New Study: Providing Incentives to Snitches Increases False Information

The results of the first behavioral study to investigate whether people will provide false secondary confessions raises significant concerns about the use of such evidence when informants are offered incentives. The study was conducted by psychological researchers at the University of Arkansas.

A "secondary confession" is a polite name for snitching. A news article on the study is here. The study is now published in the Journal of Law and Human Behavior in an article titled “Snitching, Lies and Computer Crashes: An Experimental Investigation of Secondary Confessions.”

Bottom line: "[A]n incentive increased the rate of false rather than true secondary confessions."[More...]

How the study worked:

In the psychology lab, participants engaged in a computer exercise that ended in a simulated crash of the computer and a purported loss of data. Data was analyzed from 129 participants who were paired with confederates of the researchers. After the crash, confederates either denied or “admitted” that they had caused the crash.

Some participants were given an incentive to tell whether the confederate had admitted to causing the problem. They were told that the faculty adviser would be informed and that the person who had caused the problem would be required to come back for a second session.

Participants were asked to sign a statement affirming a secondary confession of guilt. That is, they stated that the other person – the confederate – had admitted crashing the computer. Not surprisingly, participants were more likely to sign when the confederate had admitted to causing the crash. In these cases, the offer of an incentive did not increase the rate of signing. In fact, with an incentive, the rate of signing increased only when the confederate had denied causing the crash. In other words, an incentive increased the rate of false rather than true secondary confessions.

Not only did incentives increase the rate of false secondary confessions, but also participants were less likely “to see freely admitted misdeeds as unintentional.” That is, some of those who signed statements also amended the statements, excusing the crash as a mistake due to external factors such as the speed of the test. The researchers found that offering an incentive for secondary confessions “eliminated this ‘honest mistake’ pattern.” (my emphasis)

The study authors' suggestions:

“The concern is partly based on confessions being assumed to be the end-all and be-all of trial evidence, when at least in the case of secondary confessions they should be treated as hearsay,” Swanner said.

She and Beike suggested several safeguards, including video recordings of all interviews and interrogations of informants and suspects as well as pretrial hearings and expert testimony to allow jurors to better assess the validity of secondary confessions entered as evidence.

“It is essential for jurors, prosecutors and judges to be informed about the potentially biasing nature of incentives to confess,” they concluded. “Snitches may indeed lie or come to believe a falsehood about another to be the truth. Jurors must be able to consider this possibility as they make their verdicts.”

< New Immigrant Detention Report | Valuing Toxic Assets >
  • The Online Magazine with Liberal coverage of crime-related political and injustice news

  • Contribute To TalkLeft

  • Display: Sort:
    this study, (none / 0) (#1)
    by Bemused on Wed Mar 25, 2009 at 12:20:44 PM EST
      at least according to the article, is not really directed that all "snitching." "Secondary confession" is actually used not as sa synonym for "snitching" but to describe the situation where a subject relates the "primary confession" of a third party.

       This is distinguishable from the very common scenario where the snitch claims to be a percipient witness to  an act of another.

      The broad interpretative findings likely apply with similar force to other forms of snitching (after all it's common sense that, statistically speaking, people given an incentive to say what the incentive givers want are more likely to do so), but insofar as the researcher suggest hearsay rules are involved, that can't really be the case in with snitch testmony claiming direct personal knowledge of the act as opposed to having been told of the act by the aCTOR.

    perhaps i totally misunderstood (none / 0) (#6)
    by cpinva on Wed Mar 25, 2009 at 03:27:29 PM EST
    the basis of this study, but the distinct impression i got was that it was oriented to the "jailhouse" variety "snitch"; they claim to have either overheard or been told directly by the accused, a confession to whatever it is they've been accused of. for a hamburger (sentence reduction) today, they will gladly testify next tuesday.

    these are not the "acts" themselves, merely a "confession" to said prior acts. the "snitch" witnessed nothing, but was allegedly told about it, by the accused.

    perhaps i am mistaken.

    that the addition of an "incentive" might prove enticing comes as no great surprise, that's how it works outside of prison and the courtroom as well.


    It would be applicable (none / 0) (#7)
    by Bemused on Wed Mar 25, 2009 at 03:39:38 PM EST
     to anyone in or out of jail who is asked to relate what someone else said about an event. That is but one breed of snitch and I assume the study itself was careful to so limit it.

      This post and the article seem to apply it more broadly to the panoply of snitching. the most comon form of snitching isn't the guy who says, "the defendant conferssed to me in jail." It's not even more broadly guys who say I heard him say he did it when we were out one night. It's far more prevalent to have the snitch say I saw him do it (often because the snitch was also doing it, and equally often because snitch set upo the defendant as an agent for the cops).


    Wow! Who knew? (none / 0) (#2)
    by hairspray on Wed Mar 25, 2009 at 12:34:55 PM EST
    This seems like a no brainer.  I am glad there is finally some objective proof here.  Now if they would just use the data they already have on eye-witness accuracy we would be rolling.

    A significant distinction (none / 0) (#9)
    by Bemused on Wed Mar 25, 2009 at 05:05:52 PM EST
     is that with eye witness testimony there are studies showing that many people not only believe eye witness terstimony is inherently credible, they often believe it is the most credible category of evidence, and we have research tending to show it is in fact often very unreliable. Therefore, expert testimony on observation and recall is helpful to the jury to explain something beyond common knowledge.

      I'm not convinced that many people believe snitch testimony generally or "secondary confession" testimony more specifically is inherently credible and rather believe most people agree with the premise of the instructions that such testimony is tobe received with caution and carefully scrutinized. So, the argument for allowing expert testimony to tell people snitches often lie when paid to snitch is harder to sell as something helpful to the jury.


    what they have to understand (none / 0) (#3)
    by bocajeff on Wed Mar 25, 2009 at 01:08:40 PM EST
    is that there are no criminals. Just crime.

    Three cops dead (none / 0) (#4)
    by daryl herbert on Wed Mar 25, 2009 at 02:50:15 PM EST
    a violent felon gunned them down.

    In the aftermath, his DNA was tested, and it turns out he's also a rapist.

    There are definitely criminals, and they should be locked away until they are too old to hurt us.  The problem is sorting them out from the people who will behave.  But there are definitely hopeless cases, and if we aren't locking them away for life, then we're failing.

    And the result of this failure is more rapes and more murders.


    Mistakes will be made.... (5.00 / 1) (#5)
    by kdog on Wed Mar 25, 2009 at 03:07:29 PM EST
    regardless, I think the question is do you want society to err on the side of redemption and liberty or err on the side of condemnation and tyranny.

    I think its a no-brainer, the world will be f*cked up either way, but at least we can hold our heads high erroring on the side of redemption and liberty....as opposed to the general shame I feel now when I look at our criminal justice system at large.


    daryl, your comment is off topic (none / 0) (#8)
    by Jeralyn on Wed Mar 25, 2009 at 04:58:04 PM EST
    do not try to hijack the thread. It's about informants and a study.

    What's interesting, (none / 0) (#10)
    by NYShooter on Wed Mar 25, 2009 at 09:09:28 PM EST
    If not surprising, is the embellishment the "snitch" adds to his story.

    I always say to my wife, children, or others, when they're relaying a story (gossip) to me;  "If you weren't there, and didn't see/hear it with your own eyes/ears, you don't know anything.

    Isn't it true? When someone is relaying an event to you, don't you notice their exaggerations, facial contortions, emphasis on this or that; in other words embellishing.

    Jailhouse snitching, to me anyway, falls below hearsay and polygraph results. How a judge should treat it, however, is beyond my  expertise to say.

    by rule (none / 0) (#11)
    by Bemused on Thu Mar 26, 2009 at 07:32:30 AM EST
    You start with what is Rule 601 in most codes, stating that "Every person is competent to be a witness except as other wise provided in these rules"

     Rule 602 states that with an exception for experts) a witness may not testify unless he "personal knowledge." (the "personal knowledge of third-party statements is knowledge the third party made the statement not of the facts related in the statement, which is what creates the problems)

      Then the hearsay rules (800 series) govern the admissibility of whether a witness may testify to what 3rd parties said.

      The out-of-court statement of a defendant is simply defined out of hearsay by Rule 801 (d)(2) which simply says such a statement is not hearsay. The rule actually goes further than that by attributing to the defendant statements of agents ob behalf of the defendant and "co-conspirators" of the defendant (this one is greatly exploited/abused by prosecutors)

      Then you have a laundry list of exceptions to hearsay rules (different ones apply based on whether the person who made the statement is avaulable or unavailable) which still consider the testimony hearsay but allow it. then you have rule 807, the residual exception which says in essence the court can allow hearsay that it determines has equivalent "circumstantial gurantees" of trustworthiness and thet types of hearsay expressly excluded.

      Also, "a statement" is not a a statement in law unless it is offered to prove the truth of the matter asserted in the statement. so words spoken whch are offered. for example, why someone did something (as opposed to word represented as asserting a fact is true)  would not be hearsay.

       So, swallowing all that,  a "secondary confession" in which a witness claims the defendant made a statement to him is not considered hearsay under current rules and the "snitch" is a person competent to testify because his testimony is not excluded by any other rules.

      The proposal in the study would be a huge departure from traditional practice and centuries of jurisprudence.

      Currenttly, we rely on cross-examinations (and sometimes collateral witnesses) to attack the credibility of snitches  and then argue it like mad in closing argument. the court also. of course, instructs the jury it should receive the testimony with caution and judge its credibility carefully (if the snitch is either paid, given or expecting leniency or is an accomnplice the jury is specifically instructed these considerations may influence the testimony.