At the Intercept, Jordan Smith reports on a bill from Texas that would end the use of jailhouse informants in death penalty cases.

The Northwestern Law School’s Center on Wrongful Convictions found in 2005 that false snitch testimony is the leading cause of wrongful convictions in capital cases in the modern era of the death penalty, implicated in nearly half of the 111 death row exonerations at that time. (As of May 4, there have been 153 death row exonerations.)

“It’s astounding, it really is,” says Alexandra Natapoff, a professor at Loyola Law School in Los Angeles and the nation’s leading expert on the use of informants. “Criminal snitching is an enormous problem for our justice system, in part because it’s an enormous source of error.” . . .

But a bill pending before the Texas Legislature would address the problem by banning all incentivized informant testimony in death penalty cases. Its chances are slim, but if passed, the law would make Texas the first state in the nation to eliminate the use of all compensated testimony in capital cases. The bill would, theoretically, still allow certain informant testimony, from jailhouse snitches, but only if the defendant’s statements to the snitch are recorded — a measure that is akin to a model policy proposal supported by Northwestern’s Center on Wrongful Convictions and the Innocence Project.

If the bill passes, I predict you’ll see the use of informants in these cases dwindle to almost never. The reason: Most jailhouse snitches are lying. Informant testimony has become such a critical tool for prosecutors precisely because it allows them to put on testimony that is a) damning, b) easy to manufacture and c) allows b) to happen while giving them plausible deniability. This isn’t to say that all prosecutors manufacture evidence by using jailhouse informants. It is to say that the way informants are treated by the courts makes it very easy to do so.

The whole concept of jailhouse informants defies credulity. The very idea that people regularly confess to crimes that could put them in prison for decades or possibly even get them executed to someone they just met in a jail cell and have known for all of a few hours is and has always been preposterous. Not to mention the fact that these are people whose word prosecutors wouldn’t trust under just about any other circumstance. When informants have later recanted testimony or claimed that police or prosecutors browbeat them into lying, a DA’s office will quickly point to the informants’ criminal records and lack of trustworthiness. But when they’re helping to win a conviction, their word is gold.

Many prosecutors also feel that they’re under no obligation to investigate the veracity of an informant’s testimony. Here are a couple of examples of how that can play out at trial:

In one case I reported on several years ago, a parade of jailhouse informants all claimed to have sold drugs to a family. The totality of the testimony was implausible. If all the informants were telling the truth, this family that lived in a modest home in a small central Louisiana town would have been the biggest kingpins in the South. Yet the U.S. attorney stated at trial that it didn’t matter if he personally believed the informants. It only mattered what the jury believed. After the family was convicted, it was revealed that the informants had lied. Yet the U.S. attorney went on to use some of the same informants in other trials.

In another trial in Chesapeake, Va., a few years ago, Ryan Frederick stood accused of murder for shooting and killing a police officer during a drug raid on his home. Frederick claimed he thought the police were criminal intruders. There were two abuses of informants in that case. First, the police had probable cause for the raid only after one of their informants broke into Frederick’s home to see the pot plants he was growing. The police claimed they had no knowledge of this, although there was some evidence that this had happened before. More troubling, during Frederick’s trial, Commonwealth’s Attorney Paul Ebert put on a jailhouse informant named Jamaal Skeeter, who claimed Frederick admitted he knew the raiding cops were police, boasted about killing the officer and mocked the officer’s widow. Ebert was named special prosecutor in the case and brought in from Northern Virginia. Frederick apparently made these statements immediately after meeting Skeeter, during their one hour per day of recreation at the jail. That in itself seemed improbable. But it seemed especially unlikely given the police station videos of Frederick sobbing with regret and vomiting upon learning that the officer he shot had died.

Then, something extraordinary happened. In mid-trial, another prosecutor in Virginia stepped forward to inform Frederick’s attorney that Skeeter had a long history of lying and that no local prosecutor trusted him. This would seem like a pretty easy thing for Ebert to have discovered on his own. But again, while prosecutors are in theory ethically obligated to only put on evidence they believe is true, there’s no way to enforce that obligation. In fact, there’s some incentive for them not to investigate their informants’ claims. Doing so could arguably put them in the role of investigators, or police. So long as they’re performing prosecutor duties, prosecutors are protected from civil liability by absolute immunity. But once they take on duties normally conferred to cops, they’re protected only by the qualified immunity given to police. (Qualified immunity is still tough to get past, but it’s also a step down from absolute immunity.)

And of course, all of this is before we even get to the problem of prosecutors giving informants time off their own sentences in exchange for their testimony.

This Texas bill is a good first step. It’s a pretty significant first step. But if we can admit that the use of jailhouse informants is fraught enough with problems that we want to keep the practice out of death penalty cases, I’m not sure why we’d be okay with allowing it in cases that could result in decades-long prison sentences.