The Supreme Court Case of Illinois v. Caballes

See the above article. In particular, read the section "Justice Souter's Odd Dissent".  Part of this section reads as follows:

"Suppose, for example, that police dogs have an average false positive rate of one in four. That simply means that when a dog alerts, the human officer knows there is a seventy-five percent chance that illegal drugs are present. The dog sniff still reveals no information whatsoever about other, legitimately private, activities of the target of the sniff."

The alert by the dog was considered to provide the probable cause necessary to allow the cops to search the guy's trunk. It seems to me that the unlawful search was not the dog's sniff, but rather the opening of the trunk by the police officers. When any reasonable person takes the time to sit down and seriously look and the numbers, he/she will see that this is so.

Suppose I live in a community of one hundred people. If marijuana averages 1%, then it is perfectly reasonable to expect that one member of that community uses marijuana and might have some in his car. Does this knowledge constitute probable cause to search the cars of all the people in that community? Most people would say no.

Suppose a dog with a false-positive rate of one in ten sniffs all the cars of the people in that community and alerts on ten of the one hundred cars. A new subset of that community has been created, consisting of the owner's of the ten cars on which the dog alerted. It would be perfectly reasonable to expect that one member of this subset of ten people uses marijuana and transports it in his car. Does this knowledge constitute probable cause to search the cars of all ten people in this subset?

Before the sniff, there is a 1% chance of finding drugs in any randomly selected car. After earning an 'alert' by the dog, there is a 10% chance of finding drugs in the car on which the dog alerted. I don't know about you, but the word "probable" usually means better than 50% in my world!

Besides, the argument that police dogs are not used ranomly is simply untrue. Although they may not be used randomly in routine traffic stops, they are certainly used randomly in airports and in schools. Drug raids are becoming increasing common in the governement controlled 'public' schools in the United States. Consider what happened just a couple of years ago in my state of South Carolina:

In this case, the police burst through the doors of a high school with guns drawn. Those kids (107) who were unlucky enough to be in the hallway at the time were subjected rough handling by the police. Dogs were brought in to search the school and the book bags of those kids who were in the hallway at the time of the raid. The dogs 'alerted' to twelve of the bags, and the police used these alerts as probable cause to search those twelve bags. No drugs were found in the bags or anywhere else in the school.

So dogs have a really great sense of smell. So what? No matter how well a police dog is trained, he is still just a dog and has absolutely no way of telling anyone just what it is that he smells. Is it marijuana? Is it a fart? Is it the scent of a female dog? Is it beef jerky? Or is the dog merely alerting because he thinks that this is what his master wants? Dogs really don't communicate with us very well, and vice-versa. As hard as it is for a 'smart' human to divine what a dog is telling him, imagine how difficult it for the relatively unintelligent dog to divine what it is that his master wants of him.

But still, what the Hell do they smell? Although marijuana is relatively aromatic, most other substances that dogs are allegedly trained to alert to are not aromatic at all. Dogs alert to cocaine ... but they don't even smell cocaine: Cocaine is a local anesthetic and as such blocks the transmission of olfactory nerve impulses to the brain.

A guy named Ken Furton who routinely serves as an expert witness for the prosecution in drug dog related arrests believes quite strongley that the primary substance that gives rise to cocaine alerts is actually methyl benzoate - a degradent and likely metabolite of cocaine. But methyl benzoate is common as hell. It's found in fruits and cloves, and is used in flavors, perfumes, and fragrances. Doesn't this sort of mean that a dog that alerts to cocaine is essentially useless as a drug-detection device?

They're good, but they aren't that good, and treating these creatures as infallible results in some pretty unfortunate misconceptions. The truth is that these dogs are 'certified' by passing a minimum number of tests under conditions very unlike the 'real world'. Yet, the concept of a false-positive seems lost on the K9 cop. Instead, these guys view all alerts as true alerts, and therefore perceive their dogs as never erring.

When dogs alert falsely during school 'raids' (often called 'lock-downs' ... are they schools or prisons?), the unfortunate child who is the recipient of the false positive inevitably is sent home with a note informing the parent that the child has been 'in contact' with drugs. When dogs falsely alert to currency, the currency is inevitably seized as 'drug money', never mind the fact that dogs apparently alert to most currency, giving rise to the myth that most currency is contaminated by drugs. This may very well be true of course. But then again, most currency is undoubtedly contaminated by chocolate, fecal matter, urine, sweat, and snot. Am I to believe that the dog can detect that trace quantity of cocaine (I mean methyl benzoate) in the midst of all this other crap?

On close scrutiny, drug dogs do not seem to be particularly reliable. Yet they are used by the police and other agents of the government to harass innocent civilians. This subject, I think, needs to return to the Supreme Court.