A few years ago, in Kyllo v. United States, the Supreme Court said police looking for the heat signature of marijuana grow lights need a warrant to examine the exterior of a home with a thermal imager. A few months ago, in Illinois v. Caballes, the Court said police don't need a warrant, probable cause, or even a vague suspicion to walk a drug-sniffing dog around a car. But what about a drug-sniffing dog outside a house? As Nick notes, the justices aren't saying. Yesterday they declined to hear a case in which a dog sniffed out methamphetamine in a man's garage while standing in his driveway.
We are left to wonder exactly what criteria transform surveillance into a "search" within the meaning of the Fourth Amendment. Homes generally get more protection than cars, so maybe the crucial issue is the property being searched–er, examined. (What about a guy who lives in his car?) Or maybe, as Harris County, Texas, District Attorney Charles A. Rosenthal argued in this case, what matters is the use of "technology" that enables police to detect what they otherwise could not see/smell/hear. According to Rosenthal, trained dogs don't count as technology.
The Supreme Court has suggested that it's the specificity of the surveillance technique that matters. In the 1983 case United States v. Place, it said a dog's sniff of luggage "discloses only the presence or absence of narcotics, a contraband item," while in Kyllo it noted that thermal imaging potentially reveals much more information than the presence of an indoor marijuana garden. In his Caballes dissent, Justice David Souter argued that the Court puts too much faith in the accuracy of dog alerts, which may (for example) indicate the presence of food, plastic bags, and other items used in training as well as contraband. Over at the Volokh Conspiracy, Orrin Kerr has noted another problem with a distinction based on the breadth of information revealed:
The Fourth Amendment traditionally has focused on how the surveillance occurred, rather than the nature of the information obtained. Under the traditional approach, the government could not invade your property without a warrant no matter what information it wished to obtain. Under the rationale followed by the Court [in Kyllo], the government may be free to invade your property so long as they only obtain "non private" information. This is particularly troubling in the context of computer searches and seizures. Can the police send a computer virus to your computer that searches your computer for obscene images, or images of child pornography, and then reports back to the police whether such images are on your computer–all without probable cause, or even any suspicion at all? The traditional answer would have been no: the police cannot enter your private property to search even for non-private stuff. But thanks to the increasing focus on the nature of the information rather than how the information is obtained, it's no longer so clear.