented and confused-sounding Sandy on the morning after she disappeared (eight hours after Coble swore he saw her body being dumped). Evidence also showed that Sandy owned multiple pairs of eyeglasses and that one of them could be missing.

GUILTY VERDICT: Jurors were swayed by the friend who swore Sandy wanted a divorce, and by Dunn’s “suspicious” conduct, particularly the phone call to his lawyer.

New findings during research for Mean Justice
(Principal sources: Court records, police reports, interviews with witnesses):

• Withheld evidence. A 1991 sheriff’s report documents how Jerry Coble had begged a Kern County sheriff’s detective to make him an informant — before Sandy Dunn died. Coble said he would do “whatever it takes” not to go to prison, and he offered to set up a variety of criminals in exchange for freedom. The detective rejected this offer because of Coble’s known lack of credibility. This report was never given to the defense. The detective who rejected Coble as an informant was kept “out of the loop” when Coble was offered freedom in exchange for testimony against Dunn; that detective now says he would have objected strenuously to such a plea bargain had prosecutors asked him.
• Withheld evidence II. Before Pat Dunn’s case came to trial, Coble attempted to cash a forged payroll check using stolen ID. This violation of his plea bargain, probation and parole should have landed him in prison. Yet he remained free. Authorities sat on the case until after Dunn was convicted. Then it was mysteriously dropped. The defense was never told.
• Withheld evidence III. Local officials who knew the Dunns — and who testified against Pat at trial — secretly lobbied prosecutors to charge Pat Dunn with murder, at one point threatening to launch a media campaign against the DA if he failed to comply. This was kept secret; at trial, these same officials were presented as disinterested, unbiased witnesses. A journal kept by one of them chronicled this lobbying, as well as the lead homicide detective's role in orchestrating the campaign. Portions of the journal were disclosed to the defense — minus any mention of the lobbying.
• Withheld evidence IV. The Dunn’s housecleaner told the trial prosecutor Sandy never wore jewelry during her walks. Not only did the prosecutor neglect to disclose this, he told the jury that the opposite was true, and used the lack of missing jewelry as powerful evidence of Dunn’s guilt.
• Scientific evidence known to prosecutors but not to the defense. Sandy's body was partially buried. The DA knew from a recent case that the insects found on autopsy could only be present if the body were completely exposed to the elements in daylight before burial. But the prosecution's timeline in the case required Dunn to stab Sandy, drive off with the body, bury it and return home entirely in darkness.
• Witness recantations/new testimony. The one witness who claimed Sandy planned to divorce Pat now says she was mistaken. The other friend who spoke of vicious fights between the Dunns (and who led the lobbying campaign to have Dunn prosecuted) now admits she failed to mention an even more salient fact: that Sandy said she would never divorce her husband. Several other witnesses have come forward with stories of Sandy's failing mental faculties.
• Errors. A review of trial transcripts shows numerous false statements and errors by the prosecution during the trial, the worst of which concerns the legal basis for admitting the testimony of Dunn’s civil attorney. The prosecutor had argued, and the trial judge agreed, that Dunn asked his attorney to do something illegal, negating any claim of attorney-client confidentiality. A review of testimony in the case, however, shows no evidence of any illegal requests by Dunn. The attorney should not have testified, and the prosecutor should not have been allowed to tell jurors that Dunn solicited his attorney to help him gain illegal control of Sandy’s money, for it was not true.

top