Ripping Off the White Sheet »« More Material for the Book

Balko on the Drug Sniffing Dog Cases

The Supreme Court heard oral argument in two cases out of Florida involving drug-sniffing dogs last week and Radley Balko, the single best criminal justice reporter in the country, has an article at the Huffington Post about those cases, the precedents and how the outcome could further erode the 4th amendment. He points out something that may be far more important than it first appears:

One theme we continue to see in cases like these is that the Supreme Court lineup is woefully lacking experience in the actual practice of criminal law. Of the nine justices, only Sonia Sotomayor and Samuel Alito have any such experience, both as prosecutors. The court hasn’t had a justice with any real criminal defense experience since Thurgood Marshall retired in 1992. That’s worth restating: There hasn’t been a single voice on the Supreme Court with any real criminal defense experience in more than 20 years.

This is a remarkable and significant experience gap that isn’t often discussed, but has a profound effect on the context in which the Court attempts to maintain the delicate balance between liberty and security. It’s one thing to opine on these issues in law journals and lectures. It’s quite another to have real knowledge of how the law plays out in the real world. The gap was on full display in this week’s drug dog cases.

He also notes, as I have before, that the court has consistently expanded the uses of drug sniffing dogs and the erosion of the 4th Amendment by ruling, in every case so far, that such a search is not, for constitutional purposes, a search. In 1983 the court ruled that a dog sniffing for drugs in your luggage at the airport was not a search. Then in 2005, they extended that to searches of your car (despite the fact that such a search requires that they detain you, sometimes for long periods of time).

But the real problem is that while the court says the use of drug-sniffing dogs does not constitute a search and therefore does not require that the police have probable cause to initiate it, they’ve also ruled that if the dog “alerts” to the presence of drugs, that alert does constitute probable cause and thus give the police the right to perform a full search of your car. This might be acceptable if such an alert is real evidence of the presence of drugs, but studies have repeatedly shown that the dogs are wildly inaccurate. Balko describes one study that I’ve discussed several times, but he notes something I didn’t know about it — that the researchers had deliberately set a trap that demonstrated how much the expectation of the dog’s handler influences whether the dog alerts.

That assumption was wrong at the time, and it has been repeatedly proven wrong since. For example, in a survey of drug dogs used by police departments in suburban Chicago published last year, the Chicago Tribune found that when a police dog alerted to the presence of drugs during a traffic stop, a subsequent search turned up narcotics just 44 percent of the time. In stops involving Hispanic drivers, the dogs’ success rate dropped to 27 percent.

This raises some interesting questions: Why are drug dogs more likely to submit an innocent motorist to the indignity of a thorough roadside search if the motorist happens to be Hispanic? Are drug dogs racist? Do they racially profile? Of course not. But their handlers probably do.

Consider another study conducted by Lisa Lit, a neurologist and former dog handler at the University of California-Davis. Lit brought 18 dog/handler teams currently operating in law enforcement agencies to an empty church. Each team conducted eight searches, each lasting about five minutes. If they were accurate, none of the dog/handler teams should have alerted in any of the searches. There were no drugs or explosives anywhere in the church.

But Lit had set some traps. The handlers were told that each search could have as many as three “target scents” — drugs for the drug dog teams, or explosives for the explosive dog teams. The handlers were told that in some cases hot packages were indicated by a piece of red paper. These red paper packages were designed to trick the handlers. Lit also set a trap for the dogs: Some of the packages contained unwrapped sausages.

The results were striking. The dogs falsely alerted in 123 of the 144 total searches. Because some dogs falsely alerted more than once in the same search, the total number of false alerts was 225. The dogs correctly completed the search without an alert just 21 times, for a success rate of 14.5 percent.

But here’s the more interesting part: The dogs were about twice as likely to falsely alert at the packages designed to trick their handlers than they were at the packages stuffed with sausages.

The dogs are reacting to the subconscious clues given off by the handlers because they’ve been bred to do exactly that, not to detect drugs. And he points out how at least one justice still doesn’t get that, or understand the issue at all:

Since there are no national drug dog standards, most certification programs for drug dogs are given by police departments themselves. Garre argued that when a police department declares a dog to be “certified,” and worthy of use in the field, the courts should simply take the department’s word for it. The argument basically boiled down to, “just trust us.”

When the attorney for the defendant pointed out the absurdity of that argument, he received an odd but revealing grilling from Justice Antonin Scalia.

“What are the incentives here? Why would a police department want to use an incompetent dog?” Scalia asked. “Is that any more likely than that a medical school would want to certify an incompetent doctor? What incentive is there for a police department?”

Scalia came back to the point later, this time with a heavy dollop of sarcasm. “So let’s get dogs that, you know, smell drugs when there are no drugs. You really think that’s what’s going on here?” he asked. “It seems to me they have every incentive to train the dog well.”

As any defense attorney will tell you, however, there are plenty of incentives for police departments to have improperly trained dogs. A drug dog that’s prone to false alerts gives police more opportunities to search. That means more opportunities to find evidence of crimes not related to drugs — untaxed cigarettes, for example, or counterfeit passports.

What’s more, in many states, asset forfeiture laws allow police to seize property on little more than a drug dog’s alert. If you’re carrying a lot of cash and get pulled over, a police dog alerting to the presence of drugs in your car can be enough for the cop to take your cash, even if a subsequent search doesn’t turn up any actual drugs. In many cases, a drug dog’s alert to the presence of a drug on the cash itself has allowed police to seize the cash, even though nearly all U.S. currency contains traces of drugs. If a drug dog’s alert allows police to seize property and cash that then goes back to the police department, that would certainly be a disincentive to hold your department’s dogs to the highest possible standards.

But it should also be pointed out that there doesn’t need to be any incentive at all. The police officer doesn’t have to be acting consciously to trigger the dog’s alert. All he has to be doing is showing a subconscious expectation that drugs will be found, as the study discussed above shows. The officer likely isn’t even aware they’re doing it, but they’re more likely to focus the dog’s attention where they think he’s most likely to find drugs — which usually means in the cars of dark-skinned people, as study after study of police behavior shows. None of this needs to be overt or conscious.

I’m not encouraged about these two cases. I think it’s likely we’re going to get a 5/4 decision further eroding our civil liberties and weakening the 4th Amendment.

Comments

  1. Chiroptera says

    The dogs are reacting to the subconscious clues given off by the handlers because they’ve been bred to do exactly that, not to detect drugs.

    Good heavens! This is the sort of thing that gets taught in a first year college psychology course. Allegedly “magic” animals aren’t really magic but respond to their handlers’ unconscious cues.

    (By the way, I think Ed meand “trained,” not “bred.”

  2. greg1466 says

    The dogs were about twice as likely to falsely alert at the packages designed to trick their handlers than they were at the packages stuffed with sausages.

    The dogs are reacting to the subconscious clues given off by the handlers because they’ve been bred to do exactly that,…

    This should be patently obvious to any long time dog owner. Anyone who has a lot of experience with dogs knows that a dogs primary mission in life is to please their owner.

  3. greg1466 says

    @Chiroptera

    No, I think Ed actually means bred. Canis lupus familiaris has been selective bred for the trait to please it’s owner for something like 15,000 years now.

  4. eric says

    “What are the incentives here? Why would a police department want to use an incompetent dog?” Scalia asked

    While I think Balko is on the right track with his list, I think its worth pointing out that the counterargument aganist the police does not depend on proving they had some specific, alternate incentive. First, the legal validity of the test should depend mostly on its accuracy, not the motives of the police using it. If its inaccurate, it doesn’t matter whether they had good incentives or bad ones, its still inaccurate and should not be used. I do not want a cop breaking into my house based on a coin flip. I could care less whether he honestly believes a coin flip is a good system or he’s just using it as an excuse – a coin flip is an illegitimate tool either way.

    Second, the fact that police want to use inaccurate detection and forensic tests – which includes dogs – can be demonstrated independently of knowing what their true incentive is. The fact is amply demonstrated by prosecutorial and police resistance to correct blinding, independent assessment, and methodological improvement. No one who was truly interested in accuracy would actively resist suggested improvements from the forensic science community. They do. Therefore, they are not really interested in accuracy but something else. We can speculate about what that something else is, but the fact will remain that resistance to good science and methodological improvements is damning in and of itself.

  5. Chiroptera says

    eric, #5: “What are the incentives here? Why would a police department want to use an incompetent dog?” Scalia asked

    And here we see the typical mind set of the politically active conservatives in the US.

    It doesn’t make sense to him, therefore empirical evidence doesn’t count.

  6. gshelley says

    From what I have seen before (in the eyewitnesses cases), Sottomayer is very receptive to the actual facts and the science (kind of the opposite of Thomas, who in the other case to come before them recently was the lone dissent, ignoring the mounds of evidence and allowing himself to be impressed by the eyewitness statement)
    So she should go for it. No idea on the others

  7. otrame says

    My little terrier used to love to answer the door with me. She even loved Halloween, which drives most dogs bonkers. I answered the door to a salesman once. I was getting a bit of a bad vib off him, but not enough to make me particularly nervous. Jill was there, wagging her tail and grinning at him like she always did. Then, as I was closing the door, he leaned forward slightly and lifted his hand. I don’t think he was actually going to do anything. But the thought flashed into my mind that he was going to try to stop me closing the door.

    It was a good thing I had a hand on her collar. She hit the end of my arm snarling. Her whole body was making it clear she intended to take him. It was not at threat display. No barking. She meant business. If I hadn’t held on he would have gotten bitten badly. The guy backed off and practically ran. I hadn’t said a word. At most, I stiffened. I suppose I also gave off a change in scent, but I am not sure there was enough time between my sudden concern and her action for it to have been scent that told her I was afraid. I closed the door. It took me half an hour to get her away from it.

    So, yes, dogs are very much attuned to their people. I doubt drug sniffers are any different.

  8. says

    Yet another reason to thank all the Gods that Mutt Romney didn’t get elected on Tuesday.

    And Scalia? He’s once again proven he doesn’t give a shit about anything other than macho posturing and punishing people he hates, with as little mental effort as he can get away with. I can’t think of a less competent SCOTUS appointee — even two of Nixon’s worst jokes, Burger and Blackmun, managed to grow into their jobs and show some integrity.

  9. apovtak says

    Hi Ed – I really enjoy your blog, and I think you have an excellent grasp on constitutional law. But I think there’s an issue in this case that I think hasn’t been discussed yet, and one that I think is getting overlooked in your (and Balko’s) analysis of these cases is Kyllo v. US, 533 US 27 (2001) (if you need the cite). There, Scalia wrote for the majority that a warrantless thermal imaging search of a home (detecting heat, which then led to an arrest for growing marijuana in the home) was a search under the 4th Amendment and thus was unconstitutional.

    I think a fair reading of that case shows that the Court is more willing to rule against the government in terms of searches when the actual home is involved. I wouldn’t be surprised if that distinction could garner a majority on the Court this time around. One can make the argument that the method of search is irrelevant, as long as the home is involved. I haven’t had a chance to look at the arguments in the Florida case, and it seems that oral arguments didn’t really address that (well, at least from the Balko article, anyway). Oral arguments, though, don’t necessarily mandate the ensuing opinion (like CJ Roberts in the recent health-care case). Of course, Scalia might be indicating that he’s deferring to the assumed accuracy of drug dog searches (which, of course, is ably picked apart by Balko), but I wouldn’t discount the fact that Scalia (and Thomas) have been willing in the past to draw the line on unreasonable searches at the door of the home.

  10. sailor1031 says

    I agree that police don’t necessarily need an incentive; invalid results can simply be due to rampant police incompetence and unprofessionalism – which is at epidemic proportions in the USA.

    But corruption may also be a fundamental issue. Our former sheriff is currently under a 25 count indictment some of which involve converting forfeited property (money, boats) to his own use….. The former sheriffs of two neighboring counties are also under similar indictments…….nice little racket the three of them had going – over a period of many years!

  11. whheydt says

    When I read the reports of the oral arguments over these cases, there was one point that struck me, and that (so far as i know) the lawyer in question didn’t field in an effective way…

    One of the Justices asked what the difference was between a cop walking down the street and sniffing the air and a drug dog doing the same?

    What no one appeared to have thought of answering is, you can cross examine the cop at trial, but you can’t cross examine the dog.

  12. comfychair says

    Another way “the law plays out in the real world” which I haven’t seen mentioned in this or previous sniffer dog discussions, and one that has been used against me personally, is that the dog doesn’t have to alert to anything.

    I saw a cop walk a dog around my car 3 times, with the dog not doing anything except being interested in everything except my car. Sniffing towards the other side of the street and being dragged back over to the car, sniffing the pavement, sniffing the handler’s shoes, etc.

    And after three times round the car, the officer says “Sir the dog has alerted to the presence of narcotics in the vehicle, I now have probable cause to search your vehicle.”

    Of course, there was nothing in the car, and never has been. They were just determined to search my car, especially after I refused consent, and what they did could have just as easily (and legally, apparently) been done with a toy stuffed dog.

  13. badandfierce says

    It may be erroneous to assume the dogs aren’t themselves racist, in a certain sense. Human races may be meaningless to dogs as a whole, but individual dogs will pick up cues from their humans, and seem to be perfectly capable of acknowledging skin color as a variant to watch out for. I can’t find any scholarly articles addressing the point, and anecdotes certainly aren’t data, but it’s generally accepted in animal-keeping circles. So even if the current handler isn’t directing the dog toward alerting, any previous handler with racist inclinations could be contaminating the success rate.

  14. Ben P says

    The dogs are reacting to the subconscious clues given off by the handlers because they’ve been bred to do exactly that, not to detect drugs.

    I take pretty serious issue with this statement because it is seriously misleading. Its not exactly true or untrue, but it misstates the problem. The problem is one of training, and not working the dog correctly.

    First, there really are no such things as dogs “bred for drug sniffing” much less dogs bred to read subconscious cues. Some dog breeds have more sensitive noses, and some are more intelligent than others and have a higher drive than others. Working dogs as a whole tend to be more intelligent and have a higher drive. Working strains within breeds have specific qualities you want.

    Second, like I said above the problem is one of training, both training the dogs properly and perhaps more importantly here, training the handlers properly.

    I work with german shepards on a competition level. My current girl is up to Sch2. Schutzhund training is effectively a “base layer” of training similar to that which most police and military dogs get. It has three main components, tracking, obedience and protection. Training as a “sniffing” dog is similar to tracking training.

    Training a dog to the level many police dogs work at takes literally a year or more of day in day out work. But more importantly, it requires a handler that knows how to work the dog.

    That’s where most police departments fail miserably. Most departments don’t have the funds to train dogs of their own, so they purchase already trained dogs and run their “K9 officers” through a weeklong training course on dog handling.

    This is the problem, and it also gets to the grain of truth in what you said. The dogs are quite a bit smarter than you give them credit for and yes, they do pick up on subconcious cues quite well. However, the important thing to remember about dog handling is that the dog isn’t thinking the same thing as you. A drug dog doesn’t actually give a damn about the drugs, it’s been trained to associate the smell of drugs with finding a reward (active indication) or to associate performing some behavior when it smells drugs with a reward (passive indication).

    However, it is EXTREMELY easy for a imprecise handler to muck this all up. It only takes a couple of repeats of a bad habit, like praising a dog for indicating regardless of whether drugs are found to corrupt a dogs training. They start associating some cue other than a drug smell with a reward, and when they’re trying to figure out that cue, they’re a lot more perceptive than you are. Tones of voice, posture, tenseness in lead handling. This is precisely the clever hans effect. It doesn’t take long before you have a dog that isn’t trained to find drugs, so much as it’s trained to find drugs, and to pretend it found drugs when the officer gives a cue, whether intended or subconconsious. A bad handler praises the dog either way and confuses the dog more.

  15. Ben P says

    It may be erroneous to assume the dogs aren’t themselves racist, in a certain sense. Human races may be meaningless to dogs as a whole, but individual dogs will pick up cues from their humans, and seem to be perfectly capable of acknowledging skin color as a variant to watch out for. I can’t find any scholarly articles addressing the point, and anecdotes certainly aren’t data, but it’s generally accepted in animal-keeping circles.

    Virtually all the dog professionals I know treat this as a socialization problem. Picking up cues from an owner may reinforce it, but more often than not the problem is that a dog was never exposed to, just for example, black people, and will react to them much differently than people that it sees as familiar.

    It’s not that the dog is racist per se, it’s not really any different from the fact that a dog raised from a puppy with cats will tolerate cats entirely, even new cats, whereas an adult dog that’s never seen a cat before will very often react with hostility or fear to the cat.

  16. Ben P says

    This should be patently obvious to any long time dog owner. Anyone who has a lot of experience with dogs knows that a dogs primary mission in life is to please their owner.

    Again, not precisely untrue, but just not accurate. Domestic dogs were bred from wolves and wild desert dogs for certain traits including friendliness and lack of strong dominance behavior.

    It is true that dogs’ natural social “pack” behavior gets channeled through this lens into trying to please its owners, but this isn’t innate. I think you only have to watch any of a various number of “bad dog” or dog training shows to realize this. The common theme running through most of these shows is that very rarely are there truly bad dogs, but there are lots of bad owners. The dogs have their own motivations and social behavior and while they may well be a member of your family, they’re not little people and don’t think the same way.

  17. lofgren says

    The common theme running through most of these shows is that very rarely are there truly bad dogs, but there are lots of bad owners.

    Most of the episodes that I have seen have the trainer teaching the owner to make it more clear what does and doesn’t please him. The dog wants to be helpful and useful and generally please the owner, but the owner does not know how to communicate with the dog which behaviors the owner likes and which he doesn’t. Because of this, the dog opts for those behaviors that get him the most attention. The dog can’t magically know what pleases you, but once you figure out how to communicate that to the dog, the dog will usually try to do those things. Of course like all animals that drive competes with others, but it’s definitely one that we have selected for in almost every extant breed.

  18. Ben P says

    Because of this, the dog opts for those behaviors that get him the most attention. The dog can’t magically know what pleases you, but once you figure out how to communicate that to the dog, the dog will usually try to do those things. Of course like all animals that drive competes with others, but it’s definitely one that we have selected for in almost every extant breed.

    You say it, but I’m not sure you recognize it.

    The dog, as a pack animal, really likes attention. There are few things more painful (psychologically, not literally) to a dog than isolation and boredom. Attention and play, is itself a reward to the dog and can be used to condition behaviors.

    That’s why I’m not saying it’s wrong to say a dog wants to please its owners, it’s just not really accurate. The dog is doing what it does because it thinks it will get something it likes, not because it wants to make you happy in some abstract way.

    Bringing this out of the abstract and into police dogs. The proper training is that a dog gets rewards when it actually finds drugs, not when it indicates in any other scenario.

    But for a handling police officer who’s not very vigilant about how they act around the dog, it’s very natural to praise the dog for indicating regardless of whether it’s actually found something, particularly if nothing is there or searching will take place much later.

    So the dog indicates, the cop says “good boy,” and pets the dog, before drugs are ever found, and the dog thinks “how do I get him to do that again?”

  19. lofgren says

    I think dogs are capable of recognizing “happiness” in an abstract way. They know when you are happy and they know when you are sad, and part of their pack drive is to keep those around them happy. Yes, this competes with their desire for attention, for excitement, for purpose, and for food, sleep, air, etc. But I definitely believe it is there. Basically a dog who gets attention and thinks he is making his master happy will be better adjusted than a dog who gets attention but gets mixed signals from his master or thinks he is making his master angry, all else being equal.

    I think humans in general gravely underestimate the ability of social animals in general to recognize abstract emotions. My dog definitely recognizes when I am upset, whether it is her fault or not, and she will often run through her litany of tricks to see if any of them please me. Again, I’m not saying that dogs don’t have a drive for attention and excitement, as you say. I’m saying that, in addition to those, they also have sufficient empathy to want the people around them to be happy, and if they are sufficiently socialized with humans then they can recognize the signs that a human, or at least their master, is not happy.

Leave a Reply