On June 22, in Turner vs. United States, the Supreme Court, by a 6 to 2 vote, affirmed the murder convictions of seven men. Unlike most Supreme Court decisions, Turner went largely unnoticed.
It deserves more attention, but not because it announced a new legal rule. Instead, Turner reaffirmed a terrible old rule that has done great harm to the accuracy of criminal trials, and will continue to do so.
The crime in the Turner case was horrific: in 1984, a middle-aged woman was grabbed off the street in Washington D.C., beaten, sexually assaulted and killed in an alley. The defendants were convicted on the theory that they were part of a group of a dozen or more who committed this atrocity. No DNA, fingerprints or physical evidence of any other kind connected any of the defendants to the crime.
Twenty-six years later, the defendants’ attorneys learned that prosecutors had concealed a laundry list of evidence that would have helped their defense at trial.
In particular, one witness identified a man at the scene of the crime as James McMillian, a local resident who was arrested several weeks later for beating and robbing two other neighborhood women, and was later convicted for robbing, sodomizing, and murdering a third young woman in an alley. And another witness testified that he heard moans—apparently from the victim—coming from inside a garage that was too small for a crime with 12 or even six perpetrators.
The Justice Department agreed that this evidence should have been disclosed at trial. The only issue was whether concealing that evidence made the trial fundamentally unfair, and requires a new trial.
In Brady v. Maryland, in 1964, the Supreme Court ruled that the government is constitutionally obligated to disclose evidence that is favorable to the defense in a criminal trial if that evidence is “material” to the case.
Later cases held that evidence is only “material” under Brady if there is a “reasonable probability” that the outcome of the trial would have been more favorable to the defendant if the evidence had been disclosed. Otherwise, under Brady, the prosecution may conceal favorable evidence from the defense and the trial court.
This standard is impossible to apply.
A prosecutor has to decide whether to disclose favorable evidence before the trial begins. At that point, she does not know what her own witnesses will say under oath (there are many surprises) let alone what the defense might put on. How can she possibly know before trial whether undisclosed evidence might tip the jury’s decision at the end of that trial? And who would trust a lawyer to make that decision about a case she herself is litigating?
Trial lawyers often believe, unrealistically, that their cases are airtight. In this setting, self-confidence is self-serving: it can lead prosecutors to decide that it’s OK to hide troubling evidence, which makes their job a bunch easier.
When a prosecutor does hide evidence, chances are nobody will ever know. If somehow it does come out, a court reviewing the case faces the same impossible question—what might have happened at trial if these facts had been known to the defense?—with an added twist: Judges are extremely reluctant to reverse jury verdicts and order new trials.
Not surprisingly, they usually conclude that concealed evidence was “immaterial” and therefore never had to be disclosed in the first place.
That’s just what the Supreme Court did in Turner. The majority points out that the hidden evidence is only exculpatory if McMillian committed the crime alone and not as another member of a large group. But none of the defendants disputed the prosecution’s claim that the victim was attacked by a group. Instead, each said that he was not involved, but his co-defendants might have been—and two additional defendants pled guilty and testified for the government in return for reduced sentences.
In that context, the majority concludes that disclosing the hidden evidence would have made no difference; presumably because there was little or no doubt that the defendants were in fact guilty.
However, as Justice Kagan points out in her dissent, no defendant argued that the murder was the work of a single person because they had no idea that there was evidence to support that claim. If they had known what we now know, the trial might have been totally different, including what was disputed and who testified. The two defendants who pled guilty might not have done so, and all of the defendants might have been acquitted—perhaps because they are innocent.
Is that a pipe dream?
Consider two disturbingly similar cases:
- In 1989, five teenagers in New York confessed to being part of a gang that attacked and raped the victim in the Central Park Jogger case, but each denied actually raping her. They were all convicted in 1990—but exonerated 12 years later when DNA confirmed the confession by an older serial rapist and murderer that he committed the crime alone.
- Also in 1989, six defendants were convicted of jointly assaulting, raping and murdering an elderly woman in Beatrice, Nebraska. Four of them had confessed, and five pled guilty in return for reduced sentences. In 2009, all six were exonerated when DNA evidence proved the crime was committed by a lone man from Oklahoma whom none of the innocent defendants knew.
Nobody doubted the guilt of these eleven defendants when they were convicted—even through, as in Turner, no physical evidence connected them to the crime. All but two had confessed; several pled guilty; none argued that the crime in their case was the work of a single person. All that would have changed if the defense had any information identifying the real criminals.
Concealing exculpatory evidence is not a rare problem. In 2013, then-Chief Judge Alex Kozinski of the Ninth Circuit Federal Court of Appeals wrote that “There is an epidemic of Brady violations abroad in the land,” because courts have defined “materiality” so narrowly that prosecutors rarely if ever have to face consequences if they hide evidence that favors defendants.
The National Registry of Exonerations (which I founded) lists 2,061 innocent defendants who were convicted of crimes in the United States and later exonerated, from 1989 on, including 802 who were convicted of murder. In half of the murder exonerations (398/802) the prosecution concealed exculpatory evidence at trial, including 17 cases with innocent murder defendants who pled guilty, 57 with defendants who falsely confessed, and 91 with supposed accomplices who confessed and falsely implicated the defendants.
And of course, as Judge Kozinski has noted, the great majority of similar cases are never detected, and the defendants, if they are innocent, are never exonerated.
I don’t know whether the defendants in the Turner case are innocent or guilty, or how their trial would have turned out if all the evidence had been fairly presented. Unlike the Supreme Court, I don’t believe we can tell by looking back and guessing 25 years after the fact.
Our best hope for avoiding tragic mistakes is to present all the evidence that matters the first time around.
Why not eliminate the “materiality” requirement entirely and treat access to exculpatory evidence like other aspects of a criminal defendant’s constitutional right to a fair trial? If exculpatory evidence is concealed, it’s a violation of the Constitution, period.
This would not mean that every violation requires a new trial. Courts often uphold convictions despite constitutional violations, because they are convinced beyond a reasonable doubt that the violation had no impact on the outcome. But that’s a tougher exception to meet, and it includes a critical message: Hiding exculpatory evidence is always a violation of the constitution.
Justice Kagan concedes in her dissent that the majority’s conclusion in Turner “is not indefensible.”
That’s right, unfortunately, given the rule the Supreme Court applied. But that rule itself is indefensible, and should be changed.
Samuel Gross is professor of Law at the University of Michigan, and the founder and Senior Editor of the National Registry of Exonerations. He welcomes readers’ comments.