Snitching markets in everything

The prisoners in Atlanta’s hulking downtown jail had a problem. They wanted to snitch for federal agents, but they didn’t know anything worth telling.  Fellow prisoner Marcus Watkins, an armed robber, had the answer.

For a fee, Watkins and his associates on the outside sold them information about other criminals that they could turn around and offer up to federal agents in hopes of shaving years off their prison sentences.  They were paying for information, but what they were really trying to buy was freedom.  “I didn’t feel as though any laws were being broken,” Watkins wrote in a 2008 letter to prosecutors.  “I really thought I was helping out law enforcement.”

That pay-to-snitch enterprise — documented in thousands of pages of court records, interviews and a stack of Watkins’ own letters — remains almost entirely unknown outside Atlanta’s towering federal courthouse, where investigators are still trying to determine whether any criminal cases were compromised.  It offers a rare glimpse inside a vast and almost always secret part of the federal criminal justice system in which prosecutors routinely use the promise of reduced prison time to reward prisoners who help federal agents build cases against other criminals.

Snitching has become so commonplace that in the past five years at least 48,895 federal convicts — one of every eight — had their prison sentences reduced in exchange for helping government investigators, a USA TODAY examination of hundreds of thousands of court cases found.  The deals can chop a decade or more off of their sentences.

That is from USA Today, from this source.  Hat tip goes to @matt_levine.


Wait, what is the problem woth this?

Legally the prosecutors can't use that information in court. And even if it isn't used directly in court, it's possible that information arising from it could be thrown out also, though I'm guessing judges will generally give the officers a pass if they didn't know.

Could you clarify?

What information can't they use in court?


First day of law school:

"Every legal question can be correctly answered with two words: It depends."

Determining what the legal question depends on is what lawyers get paid for. The rest of the work could be done by someone with modest talents.

That's not really true. Confidential informants (CIs) are widely used and their information is generally most useful and legally permissible at the reasonable suspicion level. Usually this information is corroborated to build probable cause. Confidentiality cannot be assured if the CI will be used as a witness.

Courts can order the production of transcripts from CIs when probable cause is in dispute, but this can be redacted to maintain anonymity to the extent possible.

Information for cash is usually admissible under the same circumstances as a reward or tip hotline. As you suggest, evidentiary rulings are unfriendly to statements in interest, that is, for pay or for 'flipping'. Corroboration often overcomes this defect. You can also collect information from an informant with a 'dirty team' who passes along select information to a 'clean team' which discovers evidence for probable cause.

Statements against interest are almost always admissible and presumed to be truthful. The character and motives of the witness may be called into question. In a recent case, an admission by an accomplice to his cell mate was admissible against the other party to the crime.

In deference to privacy laws, there are exhaustion requirements that make CIs a last resort. Courts also apply a proportionality test to weigh the value of the evidence against the interests of privacy.

It can be very complicated, situations vary, and case law evolves. If I were a defense attorney, I would raise the question.

If a non-felon witness was given $10K to testify would you allow ANY of his testimony?

Yes, it happens all the time. People often get rewards for information.

But as I said, the greatest usefulness of tips is at the earliest stages of investigation when we need a name and reasonable suspicion.

At trial, the defense counsel will challenge the evidence because a statement in your own interest is not presumed to be true. Your testimony is presumed true but it is rebuttable. If the only evidence we have is from a paid witness, we aren't going to be able to prove guilt.beyond a reasonable doubt.

Evidentiary rules are not kind in these cases, so there has to be a strong case just to go to trial. I understand and share your concerns and to a large extent they are addressed. Prosecutors don't plan to take losing cases to court. To the credit of our system, prosecutors often determine that there isn't sufficient evidence before the suspect is Indicted. Remember there is going to be a grand jury hearing. We don't take losers to grand juries either; it's a waste of time.

One of many reasons I quit as a prosecutor. We made too many sacrifices of justice for expediency. Too many deals with the Devil just to keep up the batting average and keep down the costs.

The hard cases were the people most worthy of punishment, and the easy caseswere the ones most deserving of leniency. And the police learned to respond to the incentives we gave them.

Justice dies in settlements and plea bargains. The adversarial system has degraded into polar extremes of posturing demonstrations or conniving gamesmanship.

If you want to know who a system serves, look for the people who never lose.

Like one of my students once told me after explaining the Prisoner's Dilemma in my law and game theory course: "The house always wins"

I wonder whether there are true arbitrage opportunities here, i.e. without negative externalities for the criminal being snitched on. Are there cases in which criminals sell information about themselves? This could make sense if the amount of time shaved off of the snitch's sentence is greater than the amount of time served by the snitch-ee. And of course we should be thinking on the margins here: how bad is the 10th year of a 10-year sentence relative to the first year of a 1-year sentence? The possible effect on ex-ante incentives is also interesting, especially given the savings in investigative costs for the police/prosecutors.

Are there cases in which criminals sell information about themselves?

Isn't that essentially what a plea bargain is?

If this is a large increase in the supply of information there should be, ceteris paribus, a decrease in the price of information. Any evidence of this?

Not really for several economic reasons. First, some people have exclusive ownership over information, e.g. a witness, so they can charge monopoly or oligopoly prices. Second, information can be a public good, so demand is determined by the vertical summation of individual demand curves.

But yes you have a good point in some or many or most circumstances. For example, Robert Hanssen or Aldritch Ames, I forget which, sold top secret information for about $1 per page.

How much longer before someone actually pays for even generating such information?

Your concern is definitely valid, but without corroboration, most of this information is legally useless. At best it raises reasonable suspicion which could lead to a Terry Stop or an 'In plain sight" finding. I couldn't get a search warrant based on any of these tips by themselves.

This information is most useful when it leads to information that established probable cause; and then you hope to find sufficient evidence after you get the warrant and conduct the search. Many searches lead to arrests for minor offenses which, as I said before, could result in miscarriage of justice and over-punishment. For example, suppose the police and the DA believe someone is guilty of murder and there is enough evidence for a search warrant. The police don't find sufficient evidence for murder, but they find evidence of a lesser crime such as possession of controlled substance with intent to distribute. If the defendant has priors, that could be a five year ride, minimum. Possession and intent to distribute are often proved by circumstantial evidence. The point is that most LEOs and DAs go home thinking that the murderer got at least five years that was coming to him. To me, if he's guilty that isn't justice and if he's not guilty that isn't justice. And if he's not the murderer, then the real murderer is at large.

There are way too many compromises between Type I and Type II errors in our legal system for my taste. Some people think that it represents balance. I think it represents injustice.

Homicide has the highest clearance rate among serious crimes (64%), but out of 15,000 murders per year, that means at least 5400 get away with it. However, clearances include acquittals, plea bargains, and deaths of the alleged perpetrator, so the number is a bit inflated as a measure of justice.

The clearance rate has dropped from 72% since 1980. So the need for this type of inside information is greater than it ever was. I'm not sure whether the criminals are getting better or the police and prosecutors are getting worse. I guess I could blame part of it on CSI - at least since 2000. :)

What I meant was pay someone by proxy, relatively anonymously to execute a hit and then the prisoner gets to snitch. Solid case. Your corroborating evidence is sure to be found.

The price of the hit would have to be less than the expected reward. But I see where you are going with it.

At first, I thought you were saying that people would pay others to snitch to get rid of an undesirable person.

Say you are a drug boss doing 10 years. What'd the going price of a hit? I think the reward would amply compensate, at least for the richer class of criminal.

Since there were 23k murders in 1980, a clearance rate of 2 percent implies that 6400 got away with it From that standpoint, 5400 is an improvement. Has there been a change in the composition of homicides such that there are relatively fewer easier to clear murders?

No, that's not right. The number who "got away with it" went down only because there were fewer homicides. I got away with murder last year because I didn't commit any murders. It's vacuously true.

The clearance rate is the correct measure of comparison for judging the effectiveness of law enforcement - assuming the data is accurate.

If one were to adopt your line of reasoning, we should also divide by the population because the number of murderers who got away with it are a smaller proportion of the population.

However, your second question is not only relevant but timely because I just got this article today. It's about ipads, but it has some relevant information

I have been a prosecutor in decades, so I cannot really tell you what has happened to law enforcement. I keep up with some buddies who were on the force in Chicago, but we don't talk about that stuff when the Bears or Bulls are playing.

I assume that organized crime is getting better at what they do. I suspect turf wars have been reduced both from clearer territories and increased police engagement. It's likely that the clearance rate and the homicide rate are simultaneously determined. If there are a lit of murders, there are lots of witnesses. I presume that as the number of murders declines, the average skill of the murderer goes up.

As the article states, dockets are full and DAs have their hands full. I was being facetious early about CSI, but I think it has had an effect and the Innocence Project has, at least in my state, made proving murder more difficult. There was also a moratorium on the death penalty by a corrupt governor trying to save his hide.

There is documented evidence of the NYPD underreporting crimes or reporting them as less serious crimes. I might guess the same thing goes on elsewhere. This recession is rather peculiar in that crime went down, and I'm not sure why.

From your link:

(Some kinks in my technological fantasy may have to be worked out: the Police Department says that a standard Internet connection is not currently secure enough for its communications.)

*shakes my head* It's trivial to keep a conversation private over a public channel these days. The easiest thing is to just use a web app over HTTPS which gets you most of the way there. (I'd suggest about a dozen more small tweaks too technical for a blog comment.)

And "iPads" (which the article seems to use as a generic term for "tablets") are much easier to secure than general purpose laptops. Consumers can get a 3G data plan for a tablet for $10 a month with some bundling; a police force buying in bulk can do way better.

Someone proposed a market for terrorist acts, perhaps we need a market for criminal ones as well?

It could offer options on future crimes as well as on likely guilty parties. Perhaps even on chances of conviction.

It might not work, but it would be interesting. Then they couldn't

Sorry, premature posting, I will try to think about baseball....

Then they couldn't trade information for leniency, but they could for cold hard cash.

No doubt we would see a return of Thief Taking where smart people got dumber people to commit crimes just so they could clean up in the market.

For most criminal acts, there is a market. It's called insurance.

For crimes like murder and rape, you would have to have a futures contract for whom? A victim? A perpetrator? A murder in a particular neighborhood during a particular time?

The market would be too illiquid to be useful. The presumption in the futures market for terrorism was that people with inside information would be active traders. Criminals are disproportionately unbanked, so it's doubtful they would be involved in a futures market. Who would make a market for this? Counterparty risk is way too high.

The public wasn't too kind when DARPA invented the former, so I doubt the latter would be very popular.

Snitching can be severely punished though - just ask any PR person what happens when they tell the truth.

This post illustrates how you can be influenced by the choice of words.

Take the word"snitching", for example.

What is the connotation of the word "snitching": praise or disapproval?

We need to find a new praiseful word for snitching to encourage snitching on criminal acts.

A true prisoners dilemma.

Actually, Bill, the word 'snitch' is used by the criminal elements. You might see someone wearing a "Don't Snitch" t-shirt in the bad parts of town. There eas a Stop Snitchin' campaign in Baltimore and a hip hop song by Mac Dre.

The campaign is a not so subtle threat to murder CIs.

CIs are essentially equivalent to HUMINT or double agents in the intelligence lingo. It is just as dangerous, often unreliable, and its legality is often questioned. Under the exhaustion principle, use of CIs has to be a last resort for law enforcement. But in an environment of organized crime, that's not a high hurdle to leap.

Actually, that was my point. Using the word "snitch" is designed to stop snitching.

If you want to encourage snitching, you need another word.

"confidential informant"

Comments for this post are closed