Is a dog-sniff reliable enough to convict an alleged criminal?

Dogs have been celebrated since antiquity for their ability to sniff a particular odor and lead humans to its source, reported Science. But the domesticated canine’s transformation into crime-fighting companion emerged much more recently, as U.S. police launched K-9 training programs and a thriving cottage industry of private firms, which often aid law enforcement, emerged. 

Today, police use dogs to track fugitives and search for missing persons, obtain probable cause (that is, legal justification to get a search warrant), and find substances, particularly illegal drugs. In what are known as scent lineups, agencies use trained canines to match evidence collected at a crime scene to the scent of a suspect or body. Increasingly, testimony from dog handlers has also served as direct evidence of guilt—accepted in lieu of an actual corpse, drug stash, or other physical evidence of a crime.

Yet critics worry that the criminal legal system has embraced a technique profoundly lacking in scientific validation. Dog-sniff evidence has led to wrongful convictions, and studies show human biases skew animal behavior. Almost no published research indicates just what dogs detect or how they do it. Defendants and their lawyers can’t cross-examine a dog, which means the accused cannot scrutinize the evidence or readily confront their accusers, a right enshrined in the U.S. Constitution.

“It’s not enough to say I have this amazing expert with an incredible nose who can distinguish between scents,” says Binyamin Blum, an evidence scholar at the University of California (UC) Hastings College of the Law, who contends such testimony short-circuits the safeguards in place to discriminate between junk science and real science. “You have to explain exactly what their method is.”

No comprehensive database exists about exoneration that involved a dog-sniff, but according to the National Registry of Exonerations, a project hosted by the University of Michigan Law School, at least 17 innocent people have been freed after dog-sniff evidence erroneously sent them to prison. Other convictions have been upheld—even after errors came to light.

Dana Delger and M. Chris Fabricant, attorneys from the Innocence Project’s strategic litigation department have did not question whether dogs could detect odors imperceptible to humans. Rather, they argued, “What is at issue is what has never been proven with any degree of scientific reliability: the ability of a dog to detect the residual scent of a particular object, including human remains, at a specific location days, weeks, months, or even more than a year after that object was removed.” A sniff could be used to corroborate, but they argued a dog’s indications alone should not be used to prove a person’s guilt.

Fabricant and Delger argued that a dog’s behavior may reflect a handler’s expectations, pointing to a 2011 study in Animal Cognition by Lisa Lit, then at UC Davis. Lit found handlers “cued” dogs into making false indications. In one test, Lit showed up each morning with evidence bags containing cannabis and gunpowder, explaining to 18 teams that those target scents might be present inside a church. No target odors were present, and yet dogs positively indicated 85% of the time, handlers said, suggesting the dogs served as loyal companions first and objective scent detectors second.

Similar issues came to light more recently in letters criticizing a 2018 study published in Forensic Science International. The experiment was designed to test the long-standing belief that dogs could “mantrail” unfamiliar suspects by their odor. In the study, dogs correctly trailed people 82% of the time. But earlier this year, several critics pointed out methodological shortcomings stemming from handler and experimenter bias, and the journal’s editors added a cautionary “expression of concern.”

To read more CLICK HERE