As Radley Balko noted this morning, a recent Chicago Tribune investigation found that drug-sniffing dogs used in traffic stops by Illinois police departments were wrong more often than they were right. In 56 percent of the cases where the dogs "alerted" to cars, police found no drugs or drug paraphernalia. The dogs' defenders say they may be detecting traces of drugs that used to be in the cars, but even they concede that is not the whole story. As detailed in the Tribune article, poor training of dogs and their handlers, coupled with cops' unconscious signals to the animals, seems to account for a large portion of these fruitless searches. "The dogs are only as good as the handlers," one expert tells the Tribune. A Republican state legislator (and former prosecutor) who wants to create certification standards for drug-detecting dogs calls them "probable cause with four legs."

That's because the Supreme Court has said a dog sniff, though it does not qualify as a "search" within the meaning of the Fourth Amendment, can be used to justify a physical inspection that does. In the case of a traffic stop, that means detaining a motorist for a half-hour or more while cops pull apart his car, looking in the trunk and the glove compartment, checking under the rugs, digging into the seats, going through bags, and attempting to locate hidden compartments. This is a pretty inconvenient, time-consuming, privacy-invading, and humiliating ordeal to impose on someone if the only basis is a tip from an animal who is usually unreliable.

The Supreme Court's decisions in this area assume otherwise. "A dog sniff conducted during a concededly lawful traffic stop that reveals no information other than the location of a substance that no individual has any right to possess does not violate the Fourth Amendment," wrote Justice John Paul Stevens for the majority in Illinois v. Caballes, a 2005 decision upholding the use of drug-detecting dogs during routine traffic stops. That decision built on United States v. Place, a 1983 ruling that said "subjecting luggage to a 'sniff test' by a well-trained narcotics detection dog does not constitute a 'search' within the meaning of the Fourth Amendment" because it "discloses only the presence or absence of narcotics, a contraband item."

These decisions assume that dogs are furry machines that virtually never malfunction, indicating the presence or absence of drugs with something like 100 percent accuracy. But as Justice David Souter, one of the two dissenters in Caballes, pointed out, "the infallible dog...is a creature of legal fiction." Souter cited examples from court cases of dogs with error rates of up to 38 percent. "Dogs in artificial testing situations return false positives anywhere from 12.5 to 60% of the time," he added. The Chicago Tribune study provides further evidence that Place and Caballes are based on a myth. The Supreme Court is allowing police to search people's luggage and vehicles based on a probable cause generator that may be wrong most of the time.

I discussed Caballes in a 2005 column. Julian Sanchez teased out some of its implications in a 2007 Reason cover story. More on the myth of the infallible police dog here and here.