SCOTUS Issues Good Ruling on Police Dogs


After royally screwing up a ruling on the use of dogs to sniff for drugs a few weeks ago, and in the midst of all the attention paid to the marriage equality cases, the Supreme Court issued a ruling in Florida v Jardines that at least puts some limits on the use of dogs to sniff around our homes without a warrant.

At issue in the case was whether the police could bring a drug-sniffing dog up to the front door of a home to smell for drugs without a warrant. In a 5-4 decision, the court ruled that this was unconstitutional to do without a warrant. And the lineup here will probably surprise you. The majority opinion was written by Justice Scalia (who actually has a decent record on 4th Amendment cases) and was joined by Justices Thomas (!), Sotomayor, Ginsburg and Kagan. The dissent was written by Justice Alito, joined by Chief Justice Roberts, Kennedy and Breyer (what?).

That principle renders this case a straightforward one. The officers were gathering information in an area belonging to Jardines and immediately surrounding his house— in the curtilage of the house, which we have held enjoys protection as part of the home itself. And they gathered that information by physically entering and occupying the area to engage in conduct not explicitly or implicitly permitted by the homeowner. …

[W]hen it comes to the Fourth Amendment, the home is first among equals. At the Amendment’s “very core” stands “the right of a man to retreat into his own home and there be free from unreasonable governmental intrusion.” This right would be of little practical value if the State’s agents could stand in a home’s porch or side gar- den and trawl for evidence with impunity; the right to retreat would be significantly diminished if the police could enter a man’s property to observe his repose from just outside the front window. …

We therefore regard the area “immediately surrounding and associated with the home”—what our cases call the curtilage—as “part of the home itself for Fourth Amendment purposes.”

The depressing thing is that four justices, including one of the liberal lions of the court, could possibly dissent from this. But at least the court got it right. The use of drug-sniffing dogs should be eliminated entirely, something the court failed to do when they had the chance, but this is at least some minimal protection.

Comments

  1. TGAP Dad says

    I think it would be great to see the “drug-sniffing dogs” come under scrutiny similar to that currently aimed at breathalyzers. It seems that we have given them (and their handlers) a pass because, I suspect, everyone loves dogs. Is there any actual evidence out there that dogs actually deserve their reputation for detection ability, or are they more like 4-legged dowsing rods? Are there any actual peer-reviewed studies on this? Are there any standards required by the courts regarding what constitutes a dog’s alert, or do they simply trust the handler? (I mean, what if he just gives the dog a subtle command to sit, tells the subject that this is the dog’s alert, and commences a search?)

  2. Artor says

    TGAP, there’s a lot of data on the effectiveness of police sniffer dogs. The results are pretty conclusive that they are excellent at giving the response their handler wants, but not so much at actually finding drugs. In one blind test, they returned false positives consistently when their handler expected there to be drugs, but only achieve something like 40% reliability otherwise. Worse than random guessing.

  3. ibbica says

    TGAP Dad

    Is there any actual evidence out there that dogs actually deserve their reputation for detection ability, or are they more like 4-legged dowsing rods?

    Off the top of Google’s head…
    http://lawreport.org/pdf/K9sniffAsWitness.pdf

    Part of the problem is the training procedures used. Drug-sniffing dogs are tested by having them find a ‘target’ against a background, which they’re of course very good at. But there’s typically no reward for signalling “there’s nothing here!” (and AFAIK such a signal is not taught; certainly not “there’s a little bit of residual smell here but not enough to worry about”). The assumption is that there is something there to find (otherwise, why bring the dog?), and if the dog doesn’t find it they’re not rewarded. They work hard to find the slightest whiff of something that might be what they’re supposed to be looking for. Sometimes if a dog is not signalling a hit the handler will ‘plant’ a target to make sure the dog can gain a reward, but of course some handlers will be quicker than others to give the searchee the benefit of the doubt.

    “Well-trained” dogs are those that read their handlers VERY well. Handlers of sniffer dogs can swear up and down that they didn’t signal the dog to ‘hit’, and I’ll give them enough credence to believe that they’re not intentionally signalling their dog. But they most certainly do give off signals that their dogs can respond to.

  4. says

    The reliability of dogs does not appear to be the issue here. This case seems to be all about WHERE the dogs were when they “alerted.” I presume that if there’d been a valid warrant, or if the dogs “alerted” in some situation otuside the “curtilage,” the Court would not have questioned the reliability of Man’s Best Friend.

  5. wscott says

    The use of drug-sniffing dogs should be eliminated entirely,

    I think that goes too far. Definitely need more studies to identify their limitations and develop better training standards, and improved certification requirements. But they’re still a useful tool. I’d be curious to what degree bomb-sniffing and other types of “working dogs” suffer from the same issues?
    .
    There are a lot of criminal justice “beliefs” that have been used for generations that only recently have started to be tested scientifically. For example, many of the basic principles of arson investigation have turned out to be little more than old wives tales. So yes to more studies & better training; but I’m not ready to throw the baby out with the bathwater just yet. (Unless we’re going to end the drug war altogether, of course.)

  6. Phillip IV says

    And the lineup here will probably surprise you.

    IPerhaps Justice Breyer is just more of a cat person?

    theschwa @ # 4:

    I have not read the decision. Is there any chance the dissent was for procedural (standing) reasons?

    As far as I can see it wasn’t procedural. The dissenters are unhappy with the way the Court applied the trespass rule. They’re arguing that, since the cop wouldn’t have been found to be trespassing if he approached the front door without his dog, how could it suddenly be trespassing just because the dog accompanied him?

  7. gshelley says

    Is Kagan a surprise? She seems to have been fairly consistently on the more liberal side, even when the administration she had previously represented was not. In most of these cases, the Obama administration has been on the wrong side.

  8. Ben P says

    TGAP, there’s a lot of data on the effectiveness of police sniffer dogs. The results are pretty conclusive that they are excellent at giving the response their handler wants, but not so much at actually finding drugs. In one blind test, they returned false positives consistently when their handler expected there to be drugs, but only achieve something like 40% reliability otherwise. Worse than random guessing.

    It’s pretty conclusive only if you cherry pick. Several such studies are cited by Souter in the dissent in Illinois v Caballes. They find error rates from 12.5% to 60%.

    More importantly, there are two very distinct issues here. One is whether a properly trained dog can be accurate, the other is whether in field conditions with inexpert handlers dogs actually are accurate.

    I think the answer to the first question is almost indisputably yes. There are routine and regularly held competitions for detection dogs, not only drugs, but explosives and other items as well. These competitions are blind to the handlers as a matter of competitive fairness. (i.e. they are deliberately designed such that the handlers do not know the precise location of the target.).

    The one exception is that competitions very rarely involve a “null” set where there is no target whatsoever.

    However, it’s very important to note that you’re involving people who are typically experts, and who are working with their personal pet dogs, with whom they have spent hours and hours training. It is not in their interest that the dogs either fail to indicate or indicate false positives. It would come as a great shock to most competitive detection trainers to suggest that dogs are actually incapable of tracking by scent.

    Your average K9 Unit in a police department purchases trained dogs from a breeder/trainer, then K9 officers attend a week to two week training course on how to handle the dogs, then (usually, not always) regular follow up tests. Very few police departments invest the money to develop a program from scratch.

    Then the officers get into the field, and they are routinely in situations where false positives are not a negative, but a positive. if the dog indicates they can search, and the training can, and studies show, often does, break down. The dogs get rewarded for positive indications, not merely accurate ones, and start associating some signal other than only the scent of drugs for an indication. Then you end up with officers like the one who testified in Commonwealth. (cited above)

    Q: do you keep records of the effectiveness of your dog?
    A: yes sir I do
    Q: Do you know how often your dog gives false positives?
    A: he doesn’t give any false positives, we’re just unable to verify the alerts he gives at the time.

    Police and prosecutors say this with a straight face, it’s absurd, but also has a hint of truth. In field conditions you don’t know that drugs aren’t absent, only that you don’t see any. That’s not precisely the same thing. .

    Are there any standards required by the courts regarding what constitutes a dog’s alert, or do they simply trust the handler? (I mean, what if he just gives the dog a subtle command to sit, tells the subject that this is the dog’s alert, and commences a search?)

    Having worked with dogs and knowing what’s involved in the training, for active indication dogs the activity is difficult to mistake.

    To train an active indication dog, effectively what you are doing is training the dog to associate the smell of cocaine (or whatever) with its favorite toy. You start by training the dog to find the toy, then you scent the toy with imitation drug scent, and continue the same training. It doesn’t take long for the dog to figure out that the smell of drugs means toy. the indication behavior you get is barking, pawing, trying to get to the source of the smell.

    There’s a separate kind of training called passive indication, that trains the dog to sit or lie down when it smells the substance. That’s more commonly associate with explosive dogs (and certain breeds) but some drug dogs are trained that way as well.

  9. Ben P says

    Police and prosecutors say this with a straight face, it’s absurd, but also has a hint of truth. In field conditions you don’t know that drugs aren’t absent, only that you don’t see any. That’s not precisely the same thing. .

    To be precise, you don’t know that the scent of drugs is absent, only that you don’t see any drugs. There is some possibility of the scent of drugs being present, but anything more than residue being gone.

    This also perfectly illustrates the training breakdown.

    The training would require the dog indicating, and then only being rewarded if drugs are actually present.

    As police officers often do it (just because of human nature) the dog will indicate, then they praise/reward the dog, then they search, and if the search finds nothing, justify it to themselves by saying “well the dog could have just smelled something that we can’t see.”

  10. Michael Heath says

    I haven’t seen any opinion trends by political ideology since the Rehnquist Court. But I have a strong hunch Justice Alito is the most reliable extant justice siding with the conservative political agenda.

  11. Ben P says

    The use of drug-sniffing dogs should be eliminated entirely,

    I think this is wrong as well, although I will say that as a matter of practice, a lot of situations where a drug dog might be used should be all but invalid as evidence.

    Any situation where a dog is brought in to sniff a single target and produce an “indication/no indication” result is highly suspicious. It’s reasonable to assume the dog has already been called in because the officer thinks there’s a chance drugs will be detected, and that’s exactly the point where there is the most dangerous potential for clever hans problems.

    On the other hand, consider a dog in the back room of an airport sniffing a line of packages and suitcases, or at a border checkpoint being walked around every car that goes through? These are markedly different situations.

  12. says

    This isn’t a good ruling… it is a less than abysmal ruling.
    5-4 is way to close for something this cut and dry.

  13. says

    The Clever Hans aspect along with the asymmetrical rewards really undermines my confidence in the program in general. I don’t think dogs are ideological enough to put honesty above the demands of peer pressure. If you can devise a training program where a dog is discouraged from false positives and encouraged to signal true negatives in spite of his handler’s wishes, I’d like to hear it.

  14. wscott says

    The dissenters are unhappy with the way the Court applied the trespass rule. They’re arguing that, since the cop wouldn’t have been found to be trespassing if he approached the front door without his dog, how could it suddenly be trespassing just because the dog accompanied him?

    That’s actually a fair point. If a cop walks up to your door and smells pot, that’s probable cause; but if he has a dog with him and the dog smells pot, that’s NOT probable cause? Setting aside reliability questions (which it doesn’t sound like the court addressed), that does seem a bit inconsistent.
    .
    IANAL, but my understanding is the courts have usually held that use of technology doesn’t change fundamental rights. For example: you have a reasonable expectation of privacy in your phone conversations, so the fact that government develops new ways to make eavesdropping easier doesn’t eliminate the warrant requirement. But conversely, if you don’t have a reasonable expectation of privacy – say, a conversation in a public space – then the use of audio equipment to make eavesdropping easier does not change the lack of a warrant requirement. In other words, the tool used is generally irrelevant to the privacy rights question.
    .
    The question becomes a bit murkier in cases where someone outside is picking up “emissions” that are technically leaving your property, e.g. thermal viewers. But it’s surprising for the court to say that the government is allowed to do something, but is NOT allowed to use tools to help. (Again, setting aside the issue of whether the tool is reliable.)
    .
    I’m not saying I’m unhappy, just surprised.

  15. freemage says

    I’m not surprised at Thomas, simply because as far as I’ve ever been able to tell, he’s Scalia’s muppet. I’m sure there’s some cases where he’s demurred from ol’ Anton, but I doubt it’s many.

  16. Ben P says

    That’s actually a fair point. If a cop walks up to your door and smells pot, that’s probable cause; but if he has a dog with him and the dog smells pot, that’s NOT probable cause? Setting aside reliability questions (which it doesn’t sound like the court addressed), that does seem a bit inconsistent.

    You’re slightly wrong about the nature of the ruling here, but I’ll get to that in a second.

    I think the opinion in this case follows relatively naturally from Kyllo (the thermal viewer case).

    The police in Kyllo used a thermal sensor to view the walls of the house, the thermal sensor indicated a high temperature area inside, based on the thermal sensor + expert testimony that marijuana growing lights would produce such a reading, they got a warrant to search the house.

    The defendant argued “no that’s invalid because the sensor sees inside the house it is a search in and of itself, and you should have had a warrant before using it,. The police argued that no it’s not a search because the thermal sensor is really just viewing the outside of the house, and isn’t seeing anything a police officer couldn’t see if he could see heat.

    This case is similar in a way. The holding wasn’t that a dog on the porch isn’t probable cause, but that the police officer bringing his dog up onto the porch is itself a search, and therefore requires a warrant. The police used the same argument they used in Kyllo, the police officer could be there normally, and the dog isn’t smelling anything that the officer couldn’t smell if he had a sensitive nose.

    The supreme court took a slightly different tack.

    The police officer has implied consent to go up onto your porch for the purpose of knocking on your door, because many such people (salesmen etc) would do the same. However, this implied consent doesn’t stretch to bringing a dog to search your porch.

    beyond that, the fact that the dog is smelling something inside the house is an additional point.

  17. naturalcynic says

    From what I have seen and read about doggy learning, it would be difficult to create a raining protocol that can reward dogs properly where false positives are concerned. Dogs respond to immediate rewards and punishments, delayed rewards and punishments do not work. The most common situation where people don’t understand doggie psychology is the delayed rubbing the nose in the pee spot on the carpet after the fact. If the drug dog signals a positive, immediate reinforcements are given. If, after a search of that location, nothing is found, there is no way to convey to the dog that nothing was found.

  18. Ben P says

    From what I have seen and read about doggy learning, it would be difficult to create a raining protocol that can reward dogs properly where false positives are concerned. Dogs respond to immediate rewards and punishments, delayed rewards and punishments do not work. The most common situation where people don’t understand doggie psychology is the delayed rubbing the nose in the pee spot on the carpet after the fact. If the drug dog signals a positive, immediate reinforcements are given. If, after a search of that location, nothing is found, there is no way to convey to the dog that nothing was found.

    Again with the training vs the field.

    In training, the dog signals, you flip over the box, and the drugs are there or they aren’t there. You reward the dog for being right, no reward for being wrong. You specifically don’t reward just for an indication.

    Good detection dog handling requires not rewarding the dog for indications only. However, to keep the dog from getting confused (or disillusioned if the concept applies) by not being rewarded every time they are right, you should, very regularly, be reinforcing the training.

  19. Phillip IV says

    fastlane @ # 19:

    I’d be more interested in the dissent in this case. Anyone got a link?

    Here you go, dissent starts on page 18 of the document.

  20. wscott says

    @ Ben P – thanks. I haven’t had time to read the actual ruling yet. So it sounds like they came at it from a trespass angle, rather than a privacy angle as such? Cop on porch = no search; cop on porch + dog = search? Was that from the majority opinion, or the dissent or both? Still seems like a fine line, but less silly than I thought.

    I was thinking of the thermal imager case too. But in this case, the ability to smell pot is something an officer is already capable of doing; the dog just does it better. Whereas “the officer could see the heat if he had superpowers” argument was significantly more absurd. Again, I’m not displeased with the ruling; just a little surprised.

  21. says

    The use of drug-sniffing dogs should be eliminated entirely…

    If we ever find a more reliable means of detecting drugs, explosives, or other contraband, I’ll be all for that. (And given the pace of technological advancement, I wouldn’t rule it out.) At the very least, we should understand that, when the Founders wrote that bit about “No warrants shall issue but upon Probable Cause,” we can be pretty sure they meant — and most reasonably educated people understand it to mean — probable cause as determined by sentient humans capable of legal and moral reasoning; not dogs, cats, badgers, hedgehogs, or even crows or eagles or my ex-wife’s grumpy old calico (who wouldn’t have “alerted” to anything if it meant losing the lap she was sleeping on).

Leave a Reply