* The Washington Post’s Radley Balko takes on a topic we discussed yesterday: The future of drug dogs if marijuana is legalized. Here’s part of his take…
I’ve written quite a bit about drug dogs in Illinois, and it turns out they’re pretty terrible at detecting drugs. In 2011, the Chicago Tribune published a review of drug dog searches conducted over three years by police departments in the Chicago suburbs. Just 44 percent of dog alerts led to the discovery of actual contraband. For Hispanic drivers, the success rate dipped to 27 percent. The following year, I obtained the records of an Illinois State Police drug dog for an 11-month period in 2007 and 2008. In nearly 30 percent of cases where the dog “alerted” no drugs at all were found. In about 75 percent of cases, the dog alerted either to no drugs or to what police officers later described as “residue,” which basically means no measurable quantity of a drug and not a significant-enough amount to merit criminal charges. Only 10 percent of the alerts resulted in a seizure of a large-enough quantity of drugs to charge someone with a felony.
This is pretty consistent with statistics from other states, as well as one fascinating academic study, which have shown that drug dogs are far more likely to merely confirm the hunches and suspicions of their handlers than they are to independently detect illicit drugs. The dogs’ high error rates often make them no more accurate than a coin flip. The problem of course is that the entire purpose of the Fourth Amendment is to protect us from searches based solely on a government official’s hunch or suspicion. There’s a reason some legal scholars call drug dogs “probable cause on a leash.”
The K9 trainers I’ve interviewed over the years have told me that drug dogs could actually be trained to only alert when there is a significant quantity of an illicit drug — that is, to ignore “residue.” The reason they aren’t is that police departments don’t want them trained that way. They want dogs that alert as often as possible. They want the dogs to err on the side of false alerts.
Why would police want a dog that falsely alerts? That’s the exact question the late Supreme Court Justice Antonin Scalia asked in a drug dog case a few years ago. The answer is incentives. Searches can lead to evidence of other illegal activity. One incentive is that police officers, particular those in drug enforcement, often evaluated based on the raw numbers of arrests. More searches mean more opportunities to make arrests.
But the more important incentive is civil-asset forfeiture. If the police find even the slightest bit of pot, sometimes even just residue, they can often justify taking a driver’s cash, jewelry or even the car itself. The owner of the property — even if completely innocent — then must endure a number of legal and procedural barriers to getting the property back.
*** UPDATE *** Thanks to a commenter for pointing this out…
The police dog trainer who said he believed drug-sniffing dogs would have to be euthanized has retracted his statement, according to his police chief.
Decatur Police Detective Chad Larner told the Bloomington Pantagraph earlier this week that if marijuana is made legal, the dogs trained to detect narcotics would have to be euthanized because they could not be retrained.
However, many experts have stated this is not the case, and Decatur Police Chief James Getz Jr. told the Chicago Tribune on Wednesday that Larner wishes he hadn’t made the statement, calling it “a bad choice of words.”
- Ali Nagib - Thursday, May 10, 18 @ 10:00 am:
In fairness to the gentleman whose comments about euthanizing police dogs led to all this attention, his statement has apparently been retracted:
http://www.bnd.com/news/local/article210810369.html
- Collinsville Kevin - Thursday, May 10, 18 @ 10:11 am:
Instead of being killed, perhaps they could be put out to pasture.
- @misterjayem - Thursday, May 10, 18 @ 10:11 am:
You can read about that study here: https://www.economist.com/blogs/babbage/2011/02/animal_behaviour
Spoiler alert: 15% accuracy.
– MrJM
- Payback - Thursday, May 10, 18 @ 10:18 am:
Related to me by an Air Force veteran and family friend who I had known since childhood circa 1982, it was common knowledge in the enlisted barracks that the dog handlers could signal the dog to false “alert” if they wanted to, by jerking on the leash in combination with verbal commands, etc.
His bunkmate was one of the dog handlers, and the guys in his barracks would get at least a 24 hour notice if a search was coming down. This lead to a humorous side effect, the guys would dump their bags of pot out the window. Around the border of the barracks building their was a flower bed, and the pot seeds would grow up, which an inspecting officer noticed, wondering how in the world marijuana plants could be growing in his garden on Okinawa!
Drug detection dogs are just another tool of the police state based on lies and deception, like so many aspects of modern law enforcement: unmarked cars, plainclothes police, roadblocks, etc. Don’t ever count on cops to go by the Constitution, most of them haven’t read it since high school.
- a drop in - Thursday, May 10, 18 @ 10:19 am:
-Larner wishes he hadn’t made the statement, calling it “a bad choice of words.”-
Is that what they are calling falsehoods these days?
- @misterjayem - Thursday, May 10, 18 @ 10:20 am:
“a bad choice of words.”
But when they testify about their drug dogs signaling an alert, ya better gotta believe ‘em.
– MrJM
- Mama - Thursday, May 10, 18 @ 10:29 am:
“Spoiler alert: 15% accuracy. ”
I am shocked! I thought the police dogs would have at least a 50% success rate. Is this low success rate due to bad training or to many variables?
- Homer J. Quinn - Thursday, May 10, 18 @ 10:48 am:
cops lie. even the four-legged ones, even the detectives and chiefs.
- @misterjayem - Thursday, May 10, 18 @ 10:48 am:
“Is this low success rate due to bad training or to many variables?”
In the study, it was due to the handlers genuinely believing there was something that the dog *should* alert to, when there wasn’t, and the dog acting in response to the handlers’ genuine but incorrect belief.
In other words, the handlers were (incorrectly) alerting the dogs rather than the dogs alerting the handlers.
– MrJM
- Generic Drone - Thursday, May 10, 18 @ 11:07 am:
So how many illegal searches took place? How many cases should be overthrown due to false searches? If lie dectectors can’t be used due to inaccurate readings, then why are judges supporting these searches to be allowed. Why are no defense attorneys challenging these?
- Anon - Thursday, May 10, 18 @ 12:56 pm:
I spent 2 hours with a Sangamon County Deputy who had the dog walk around my car and apparently hit. I was driving my parents car and I know they don’t do any drugs.
Well needless to say the cops searched through everything and didn’t find a thing but the cop absolutely thought I did so he didn’t let up and I am fairly sure the dog didn’t hit it just was mirroring the cops excitement.
In the end the cop did NOT appoligize for wasting my time. I sent a complaint in but nothing ever came of it.
I’m all for police having dogs but we need to stop this war on marijuana and start catching real criminals.
- Han's Solo Cup - Thursday, May 10, 18 @ 12:59 pm:
What the story leaves out is that the residual odor of cannabis and other drugs can stay in a vehicle or on a person’s clothes for weeks. Hence, a positive alert which results in no drugs found.
- Demoralized - Thursday, May 10, 18 @ 1:28 pm:
I had a dog hit on my suitcase at the airport once. I had some licorice in the pocket the dog hit on. Perhaps that’s why I always feel so happy when eating licorice?
- wondering - Thursday, May 10, 18 @ 1:38 pm:
Much cheaper to cage the dogs than the people.
- Amalia - Thursday, May 10, 18 @ 2:06 pm:
More fodder for the late night comedy shows.
- Da Big Bad Wolf - Thursday, May 10, 18 @ 4:41 pm:
So, the dogs being expensive and all, couldn’t we substitute Enee meenie meiney mo, and save the state money with the same amount of accuracy?