Even if a drug-sniffing dog is wrong about what he smells almost half of the time, police and courts are still allowed to rely on that dog to trigger a constitutional search of a car. That's what the U.S. Seventh Circuit Court of Appeals said this week.

In this particular case, police officers in Bloomington, Illinois, searched a car for drugs and found 20 kilograms of cocaine. As a result, the driver of the car, Larry Bentley Jr., is serving 20 years in prison for drug possession. The search was legal because the drug-sniffing dog, Lex, had been alerted to the presence of drugs. 

U.S. Army Corps of Engineers / Flickr - flickr.com

On appeal of his case, Bentley argued that the dog's alert was not reliable enough to trigger the probable cause required for a constitutional search of his car. For one, Lex wasn't very good at drug sniffing. In testing, he was accurate only 59 percent of the time -- not much better than a coin flip, as the court admitted. Lex was also trained in a way that encouraged him to act as if he'd found drugs. That was because the dog was given a treat, "a rubber hose stuffed with a sock," each time he alerted. So, understandably, Lex alerted a lot. Ninety-three percent of the time to be exact.

Even though the court agreed that Lex wasn't a great drug-sniffer and had problematic training, the court still said Lex could be relied upon to trigger a constitutional search not in violation of the Fourth Amendment. The court said that 59 percent "is good enough to support a finding of his reliability and thus to allow his alert to constitute a significant piece of evidence supporting the ultimate conclusion of probable cause."

We have low standards for drug-sniffing dogs.

Why is the bar so low? Because it was set low by the U.S. Supreme Court. In 2013, in a unanimous decision written by Justice Elena Kagan, the court ruled that law enforcement need only show that a dog has been trained or certified in order to use that dog's alert to trigger a constitutional search. The problem, according to libertarian outlet Reason, is that there is no universal standard for training dogs. And you can tell by the results. In 2011, the Chicago Tribune looked at three years worth of dog-sniffing data from Chicago's suburban police departments. They found that over those three years, "only 44 percent of those alerts by the dogs led to the discovery of drugs or paraphernalia."

What's worse, according to the Washington Post's Radley Balko, is that law enforcement has an incentive to use the low bar of drug-sniffing dogs to create opportunities for more searches and also more asset seizures.

"A dog prone to false alerts means more searches, which means more opportunities to find and seize cash and other lucre under asset forfeiture policies," Balko wrote this week in the Post. "In fact, a drug dog’s alert in and of itself is often cited as evidence of drug activity, even if no drugs are found, thus enabling police to seize cash, cars and other property from motorists."

When drug-sniffing dogs lead to anal probes.

In two high-profile cases in New Mexico, searches triggered by false positives from police-sniffing dogs led police to conduct anal probes on suspects, both of whom had no drugs. There was the case of Daniel Eckert in 2013. He was stopped in a Wal-Mart parking lot after a dog was falsely alerted to drugs. Eckert was brought back to a hospital and subjected to multiple anal searches. He sued the city of Deming and Hidalgo County and was awarded $1.6 million. Another victim, Timothy Young, was also anally probed in Hidalgo in 2012 after a false-positive dog alert. Reason put together a video about his case:

Using drug dogs in schools.

This is not a new problem, especially in New Mexico. In 2005, CBS' "60 Minutes" covered the story of Lezley Whipple of Lordsburg, New Mexico. Whipple attended a high school where a drug-sniffing dog made rounds in the student parking lot, sniffing for contraband. One day, the dog sensed something in Whipple's car and alerted. The problem was that this dog also had a pretty poor accuracy rate -- having been falsely alerted to drugs 71 times of out 72 total alerts. Nevertheless, the school relied on the dog and removed Whipple from class that day. They also began to search every student.

"There would be an announcement over the intercom. And they would say, 'We're on lockdown. Nobody's to leave the building or the room.' And the teachers would shut the door and we all had to sit in our seats," Whipple told 60 Minutes. "And we'd just wait there in class. Class stopped actually. And then the dog would go up and down the rows."

Whipple had no drugs in her car. Her parents sued, and the district eventually agreed to stop using the dog, despite their initial plans bring drug-sniffing dogs into elementary schools.

"Personally, I was offended by the concept of drug dogs coming inside of the classroom where the kids are forced to remain in their seats and sniffing the kids," said Whipple's attorney Jane Gagne. "And legally speaking, the Fourth Amendment of our constitution prohibits something like that, or should prohibit something like that."