You might think that, from an ethical standpoint, prosecutors should believe that the evidence they put in front of juries is actually true and accurate. After all, they’re state employees, paid with public funds, and they’re entrusted with enormous power. At the very least, we should expect them not only to refrain from putting forth evidence they know to be false but also to exercise some due diligence to ensure that everything they tell jurors is true.
As it turns out, that just isn’t the case. Two recent stories at the Marshall Project illustrate the problem. The first, from Andrew Cohen, concerns Paysun Long, an Illinois man convicted for a 2001 murder at a housing development in Peoria. There was no physical evidence linking Long to the crime, only the testimony of four eyewitnesses, two of whom later recanted. Incredibly, prosecutors showed jurors video of the two recanting eyewitnesses’ original claims, without telling them that those witnesses had since changed their stories. The defense then had to call those witnesses so that the jury could hear their recantations. Long’s first conviction was later overturned because a prosecutor told the jury that the recanting witnesses had done so out of fear, which wasn’t true.
At the second trial, it was more of the same. Prosecutors again showed jurors video of the recanting witnesses’ testimony, and again refused to tell them they had recanted. One of the recanting witnesses changed her story again, this time implicating Long. She then lied on the witness stand, claiming she never said at Long’s first trial that she had recanted. In short, prosecutors allowed the jury to hear evidence they knew was false or misleading and simply left it to Long’s defense attorneys to correct the record.
As Cohen points out, there’s plenty of precedent laying out a prosecutor’s obligation to correct witnesses who give false statements. But that precedent has no enforcement mechanism. And if the courts don’t want to enforce it, a defendant is out of luck. From Cohen:
When Long appealed up the ladder in Illinois the appellate judges there agreed that prosecutors were required to correct [the witness] Irby’s false testimony but that this was a “harmless error” because of other evidence in the case (the other evidence being, remember, the shaky testimony of the only other eyewitness who never recanted). So Long’s conviction and sentence were ratified by the state. He then took his case to federal court and argued that Illinois’s rulings against him were so “contrary” to the Supreme Court’s precedents in Napue and Giglio that the feds had a duty to rescue him from the result of the second trial.
Long appealed in federal court, where he lost.
For the five judges in the majority, it came down to this: Those old cases, the ones in which the justices declared that prosecutors had an affirmative obligation to tell jurors about lying witnesses, didn’t really mean what they said. If the false testimony wasn’t elicited by prosecutors, or if the truth wasn’t known to the defense during trial, or if the prosecutor asked jurors to rely on the false testimony, or if the jury never learned the truth then, yes, perhaps, the Illinois courts had gotten it wrong about Long’s case and he was entitled to relief.
But that’s not what happened here, the 7th Circuit’s majority ruled. Even if prosecutors remained silent during Long’s trial defense attorneys pointed out Irby’s perjurious testimony. And Long’s prosecutors didn’t exactly rely on Irby’s testimony so much as muddy its import during closing argument. Maybe the prosecutor’s failure or refusal to correct Irby’s testimony actually helped Long because it allowed his attorneys to be the ones to portray Irby as a perjurer, the majority argued, glossing over the likelihood that the perjury would have been more credible if admitted by the prosecutor.
Long’s attorneys wouldn’t tell me if they plan to ask the Supreme Court to hear the case. But it seems an obvious avenue of appeal, giving the justices an opportunity to reaffirm the principle they announced in 1959 and in 1972, that the government has a constitutional and ethical obligation to ensure that it speaks out against perjury when it unfolds in the middle of a trial in front of the jury.
In the second article, from Ken Armstrong, two men who participated in an armed robbery that resulted in a woman being shot to death were both tried for murder. The same prosecutor then argued at each man’s trial that he — and not the other — was the one who pulled the trigger. Both things can’t possibly be true. Which means that the prosecutor knew he was presenting false evidence in at least one of the trials. He may not have known which one. But that’s beside the point.
Armstrong found this sort of thing happens more often than you might expect.
There’s no saying exactly how often. But, in a recent canvass of court rulings, I turned up more than four dozen cases, from California to Massachusetts, in which the defense attorney argued in an appeal that the prosecution had told conﬂicting stories about the crime. Prosecutors have oﬀered contradictory theories about which defendant stabbed someone with a knife, or chopped a woman’s skull with a hatchet, or held a man’s head underwater. The most common scenario involves a fatal shot: the prosecutor puts the gun in the hand of one defendant, then another. Under the legal principle of accomplice liability, a defendant can be convicted of murder without being the killer. But, if the prosecutor says that a defendant pulled the trigger, it’s easier to ask a judge or a jury for a death sentence. At least twenty-nine men have been condemned in cases in which defense attorneys accused prosecutors of presenting contradictory theories. To date, seven of those twenty-nine have been executed.
The courts have frowned on this tactic, but rarely overturn convictions when it’s used. Again from Armstrong:
The descriptions applied by judges include “unseemly,” “unseemly at best,” “troubling,” “deeply troubling,” and “mighty troubling.” “The state cannot divide and conquer in this manner,” a federal appeals-court judge wrote in one Georgia case, in which the court threw out a defendant’s conviction on other grounds. “Such actions reduce criminal trials to mere gamesmanship and rob them of their supposed purpose of a search for truth.”
In 2004, the Sixth Circuit Court of Appeals overturned John David Stumpf’s conviction in the murder of Mary Jane Stout, writing, “Inconsistent theories render convictions unreliable.” The state appealed, and on April 19, 2005, the Supreme Court heard oral arguments in Stumpf’s case. Justice David Souter said, of the prosecution’s contradictory theories, “It has to be the case that one of those arguments, if accepted, would lead to a false result.” Souter asked how the use of conﬂicting arguments could square with due process. Justice Antonin Scalia said that he saw no such problem: “Due process doesn’t mean perfection. It doesn’t mean that each jury has to always reach the right result.” …
Two months later, the Supreme Court issued a unanimous opinion, written by Justice Sandra Day O’Connor, aﬃrming Stumpf’s conviction while avoiding the due-process question. Under Ohio’s law on aiding and abetting, Stumpf could have been convicted of aggravated murder no matter who ﬁred the gun. The question, the Court determined, was whether the prosecution’s inconsistency should invalidate Stumpf’s death sentence. The Sixth Circuit had not tackled that issue, so the Supreme Court sent the case back for an answer. To this day, the Supreme Court has not ruled squarely on the validity of conﬂicting prosecution theories.
I encountered this problem several years ago while reporting a story about a family in Church Point, La., who had been convicted in federal court of running a massive drug operation. Ann Colomb and three of her sons were convicted on conspiracy charges, due mostly to the testimony of more than two dozen informants, all of whom claimed to have sold crack to the family. The informants began coming forward after a federal prosecutor inadvertently left his file on the case at a federal prison while interviewing a witness. All requested time off from their own sentences in exchange for their testimony.
The federal prosecutors should have known that there was simply no way the informants’ testimony could have been true. First was the matter that all of them came forward with specific information about the family only after the prosecutor’s file had been shared throughout the federal prison. Second, if you added up the amount of crack the witnesses claimed to have sold to the family, the Colombs would have been purchasing about $500,000 in crack each month, enough to make them regional kingpins. Yet the Colombs lived in a modest, working-class home and showed no signs of that sort of wealth. Third, the federal government couldn’t produce a single witness who claimed to have bought drugs from the family, only the 30 or so who claimed to have sold it to them.
Yet when challenged on all of this, the federal prosecutor said in open court that it didn’t matter if he personally believed in the accuracy of the evidence he was presenting and all that mattered was that the jury believed it. The Colomb convictions were eventually overturned and the family was freed after additional inmates in the same prison later came forward to say that the other inmates had offered to sell them copies of the Colomb case file. The federal prosecutors were never sanctioned.
When I inquired with legal ethics experts about this, I was surprised to learn that most state bars have no requirement that prosecutors believe that the evidence they present is truthful. The only requirement is that they believe in the guilt of the person they’re putting on trial. Of course, prosecutors can’t present evidence they know for certain to be false. But as the examples above demonstrate, with anything short of absolute certainty, the courts usually give prosecutors the benefit of the doubt.
In fact, there’s a pretty strong incentive for prosecutors to avoid spending much time or resources investigating the veracity of the evidence they plan to present at trial. So long as they’re acting in the role of a prosecutor, the doctrine of absolute immunity protects them from any civil liability resulting from a wrongful conviction. It’s a shield that’s basically impenetrable. Courts have ruled that even if prosecutors knowingly break the law — such as detaining defense witnesses in order to prevent them from testifying at trial — those who were harmed by the prosecutors’ behavior have no recourse.
But the Supreme Court has carved out one notable exception. When prosecutors step outside their role as prosecutors and engage in activities more akin to those of a police investigator, they may lose absolute immunity. Even here, they’re protected by the doctrine of qualified immunity — the same immunity afforded to police officers. The line between prosecutor and investigator is a fuzzy one, and there hasn’t been a lot of case law defining exactly where it lies. So if you’re a prosecutor, the more time and energy you put into investigating, say, eyewitness statements to police or the veracity of an informant’s claims, the more you risk subjecting yourself to the exception to absolute immunity. Put another way, the more you do to ensure that the person on trial actually committed the crime, the more you subject yourself to the risk of liability should that person turn out to be innocent. That’s a textbook perverse incentive.
Of course, in most of the cases above, the prosecutors didn’t need to spend much time at all investigating to know that the evidence they presented was false. Once you’ve told two different juries two contradictory theories about who pulled the trigger during a robbery, you pretty well know that you’ve lied to one of those juries.
The absolute immunity/qualified immunity problem is a very real one. We should be structuring incentives to ensure due diligence, accuracy and truthfulness from prosecutors. But at the moment, we can’t even hold prosecutors accountable when they both present false evidence and had to have known that the evidence was false when they presented it. If they can get away with that, we’re a long way from demanding that they actually investigate the veracity of evidence themselves before putting it in front of jurors.