The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
The Fourth Amendment and the "Instinctive" Drug Detection Dog
This one is a search.
Lower courts are divided on the Fourth Amendment implications of a drug detection dog that jumps into a car on its own and then alerts to illegal drugs. I thought I would offer some thoughts on the problem. In my view, unprompted entry should be deemed a Fourth Amendment search. This post explains why.
First, some context. It's settled law that use of the drug sniffing dog to sniff in the area outside a car is not a search. See Illinois v. Caballes, 543 U.S. 405 (2005). It's equally clear that a search of a car occurs if the officer directs a drug detection dog to physically enter the car, the dog enters, and then the dog alerts. What courts struggle with is how to treat the dog that jumps into the car unprompted. If the officer wants the dog to stay out of the car, and it's the dog's own idea to enter the car, do you say that the dog's entry is a Fourth Amendment search attributable to the government? Or is the dog sort of its own independent actor whose instinctive conduct is not attributable to the government, and therefore its subsequent alert is not a search? See, e.g., United States v. Sharp, 689 F.3d 616(6th Cir. 2012) (holding the latter).
It seems to me that this is a question that should have a straightforward answer: An unprompted and instinctive entry by the dog, followed by an alert, is a government search. The drug detection dog is a technological tool that the government trains and brings to the scene to detect narcotics. Like most tools, it is imperfect. It doesn't always go where the government wants it to go. But it makes no sense to say that the dog loses its government character or is somehow not engaged in government action when, being led around a car, it jumps into it. If the dog enters unprompted and alerts inside the car, the government certainly uses that alert for investigatory purposes just like it would if the officer had directed the dog to enter the car. Whether the officer prompted the dog to enter shouldn't matter.
The key precedent that comes to mind is United States v. Karo, 468 U.S. 705 (1984). In Karo, the government secretly put a radio beeper in a can of chemicals sold to a narcotics ring, and they watched to find out where the bad guys carried out their crime by watching where the can went using the beeper. As long as the can stayed on public roads, there was no search under a prior case, United States v. Knotts, 460 U. S. 276 (1983). But what made Karo different is that someone—presumably one of the bad guys—brought the can into a house. The beeper thus registered its location from inside the house instead of on a public road. Karo deemed that alert from inside the house a search for the simple reason that the beeper was transmitting information from inside the home.
But wait, the government protested in Karo: They weren't in control of where the beeper went. The bad guys had brought the beeper into the house, not the government, so it couldn't predict when a search would occur and a warrant would be needed; "they have no way of knowing in advance," the government noted, "whether the beeper will be transmitting its signals from inside private premises." The Supreme Court was unmoved by this argument, basically telling the government that the unpredictability of the beeper's location was its own problem to deal with: "The argument that a warrant requirement would oblige the Government to obtain warrants in a large number of cases is hardly a compelling argument against the requirement."
It seems to me that drug-sniffing dogs in these "instinctive act" cases are a lot like the radio beeper in Karo. In both situations, the government has introduced surveillance tools to obtain information not otherwise known using normal human senses. As long as the surveillance tools stay outside the protected area, their use is not a search under Caballes and Knotts. In both situations, that Fourth Amendment rule gives agents a reason to want the tools to go near places but not go inside them. But in both situations, agents don't have perfect control, and they run the risk that the devices will go in the protected areas of the homes (in Karo) or car (in the dog cases).
Under Karo, if a suspect brings the the beeper into a home, the Fourth Amendment protection changes. Because the beeper is now in the house, a search occurs. I don't know why there would be a different result using the technological tool of a dog as compared to a technological tool of a beeper, especially given that the government has so much more control over where the dog goes as compared to where the beeper went in Karo.
Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Comments may only be edited within 5 minutes of posting. Report abuses.
Please to post comments
Isn't, realistically, the bigger problem with drug dogs the Clever Hans effect? The dog is better at figuring out what it's handler wants, than it is at detecting drugs, and if the handler wants it to enter the building, it will.
The dog isn't really a technological tool, IOW.
I'd go one step further and argue that the handler ordered the dog to jump into the vehicle via nonverbal cues.
Wouldn't the state have the burden to prove the officer didn't, and how the hell would the state ever prove that.
Thank you. I was about to bring up Clever Hans. I've been in the Skeptic movement way longer than here.
A hundred years ago, an amazing horse named Clever Hans was doing rounds, answering math questions. He's tap his hoof out the correct number of times. The idea was perposterous, but how to prove it was fake?
They obtained films with sound, and, with the sound off, thus not even knowing the question, investigators could arrive at the correct answer by noticing the subtle cues by the trainer, which occured when the trained horse, tapping its foot, arrived at the correct answer, so it stopped.
It's possible, to perhaps give more credit than is due, the handler, horse or drug dog, doesn't even realize what they're doing as part of the effect.
One could theoretically run a similar trial to see if an investigator could "sniff" drugs in a car or bag just by watching videos of the K9 handler during positive and negative test runs.
Exactly right.
And any handler who knows what he's doing could train the dog to respond to subtle cues which onlookers would easily miss but the dog know means "jump into the car" or even "act like you're detecting drugs whether you really are or not".
The 4th amendment safeguards are there to protect members of the public and should not be circumvented by using authorities who use trained animals.
Exactly. A skilled trainer can get a dog to do whatever the trainer wants it to do.
But, if what Prof Kerr argues for here becomes settled law, the trainer will not want to direct the dog to enter the car since that would be an unwarranted search.
Expanding on Brett's point, given how easy it is to manipulate a drug-sniffing dog, why should this be grounds for probable cause? "Yeah, we got a guy with a dog on staff who will give us probable cause anytime we want it."
Bingo. The solution to this problem is that a drug sniffing dog IS A SEARCH, because ITS A SEARCH.
SCOTUS got this wrong.
There are two "good" articles on drug-sniffing dogs and Clever Hans. I've left off the http to sneak them past Reason's comment filter.
shunauto.com/article/how-often-do-drug-sniffing-dogs-falsely-indicate-a-car
http://www.bjjohnsonlaw.com/probable-cause-or-probable-fiction-the-legal-myth-of-the-drug-sniffing-dog
It's a huge practical problem, but only Souter seemed to care. Florida v. Harris seemed to kill any real argument challenging drug dog reliability.
The claim the behavior is instinctive or initiated by the dog is simply not supportable by evidence. Dogs in the wild do not react to drugs this way. They behave this way solely because they are trained to do this by the government. Because dogs on their own would never engage in this behavior, it is caused by the government. The dog is simply not acting on its own. It is acting under government direction.
This is so whether or not the handler gave unconscious cues in the specific instance. The entire general tendency to engage in this behavior is something the dog is doing only because of the government. It is not something a dog would ever do on its own.
"It is not something a dog would ever do on its own."
Don't know much about dogs, do you?
A wild dog might jump in a car. But specifically in cars that contain specific drugs and not other cars?
Depends on the Dog, some like Drugs, some don’t. When I used to smoke (the) Pot (Had to, the Glaucoma) our Pomeranian would roll around like a puppy in the flecks of Bud that would get on the carpet.
Got to where she quit her job, sat around eating doggy treats and watching “All dogs go to Heaven” all day.
Much better now.
Mitzi I mean, sometimes the Glaucoma flares up and I have to “treat”it.
Funny it only flares up in Colorado, California, Oregon….
The dog is acting on its own, but has a general motivation to please his handlers.
IOW, it's not so much that the dog is a tool, but an agent of the government.
By definition, a dog cannot have agency. I think you would have to analyze the dog as a tool, sort of an AI. As Readery said a dog only acts this way because of its training.
The entry should be attributable to the instructions given to it by its handlers, or better stated, the failure of the dog to conform its behavior to constitutional norms should be attributable to the government.
It's not that far removed from bringing kids along with you. Would the deputy not be responsible if his kid ran into a car and started searching it? You know how kids are, they just have a mind of their own.
There are no bad dogs, only bad trolls.
Bad Rex! BAD!
What? You found something?
Well, since we're all here we might as well take a look...
Yeah; pretty much. In your 3-line play, I think you summed it up nicely. (It's pretty rare, here at the VC, to get a comment thread where EVERYONE on the entire political and ideological spectrum looks at an issue and immediately concludes, "Um, yup...this stinks to high heaven.")
If I am inside the vehicle the dog jumps into and I punch the dog in the nose, do I commit ordinary animal cruelty or the more serious crime of assault on a police animal? I'm sure the prosecution will say the dog was on the clock.
Animal Cruelty is a predictor of significant Psychopathic behavior, most of your really horrible Criminals, Hannibal Lecter(Late/Great), Ted Bundy, James Comey started that way. Might want to make sure Dexter Morgan doesn’t see you.
Frank
Punch the dog - animal cruelty.
Shoot the dog - promotion to Secretary of Homeland Security.
I was going to say Youth-enizing an Animal is different, but it’s not, it’s worse, hey at least now she’s Youth-entizing Ill-legals.
Why wouldn't it be self defense?
Or is the dog sort of its own independent actor whose instinctive conduct is not attributable to the government, and therefore its subsequent alert is not a search?
There's also the issue that comes up here from time to time, where a laptop repair shop discovers illegal pictures, shall we say, and calls the police.
The principle described was if the shop found it themselves, ok. Same if a cleaning crew sees a huge bag of drugs on the coffee table. But if there's a "too cosy" relationship with police, then the shop becomes a functional reporting arm of the government. I'll leave it to the lawyers for the details.
This is an even bigger deal as gigantinc corporations like Apple and Microsoft could trivially scan and report to police, in an industrial strength automater spy pipeline the likes of which George Orwell could scarcely imagine.
It seems clear, in this dog instance, there's simply a too cosy relationship.
A humorous descriptor was not just pay, but shop owners regularly attending police balls. I dare say the K9s regularly make appearances as well, to say nothing of rewards for reporting, good boy, here's a Milk Bone MaroSnack!
100%. The dog is an instrument of the government, trained to do exactly that. Who cares if it does it at the "wrong" time? Plus, as a practical matter, allowing such "accidental" entries just encourages cops to "accidentally" let the dog into the car.
The dog is a cop. A trained cop.
The dog isn't acting 'on his own' he is doing cop stuff.
No different than if a human cop "jumps" into your house without a warrant "on his own".
Alito would find for the cops, of course.
My thinking is similar to what others have already said:
When the government uses a police dog, it’s effectively extending its reach. If that dog sticks its nose past the door and into someone’s car, that’s a search. It doesn’t matter whether the dog jumped in on its own. Choosing to use an animal that might do that is the government’s choice, so the government bears the risk.
People are entitled to draw a bright line at the threshold. Outside is one thing, inside is another. If the courts bless “self-entry,” we create a loophole around warrants: train the dog to be eager, look the other way, and pretend it was an accident, i.e. Clever Hans.
A simple rule: cross the line and it’s a search. This matches common sense and long-standing law about doors and thresholds. Keep the nose out, or get a warrant.
The search also fails under the test I'd prefer: is this behavior what we'd permit from any random person passing by.
Can someone walking their dog let it sniff at your tire? Yes.
Can someone walking their dog let it jump through your car window or walk on your hood? No.
Yet if a rando walking down the street actually broke your car window, searched, found drugs, and reported them, those drugs would not be suppressed under the 4A because the person is not a government actor.
On the other hand, I don't have any drugs, and the rando would have committed at least a serious misdemeanor and would also have civil liability for the damage. Unless he was literally working for the police they'd be unsympathetic to any claim he was helping fight the war on drugs.
Deterrence Matrix:
Innocent searchee, rando searcher: Real criminal and civil liability.
Guilty searchee, rando searcher: Mostly theoretical criminal and civil liability, since I'm not likely to be going to the cops and saying he took my drugs.
Innocent searchee, LEO searcher: Mostly theoretical criminal and civil liability, since cops are unlikely to arrest each other and are slathered in QI.
Guilty searchee, LEO searcher: Exclusionary rule.
The third one is the problem. Insufficient deterrence for the core group that should be protected.
I like that addition.
I like this logic.
I would go further and argue that Caballes was wrongly decided - that using a technological tool (including a dog) to look for things that humans senses can't detect is a search of the interior even when you're only detecting things that "leak out" of the interior.
Yes, that means that every use of a drug-sniffing dog is a search requiring a warrant. That's a feature, not a bug.
Your correct characterization of it is what makes it not a search. The dog is sniffing outside the exterior of the car. That is a place where nobody has a property right or a reasonable expectation of privacy.
Regardless of the means or technology used a person does not have standing to complain that his rights were violated by any search in that area.
Now I would agree with the gloss that Rodriguez puts on it. If the police force you to keep your car there for a delayed period of time so that they can conduct such a search, then they have improperly seized you.
Isn't there a similar test as to what Prof Kerr is asking for in United States v Jones, 565 U.S. 400, 406 (2012)? In Jones, the Court held that a Fourth Amendment search occurs 'whenever a government actor physically intrudes into a constitutionally protected space with an intent to obtain information.'
Clearly use of the drug dog is intended to obtain information (whether the dog can detect odor of contraband). Warrantless searches w/out consent of a vehicle and without PC are generally unreasonable, are they not? The vehicle exception gives police lots of leeway no doubt. But the reason the cops use the dog is because they don't have reasonable suspicion or PC to search without consent because if they did, the dog would not be needed to provide it.
Its kinda bullshit to me that a routine traffic stop can be transformed into a drug investigation when there is no information known to the cop (before the k-9 sniff) that the stop IS a drug investigation; but the precedent is what it is. Cops run drivers license, see past association of criminal activity or drug possession...call the k-9 when other cases say that past criminal history alone is NOT present PC to believe the person is breaking the law today. It's a jumbled mess.
You are correct as a statement of the current law. You (and SCOTUS) are badly wrong as a normative statement of what the law should be. I argue that you do have a reasonable expectation of privacy from collecting evidence of the insides of your property based on emanations to the outside that can only be perceived through the use of advanced technology (including dogs). Any lower standard makes your "reasonable expectation" of privacy a function of available technology that eventually erodes to nothing.
Consider - a century ago, walls blocked visible light making walls and drapes sufficient to require a warrant to look inside. We invented infrared detectors and can now look through walls - and consider that legally acceptable because it's the heat "escaping" and because that was relatively low-resolution, expensive and rarely used. Now it's ubiquitous and easy, though still low-resolution. What happens when neutrino detectors get cheap?
I agree with the consensus here*. Dogs, as others have pointed out, are astonishingly sensitive to human cues.
A drug-sniffing dog should be regarded, for these purposes, no differently than any other tool the cop might use.
And of course that sensitivity leads to the problem of the "alert" being a response to the cop's very subtle instruction, rather than to the actual presence of drugs.
*Should we mark this day and this post somehow, as it seems to be setting a record for highest degree of consensus in the history of the VC?
"Regardless of the means or technology used a person does not have standing to complain that his rights were violated by any search in that area."
Well, Kyllo says they can't use thermal imaging. Apparently dogs are different.
I agree that seems like an outlier. Using a device not in the general public use is somehow different. A dog is in the general public use.
BUT, a drug sniffing dog is not in the general public use. Kyllo is really not doctrinally consistent.
On a recent Business Flight, sitting in the last row of Delta First Class, I noted a TSA Officer was traveling home with her Bomb Sniffing Dog.
In the first row of Delta Comfort she had 1 seat for herself, 2 for her friend.
Dog slept the whole way, unlike most of the hellions parents insist on inflicting on everyone.
Frank
As an analogy to the dog situation, suppose a police camera drone malfunctioned through no fault of the operator, entered a space where its warrantless entry would be prohibited by the Fourth Amendment, and transmitted images giving evidence of crimes in progress. Since the police didn't deliberately direct the drone, and in fact didn't have control of it, would that not count as a search for Fourth Amendment purposes?
For that matter, what happens if a police car blows a tire, spins out of control, and crashes through the high fence around my yard, allowing the car's occupants to see the illegal activities going on in that yard? The police didn't intentionally enter the yard, so would their entry and observation be regarded as a search?
If we stick with the principle that the purpose of the exclusionary rule is to deter police misconduct it would be difficult to say that the car crashing through the fence would be something the police should be punished for.
The drone example is different. There is enough suspicion there that police will make a "mistake" that always has drones (and dogs) going onto private property.
In the car example, the police weren't conducting a search. In the drone and dog examples, they were, and they should be held to the standard that requires their searches to be constitutional.
If we stick with the principle that the purpose of the exclusionary rule is to deter police misconduct it would be difficult to say that the car crashing through the fence would be something the police should be punished for.
But are the police actually punished? I mean, the evidence is suppressed, but does that affect the individual cop?
Given the ambiguity around some of these cases/examples, isn't it a good idea to "build a fence" around the 4th Amendment?
IMHO, if a handler cannot control the dog well enough to prevent it from jumping through a window into a car, how reliable can we say the dog is at detecting contraband?
I agree with the general opinion here.
I also think Illinois v. Caballes was wrongy decided. I think the dissents have the better case.
The overall idea that drug sniffing dogs did not raise 4A concerns was a misguided enterprise. They violate privacy and are skilled agents of the state, imperfect agents (even good boys and girls) at that. Reasonableness and other rules should apply.
These questions remind me that lawyers are wired differently, or perhaps just broken during law school.
How could the intrusion of a government agent/instrument/asset be anything other than an act of the government? The dogs are typically wearing a fcking uniform. THEY AWARD THEM MEDALS.
The court should abandon the privacy nonsense on the Fourth Amendment for a simple test: if the government is looking for something, it's a search.
If they go digging in a public landfill for something someone discarded, it's a search. If they pick through someone's trash bin, it's a search.
Those might be searches, sure, but they're not the kind of search the Fourth Amendment is concerned with and the police don't need a warrant for them.
The problem is that, under the text of the Constitution, the 4th Amendment only applies to searches of "persons, houses, papers, and effects." I don't know if the maximalist view of what is a search is really simple, but even if it is, what do you do with that textual limit to searches of persons, houses, papers, and effects?
Just getting to the part early in The Digital Fourth Amendment where you discuss Katz. A phone call certainly isn't a person or a house. Is it an extension of papers and effects? Or was the theory that the phone booth is a temporary house and the wiretap was an effective intrusion into that house?
I suppose I should read the whole thing before asking...
But if the police search the landfill, and come with my "papers and effects," and search them, why isn't that a violation?
The 4th Amendment, after all, does not say anything about where the papers and effects mighty be located. Is the idea that once I have dumped them they are no longer mine? I'm not sure that's right, at least for 4A purposes.
"[E]ffects" is a very broad term. The words chosen in the Fourth Amendment encompassed essentially everything the government might be expected to search in that period when it was adopted.
The "privacy nonsense" is a core reason why the Fourth Amendment is there. The people behind it were concerned about privacy. Over time, it was a major concern.
Privacy is not be the only thing at issue here, but it is far from "nonsense" when determining what is reasonable or unreasonable.
A "public landfill," as noted in the replies, is not what the text of the amendment covers. A better borderline case would be personal trash. Do the police need a warrant to search a person's trash?
See, e.g., California v. Greenwood. The dissent makes some good points, and some state courts follow them.
Is the dog going into the car not a trespass?