The use of drug detecting dogs in law enforcement is ubiquitous across the country. They are a popular tool among police agencies, because a drug dog’s “alert” provides the probable cause necessary to legally search a vehicle without warrant or permission. But are these alerts a response to the smell of illegal drugs or a response to unconscious cues from a handler?
As noted in an article on TechDirt.com, dogs like to please their handlers. They are also highly sensitive to behavior that leads to reward. While there is no denying the power of a dog’s sniff, there is a worrisome shortage of data on what, exactly, a drug dog is doing when it alerts.
According to the TechDirt.com article, “there’s a deliberate dearth of data when it comes to drug-sniffing dog fallibility. Tracking this data would undercut the dogs’ raison d’etre: to act as probable cause for warrantless searches.” Of course, the lack of hard data “makes challenging drug dog ‘alerts’ in court almost impossible.”
One study, reported more than seven years ago by NPR, showed that drug dogs tend to respond to handler cues more than to actual drugs. Researcher Lisa Lit’s study was roundly denounced by law enforcement, which refused to provide any further assistance to researchers. Lit referred to her finding as a “career killer,” for its likelihood of disproving popular theory on drug dogs.
Fred Helfers, of the Pacific Northwest Police Detection Dog Association, aims to ensure that handler bias is eliminated from the process of K9 drug detection. Helfers’ recertification process uses a scientifically rigorous method. When drug dogs are tested, the location of drugs is determined by the roll of a die, and the handler is not informed of the location. Sometimes there are no drugs at all. Helfers said that the lack of drugs has tripped some teams up. “There were some teams that failed that sequence,” he said, “because they didn’t trust their dog.”
What they didn’t trust was that there were no drugs, a situation that suspicious officers encounter daily. But they worked their dogs until they alerted to nonexistent drugs. “I think they ‘overworked’ the car,” Helfers said. “Instead of going around once or twice and trusting their dog and watching their dog work, maybe they’d seen something that wasn’t there.”
Some K9 officers are so firm in their belief that their dogs are alerting to drugs (and not them) that they don’t accept any other possibility—even when drugs aren’t found after a dog alerts. Case in point is K9 Officer Gunnar Fulmer with the Walla Walla Police Department. Fulmer declares that the dogs “are mainly used to confirm what we already suspect.” He boasts that “about 99 percent of the time we get an alert.” He attributes that suspiciously high figure to the fact “it’s because we already know what’s in the car; we just need that confirmation to help us out with that.” He clearly exhibits confirmation bias, but he doesn’t even realize it.
Police resist and reject empirical studies on the accuracy of drug dog alerts because they “want to have free rein to allow their hunches to develop into warrantless searches with the assistance of animals prone to responding to handlers’ cues, rather than the existence of contraband.” The limited data available show that drug dogs (or more accurately) their handlers “aren’t to be trusted—not without more data.” In light of the available evidence, drug dogs should not be considered “probable cause on four legs.” As they are currently being used, “they’re walking confirmation bias—self-serving tools of civil liberties circumvention.”