Report on Unreliability of Drug Detecting Dogs

The Chicago Tribune has completed a review of three years of data on drug detection dogs used in traffic stops. It finds the dogs were wrong more often than they were right. Only 44% of positive alerts led to the recovery of drugs or paraphernalia. The success rate dropped to 27% for hispanic drivers.

Most courts hold that during a traffic stop, if a properly trained dog alerts, it amounts to probable cause to search the vehicle. The problem seems to be that many of the dogs, and their handlers, are not properly trained. Few states require certification and mandatory training and there is a paucity of "best practices" or even uniform standards. [More...]

Even advocates for the use of drug-sniffing dogs agree with experts who say many dog-and-officer teams are poorly trained and prone to false alerts that lead to unjustified searches. Leading a dog around a car too many times or spending too long examining a vehicle, for example, can cause a dog to give a signal for drugs where there are none, experts said.

There are also issues with dogs responding to a handler’s cues rather than to actual drug odors.

For more, check out "Those Doggone Sniffs Are Often Wrong: The Fourth Amendment Has Gone To The Dogs."

< Thursday Afternoon Open Thread | Budget Cuts? 1,000 Additional Marines to be Sent to Afghanistan >
  • The Online Magazine with Liberal coverage of crime-related political and injustice news

  • Contribute To TalkLeft

  • Display: Sort:
    It is positively rampant (5.00 / 1) (#2)
    by Militarytracy on Thu Jan 06, 2011 at 08:16:40 PM EST
    that the K-9 dogs that police departments try to claim are drug dogs are not trained in actual drug detection outside of a weekend here and there.  Dogs don't have the frontal lobes of human beings, they must always be schooled in what they are "expert" in....they don't retain training well long term that isn't part of their genetic imprint (my dogs will always instinctively herd but they don't care what pot or cocaine smells like right now...it isn't in their nature) and it varies from dog to dog. For a dog to be certified and used in arson investigations, the process and training is grueling and constantly on going.  The certification process is constant and grueling too.  How can a drug dog's training and certification be any different in order to have anything that it does hold up in a court of law?  I still don't understand any of this.

    Apparently the dogs at the border are (5.00 / 1) (#7)
    by oculus on Thu Jan 06, 2011 at 10:44:44 PM EST
    not trained to detect Cuban cigars.

    "tell" thank you, pain meds are working (5.00 / 1) (#8)
    by republicratitarian on Thu Jan 06, 2011 at 11:05:32 PM EST
    I couldn't remember the word. I blew out my knee playing basketball last night. Appointment with an orthopedic surgeon tomorrow.

    Ouch (none / 0) (#9)
    by Militarytracy on Fri Jan 07, 2011 at 08:20:25 AM EST
    I have one knee giving me minor problems here and there.  They have made some really amazing progress in knee repair though.  Good luck with your healing and take good care of yourself.

    Sounds too much (none / 0) (#1)
    by Zorba on Thu Jan 06, 2011 at 08:11:45 PM EST
    like a case of canine Clever Hans (and a bunch of them, at that).

    Kind of discouraging (none / 0) (#3)
    by magster on Thu Jan 06, 2011 at 08:20:22 PM EST
    I was hoping dogs might be a less invasive remedy than the full body scans at airport security.  

    If you want to spend the money training (none / 0) (#4)
    by Militarytracy on Thu Jan 06, 2011 at 08:27:15 PM EST
    and training and training like the military does, they can be.  But every day is training day and certification day...every single day.  It can be draining on a system not designed for and prepared to pay for constant training.  The military has some impressive dogs doing it but that is all these dogs and their handlers do 24/7, they don't make traffic stops and worry about writing tickets or even do "bite work" unless they are superstars who can handle that along with detection....but that is a rockstar dog....that would be as common as a Cam Newton.

    I'm curious what percentage (none / 0) (#5)
    by republicratitarian on Thu Jan 06, 2011 at 09:53:47 PM EST
    Of the positive hits come from cues from the cop or from a cop just outright lying and saying that the dog "gave" a sign.

    So (none / 0) (#10)
    by jbindc on Fri Jan 07, 2011 at 12:37:28 PM EST
    The success rate dropped to 27% for hispanic drivers.

    Racist drug dogs?

    No (none / 0) (#11)
    by Zorba on Fri Jan 07, 2011 at 01:14:56 PM EST
    Racist handlers.  Did you read Militarytracy's comments on the training of the dogs?

    "probable cause" is not certainty (none / 0) (#13)
    by diogenes on Fri Jan 07, 2011 at 05:45:12 PM EST
    "Most courts hold that during a traffic stop, if a properly trained dog alerts, it amounts to probable cause to search the vehicle. "

    If preponderance of the evidence is at least 51% likelihood of being so, then 27-44% correct surely represents probable cause from a legal standpoint.  

    Re: Probable Cause (none / 0) (#14)
    by Harry Saxon on Fri Jan 07, 2011 at 05:49:37 PM EST
    From www.lectlaw(dot)com

    A reasonable belief that a person has committed a crime. The test the court of appeals employs to determine whether probable cause existed for purposes of arrest is whether facts and circumstances within the officer's knowledge are sufficient to warrant a prudent person to believe a suspect has committed, is committing, or is about to commit a crime. U.S. v. Puerta, 982 F.2d 1297, 1300 (9th Cir. 1992). In terms of seizure of items, probable cause merely requires that the facts available to the officer warrants a "man of reasonable caution" to conclude that certain items may be contraband or stolen property or useful as evidence of a crime. U.S. v. Dunn, 946 F.2d 615, 619 (9th Cir. 1991), cert. Denied, 112 S. Ct. 401 (1992).

    Sounds more like 51% to me, doc.

    Click Me