Imagine that a police officer, after taking it upon himself to search someone's car, is asked to explain why he thought he would find contraband there. "A little birdie told me," he replies.
Most judges would react with appropriate skepticism to such a claim. But substitute "a big dog" for "a little birdie," and you've got probable cause.
Or so says the U.S. Supreme Court, which last week unanimously ruled that "a court can presume" a search is valid if police say it was based on an alert by a dog trained to detect drugs. The Court thereby encouraged judges to accept self-interested proclamations about a canine's capabilities, reinforcing the alarmingly common use of dogs to justify invasions of privacy.
Drug-detecting dogs are much less reliable than widely believed, with false-positive error rates as high as 96 percent in the field. A 2006 Australian study found that the rate of unverified alerts by 17 police dogs used to sniff out drugs on people ranged from 44 percent to 93 percent.
Police and prosecutors commonly argue that when a dog alerts and no drugs are found "the dog may not have made a mistake at all," as Justice Elena Kagan put it, writing for the Court. Instead it "may have detected substances that were too well hidden or present in quantities too small for the officer to locate."
This excuse is very convenient—and completely unfalsifiable. Furthermore, probable cause is supposed to hinge on whether there is a "fair probability" that a search will discover evidence of a crime. The possibility that dogs will react to traces of drugs that are no longer present makes them less reliable for that purpose.
So does the possibility that a dog will react to smell-alike odors from legal substances, distractions such as food, or cues from their handlers. Given all the potential sources of error, it is hard to assess a dog's reliability without looking at its real-world track record.
That is why the Florida Supreme Court, in the 2011 decision that the U.S. Supreme Court overturned, said police should provide information about a dog's hits and misses. "The fact that the dog has been trained and certified," it said, "is simply not enough to establish probable cause," especially when, as in most states, there are no uniform standards for training or certification.
Kagan, by contrast, minimized the significance of a dog's success at finding drugs in the field. She said police testing in artificial conditions is a better measure of reliability, even though handlers typically know where the drugs are hidden and can therefore direct the animals to the right locations, either deliberately or subconsciously.
Instead of requiring police to demonstrate that a dog is reliable, this decision puts the burden on the defense to show the dog is not reliable through expert testimony and other evidence that casts doubt on the training and testing methods used by police. But experts are expensive, and police control all the relevant evidence.
Police even determine whether the evidence exists. Many departments simply do not keep track of how often dog alerts lead to unsuccessful searches, and this decision will only encourage such incuriosity.
The Court previously has said that police may use drug-sniffing dogs at will during routine traffic stops and may search cars without a warrant, based on their own determination of probable cause. Now that it has said a dog's alert by itself suffices for probable cause, a cop with a dog has the practical power to search the car of anyone who strikes him as suspicious.
Even the question of whether a dog did in fact alert may be impossible to resolve if there is no video record of the encounter, which is often the case. As Florida defense attorney Jeff Weiner puts it, the justices "have given law enforcement a green light to do away with the Fourth Amendment merely by uttering the magic words, 'My dog alerted.'"