Falsified informant testimony accounts for nearly half of all wrongful convictions in capital cases nationwide, according to data from Northwestern University Law School’s Center on Wrongful Convictions. Since 1973, 129 innocent people were released from death row — more than 50 of whom were sentenced to death based partly or wholly on false informant testimony, according to the Center.

(emphasis added) Fifty people condemned to die on the basis of a lie. How many more death sentences, or life sentences, or 30 year sentences, have been imposed on the word of a snitch?

“We have the most data on capital and homicide convictions because they are the most high profile,” [Alexandra Natapoff, an associate professor of law at Loyola University] says, “so we have no idea how many wrongful convictions there are in larceny cases or assault cases or any other because nobody is paying any attention to those.” ...

“The government’s use of criminal informants is largely secretive, unregulated and unaccountable,” she says. “This lack of oversight and quality control leads to wrongful convictions, more crime, disrespect for the law and sometimes even official corruption.”

She continues: “If the criminal system can’t get homicide cases right, then it’s very unlikely that we’re getting other things right.”

The war on drugs is partly to blame. It's easy to arrest street level dealers, but to find the big players, lazy police officers rely on the word of snitches. Evidence that is unreliable by nature becomes the heart of large drug prosecutions.

With the expansion of the “war on drugs” during the crack epidemic of the late 1980s, police began to abandon traditional investigative work in favor of insider cooperation. Cops say it’s almost impossible to make a drug case any other way. But critics say the practice has led to a “dumbing down” of police work across the board.

While prosecutors like to talk about "climbing the ladder" -- squeezing a petty dealer to give up his supplier, then squeezing the dealer to give up his source, and so on -- the criminals with the best information are the ones at the top. And since they have the most to sell, they often get the best deals. Sometimes the small dealer who refuses to rat out anyone else ends up with the longest sentence -- longer than the sentences given to snitches who committed more serious crimes.

You might think that a fair system of justice wouldn't let the government buy testimony. After all, if defense attorneys did that, they would be arrested for bribing witnesses. But fairness has little to do with snitch testimony.

The Supreme Court established the constitutional basis for using paid informants in 1966 with U.S. v. Hoffa, which decided that rewarding a witness for testimony does not violate due process. In its opinion, the court wrote: “The established safeguards of the Anglo-American legal system leave the veracity of a witness to be tested by cross examination, and the credibility of his testimony to be determined by a properly instructed jury.”

But Natapoff says, in practice, those mechanisms are deeply ineffective at protecting defendants from lying informants. “Let’s say the government does disclose [compensation] and the jury knows about it,” she says. “You would think, and the Supreme Court certainly thinks, that that will make a difference. Well, psychological research has found that it makes almost no difference, that jurors ignore the fact that the witness is compensated.”

Possible solutions:

According to the American Bar Association (ABA), 18 states now require corroboration of an accomplice’s statements. Those that require corroboration for other forms of incentivized witnesses, however, are few and far between.

Illinois currently mandates corroboration in capital cases, and courts in Nebraska and Oklahoma have required corroboration for jailhouse snitches. Texas, meanwhile, has a different requirement, not for jailhouse snitches, but for undercover drug operatives working for the police. ...

[Howard University law professor Andrew] Taslitz, who serves in the ABA’s Criminal Justice Section, says more transparency is needed during the discovery phase. For example, he’d like to see defendants who are negotiating a plea agreement have access to the information and witnesses being used against them.

In a 2005 ABA resolution that Taslitz helped write, the association urged federal, state and local authorities to require that informants meet certain standards of credibility and that courts mandate corroboration in all cases that involve jailhouse snitches.

A better solution: police and prosecutors should stop getting into bed with snitches. If a potential informant wants to be a good citizen, tell him he gets nothing for his cooperation. And not with a wink and a nudge; snitches should never be rewarded, ever. If the potential informant still wants to talk, maybe he's credible. If he wants to be paid or released from prison or doesn't want to be charged for a crime, don't believe him. If that makes it a little more difficult to prosecute crime, it's a small price to pay to save the innocent from unjust incarceration.