The Myth of the Infallible Police Dog
In Caballes v. Illinois, the 2005 Supreme Court decision that upheld the use of drug-sniffing canines during routine traffic stops, dissenting Justice David Souter noted that "the infallible dog…is a creature of legal fiction." Since false "alerts" seem to be fairly common, Souter warned, it's not safe to assume that signals from police dogs reliably indicate the presence of illegal substances, a premise underlying the Court's conclusion that a dog sniff does not count as a "search" for Fourth Amendment purposes. Now the myth of the infallible police dog is receiving attention in a new, even more troubling context: the "dog-scent lineup," in which a dog is expected to match a suspect's smell to olfactory evidence from a crime scene. Such tests, which can be compromised by cross-contamination of samples or by involuntary cues from handlers or detectives, do not merely provide pretexts for otherwise illegal searches; they can put innocent people in jail.
A front-page story in today's New York Times highlights several such cases in Texas. One man spent eight months in jail after a dog-scent lineup falsely implicated him in a triple homicide. Another was locked up for nine months after a police dog allegedly connected him to a series of robberies, even though security camera footage showed someone else. In both cases, the dog handler was Fort Bend County Sheriff's Deputy Keith A. Pikett, who consults widely throughout the state. The Innocence Project of Texas estimates that 15 to 20 people are in prison based mainly on Pikett's testimony, even though the FBI cautions that dog-sniff tests "should not be used as primary evidence." Rex Easley, a lawyer representing a man who was falsely accused of murdering his neighbor based on one of Pikett's lineups, calls the deputy a "charlatan" who "devised an unreliable dog trick to justify local police agencies' suspicions." The head of a British police dog unit who reviewed footage of the lineup that implicated Easley's client told the Times, "If it was not for the fact that this is a serious matter, I could have been watching a comedy." Pikett declined to be interviewed for the Times story.
I discussed Caballes in a 2005 column. Julian Sanchez considered the implications of its reasoning in a 2007 Reason article. This year Radley Balko noted the incredible forensic feats performed by police dogs under the supervision of the late Florida canine handler John Preston.